North American Chapter of the Association for Computational Linguistics (2024)

Volumes

up

pdf (full)
bib (full) Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)

pdf bib
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)
Kevin Duh | Helena Gomez | Steven Bethard

pdf bib
Named Entity Recognition Under Domain Shift via Metric Learning for Life Sciences
Hongyi Liu | Qingyun Wang | Payam Karisani | Heng Ji

Named entity recognition is a key component of Information Extraction (IE), particularly in scientific domains such as biomedicine and chemistry, where large language models (LLMs), e.g., ChatGPT, fall short. We investigate the applicability of transfer learning for enhancing a named entity recognition model trained in the biomedical domain (the source domain) to be used in the chemical domain (the target domain). A common practice for training such a model in a few-shot learning setting is to pretrain the model on the labeled source data, and then, to finetune it on a hand-full of labeled target examples. In our experiments, we observed that such a model is prone to mislabeling the source entities, which can often appear in the text, as the target entities. To alleviate this problem, we propose a model to transfer the knowledge from the source domain to the target domain, but, at the same time, to project the source entities and target entities into separate regions of the feature space. This diminishes the risk of mislabeling the source entities as the target entities. Our model consists of two stages: 1) entity grouping in the source domain, which incorporates knowledge from annotated events to establish relations between entities, and 2) entity discrimination in the target domain, which relies on pseudo labeling and contrastive learning to enhance discrimination between the entities in the two domains. We conduct our extensive experiments across three source and three target datasets, demonstrating that our method outperforms the baselines by up to 5% absolute value. Code, data, and resources are publicly available for research purposes: https://github.com/Lhtie/Bio-Domain-Transfer .

pdf bib
Text Diffusion Model with Encoder-Decoder Transformers for Sequence-to-Sequence Generation
Hongyi Yuan | Zheng Yuan | Chuanqi Tan | Fei Huang | Songfang Huang

The diffusion model, a new generative modeling paradigm, has achieved great success in image, audio, and video generation.However, considering the discrete categorical nature of the text, it is not trivial to extend continuous diffusion models to natural language. In this work, we propose SeqDiffuSeq, a text diffusion model, to approach sequence-to-sequence text generation with an encoder-decoder Transformer architecture.To improve the generation performance, SeqDiffuSeq is equipped with the self-conditioning technique and our newly proposed adaptive noise schedule technique. Self-conditioning enables SeqDiffuSeq to better use the predicted sequence information during the generation process.The adaptive noise schedule balances the difficulty of denoising across time steps at the token level.Experiment results illustrate the improved performance on five sequence-to-sequence generation tasks compared to other diffusion-based models regarding text quality and inference time.

pdf bib
An Interactive Framework for Profiling News Media Sources
Nikhil Mehta | Dan Goldwasser

The recent rise of social media has led to the spread of large amounts of fake and biased news, content published with the intent to sway beliefs. While detecting and profiling the sources that spread this news is important to maintain a healthy society, it is challenging for automated systems.In this paper, we propose an interactive framework for news media profiling. It combines the strengths of graph based news media profiling models, Pre-trained Large Language Models, and human insight to characterize the social context on social media. Experimental results show that with as little as 5 human interactions, our framework can rapidly detect fake and biased news media, even in the most challenging settings of emerging news events, where test data is unseen.

pdf bib
Assessing Logical Puzzle Solving in Large Language Models: Insights from a Minesweeper Case Study
Yinghao Li | Haorui Wang | Chao Zhang

Large Language Models (LLMs) have shown remarkable proficiency in language understanding and have been successfully applied to a variety of real-world tasks through task-specific fine-tuning or prompt engineering. Despite these advancements, it remains an open question whether LLMs are fundamentally capable of reasoning and planning, or if they primarily rely on recalling and synthesizing information from their training data. In our research, we introduce a novel task—Minesweeper—specifically designed in a format unfamiliar to LLMs and absent from their training datasets. This task challenges LLMs to identify the locations of mines based on numerical clues provided by adjacent opened cells. Successfully completing this task requires an understanding of each cell’s state, discerning spatial relationships between the clues and mines, and strategizing actions based on logical deductions drawn from the arrangement of the cells. Our experiments, including trials with the advanced GPT-4 model, indicate that while LLMs possess the foundational abilities required for this task, they struggle to integrate these into a coherent, multi-step logical reasoning process needed to solve Minesweeper. These findings highlight the need for further research to understand the nature of reasoning capabilities in LLMs under similar circumstances, and to explore pathways towards more sophisticated AI reasoning and planning models.

pdf bib
TelME: Teacher-leading Multimodal Fusion Network for Emotion Recognition in Conversation
Taeyang Yun | Hyunkuk Lim | Jeonghwan Lee | Min Song

Emotion Recognition in Conversation (ERC) plays a crucial role in enabling dialogue sys- tems to effectively respond to user requests. The emotions in a conversation can be identi- fied by the representations from various modal- ities, such as audio, visual, and text. How- ever, due to the weak contribution of non-verbal modalities to recognize emotions, multimodal ERC has always been considered a challenging task. In this paper, we propose Teacher-leading Multimodal fusion network for ERC (TelME). TelME incorporates cross-modal knowledge distillation to transfer information from a lan- guage model acting as the teacher to the non- verbal students, thereby optimizing the efficacy of the weak modalities. We then combine multi- modal features using a shifting fusion approach in which student networks support the teacher. TelME achieves state-of-the-art performance in MELD, a multi-speaker conversation dataset for ERC. Finally, we demonstrate the effec- tiveness of our components through additional experiments.

pdf bib
Effective and Efficient Conversation Retrieval for Dialogue State Tracking with Implicit Text Summaries
Seanie Lee | Jianpeng Cheng | Joris Driesen | Alexandru Coca | Anders Johannsen

Few-shot dialogue state tracking (DST) with Large Language Models (LLM) relies on an effective and efficient conversation retriever to find similar in-context examples for prompt learning. Previous works use raw dialogue context as search keys and queries, and a retriever is fine-tuned with annotated dialogues to achieve superior performance. However, the approach is less suited for scaling to new domains or new annotation languages, where fine-tuning data is unavailable. To address this problem, we handle the task of conversation retrieval based on text summaries of the conversations.A LLM-based conversation summarizer is adopted for query and key generation, which enables effective maximum inner product search. To avoid the extra inference cost brought by LLM-based conversation summarization, we further distill a light-weight conversation encoder which produces query embeddings without decoding summaries for test conversations. We validate our retrieval approach on MultiWOZ datasets with GPT-Neo-2.7B and LLaMA-7B/30B. The experimental results show a significant improvement over relevant baselines in real few-shot DST settings.

pdf bib
Promptly Predicting Structures: The Return of Inference
Maitrey Mehta | Valentina Pyatkin | Vivek Srikumar

Prompt-based methods have been used extensively across NLP to build zero- and few-shot label predictors. Many NLP tasks are naturally structured: that is, their outputs consist of multiple labels which constrain each other. Annotating data for such tasks can be cumbersome. Can the promise of the prompt-based paradigm be extended to such structured outputs? In this paper, we present a framework for constructing zero- and few-shot linguistic structure predictors. Our key insight is that we can use structural constraints—and combinatorial inference derived from them—to filter out inconsistent structures predicted by large language models. We instantiated this framework on two structured prediction tasks, and five datasets. Across all cases, our results show that enforcing consistency not only constructs structurally valid outputs, but also improves performance over the unconstrained variants.

pdf bib
On Linearizing Structured Data in Encoder-Decoder Language Models: Insights from Text-to-SQL
Yutong Shao | Ndapa Nakashole

Structured data, prevalent in tables, databases, and knowledge graphs, poses a significant challenge in its representation. With the advent of large language models (LLMs), there has been a shift towards linearization-based methods, which process structured data as sequential token streams, diverging from approaches that explicitly model structure, often as a graph. Crucially, there remains a gap in our understanding of how these linearization-based methods handle structured data, which is inherently non-linear.This work investigates the linear handling of structured data in encoder-decoder language models, specifically T5. Our findings reveal the model’s ability to mimic human-designed processes such as schema linking and syntax prediction, indicating a deep, meaningful learning of structure beyond simple token sequencing. We also uncover insights into the model’s internal mechanisms, including the ego-centric nature of structure node encodings and the potential for model compression due to modality fusion redundancy. Overall, this work sheds light on the inner workings of linearization-based methods and could potentially provide guidance for future research.

pdf bib
Extractive Summarization with Text Generator
Thang Le | Anh Tuan Luu

Standard extractive systems suffer from the lack of gold training signals since existing corpora solely provide document and human-written summary pairs while disregarding extractive labels. As a result, existing methods resort to imperfect pseudo-labels that are both biased and error-prone, thereby hindering the learning process of extractive models. In contrast, text generators which are commonly employed in abstractive summarization can effortlessly overcome this predicament on account of flexible sequence-to-sequence architectures. Motivated to bypass this inherent limitation, we investigate the possibility of conducting extractive summarization with text generators. Through extensive experiments covering six summarization benchmarks, we show that high-quality extractive summaries can be assembled via approximating the outputs (abstractive summaries) of these generators. Moreover, we find that the approximate summaries correlate positively with the auxiliary summaries (i.e. a better generator enables the production of better extractive summaries). Our results signify a new paradigm for training extractive summarizers i.e. learning with generation (abstractive) objectives rather than extractive schemes.

pdf bib
Self-generated Replay Memories for Continual Neural Machine Translation
Michele Resta | Davide Bacciu

Modern Neural Machine Translation systems exhibit strong performance in several different languages and are constantly improving. Their ability to learn continuously is, however, still severely limited by the catastrophic forgetting issue. In this work, we leverage a key property of encoder-decoder Transformers, i.e. their generative ability, to propose a novel approach to continually learning Neural Machine Translation systems. We show how this can effectively learn on a stream of experiences comprising different languages, by leveraging a replay memory populated by using the model itself as a generator of parallel sentences. We empirically demonstrate that our approach can counteract catastrophic forgetting without requiring explicit memorization of training data. Code will be publicly available upon publication.

pdf bib
Measuring and Improving Chain-of-Thought Reasoning in Vision-Language Models
Yangyi Chen | Karan Sikka | Michael Cogswell | Heng Ji | Ajay Divakaran

Vision-language models (VLMs) have recently demonstrated strong efficacy as visual assistants that can parse natural queries about the visual content and generate human-like outputs. In this work, we explore the ability of these models to demonstrate human-like reasoning based on the perceived information. To address a crucial concern regarding the extent to which their reasoning capabilities are fully consistent and grounded, we also measure the reasoning consistency of these models. We achieve this by proposing a chain-of-thought (CoT) based consistency measure. However, such an evaluation requires a benchmark that encompasses both high-level inference and detailed reasoning chains, which is costly. We tackle this challenge by proposing an LLM-Human-in-the-Loop pipeline, which notably reduces cost while simultaneously ensuring the generation of a high-quality dataset. Based on this pipeline and the existing coarse-grained annotated dataset, we build the CURE benchmark to measure both the zero-shot reasoning performance and consistency of VLMs. We evaluate existing state-of-the-art VLMs, and find that even the best-performing model is unable to demonstrate strong visual reasoning capabilities and consistency, indicating that substantial efforts are required to enable VLMs to perform visual reasoning as systematically and consistently as humans. As an early step, we propose a two-stage training framework aimed at improving both the reasoning performance and consistency of VLMs. The first stage involves employing supervised fine-tuning of VLMs using step-by-step reasoning samples automatically generated by LLMs. In the second stage, we further augment the training process by incorporating feedback provided by LLMs to produce reasoning chains that are highly consistent and grounded. We empirically highlight the effectiveness of our framework in both reasoning performance and consistency.

pdf bib
Building Knowledge-Guided Lexica to Model Cultural Variation
Shreya Havaldar | Salvatore Giorgi | Sunny Rai | Thomas Talhelm | Sharath Chandra Guntuku | Lyle Ungar

Cultural variation exists between nations (e.g., the United States vs. China), but also within regions (e.g., California vs. Texas, Los Angeles vs. San Francisco). Measuring this regional cultural variation can illuminate how and why people think and behave differently. Historically, it has been difficult to computationally model cultural variation due to a lack of training data and scalability constraints. In this work, we introduce a new research problem for the NLP community: How do we measure variation in cultural constructs across regions using language? We then provide a scalable solution: building knowledge-guided lexica to model cultural variation, encouraging future work at the intersection of NLP and cultural understanding. We also highlight modern LLMs’ failure to measure cultural variation or generate culturally varied language.

pdf bib
Adaptive Rank Selections for Low-Rank Approximation of Language Models
Shangqian Gao | Ting Hua | Yen-Chang Hsu | Yilin Shen | Hongxia Jin

Singular Value Decomposition (SVD) or its weighted variants has significantly progressed in compressing language models. Previous works assume the same importance for all operations and assign the same number of ranks for different layers in a language model. However, such a uniform rank selection is sub-optimal since different operations (layers) have non-uniform demand in capacity. In other words, a desired SVD strategy should allocate more ranks for important operations and vice versa. However, a globally-optimized selection of ranks for neural networks is still an open problem, and this is a non-trivial challenge since the selection is discrete. In this work, we propose a novel binary masking mechanism for optimizing the number of ranks in a differentiable framework. Our strategy uses a novel regularization to enable the masking to comply with the SVD property where the ranks have sorted singular values. The experiments examined both types of language models, encoder-only and decoder-only models, including large language models like LLaMA. Our compressed model achieves much better accuracy than previous SVD and their SOTA variants. More interestingly, our method retains significantly better accuracy with zero or limited fine-tuning, proving the substantial advantage of adaptive rank selection.

pdf bib
An Empirical Study of Consistency Regularization for End-to-End Speech-to-Text Translation
Pengzhi Gao | Ruiqing Zhang | Zhongjun He | Hua Wu | Haifeng Wang

Consistency regularization methods, such as R-Drop (Liang et al., 2021) and CrossConST (Gao et al., 2023), have achieved impressive supervised and zero-shot performance in the neural machine translation (NMT) field. Can we also boost end-to-end (E2E) speech-to-text translation (ST) by leveraging consistency regularization? In this paper, we conduct empirical studies on intra-modal and cross-modal consistency and propose two training strategies, SimRegCR and SimZeroCR, for E2E ST in regular and zero-shot scenarios. Experiments on the MuST-C benchmark show that our approaches achieve state-of-the-art (SOTA) performance in most translation directions. The analyses prove that regularization brought by the intra-modal consistency, instead of the modality gap, is crucial for the regular E2E ST, and the cross-modal consistency could close the modality gap and boost the zero-shot E2E ST performance.

pdf bib
Unleashing the Emergent Cognitive Synergy in Large Language Models: A Task-Solving Agent through Multi-Persona Self-Collaboration
Zhenhailong Wang | Shaoguang Mao | Wenshan Wu | Tao Ge | Furu Wei | Heng Ji

Human intelligence thrives on cognitive synergy, where collaboration among different minds yield superior outcomes compared to isolated individuals. In this work, we propose Solo Performance Prompting (SPP), which transforms a single LLM into a cognitive synergist by engaging in multi-turn self-collaboration with multiple personas. A cognitive synergist is an intelligent agent that collaboratively combines multiple minds’ strengths and knowledge to enhance problem-solving in complex tasks. By dynamically identifying and simulating different personas based on task inputs, SPP unleashes the potential of cognitive synergy in LLMs. Our in-depth analysis shows that assigning multiple fine-grained personas in LLMs improves problem-solving abilities compared to using a single or fixed number of personas. We evaluate SPP on three challenging tasks: Trivia Creative Writing, Codenames Collaborative, and Logic Grid Puzzle, encompassing both knowledge-intensive and reasoning-intensive types. Unlike previous works, such as Chain-of-Thought, that solely enhance the reasoning abilities in LLMs, experimental results demonstrate that SPP effectively reduces factual hallucination, and maintains strong reasoning capabilities. Additionally, comparative experiments show that cognitive synergy only emerges in GPT-4 and does not appear in less capable models, such as GPT-3.5-turbo and Llama2-13b-chat, which draws an interesting analogy to human development. Code, data, and prompts can be found at: https://github.com/MikeWangWZHL/Solo-Performance-Prompting.git.

pdf bib
FPT: Feature Prompt Tuning for Few-shot Readability Assessment
Ziyang Wang | Sanwoo Lee | Hsiu-Yuan Huang | Yunfang Wu

Prompt-based methods have achieved promising results in most few-shot text classification tasks. However, for readability assessment tasks, traditional prompt methods lack crucial linguistic knowledge, which has already been proven to be essential.Moreover, previous studies on utilizing linguistic features have shown non-robust performance in few-shot settings and may even impair model performance.To address these issues, we propose a novel prompt-based tuning framework that incorporates rich linguistic knowledge, called Feature Prompt Tuning (FPT). Specifically, we extract linguistic features from the text and embed them into trainable soft prompts. Further, we devise a new loss function to calibrate the similarity ranking order between categories. Experimental results demonstrate that our proposed method FTPnot only exhibits a significant performance improvement over the prior best prompt-based tuning approaches, but also surpasses the previous leading methods that incorporate linguistic features. Also, our proposed model significantly outperforms the large language model gpt-3.5-turbo-16k in most cases. Our proposed method establishes a new architecture for prompt tuning that sheds light on how linguistic features can be easily adapted to linguistic-related tasks.

pdf bib
Self-Prompting Large Language Models for Zero-Shot Open-Domain QA
Junlong Li | Jinyuan Wang | Zhuosheng Zhang | Hai Zhao

Open-Domain Question Answering (ODQA) aims to answer questions without explicitly providing specific background documents. This task becomes notably challenging in a zero-shot setting where no data is available to train tailored retrieval-reader models.While recent Large Language Models (LLMs) like GPT-3 have demonstrated their effectiveness in zero-shot ODQA using direct prompting methods, these methods still fall short of fully harnessing the potential of LLMs when implicitly invoked.In this paper, we propose a Self-Prompting framework to explicitly utilize the massive knowledge encoded in the parameters of LLMs and their strong instruction understanding abilities. Concretely, we prompt LLMs step by step to generate multiple pseudo QA pairs with background passages and explanations entirely from scratch.These generated elements are then utilized for in-context learning. Experimental results show that our method significantly surpasses previous state-of-the-art zero-shot methods on three widely-used ODQA datasets and even achieves comparable performance with various customized fine-tuned models on full training data. Our code is available at https://github.com/lockon-n/self-prompting.

pdf bib
Head-to-Tail: How Knowledgeable are Large Language Models (LLMs)? A.K.A. Will LLMs Replace Knowledge Graphs?
Kai Sun | Yifan Xu | Hanwen Zha | Yue Liu | Xin Luna Dong

Since the recent prosperity of Large Language Models (LLMs), there have been interleaved discussions regarding how to reduce hallucinations from LLM responses, how to increase the factuality of LLMs, and whether Knowledge Graphs (KGs), which store the world knowledge in a symbolic form, will be replaced with LLMs. In this paper, we try to answer these questions from a new angle: How knowledgeable are LLMs?To answer this question, we constructed Head-to-Tail, a benchmark that consists of 18K question-answer (QA) pairs regarding head, torso, and tail facts in terms of popularity. We designed an automated evaluation method and a set of metrics that closely approximate the knowledge an LLM confidently internalizes. Through a comprehensive evaluation of 16 publicly available LLMs, we show that existing LLMs are still far from being perfect in terms of their grasp of factual knowledge, especially for facts of torso-to-tail entities.

pdf bib
kNN-ICL: Compositional Task-Oriented Parsing Generalization with Nearest Neighbor In-Context Learning
Wenting Zhao | Ye Liu | Yao Wan | Yibo Wang | Qingyang Wu | Zhongfen Deng | Jiangshu Du | Shuaiqi Liu | Yunlong Xu | Philip Yu

Task-Oriented Parsing (TOP) enables conversational assistants to interpret user commands expressed in natural language, transforming them into structured outputs that combine elements of both natural language and intent/slot tags. Recently, Large Language Models (LLMs) have achieved impressive performance in synthesizing computer programs based on a natural-language prompt, mitigating the gap between natural language and structured programs. Our paper focuses on harnessing the capabilities of LLMs for semantic parsing tasks, addressing the following three key research questions: 1) How can LLMs be effectively utilized for semantic parsing tasks? 2) What defines an effective prompt? and 3) How can LLM overcome the length constraint and streamline prompt design by including all examples as prompts? We introduce k Nearest Neighbor In-Context Learning (kNN-ICL), which simplifies prompt engineering by allowing it to be built on top of any design strategy while providing access to all demo examples. Extensive experiments show that: 1) Simple ICL without kNN search can achieve a comparable performance with strong supervised models on the TOP tasks, and 2) kNN-ICL significantly improves the comprehension of complex requests by seamlessly integrating ICL with a nearest-neighbor approach. Notably, this enhancement is achieved without the need for additional data or specialized prompts.

pdf bib
ARES: An Automated Evaluation Framework for Retrieval-Augmented Generation Systems
Jon Saad-Falcon | Omar Khattab | Christopher Potts | Matei Zaharia

Evaluating retrieval-augmented generation (RAG) systems traditionally relies on hand annotations for input queries, passages to retrieve, and responses to generate. We introduce ARES, an Automated RAG Evaluation System, for evaluating RAG systems along the dimensions of context relevance, answer faithfulness, and answer relevance. By creating its own synthetic training data, ARES finetunes lightweight LM judges to assess the quality of individual RAG components. To mitigate potential prediction errors, ARES utilizes a small set of human-annotated datapoints for prediction-powered inference (PPI). Across eight different knowledge-intensive tasks in KILT, SuperGLUE, and AIS, ARES accurately evaluates RAG systems while using only a few hundred human annotations during evaluation. Furthermore, ARES judges remain effective across domain shifts, proving accurate even after changing the type of queries and/or documents used in the evaluated RAG systems. We make our code and datasets publicly available on Github.

pdf bib
DEMO: A Statistical Perspective for Efficient Image-Text Matching
Fan Zhang | Xian-Sheng Hua | Chong Chen | Xiao Luo

Image-text matching has been a long-standing problem, which seeks to connect vision and language through semantic understanding. Due to the capability to manage large-scale raw data, unsupervised hashing-based approaches have gained prominence recently. They typically construct a semantic similarity structure using the natural distance, which subsequently guides the optimization of the hashing network. However, the similarity structure could be biased at the boundaries of semantic distributions, causing error accumulation during sequential optimization. To tackle this, we introduce a novel hashing approach termed Distribution-based Structure Mining with Consistency Learning (DEMO) for efficient image-text matching. From a statistical view, DEMO characterizes each image using multiple augmented views, which are considered as samples drawn from its intrinsic semantic distribution. Then, we employ a non-parametric distribution divergence to ensure a robust and precise similarity structure. In addition, we introduce collaborative consistency learning which not only preserves the similarity structure in the Hamming space but also encourages consistency between retrieval distribution from different directions in a self-supervised manner. Extensive experiments on several widely used datasets demonstrate that DEMO achieves superior performance compared with various state-of-the-art methods.

pdf bib
SeaEval for Multilingual Foundation Models: From Cross-Lingual Alignment to Cultural Reasoning
Bin Wang | Zhengyuan Liu | Xin Huang | Fangkai Jiao | Yang Ding | AiTi Aw | Nancy Chen

We present SeaEval, a benchmark for multilingual foundation models. In addition to characterizing how these models understand and reason with natural language, we also investigate how well they comprehend cultural practices, nuances, and values. Alongside standard accuracy metrics, we investigate the brittleness of foundation models in the dimensions of semantics and multilinguality. Our analyses span both open-sourced and closed models, leading to empirical results across classic NLP tasks, reasoning, and cultural comprehension. Key findings indicate (1) Many models exhibit varied behavior when given paraphrased instructions. (2) Many models still suffer from exposure bias (e.g., positional bias, majority label bias). (3) For questions rooted in factual, scientific, and commonsense knowledge, consistent responses are expected across multilingual queries that are semantically equivalent. Yet, most models surprisingly demonstrate inconsistent performance on these queries. (4) Multilingually-trained models have not attained “balanced multilingual” capabilities. Our endeavors underscore the need for more generalizable semantic representations and enhanced multilingual contextualization. SeaEval can serve as a launchpad for more thorough investigations and evaluations for multilingual and multicultural scenarios.

pdf bib
Volcano: Mitigating Multimodal Hallucination through Self-Feedback Guided Revision
Seongyun Lee | Sue Park | Yongrae Jo | Minjoon Seo

Large multimodal models suffer from multimodal hallucination, where they provide incorrect responses misaligned with the given visual information. Recent works have conjectured that one of the reasons behind multimodal hallucination is due to the vision encoder failing to ground on the image properly. To mitigate this issue, we propose a novel approach that leverages self-feedback as visual cues. Building on this approach, we introduce Volcano, a multimodal self-feedback guided revision model. Volcano generates natural language feedback to its initial response based on the provided visual information and utilizes this feedback to self-revise its initial response. Volcano effectively reduces multimodal hallucination and achieves state-of-the-art on MMHal-Bench, POPE, and GAVIE. It also improves on general multimodal abilities and outperforms previous models on MM-Vet and MMBench. Through qualitative analysis, we show that Volcano’s feedback is properly grounded on the image than the initial response. This indicates that Volcano can provide itself with richer visual information through feedback generation, leading to self-correct hallucinations. We publicly release our model, data, and code at https://github.com/kaistAI/Volcanogithub.com/kaistAI/Volcano

pdf bib
LLMs Are Few-Shot In-Context Low-Resource Language Learners
Samuel Cahyawijaya | Holy Lovenia | Pascale Fung

In-context learning (ICL) empowers large language models (LLMs) to perform diverse tasks in underrepresented languages using only short in-context information, offering a crucial avenue for narrowing the gap between high-resource and low-resource languages.Nonetheless, there is only a handful of works explored ICL for low-resource languages with most of them focusing on relatively high-resource languages, such as French and Spanish. In this work, we extensively study ICL and its cross-lingual variation (X-ICL) on 25 low-resource and 7 relatively higher-resource languages.Our study not only assesses the effectiveness of ICL with LLMs in low-resource languages but also identifies the shortcomings of in-context label alignment, and introduces a more effective alternative: query alignment. Moreover, we provide valuable insights into various facets of ICL for low-resource languages.Our study concludes the significance of few-shot in-context information on enhancing the low-resource understanding quality of LLMs through semantically relevant information by closing the language gap in the target language and aligning the semantics between the targeted low-resource and the high-resource language that the model is proficient in. Our work highlights the importance of advancing ICL research, particularly for low-resource languages.

pdf bib
Simple and effective data augmentation for compositional generalization
Yuekun Yao | Alexander Koller

Compositional generalization, the ability to predict complex meanings from training on simpler sentences, poses challenges for powerful pretrained seq2seq models. In this paper, we show that data augmentation methods that sample MRs and backtranslate them can be effective for compositional generalization, but only if we sample from the right distribution. Remarkably, sampling from a uniform distribution performs almost as well as sampling from the test distribution, and greatly outperforms earlier methods that sampled from the training distribution.We further conduct experiments to investigate the reason why this happens and where the benefit of such data augmentation methods come from.

pdf bib
Rethinking Tabular Data Understanding with Large Language Models
Tianyang Liu | Fei Wang | Muhao Chen

Large Language Models (LLMs) have shown to be capable of various tasks, yet their capability in interpreting and reasoning over tabular data remains an underexplored area. In this context, this study investigates from three core perspectives: the robustness of LLMs to structural perturbations in tables, the comparative analysis of textual and symbolic reasoning on tables, and the potential of boosting model performance through the aggregation of multiple reasoning pathways. We discover that structural variance of tables presenting the same content reveals a notable performance decline, particularly in symbolic reasoning tasks. This prompts the proposal of a method for table structure normalization. Moreover, textual reasoning slightly edges out symbolic reasoning, and a detailed error analysis reveals that each exhibits different strengths depending on the specific tasks. Notably, the aggregation of textual and symbolic reasoning pathways, bolstered by a mix self-consistency mechanism, resulted in achieving SOTA performance, with an accuracy of 73.6% on WikiTableQuestions, representing a substantial advancement over previous existing table processing paradigms of LLMs.

pdf bib
From Shortcuts to Triggers: Backdoor Defense with Denoised PoE
Qin Liu | Fei Wang | Chaowei Xiao | Muhao Chen

Language models are often at risk of diverse backdoor attacks, especially data poisoning. Thus, it is important to investigate defense solutions for addressing them. Existing backdoor defense methods mainly focus on backdoor attacks with explicit triggers, leaving a universal defense against various backdoor attacks with diverse triggers largely unexplored. In this paper, we propose an end-to-end ensemble-based backdoor defense framework, DPoE (Denoised Product-of-Experts), which is inspired by the shortcut nature of backdoor attacks, to defend various backdoor attacks. DPoE consists of two models: a shallow model that captures the backdoor shortcuts and a main model that is prevented from learning the shortcuts. To address the label flip caused by backdoor attackers, DPoE incorporates a denoising design. Experiments on three NLP tasks show that DPoE significantly improves the defense performance against various types of backdoor triggers including word-level, sentence-level, and syntactic triggers. Furthermore, DPoE is also effective under a more challenging but practical setting that mixes multiple types of triggers.

pdf bib
BookSQL: A Large Scale Text-to-SQL Dataset for Accounting Domain
Rahul Kumar | Amar Raja Dibbu | Shrutendra Harsola | Vignesh Subrahmaniam | Ashutosh Modi

Several large-scale datasets (e.g., WikiSQL, Spider) for developing natural language interfaces to databases have recently been proposed. These datasets cover a wide breadth of domains but fall short on some essential domains, such as finance and accounting. Given that accounting databases are used worldwide, particularly by non-technical people, there is an imminent need to develop models that could help extract information from accounting databases via natural language queries. In this resource paper, we aim to fill this gap by proposing a new large-scale Text-to-SQL dataset for the accounting and financial domain: BookSQL. The dataset consists of 100k natural language queries-SQL pairs, and accounting databases of 1 million records. We experiment with and analyze existing state-of-the-art models (including GPT-4) for the Text-to-SQL task on BookSQL. We find significant performance gaps, thus pointing towards developing more focused models for this domain.

pdf bib
FLAP: Flow-Adhering Planning with Constrained Decoding in LLMs
Shamik Roy | Sailik Sengupta | Daniele Bonadiman | Saab Mansour | Arshit Gupta

Planning is a crucial task for agents in task oriented dialogs (TODs). Human agents typically resolve user issues by following predefined workflows, decomposing workflow steps into actionable items, and performing actions by executing APIs in order; all of which require reasoning and planning. With the recent advances in LLMs, there have been increasing attempts to use them for task planning and API usage. However, the faithfulness of the plans to predefined workflows and API dependencies, is not guaranteed with LLMs. Moreover, workflows in real life are often custom-defined and prone to changes; hence, adaptation is desirable. To study this, we propose the problem of faithful planning in TODs that needs to resolve user intents by following predefined flows and preserving API dependencies. To solve this problem, we propose FLAP, a Flow-Adhering Planning algorithm based on constrained decoding with lookahead heuristic for LLMs. Our algorithm alleviates the need for finetuning LLMs using domain specific (plan/dependency) data, enables quick adaptation to predefined flows, and outperforms other decoding and prompting-based baselines. Further, our algorithm empowers smaller LLMs (≈7B) to perform at par larger LLMs (≈30B-40B).

pdf bib
DuRE: Dual Contrastive Self Training for Semi-Supervised Relation Extraction
Yuxi Feng | Laks Lakshmanan

Document-level Relation Extraction (RE) aims to extract relation triples from documents. Existing document-RE models typically rely on supervised learning which requires substantial labeled data. To alleviate the amount of human supervision, Self-training (ST) has prospered again in language understanding by augmenting the fine-tuning of big pre-trained models whenever labeled data is insufficient. However, existing ST methods in RE fail to tackle the challenge of long-tail relations. In this work, we propose DuRE, a novel ST framework to tackle these problems. DuRE jointly models RE classification and text generation as a dual process. In this way, our model could construct and utilize both pseudo text generated from given labels and pseudo labels predicted from available unlabeled text, which are gradually refined during the ST phase. We proposed a contrastive loss to leverage the signal of the RE classifier to improve generation quality. In addition, we propose a self-adaptive way to sample pseudo text from different relation classes. Experiments on two document-level RE tasks show that DuRE significantly boosts recall and F1 score with comparable precision, especially for long-tail relations against several strong baselines.

pdf bib
Query-Efficient Textual Adversarial Example Generation for Black-Box Attacks
Zhen Yu | Zhenhua Chen | Kun He

Deep neural networks for Natural Language Processing (NLP) have been demonstrated to be vulnerable to textual adversarial examples. Existing black-box attacks typically require thousands of queries on the target model, making them expensive in real-world applications. In this paper, we propose a new approach that guides the word substitutions using prior knowledge from the training set to improve the attack efficiency. Specifically, we introduce Adversarial Boosting Preference (ABP), a metric that quantifies the importance of words and guides adversarial word substitutions. We then propose two query-efficient attack strategies based on ABP: query-free attack (ABPfree) and guided search attack (ABPguide). Extensive evaluations for text classification demonstrate that ABPfree generates more natural adversarial examples than existing universal attacks, ABPguide significantly reduces the number of queries by a factor of 10 500 while achieving comparable or even better performance than black-box attack baselines. Furthermore, we introduce the first ensemble attack ABPens in NLP, which gains further performance improvements and achieves better transferability and generalization by the ensemble of the ABP across different models and domains. Code is available at https://github.com/BaiDingHub/ABP.

pdf bib
Embrace Divergence for Richer Insights: A Multi-document Summarization Benchmark and a Case Study on Summarizing Diverse Information from News Articles
Kung-Hsiang Huang | Philippe Laban | Alexander Fabbri | Prafulla Kumar Choubey | Shafiq Joty | Caiming Xiong | Chien-Sheng Wu

Previous research in multi-document news summarization has typically concentrated on collating information that all sources agree upon. However, the summarization of diverse information dispersed across multiple articles about an event remains underexplored. In this paper, we propose a new task of summarizing diverse information encountered in multiple news articles encompassing the same event. To facilitate this task, we outlined a data collection schema for identifying diverse information and curated a dataset named DiverseSumm. The dataset includes 245 news stories, with each story comprising 10 news articles and paired with a human-validated reference. Next, to enable consistent automatic evaluation, we conducted a comprehensive analysis to pinpoint the position and verbosity biases when utilizing Large Language Model (LLM)-based metrics for evaluating the coverage and faithfulness of summaries. Through correlation analyses, we outline the best practices for effectively using automatic LLM-based metrics on the DiverseSumm dataset. Finally, we study how LLMs summarize multiple news articles by analyzing which type of diverse information LLMs are capable of identifying. Our analyses suggest that despite the extraordinary capabilities of LLMs in single-document summarization, the proposed task remains a complex challenge for them mainly due to their limited coverage, with GPT-4 only able to cover under 40% of the diverse information on average.

pdf bib
AMRFact: Enhancing Summarization Factuality Evaluation with AMR-Driven Negative Samples Generation
Haoyi Qiu | Kung-Hsiang Huang | Jingnong Qu | Nanyun Peng

Ensuring factual consistency is crucial for natural language generation tasks, particularly in abstractive summarization, where preserving the integrity of information is paramount. Prior works on evaluating factual consistency of summarization often take the entailment-based approaches that first generate perturbed (factual inconsistent) summaries and then train a classifier on the generated data to detect the factually inconsistencies during testing time. However, previous approaches generating perturbed summaries are either of low coherence or lack error-type coverage. To address these issues, we propose AMRFact, a framework that generates perturbed summaries using Abstract Meaning Representations (AMRs). Our approach parses factually consistent summaries into AMR graphs and injects controlled factual inconsistencies to create negative examples, allowing for coherent factually inconsistent summaries to be generated with high error-type coverage. Additionally, we present a data selection module NegFilter based on natural language inference and BARTScore to ensure the quality of the generated negative samples. Experimental results demonstrate our approach significantly outperforms previous systems on the AggreFact-SOTA benchmark, showcasing its efficacy in evaluating factuality of abstractive summarization.

pdf bib
PILOT: Legal Case Outcome Prediction with Case Law
Lang Cao | Zifeng Wang | Cao Xiao | Jimeng Sun

Machine learning shows promise in predicting the outcome of legal cases, but most research has concentrated on civil law cases rather than case law systems. We identified two unique challenges in making legal case outcome predictions with case law. First, it is crucial to identify relevant precedent cases that serve as fundamental evidence for judges during decision-making. Second, it is necessary to consider the evolution of legal principles over time, as early cases may adhere to different legal contexts. In this paper, we proposed a new framework named PILOT (PredictIng Legal case OuTcome) for case outcome prediction. It comprises two modules for relevant case retrieval and temporal pattern handling, respectively. To benchmark the performance of existing legal case outcome prediction models, we curated a dataset from a large-scale case law database. We demonstrate the importance of accurately identifying precedent cases and mitigating the temporal shift when making predictions for case law, as our method shows a significant improvement over the prior methods that focus on civil law case outcome predictions.

pdf bib
ALoRA: Allocating Low-Rank Adaptation for Fine-tuning Large Language Models
Zequan Liu | Jiawen Lyn | Wei Zhu | Xing Tian

Parameter-efficient fine-tuning (PEFT) is widely studied for its effectiveness and efficiency in the era of large language models. Low-rank adaptation (LoRA) has demonstrated commendable performance as a popular and representative method. However, it is implemented with a fixed intrinsic rank that might not be the ideal setting for the downstream tasks. Recognizing the need for more flexible downstream task adaptation, we extend the methodology of LoRA to an innovative approach we call allocating low-rank adaptation (ALoRA) that enables dynamic adjustments to the intrinsic rank during the adaptation process. First, we propose a novel method, AB-LoRA, that can effectively estimate the importance score of each LoRA rank. Second, guided by AB-LoRA, we gradually prune abundant and negatively impacting LoRA ranks and allocate the pruned LoRA budgets to important Transformer modules needing higher ranks. We have conducted experiments on various tasks, and the experimental results demonstrate that our ALoRA method can outperform the recent baselines with comparable tunable parameters.

pdf bib
R-Spin: Efficient Speaker and Noise-invariant Representation Learning with Acoustic Pieces
Heng-Jui Chang | James Glass

This paper introduces Robust Spin (R-Spin), a data-efficient domain-specific self-supervision method for speaker and noise-invariant speech representations by learning discrete acoustic units with speaker-invariant clustering (Spin). R-Spin resolves Spin’s issues and enhances content representations by learning to predict acoustic pieces. R-Spin offers a 12X reduction in computational resources compared to previous state-of-the-art methods while outperforming them in severely distorted speech scenarios. This paper provides detailed analyses to show how discrete units contribute to speech encoder training and improving robustness in diverse acoustic environments.

pdf bib
InsCL: A Data-efficient Continual Learning Paradigm for Fine-tuning Large Language Models with Instructions
Yifan Wang | Yafei Liu | Chufan Shi | Haoling Li | Chen Chen | Haonan Lu | Yujiu Yang

Instruction tuning effectively optimizes Large Language Models (LLMs) for downstream tasks. Due to the changing environment in real-life applications, LLMs necessitate continual task-specific adaptation without catastrophic forgetting. Considering the heavy computational cost, replay-based Continual Learning (CL) methods are the simplest and most widely used for LLMs to address the forgetting issue. However, traditional replay-based methods do not fully utilize instructions to customize the replay strategy. In this work, we propose a novel paradigm called Instruction-based Continual Learning (InsCL). InsCL dynamically replays previous data based on task similarity, calculated by Wasserstein Distance with instructions. Moreover, we further introduce an Instruction Information Metric (InsInfo) to quantify the complexity and diversity of instructions. According to InsInfo, InsCL guides the replay process more inclined to high-quality data. We conduct extensive experiments over 16 tasks with different training orders, observing consistent performance improvements of InsCL. When all tasks have been trained, InsCL achieves performance gains of 3.0 Relative Gain compared with Random Replay, and 27.96 Relative Gain compared with No Replay.

pdf bib
Language Agnostic Code Embeddings
Saiteja Utpala | Alex Gu | Pin-Yu Chen

Recently, code language models have achieved notable advancements in addressing a diverse array of essential code comprehension and generation tasks. Yet, the field lacks a comprehensive deep dive and understanding of the code embeddings of multilingual code models. In this paper, we present a comprehensive study on multilingual code embeddings, focusing on the cross-lingual capabilities of these embeddings across different programming languages. Through probing experiments, we demonstrate that code embeddings comprise two distinct components: one deeply tied to the nuances and syntax of a specific language, and the other remaining agnostic to these details, primarily focusing on semantics. Further, we show that when we isolate and eliminate this language-specific component, we witness significant improvements in downstream code retrieval tasks, leading to an absolute increase of up to +17 in the Mean Reciprocal Rank (MRR).

pdf bib
An Examination of the Compositionality of Large Generative Vision-Language Models
Teli Ma | Rong Li | Junwei Liang

With the success of Large Language Models (LLMs), many Generative Vision-Language Models (GVLMs) have been constructed via multimodal instruction tuning. However, the performance of GVLMs in multimodal compositional reasoning remains under-explored. In this paper, we examine both the evaluation metrics ( VisualGPTScore, etc.) and current benchmarks for evaluating the compositionality of GVLMs. We identify the syntactical bias in current benchmarks, which is exploited by the linguistic capability of GVLMs. The bias renders VisualGPTScore an insufficient metric for assessing GVLMs. To combat this, we first introduce a **SyntaxBias Score**, leveraging LLMs to quantify such bias for mitigation. A challenging new task is subsequently added to evaluate the robustness of GVLMs against inherent inclination toward syntactical correctness. Using the bias-mitigated datasets and the new task, we propose a novel benchmark, namely **S**ynt**A**ctically **DE**-biased benchmark (SADE). Our study provides an unbiased benchmark for the compositionality of GVLMs, facilitating future research in this direction. Code and dataset are available at https://github.com/TeleeMa/SADE.

pdf bib
Two Heads are Better than One: Nested PoE for Robust Defense Against Multi-Backdoors
Victoria Graf | Qin Liu | Muhao Chen

Data poisoning backdoor attacks can cause undesirable behaviors in large language models (LLMs), and defending against them is of increasing importance. Existing defense mechanisms often assume that only one type of trigger is adopted by the attacker, while defending against multiple simultaneous and independent trigger types necessitates general defense frameworks and is relatively unexplored. In this paper, we propose Nested Product of Experts (NPoE) defense framework, which involves a mixture of experts (MoE) as a trigger-only ensemble within the PoE defense framework to simultaneously defend against multiple trigger types. During NPoE training, the main modelis trained in an ensemble with a mixture of smaller expert models that learn the features of backdoor triggers. At inference time, only the main model is used. Experimental results on sentiment analysis, hate speech detection, and question classification tasks demonstrate that NPoE effectively defends against a variety of triggers both separately and in trigger mixtures. Due to the versatility of the MoE structure in NPoE, this framework can be further expanded to defend against other attack settings.

pdf bib
VertAttack: Taking Advantage of Text Classifiers’ Horizontal Vision
Jonathan Rusert

Text classification systems have continuouslyimproved in performance over the years. How-ever, nearly all current SOTA classifiers have asimilar shortcoming, they process text in a hor-izontal manner. Vertically written words willnot be recognized by a classifier. In contrast,humans are easily able to recognize and readwords written both horizontally and vertically.Hence, a human adversary could write problem-atic words vertically and the meaning wouldstill be preserved to other humans. We simulatesuch an attack, VertAttack. VertAttack identifieswhich words a classifier is reliant on and thenrewrites those words vertically. We find thatVertAttack is able to greatly drop the accuracyof 4 different transformer models on 5 datasets.For example, on the SST2 dataset, VertAttackis able to drop RoBERTa’s accuracy from 94 to13%. Furthermore, since VertAttack does notreplace the word, meaning is easily preserved.We verify this via a human study and find thatcrowdworkers are able to correctly label 77%perturbed texts perturbed, compared to 81% ofthe original texts. We believe VertAttack offersa look into how humans might circumvent clas-sifiers in the future and thus inspire a look intomore robust algorithms.

pdf bib
KDMCSE: Knowledge Distillation Multimodal Sentence Embeddings with Adaptive Angular margin Contrastive Learning
Cong-Duy Nguyen | Thong Nguyen | Xiaobao Wu | Anh Tuan Luu

Previous work on multimodal sentence embedding has proposed multimodal contrastive learning and achieved promising results. However, by taking the rest of the batch as negative samples without reviewing when forming contrastive pairs, those studies encountered many suspicious and noisy negative examples, significantly affecting the methods’ overall performance. In this work, we propose KDMCSE (Knowledge Distillation Multimodal contrastive learning of Sentence Embeddings), a novel approach that enhances the discrimination and generalizability of multimodal representation and inherits the knowledge from the teacher model to learn the difference between positive and negative instances and via that, can detect noisy and wrong negative samples effectively before they are calculated in the contrastive objective. Furthermore, to overcome the limitation of modeling the variation within negative pairs, we introduce a new contrastive objective, AdapACSE (Adaptive Angular Margin Supervised Contrastive Learning for Multimodal sentence embeddings), that enhances the discriminative representation by strengthening the margin within the angular space while capturing varying semantics within the negative. Experimental results on widely used Semantic Textual Similarity (STS) benchmarks demonstrate the effectiveness of our approach.

pdf bib
The taste of IPA: Towards open-vocabulary keyword spotting and forced alignment in any language
Jian Zhu | Changbing Yang | Farhan Samir | Jahurul Islam

In this project, we demonstrate that phoneme-based models for speech processing can achieve strong crosslinguistic generalizability to unseen languages. We curated the IPAPACK, a massively multilingual speech corpora with phonemic transcriptions, encompassing more than 115 languages from diverse language families, selectively checked by linguists. Based on the IPAPACK, we propose CLAP-IPA, a multi-lingual phoneme-speech contrastive embedding model capable of open-vocabulary matching between arbitrary speech signals and phonemic sequences. The proposed model was tested on 95 unseen languages, showing strong generalizability across languages. Temporal alignments between phonemes and speech signals also emerged from contrastive training, enabling zeroshot forced alignment in unseen languages. We further introduced a neural forced aligner IPA-ALIGNER by finetuning CLAP-IPA with the Forward-Sum loss to learn better phone-to-audio alignment. Evaluation results suggest that IPA-ALIGNER can generalize to unseen languages without adaptation.

pdf bib
Think Before You Act: A Two-Stage Framework for Mitigating Gender Bias Towards Vision-Language Tasks
Yunqi Zhang | Songda Li | Chunyuan Deng | Luyi Wang | Hui Zhao

Gender bias in vision-language models (VLMs) can reinforce harmful stereotypes and discrimination. In this paper, we focus on mitigating gender bias towards vision-language tasks. We identify object hallucination as the essence of gender bias in VLMs. Existing VLMs tend to focus on salient or familiar attributes in images but ignore contextualized nuances. Moreover, most VLMs rely on the co-occurrence between specific objects and gender attributes to infer the ignored features, ultimately resulting in gender bias. We propose GAMA, a task-agnostic generation framework to mitigate gender bias. GAMA consists of two stages: narrative generation and answer inference. During narrative generation, GAMA yields all-sided but gender-obfuscated narratives, which prevents premature concentration on localized image features, especially gender attributes. During answer inference, GAMA integrates the image, generated narrative, and a task-specific question prompt to infer answers for different vision-language tasks. This approach allows the model to rethink gender attributes and answers. We conduct extensive experiments on GAMA, demonstrating its debiasing and generalization ability.

pdf bib
BeLLM: Backward Dependency Enhanced Large Language Model for Sentence Embeddings
Xianming Li | Jing Li

Sentence embeddings are crucial in measuring semantic similarity. Most recent studies employed large language models (LLMs) to learn sentence embeddings. Existing LLMs mainly adopted autoregressive architecture without explicit backward dependency modeling. Therefore, we examined the effects of backward dependencies in LLMs for semantic similarity measurements. Concretely, we propose a novel model: backward dependency enhanced large language model (BeLLM). It learns sentence embeddings via transforming specific attention layers from uni- to bi-directional. We extensively experiment across various semantic textual similarity (STS) tasks and downstream applications. BeLLM achieves state-of-the-art performance in varying scenarios. It shows that autoregressive LLMs benefit from backward dependencies for sentence embeddings.

pdf bib
Assessing Factual Reliability of Large Language Model Knowledge
Weixuan Wang | Barry Haddow | Alexandra Birch | Wei Peng

The factual knowledge of LLMs is typically evaluated using accuracy, yet this metric does not capture the vulnerability of LLMs to hallucination-inducing factors like prompt and context variability. How do we evaluate the capabilities of LLMs to consistently produce factually correct answers? In this paper, we propose MOdel kNowledge relIabiliTy scORe (MONITOR), a novel metric designed to directly measure LLMs’ factual reliability. MONITOR is designed to compute the distance between the probability distributions of a valid output and its counterparts produced by the same LLM probing the same fact using different styles of prompts and contexts. Experiments on a comprehensive range of 12 LLMs demonstrate the effectiveness of MONITOR in evaluating the factual reliability of LLMs while maintaining a low computational overhead. In addition, we release the FKTC (Factual Knowledge Test Corpus) to foster research along this line https://github.com/Vicky-Wil/MONITOR.

pdf bib
Dial-MAE: ConTextual Masked Auto-Encoder for Retrieval-based Dialogue Systems
Zhenpeng Su | Xing W | Wei Zhou | Guangyuan Ma | Songlin Hu

Dialogue response selection aims to select an appropriate response from several candidates based on a given user and system utterance history. Most existing works primarily focus on post-training and fine-tuning tailored for cross-encoders. However, there are no post-training methods tailored for dense encoders in dialogue response selection. We argue that when the current language model, based on dense dialogue systems (such as BERT), is employed as a dense encoder, it separately encodes dialogue context and response, leading to a struggle to achieve the alignment of both representations. Thus, we propose Dial-MAE (Dialogue Contextual Masking Auto-Encoder), a straightforward yet effective post-training technique tailored for dense encoders in dialogue response selection. Dial-MAE uses an asymmetric encoder-decoder architecture to compress the dialogue semantics into dense vectors, which achieves better alignment between the features of the dialogue context and response. Our experiments have demonstrated that Dial-MAE is highly effective, achieving state-of-the-art performance on two commonly evaluated benchmarks.

pdf bib
Toolink: Linking Toolkit Creation and Using through Chain-of-Solving on Open-Source Model
Cheng Qian | Chenyan Xiong | Zhenghao Liu | Zhiyuan Liu

Large Language Models (LLMs) have demonstrated remarkable progress in utilizing tools, but their closed-source nature and high inference costs pose limitations on their adaptability, necessitating a valid method that leverages smaller, open-sourced models. In this paper, we introduce Toolink, a comprehensive framework that performs task-solving by first creating a toolkit and then integrating the planning and calling of tools through a chain-of-solving (CoS) approach. We first validate the efficacy of Toolink in harnessing the model’s creativity and CoS ability on ChatGPT. Subsequently, we curate CoS-GPT, a chain-of-solving dataset designed for tool-using, and finetune the LLaMA-7B model. It results in LLaMA-CoS, a powerful open-source model with advanced tool-planning and tool-calling capabilities. Evaluation of diverse tasks from BIG-bench demonstrates its CoS ability matches that of ChatGPT while its performance surpasses the chain-of-thought approach. Further studies highlight the generalization of LLaMA-CoS to unseen tasks and showcase its capability in using toolkits not explicitly tailored for the target task, affirming its robustness in real-world scenarios. All codes and data are released.

pdf bib
Create! Don’t Repeat: A Paradigm Shift in Multi-Label Augmentation through Label Creative Generation
Letian Wang | Xianggen Liu | Jiancheng Lv

We propose Label Creative Generation (LCG), a new paradigm in multi-label data augmentation. Beyond repeating data points with fixed labels, LCG creates new data by exploring innovative label combinations. Within LCG, we introduce Tail-Driven Conditional Augmentation (TDCA), combining tail-driven label sampling and label-conditioned text generation for balanced, consistent data augmentation. Our approach has demonstrated a **100.21%** increase in PSP@1 across three datasets, successfully mitigating the long-tail effect in MLTC and markedly enhancing model performance.

pdf bib
Neurocache: Efficient Vector Retrieval for Long-range Language Modeling
Ali Safaya | Deniz Yuret

This paper introduces Neurocache, an approach to extend the effective context size of large language models (LLMs) using an external vector cache to store its past states. Like recent vector retrieval approaches, Neurocache uses an efficient k-nearest-neighbor (kNN) algorithm to retrieve relevant past states and incorporate them into the attention process. Neurocache improves upon previous methods by (1) storing compressed states, which reduces cache size; (2) performing a single retrieval operation per token which increases inference speed; and (3) extending the retrieval window to neighboring states, which improves both language modeling and downstream task accuracy. Our experiments show the effectiveness of Neurocache both for models trained from scratch and for pre-trained models such as Llama2-7B and Mistral-7B when enhanced with the cache mechanism. We also compare Neurocache with text retrieval methods and show improvements in single-document question-answering and few-shot learning tasks. We made the source code available under: https://github.com/alisafaya/neurocache

pdf bib
Unveiling the Generalization Power of Fine-Tuned Large Language Models
Haoran Yang | Yumeng Zhang | Jiaqi Xu | Hongyuan Lu | Pheng-Ann Heng | Wai Lam

While Large Language Models (LLMs) have demonstrated exceptional multitasking abilities, fine-tuning these models on downstream, domain-specific datasets is often necessary to yield superior performance on test sets compared to their counterparts without fine-tuning. However, the comprehensive effects of fine-tuning on the LLMs’ generalization ability are not fully understood.This paper delves into the differences between original, unmodified LLMs and their fine-tuned variants. Our primary investigation centers on whether fine-tuning affects the generalization ability intrinsic to LLMs. To elaborate on this, we conduct extensive experiments across five distinct language tasks on various datasets.Our main findings reveal that models fine-tuned on generation and classification tasks exhibit dissimilar behaviors in generalizing to different domains and tasks.Intriguingly, we observe that integrating the in-context learning strategy during fine-tuning on generation tasks can enhance the model’s generalization ability.Through this systematic investigation, we aim to contribute valuable insights into the evolving landscape of fine-tuning practices for LLMs.

pdf bib
A Closer Look at the Self-Verification Abilities of Large Language Models in Logical Reasoning
Ruixin Hong | Hongming Zhang | Xinyu Pang | Dong Yu | Changshui Zhang

Logical reasoning has been an ongoing pursuit in the field of AI. Despite significant advancements made by large language models (LLMs), they still struggle with complex logical reasoning problems. To enhance reasoning performance, one promising direction is scalable oversight, which requires LLMs to identify their own errors and then improve by themselves. Various self-verification methods have been proposed in pursuit of this goal. Nevertheless, whether existing models understand their own errors well is still under investigation. In this paper, we take a closer look at the self-verification abilities of LLMs in the context of logical reasoning, focusing on their ability to identify logical fallacies accurately. We introduce a dataset, FALLACIES, containing 232 types of reasoning fallacies categorized in a hierarchical taxonomy. By conducting exhaustive experiments on FALLACIES, we obtain comprehensive and detailed analyses of a series of models on their verification abilities. Our main findings suggest that existing LLMs could struggle to identify fallacious reasoning steps accurately and may fall short of guaranteeing the validity of self-verification methods. Drawing from these observations, we offer suggestions for future research and practical applications of self-verification methods.

pdf bib
Exploring Self-supervised Logic-enhanced Training for Large Language Models
Fangkai Jiao | Zhiyang Teng | Bosheng Ding | Zhengyuan Liu | Nancy Chen | Shafiq Joty

Traditional attempts to enhance the logical reasoning abilities of language models often rely on supervised fine-tuning, limiting their generalization to new tasks or domains. Large Language Models (LLMs), with their capacity to condense vast knowledge, can effectively tackle many tasks. Yet, our experiments reveal a gap in their performance on logical reasoning benchmarks when compared to state-of-the-art fine-tuning based models. To bridge this gap, we present LogicLLM, a first-of-its-kind, fully self-supervised framework for integrating logical reasoning capabilities into LLMs, and activating them via in-context learning. We apply this to two LLM series, FLAN-T5 and LLaMA, with parameter sizes from 3 billion to 33 billion. LogicLLM demonstrates its effectiveness through successful improvements on two logical reasoning benchmarks (ReClor and LogiQA-v2). Additionally, LogicLLM based on FLAN-T5-11B attains comparable results to ChatGPT, and evaluations with LLaMA-based models on three language understanding benchmarks (RACE, MMLU and Big-Bench-Hard) confirm that the improvements come without compromising the model’s general language understanding capabilities.

pdf bib
MATHSENSEI: A Tool-Augmented Large Language Model for Mathematical Reasoning
Debrup Das | Debopriyo Banerjee | Somak Aditya | Ashish Kulkarni

Tool-augmented Large Language Models (TALMs) are known to enhance the skillset of large language models (LLMs), thereby, leading to their improved reasoning abilities across many tasks. While, TALMs have been successfully employed in different question-answering benchmarks, their efficacy on complex mathematical reasoning benchmarks, and the potential complementary benefits offered by tools for knowledge retrieval and mathematical equation solving are open research questions. In this work, we present MathSensei, a tool-augmented large language model for mathematical reasoning. We study the complementary benefits of the tools - knowledge retriever (Bing Web Search), program generator + executor (Python), and symbolic equation solver (Wolfram-Alpha API) through evaluations on mathematical reasoning datasets. We perform exhaustive ablations on MATH, a popular dataset for evaluating mathematical reasoning on diverse mathematical disciplines. We also conduct experiments involving well-known tool planners to study the impact of tool sequencing on the model performance. MathSensei achieves 13.5% better accuracy over gpt-3.5-turbo with Chain-of-Thought on the MATH dataset. We further observe that TALMs are not as effective for simpler math word problems (in GSM-8K), and the benefit increases as the complexity and required knowledge increases (progressively over AQuA, MMLU-Math, and higher level complex questions in MATH). The code and data are available at https://github.com/Debrup-61/MathSensei.

pdf bib
CoUDA: Coherence Evaluation via Unified Data Augmentation
Dawei Zhu | Wenhao Wu | Yifan Song | Fangwei Zhu | Ziqiang Cao | Sujian Li

Coherence evaluation aims to assess the organization and structure of a discourse, which remains challenging even in the era of large language models. Due to the scarcity of annotated data, data augmentation is commonly used for training coherence evaluation models. However, previous augmentations for this task primarily rely on heuristic rules, lacking designing criteria as guidance.In this paper, we take inspiration from linguistic theory of discourse structure, and propose a data augmentation framework named CoUDA. CoUDA breaks down discourse coherence into global and local aspects, and designs augmentation strategies for both aspects, respectively.Especially for local coherence, we propose a novel generative strategy for constructing augmentation samples, which involves post-pretraining a generative model and applying two controlling mechanisms to control the difficulty of generated samples. During inference, CoUDA also jointly evaluates both global and local aspects to comprehensively assess the overall coherence of a discourse.Extensive experiments in coherence evaluation show that, with only 233M parameters, CoUDA achieves state-of-the-art performance in both pointwise scoring and pairwise ranking tasks, even surpassing recent GPT-3.5 and GPT-4 based metrics.

pdf bib
mEdIT: Multilingual Text Editing via Instruction Tuning
Vipul Raheja | Dimitris Alikaniotis | Vivek Kulkarni | Bashar Alhafni | Dhruv Kumar

We introduce mEdIT, a multi-lingual extension to CoEdIT – the recent state-of-the-art text editing models for writing assistance. mEdIT models are trained by fine-tuning multi-lingual large, pre-trained language models (LLMs) via instruction tuning. They are designed to take instructions from the user specifying the attributes of the desired text in the form of natural language instructions, such as “Grammatik korrigieren” (German) or “이 텍스 트를 단순화” (Korean). We build mEdIT by curating data from multiple publicly available human-annotated text editing datasets for three text editing tasks (Grammatical Error Correction (GEC), Text Simplification, and Paraphrasing) across diverse languages belonging to six different language families. We detail the design and training of mEdIT models and demonstrate their strong performance on many multi-lingual text editing benchmarks against other multilingual LLMs. We also find that mEdIT generalizes effectively to new languages over multilingual baselines. We publicly release our data, code, and trained models.

pdf bib
Navigation as Attackers Wish? Towards Building Robust Embodied Agents under Federated Learning
Yunchao Zhang | Zonglin Di | Kaiwen Zhou | Cihang Xie | Xin Wang

Federated embodied agent learning protects the data privacy of individual visual environments by keeping data locally at each client (the individual environment) during training. However, since the local data is inaccessible to the server under federated learning, attackers may easily poison the training data of the local client to build a backdoor in the agent without notice. Deploying such an agent raises the risk of potential harm to humans, as the attackers may easily navigate and control the agent as they wish via the backdoor. Towards Byzantine-robust federated embodied agent learning, in this paper, we study the attack and defense for the task of vision-and-language navigation (VLN), where the agent is required to follow natural language instructions to navigate indoor environments. First, we introduce a simple but effective attack strategy, Navigation as Wish (NAW), in which the malicious client manipulates local trajectory data to implant a backdoor into the global model. Results on two VLN datasets (R2R and RxR) show that NAW can easily navigate the deployed VLN agent regardless of the language instruction, without affecting its performance on normal test sets. Then, we propose a new Prompt-Based Aggregation (PBA) to defend against the NAW attack in federated VLN, which provides the server with a ”prompt” of the vision-and-language alignment variance between the benign and malicious clients so that they can be distinguished during training. We validate the effectiveness of the PBA method on protecting the global model from the NAW attack, which outperforms other state-of-the-art defense methods by a large margin in the defense metrics on R2R and RxR.

pdf bib
In-context Learning and Gradient Descent Revisited
Gilad Deutch | Nadav Magar | Tomer Natan | Guy Dar

In-context learning (ICL) has shown impressive results in few-shot learning tasks, yet its underlying mechanism is still not fully understood. A recent line of work suggests that ICL performs gradient descent (GD)-based optimization implicitly. While appealing, much of the research focuses on simplified settings, where the parameters of a shallow model are optimized. In this work, we revisit evidence for ICL-GD correspondence on realistic NLP tasks and models. We find gaps in evaluation, both in terms of problematic metrics and insufficient baselines. We show that surprisingly, even untrained models achieve comparable ICL-GD similarity scores despite not exhibiting ICL.Next, we explore a major discrepancy in the flow of information throughout the model between ICL and GD, which we term Layer Causality. We propose a simple GD-based optimization procedure that respects layer causality, and show it improves similarity scores significantly.

pdf bib
Corpus Considerations for Annotator Modeling and Scaling
Sarumi Oluyemi | Béla Neuendorf | Joan Plepi | Lucie Flek | Jörg Schlötterer | Charles Welch

Recent trends in natural language processing research and annotation tasks affirm a paradigm shift from the traditional reliance on a single ground truth to a focus on individual perspectives, particularly in subjective tasks. In scenarios where annotation tasks are meant to encompass diversity, models that solely rely on the majority class labels may inadvertently disregard valuable minority perspectives. This oversight could result in the omission of crucial information and, in a broader context, risk disrupting the balance within larger ecosystems. As the landscape of annotator modeling unfolds with diverse representation techniques, it becomes imperative to investigate their effectiveness with the fine-grained features of the datasets in view. This study systematically explores various annotator modeling techniques and compares their performance across seven corpora. From our findings, we show that the commonly used user token model consistently outperforms more complex models. We introduce a composite embedding approach and show distinct differences in which model performs best as a function of the agreement with a given dataset. Our findings shed light on the relationship between corpus statistics and annotator modeling performance, which informs future work on corpus construction and perspectivist NLP.

pdf bib
On Large Language Models’ Hallucination with Regard to Known Facts
Che Jiang | Biqing Qi | Xiangyu Hong | Dayuan Fu | Yang Cheng | Fandong Meng | Mo Yu | Bowen Zhou | Jie Zhou

Large language models are successful in answering factoid questions but are also prone to hallucination.We investigate the phenomenon of LLMs possessing correct answer knowledge yet still hallucinating from the perspective of inference dynamics, an area not previously covered in studies on hallucinations.We are able to conduct this analysis via two key ideas.First, we identify the factual questions that query the same triplet knowledge but result in different answers. The difference between the model behaviors on the correct and incorrect outputs hence suggests the patterns when hallucinations happen.Second, to measure the pattern, we utilize mappings from the residual streams to vocabulary space.We reveal the different dynamics of the output token probabilities along the depths of layers between the correct and hallucinated cases. In hallucinated cases, the output token’s information rarely demonstrates abrupt increases and consistent superiority in the later stages of the model.Leveraging the dynamic curve as a feature, we build a classifier capable of accurately detecting hallucinatory predictions with an 88% success rate. Our study shed light on understanding the reasons for LLMs’ hallucinations on their known facts, and more importantly, on accurately predicting when they are hallucinating.

pdf bib
“One-Size-Fits-All”? Examining Expectations around What Constitute “Fair” or “Good” NLG System Behaviors
Li Lucy | Su Lin Blodgett | Milad Shokouhi | Hanna Wallach | Alexandra Olteanu

Fairness-related assumptions about what constitute appropriate NLG system behaviors range from invariance, where systems are expected to behave identically for social groups, to adaptation, where behaviors should instead vary across them. To illuminate tensions around invariance and adaptation, we conduct five case studies, in which we perturb different types of identity-related language features (names, roles, locations, dialect, and style) in NLG system inputs. Through these cases studies, we examine people’s expectations of system behaviors, and surface potential caveats of these contrasting yet commonly held assumptions. We find that motivations for adaptation include social norms, cultural differences, feature-specific information, and accommodation; in contrast, motivations for invariance include perspectives that favor prescriptivism, view adaptation as unnecessary or too difficult for NLG systems to do appropriately, and are wary of false assumptions. Our findings highlight open challenges around what constitute “fair” or “good” NLG system behaviors.

pdf bib
Language Models Hallucinate, but May Excel at Fact Verification
Jian Guan | Jesse Dodge | David Wadden | Minlie Huang | Hao Peng

Recent progress in natural language processing (NLP) owes much to remarkable advances in large language models (LLMs). Nevertheless, LLMs frequently “hallucinate,” resulting in non-factual outputs. Our carefully-designed human evaluation substantiates the serious hallucination issue, revealing that even GPT-3.5 produces factual outputs less than 25% of the time. This underscores the importance of fact verifiers in order to measure and incentivize progress. Our systematic investigation affirms that LLMs can be repurposed as effective fact verifiers with strong correlations with human judgments. Surprisingly, FLAN-T5-11B , the least factual generator in our study, performs the best as a fact verifier, even outperforming more capable LLMs like GPT3.5 and ChatGPT. Delving deeper, we analyze the reliance of these LLMs on high-quality evidence, as well as their deficiencies in robustness and generalization ability. Our study presents insights for developing trustworthy generation models.

pdf bib
A Rationale-centric Counterfactual Data Augmentation Method for Cross-Document Event Coreference Resolution
Bowen Ding | Qingkai Min | Shengkun Ma | Yingjie Li | Linyi Yang | Yue Zhang

Based on Pre-trained Language Models (PLMs), event coreference resolution (ECR) systems have demonstrated outstanding performance in clustering coreferential events across documents. However, the state-of-the-art system exhibits an excessive reliance on the ‘triggers lexical matching’ spurious pattern in the input mention pair text. We formalize the decision-making process of the baseline ECR system using a Structural Causal Model (SCM), aiming to identify spurious and causal associations (i.e., rationales) within the ECR task. Leveraging the debiasing capability of counterfactual data augmentation, we develop a rationale-centric counterfactual data augmentation method with LLM-in-the-loop. This method is specialized for pairwise input in the ECR system, where we conduct direct interventions on triggers and context to mitigate the spurious association while emphasizing the causation. Our approach achieves state-of-the-art performance on three popular cross-document ECR benchmarks and demonstrates robustness in out-of-domain scenarios.

pdf bib
TrojFSP: Trojan Insertion in Few-shot Prompt Tuning
Mengxin Zheng | Jiaqi Xue | Xun Chen | Yanshan Wang | Qian Lou | Lei Jiang

Prompt tuning is one of the most effective solutions to adapting a fixed pre-trained language model (PLM) for various downstream tasks, especially with only a few input samples. However, the security issues, e.g., Trojan attacks, of prompt tuning on a few data samples are not well-studied. Transferring established data poisoning attacks directly to few-shot prompt tuning presents multiple challenges. One significant issue is the _poisoned imbalance issue_, where non-target class samples are added to the target class, resulting in a greater number of target-class samples compared to non-target class. While this issue is not critical in regular tuning, it significantly hampers the few-shot prompt tuning, making it difficult to simultaneously achieve a high attack success rate (ASR) and maintain clean data accuracy (CDA). Additionally, few-shot prompting is prone to overfitting in terms of both ASR and CDA. In this paper, we introduce _TrojFSP_, a method designed to address the challenges. To solve the poisoned imbalance issue, we develop a _Target-Class Shrink (TC-Shrink)_ technique, which aims to equalize the number of poisoning samples. To combat overfitting, we employ a _Selective Token Poisoning_ technique to boost attack performance. Furthermore, we introduce a _Trojan-Trigger Attention_ objective function to amplify the attention of the poisoned trojan prompt on triggers. Experiments show that our TrojFSP achieves an ASR of over 99% while maintaining negligible decreases in CDA across various PLMs and datasets. The source code of TrojFSP is available at _https://github.com/UCF-ML-Research/TrojFSP_.

pdf bib
Ensuring Safe and High-Quality Outputs: A Guideline Library Approach for Language Models
Yi Luo | Zhenghao Lin | YuHao Zhang | Jiashuo Sun | Chen Lin | Chengjin Xu | Xiangdong Su | Yelong Shen | Jian Guo | Yeyun Gong

Large Language Models (LLMs) exhibit impressive capabilities but also present risks such as biased content generation and privacy issues. One of the current alignment techniques includes principle-driven integration, but it faces challenges arising from the imprecision of manually crafted rules and inadequate risk perception in models without safety training. To address these, we introduce Guide-Align, a two-stage approach. Initially, a safety-trained model identifies potential risks and formulates specific guidelines for various inputs, establishing a comprehensive library of guidelines and a model for input-guidelines retrieval. Subsequently, the retrieval model correlates new inputs with relevant guidelines, which guide LLMs in response generation to ensure safe and high-quality outputs, thereby aligning with human values. An additional optional stage involves fine-tuning a model with well-aligned datasets generated through the process implemented in the second stage.Our method customizes guidelines to accommodate diverse inputs, thereby enhancing the fine-grainedness and comprehensiveness of the guideline library. Furthermore, it incorporates safety expertise from a safety-trained LLM through a lightweight retrieval model.We evaluate our approach on three benchmarks, demonstrating significant improvements in LLM security and quality. Notably, our fine-tuned model, Labrador, even at 13 billion parameters, outperforms GPT-3.5-turbo and surpasses GPT-4 in alignment capabilities.

pdf bib
X-PARADE: Cross-Lingual Textual Entailment and Information Divergence across Paragraphs
Juan Rodriguez | Katrin Erk | Greg Durrett

Understanding when two pieces of text convey the same information is a goal touching many subproblems in NLP, including textual entailment and fact-checking. This problem becomes more complex when those two pieces of text are in different languages. Here, we introduce X-PARADE (Cross-lingual Paragraph-level Analysis of Divergences and Entailments), the first cross-lingual dataset of paragraph-level information divergences. Annotators label a paragraph in a target language at the span level and evaluate it with respect to a corresponding paragraph in a source language, indicating whether a given piece of information is the same, new, or new but can be inferred. This last notion establishes a link with cross-language NLI. Aligned paragraphs are sourced from Wikipedia pages in different languages, reflecting real information divergences observed in the wild. Armed with our dataset, we investigate a diverse set of approaches for this problem, including classic token alignment from machine translation, textual entailment methods that localize their decisions, and prompting LLMs. Our results show that these methods vary in their capability to handle inferable information, but they all fall short of human performance.

pdf bib
Topics, Authors, and Institutions in Large Language Model Research: Trends from 17K arXiv Papers
Rajiv Movva | Sidhika Balachandar | Kenny Peng | Gabriel Agostini | Nikhil Garg | Emma Pierson

Large language models (LLMs) are dramatically influencing AI research, spurring discussions on what has changed so far and how to shape the field’s future. To clarify such questions, we analyze a new dataset of 16,979 LLM-related arXiv papers, focusing on recent trends in 2023 vs. 2018-2022. First, we study disciplinary shifts: LLM research increasingly considers societal impacts, evidenced by 20× growth in LLM submissions to the Computers and Society sub-arXiv. An influx of new authors – half of all first authors in 2023 – are entering from non-NLP fields of CS, driving disciplinary expansion. Second, we study industry and academic publishing trends. Surprisingly, industry accounts for a smaller publication share in 2023, largely due to reduced output from Google and other Big Tech companies; universities in Asia are publishing more. Third, we study institutional collaboration: while industry-academic collaborations are common, they tend to focus on the same topics that industry focuses on rather than bridging differences. The most prolific institutions are all US- or China-based, but there is very little cross-country collaboration. We discuss implications around (1) how to support the influx of new authors, (2) how industry trends may affect academics, and (3) possible effects of (the lack of) collaboration.

pdf bib
E5: Zero-shot Hierarchical Table Analysis using Augmented LLMs via Explain, Extract, Execute, Exhibit and Extrapolate
Zhehao Zhang | Yan Gao | Jian-Guang Lou

Analyzing large hierarchical tables with multi-level headers presents challenges due to their complex structure, implicit semantics, and calculation relationships. While recent advancements in large language models (LLMs) have shown promise in flat table analysis, their application to hierarchical tables is constrained by the reliance on manually curated exemplars and the model’s token capacity limitations. Addressing these challenges, we introduce a novel code-augmented LLM-based framework, E5, for zero-shot hierarchical table question answering. This approach encompasses self-explaining the table’s hierarchical structures, code generation to extract relevant information and apply operations, external code execution to prevent hallucinations, and leveraging LLMs’ reasoning for final answer derivation. Empirical results indicate that our method, based on GPT-4, outperforms state-of-the-art fine-tuning methods with a 44.38 Exact Match improvement. Furthermore, we present F3, an adaptive algorithm designed for token-limited scenarios, effectively condensing large tables while maintaining useful information. Our experiments prove its efficiency, enabling the processing of large tables even with models having limited context lengths. The code is available at https://github.com/zzh-SJTU/E5-Hierarchical-Table-Analysis.

pdf bib
S3Eval: A Synthetic, Scalable, Systematic Evaluation Suite for Large Language Model
Fangyu Lei | Qian Liu | Yiming Huang | Shizhu He | Jun Zhao | Kang Liu

The rapid development of Large Language Models (LLMs) has led to great strides in model capabilities like long-context understanding and reasoning.However, as LLMs are able to process longer contexts, it becomes more challenging to evaluate whether they have acquired certain capabilities, since the length of text (e.g., 200K tokens) they can process far exceeds what humans can reliably assess in a reasonable duration.In this paper, we propose using complex synthetic tasks as a proxy evaluation method, and present S3Eval, a Synthetic, Scalable, Systematic evaluation suite for LLMs evaluation.The synthetic nature of S3Eval provides users full control over the dataset, allowing them to systematically probe LLM capabilities by scaling text length and varying task difficulty across diverse scenarios.The strong correlation between S3Eval and real-world benchmarks demonstrates the soundness of using S3Eval for evaluation of LLMs.S3Eval provides a flexible and infinite long-context data generation method. We have generated a comprehensive dataset called S3Eval-Standard, and experimental results have shown that it poses significant challenges for all existing LLMs.

pdf bib
MMC: Advancing Multimodal Chart Understanding with Large-scale Instruction Tuning
Fuxiao Liu | Xiaoyang Wang | Wenlin Yao | Jianshu Chen | Kaiqiang Song | Sangwoo Cho | Yaser Yacoob | Dong Yu

With the rapid development of large language models (LLMs) and their integration into large multimodal models (LMMs), there has beenimpressive progress in zero-shot completion of user-oriented vision-language tasks. However, a gap remains in the domain of chartimage understanding due to the distinct abstract components in charts. To address this, we introduce a large-scale MultiModal ChartInstruction (MMC-Instruction) dataset comprising 600k instances supporting diverse tasks and chart types. Leveraging this data, we de-velop MultiModal Chart Assistant (MMCA), an LMM that achieves state-of-the-art performance on existing chart QA benchmarks. Recognizing the need for a comprehensive evaluation of LMM chart understanding, we also propose a MultiModal Chart Benchmark (MMC-Benchmark), a comprehensive human-annotated benchmark with nine distinct tasks evaluating reasoning capabilities over charts.Extensive experiments on MMC-Benchmark reveal the limitations of existing LMMs on correctly interpreting charts, even for the mostrecent GPT-4V model. Our work provides an instruction-tuning methodology and benchmark to advance multimodal understanding ofcharts. Code and data are available at https://github.com/FuxiaoLiu/MMC.

pdf bib
Visual Grounding Helps Learn Word Meanings in Low-Data Regimes
Chengxu Zhuang | Evelina Fedorenko | Jacob Andreas

Modern neural language models (LMs) are powerful tools for modeling human sentence production and comprehension, and their internal representations are remarkably well-aligned with representations of language in the human brain. But to achieve these results, LMs must be trained in distinctly un-human-like ways — requiring orders of magnitude more language data than children receive during development, and without perceptual or social context. Do models trained more naturalistically — with grounded supervision — exhibit more humanlike language learning? We investigate this question in the context of word learning, a key sub-task in language acquisition. We train a diverse set of LM architectures, with and without auxiliary visual supervision, on datasets of varying scales. We then evaluate these models’ learning of syntactic categories, lexical relations, semantic features, word similarity, and alignment with human neural representations. We find that visual supervision can indeed improve the efficiency of word learning. However, these improvements are limited: they are present almost exclusively in the low-dataregime, and sometimes canceled out by the inclusion of rich distributional signals from text. The information conveyed by text and images isnot redundant—models mainly driven by visual information yield qualitatively different from those mainly driven by word co-occurrences. However, our results suggest that current multimodal modeling approaches fail to effectively leverage visual information to build human-like word representations from human-scale data.

pdf bib
Accurate Knowledge Distillation via n-best Reranking
Hendra Setiawan

We propose utilizing n-best reranking to enhance Sequence-Level Knowledge Distillation (Kim and Rush, 2016) where we extract pseudo-labels for student model’s training data from top n-best hypotheses and leverage a diverse set of models with different inductive biases, objective functions or architectures, including some publicly-available large language models, to pick the highest-quality hypotheses as labels. The effectiveness of our proposal is validated through experiments on the WMT’21 German ↔ English and Chinese ↔ English translation tasks. Our results demonstrate that utilizing pseudo-labels generated by our n-best reranker leads to a significantly more accurate student model. In fact, our best student model achieves comparable accuracy to a large translation model from (Tran et al., 2021) with 4.7 billion parameters, while having two orders of magnitude fewer parameters.

pdf bib
AutoPRM: Automating Procedural Supervision for Multi-Step Reasoning via Controllable Question Decomposition
Zhaorun Chen | Zhuokai Zhao | Zhihong Zhu | Ruiqi Zhang | Xiang Li | Bhiksha Raj | Huaxiu Yao

Recent advancements in large language models (LLMs) have shown promise in multi-step reasoning tasks, yet their reliance on extensive manual labeling to provide procedural feedback remains a significant impediment. To address this challenge, in this paper, we propose a novel self-supervised framework **AutoPRM** that efficiently enhances the fine-tuning of LLMs for intricate reasoning challenges. Specifically, **AutoPRM** first decomposes complex problems into more manageable subquestions with a controllable granularity switch, then sequentially apply reinforcement learning to iteratively improve the subquestion solver. Additionally, we propose context-guided decoding to avoid reward tampering and guide the subquestion solver towards the solution of the holistic problem. Extensive experiments show that **AutoPRM** significantly improves performance on mathematical and commonsense reasoning tasks over SOTA. More encouragingly, **AutoPRM** can be easily integrated with other orthogonal reasoning pipelines.

pdf bib
SEMQA: Semi-Extractive Multi-Source Question Answering
Tal Schuster | Adam Lelkes | Haitian Sun | Jai Gupta | Jonathan Berant | William Cohen | Donald Metzler

Recently proposed long-form question answering (QA) systems, supported by large language models (LLMs), have shown promising capabilities. Yet, attributing and verifying their generated abstractive answers can be difficult, and automatically evaluating their accuracy remains an ongoing challenge.In this work, we introduce a new QA task for answering multi-answer questions by summarizing multiple diverse sources in a semi-extractive fashion. Specifically, Semi-extractive Multi-source QA (SEMQA) requires models to output a comprehensive answer, while mixing factual quoted spans—copied verbatim from given input sources—and non-factual free-text connectors that glue these spans together into a single cohesive passage. This setting bridges the gap between the outputs of well-grounded but constrained extractive QA systems and more fluent but harder to attribute fully abstractive answers. Particularly, it enables a new mode for language models that leverages their advanced language generation capabilities, while also producing fine in-line attributions by-design that are easy to verify, interpret, and evaluate. To study this task, we create the first dataset of this kind, QuoteSum, with human-written semi-extractive answers to natural and generated questions, and define text-based evaluation metrics. Experimenting with several LLMs in various settings, we find this task to be surprisingly challenging, demonstrating the importance of QuoteSum for developing and studying such consolidation capabilities.

pdf bib
Fine-Tuning Language Models with Reward Learning on Policy
Hao Lang | Fei Huang | Yongbin Li

Reinforcement learning from human feedback (RLHF) has emerged as an effective approach to aligning large language models (LLMs) to human preferences.RLHF contains three steps, i.e., human preference collecting, reward learning, and policy optimization, which are usually performed serially.Despite its popularity, however, (fixed) reward models may suffer from inaccurate off-distribution, since policy optimization continuously shifts LLMs’ data distribution.Repeatedly collecting new preference data from the latest LLMs may alleviate this issue, which unfortunately makes the resulting system more complicated and difficult to optimize.In this paper, we propose reward learning on policy (RLP), an unsupervised framework that refines a reward model using policy samples to keep it on-distribution.Specifically, an unsupervised multi-view learning method is introduced to learn robust representations of policy samples.Meanwhile, a synthetic preference generation approach is developed to simulate high-quality preference data with policy outputs.Extensive experiments on three benchmark datasets show that RLP consistently outperforms the state-of-the-art.Our code is available at https://github.com/AlibabaResearch/DAMO-ConvAI/tree/main/rlp.

pdf bib
A Universal Dependencies Treebank for Highland Puebla Nahuatl
Robert Pugh | Francis Tyers

We present a Universal Dependencies (UD) treebank for Highland Puebla Nahuatl. The treebank is only the second such UD corpus for a Mexican language, and supplements an existing treebank for another Nahuatl variant. We describe the process of data collection, annotation decisions and interesting syntactic constructions, and discuss some similarities and differences between the Highland Puebla Nahuatl treebank and the existing Western Sierra Puebla Nahuatl treebank.

pdf bib
COPAL-ID: Indonesian Language Reasoning with Local Culture and Nuances
Haryo Wibowo | Erland Fuadi | Made Nityasya | Radityo Eko Prasojo | Alham Aji

We present COPAL-ID, a novel, public Indonesian language common sense reasoning dataset. Unlike the previous Indonesian COPA dataset (XCOPA-ID), COPAL-ID incorporates Indonesian local and cultural nuances, and therefore, provides a more natural portrayal of day-to-day causal reasoning within the Indonesian cultural sphere. Professionally written by natives from scratch, COPAL-ID is more fluent and free from awkward phrases, unlike the translated XCOPA-ID. In addition, we present COPALID in both standard Indonesian and in Jakartan Indonesian–a dialect commonly used in daily conversation. COPAL-ID poses a greater challenge for existing open-sourced and closedstate-of-the-art multilingual language models, yet is trivially easy for humans. Our findings suggest that general multilingual models struggle to perform well, achieving 66.91% accuracy on COPAL-ID. South-East Asian-specific models achieve slightly better performance of 73.88% accuracy. Yet, this number still falls short of near-perfect human performance. This shows that these language models are still way behind in comprehending the local nuances of Indonesian.

pdf bib
IterAlign: Iterative Constitutional Alignment of Large Language Models
Xiusi Chen | Hongzhi Wen | Sreyashi Nag | Chen Luo | Qingyu Yin | Ruirui Li | Zheng Li | Wei Wang

With the rapid development of large language models (LLMs), aligning LLMs with human values and societal norms to ensure their reliability and safety has become crucial. Reinforcement learning with human feedback (RLHF) and Constitutional AI (CAI) have been proposed for LLM alignment. However, these methods require either heavy human annotations or explicitly pre-defined constitutions, which are labor-intensive and resource-consuming. To overcome these drawbacks, we study constitution-based LLM alignment and propose a data-driven constitution discovery and self-alignment framework called IterAlign. IterAlign leverages red teaming to unveil the weaknesses of an LLM and automatically discovers new constitutions using a stronger LLM. These constitutions are then used to guide self-correction of the base LLM. Such a constitution discovery pipeline can be run iteratively and automatically to discover new constitutions that specifically target the alignment gaps in the current LLM. Empirical results on several safety benchmark datasets and multiple base LLMs show that IterAlign successfully improves truthfulness, helpfulness, harmlessness and honesty, improving the LLM alignment by up to 13.5% in harmlessness.

pdf bib
OrchestraLLM: Efficient Orchestration of Language Models for Dialogue State Tracking
Chia-Hsuan Lee | Hao Cheng | Mari Ostendorf

Large language models (LLMs) have revolutionized the landscape of Natural Language Processing, but are computationally expensive. To reduce the cost without sacrificing performance, previous studies have explored various approaches to harness the potential of Smaller Language Models (SLMs) as cost-effective alternatives to their larger counterparts. Driven by findings that SLMs and LLMs exhibit complementary strengths in a structured knowledge extraction task, this work presents a novel SLM/LLM routing framework designed to improve computational efficiency and enhance task performance. In dialogue state tracking tasks, the proposed routing framework enhances performance substantially compared to relying solely on LLMs, while reducing the computational costs by over 50%.

pdf bib
Multi-Operational Mathematical Derivations in Latent Space
Marco Valentino | Jordan Meadows | Lan Zhang | Andre Freitas

This paper investigates the possibility of approximating multiple mathematical operations in latent space for expression derivation. To this end, we introduce different multi-operational representation paradigms, modelling mathematical operations as explicit geometric transformations. By leveraging a symbolic engine, we construct a large-scale dataset comprising 1.7M derivation steps stemming from 61K premises and 6 operators, analysing the properties of each paradigm when instantiated with state-of-the-art neural encoders.Specifically, we investigate how different encoding mechanisms can approximate expression manipulation in latent space, exploring the trade-off between learning different operators and specialising within single operations, as well as the ability to support multi-step derivations and out-of-distribution generalisation. Our empirical analysis reveals that the multi-operational paradigm is crucial for disentangling different operators, while discriminating the conclusions for a single operation is achievable in the original expression encoder. Moreover, we show that architectural choices can heavily affect the training dynamics, structural organisation, and generalisation of the latent space, resulting in significant variations across paradigms and classes of encoders.

pdf bib
Large Language Models Help Humans Verify Truthfulness – Except When They Are Convincingly Wrong
Chenglei Si | Navita Goyal | Tongshuang Wu | Chen Zhao | Shi Feng | Hal Daumé Iii | Jordan Boyd-Graber

Large Language Models (LLMs) are increasingly used for accessing information on the web. Their truthfulness and factuality are thus of great interest. To help users make the right decisions about the information they get, LLMs should not only provide information but also help users fact-check it. We conduct human experiments with 80 crowdworkers to compare language models with search engines (information retrieval systems) at facilitating fact-checking. We prompt LLMs to validate a given claim and provide corresponding explanations. Users reading LLM explanations are significantly more efficient than those using search engines while achieving similar accuracy. However, they over-rely on the LLMs when the explanation is wrong. To reduce over-reliance on LLMs, we ask LLMs to provide contrastive information—explain both why the claim is true and false, and then we present both sides of the explanation to users. This contrastive explanation mitigates users’ over-reliance on LLMs, but cannot significantly outperform search engines. Further, showing both search engine results and LLM explanations offers no complementary benefits compared to search engines alone. Taken together, our study highlights that natural language explanations by LLMs may not be a reliable replacement for reading the retrieved passages, especially in high-stakes settings where over-relying on wrong AI explanations could lead to critical consequences.

pdf bib
XferBench: a Data-Driven Benchmark for Emergent Language
Brendon Boldt | David Mortensen

In this paper, we introduce a benchmark for evaluating the overall quality of emergent languages using data-driven methods. Specifically, we interpret the notion of the “quality” of an emergent language as its similarity to human language within a deep learning framework. We measure this by using the emergent language as pretraining data for a downstream NLP tasks in human language—the better the downstream performance, the better the emergent language. We implement this benchmark as an easy-to-use Python package that only requires a text file of utterances from the emergent language to be evaluated. Finally, we empirically test the benchmark’s validity using human, synthetic, and emergent language baselines.

pdf bib
Evaluating Large Language Models as Generative User Simulators for Conversational Recommendation
Se-eun Yoon | Zhankui He | Jessica Echterhoff | Julian McAuley

Synthetic users are cost-effective proxies for real users in the evaluation of conversational recommender systems. Large language models show promise in simulating human-like behavior, raising the question of their ability to represent a diverse population of users. We introduce a new protocol to measure the degree to which language models can accurately emulate human behavior in conversational recommendation. This protocol is comprised of five tasks, each designed to evaluate a key property that a synthetic user should exhibit: choosing which items to talk about, expressing binary preferences, expressing open-ended preferences, requesting recommendations, and giving feedback. Through evaluation of baseline simulators, we demonstrate these tasks effectively reveal deviations of language models from human behavior, and offer insights on how to reduce the deviations with model selection and prompting strategies.

pdf bib
A Symbolic Framework for Evaluating Mathematical Reasoning and Generalisation with Transformers
Jordan Meadows | Marco Valentino | Damien Teney | Andre Freitas

This paper proposes a methodology for generating and perturbing detailed derivations of equations at scale, aided by a symbolic engine, to evaluate the generalisability of Transformers to out-of-distribution mathematical reasoning problems. Instantiating the framework in the context of sequence classification tasks, we compare the capabilities of GPT-4, GPT-3.5, and a canon of fine-tuned BERT models, exploring the relationship between specific operators and generalisation failure via the perturbation of reasoning aspects such as symmetry and variable surface forms. Surprisingly, our empirical evaluation reveals that the average in-distribution performance of fine-tuned models surpasses GPT-3.5, and rivals GPT-4. However, perturbations to input reasoning can reduce their performance by up to 80 F1 points. Overall, the results suggest that the in-distribution performance of smaller open-source models may potentially rival GPT by incorporating appropriately structured derivation dependencies during training, and highlight a shared weakness between BERT and GPT involving a relative inability to decode indirect references to mathematical entities. We release the full codebase, constructed datasets, and fine-tuned models to encourage future progress in the field.

pdf bib
Identifying Linear Relational Concepts in Large Language Models
David Chanin | Anthony Hunter | Oana-Maria Camburu

Transformer language models (LMs) have been shown to represent concepts as directions in the latent space of hidden activations. However, for any human-interpretable concept, how can we find its direction in the latent space? We present a technique called linear relational concepts (LRC) for finding concept directions corresponding to human-interpretable concepts by first modeling the relation between subject and object as a linear relational embedding (LRE). We find that inverting the LRE and using earlier object layers results in a powerful technique for finding concept directions that outperforms standard black-box probing classifiers. We evaluate LRCs on their performance as concept classifiers as well as their ability to causally change model output.

pdf bib
Benchmark Transparency: Measuring the Impact of Data on Evaluation
Venelin Kovatchev | Matthew Lease

In this paper we present an exploratory research on quantifying the impact that data distribution has on the performance and evaluation of NLP models. We propose an automated framework that measures the data point distribution across 6 different dimensions: ambiguity, difficulty, discriminability, length, noise, and perplexity.We use disproportional stratified sampling to measure how much the data distribution affects absolute (Acc/F1) and relative (Rank) model performance. We experiment on 2 different datasets (SQUAD and MNLI) and test a total of 135 different models (125 on SQUAD and 10 on MNLI). We demonstrate that without explicit control of the data distribution, standard evaluation frameworks are inconsistent and unreliable. We find that the impact of the data is statistically significant and is often larger than the impact of changing the metric. In a second set of experiments, we demonstrate that the impact of data on evaluation is not just observable, but also predictable. We propose to use benchmark transparency as a method for comparing datasets and quantifying the similarity between them. We find that the “dataset similarity vector” can be used to predict how well a model generalizes out of distribution.

pdf bib
JAMDEC: Unsupervised Authorship Obfuscation using Constrained Decoding over Small Language Models
Jillian Fisher | Ximing Lu | Jaehun Jung | Liwei Jiang | Zaid Harchaoui | Yejin Choi

The permanence of online content combined with the enhanced authorship identification techniques calls for stronger computational methods to protect the identity and privacy of online authorship when needed, e.g., blind reviews for scientific papers, anonymous online reviews, or anonymous interactions in the mental health forums. In this paper, we propose an unsupervised inference-time approach to authorship obfuscation to address the unique challenges of authorship obfuscation: lack of supervision data for diverse authorship and domains, and the need for a sufficient level of revision beyond simple paraphrasing to obfuscate the authorship, all the while preserving the original content and fluency.We introduce JAMDEC, a user-controlled, inference-time algorithm for authorship obfuscation that can be in principle applied to any text and authorship. Our approach builds on small language models such as GPT2-XL in order to help avoid disclosing the original content to proprietary LLM’s APIs, while also reducing the performance gap between small and large language models via algorithmic enhancement. The key idea behind our approach is to boost the creative power of smaller language models through constrained decoding, while also allowing for user-specified controls and flexibility. Experimental results demonstrate that our approach based on GPT2-XL outperforms previous state-of-the-art methods based on comparably small models, while performing competitively against GPT3.5 175B, a propriety model that is two orders of magnitudes larger.

pdf bib
REST: Retrieval-Based Speculative Decoding
Zhenyu He | Zexuan Zhong | Tianle Cai | Jason Lee | Di He

We introduce Retrieval-Based Speculative Decoding (REST), a novel algorithm designed to speed up language model generation. The key insight driving the development of REST is the observation that the process of text generation often includes certain common phases and patterns. Unlike previous methods that rely on a draft language model for speculative decoding, REST harnesses the power of retrieval to generate draft tokens. This method draws from the reservoir of existing knowledge, retrieving and employing relevant tokens based on the current context. Its plug-and-play nature allows for seamless integration and acceleration of any language model, all without necessitating additional training. When benchmarked on 7B and 13B language models in a single-batch setting, REST achieves a significant speedup of 1.62 × to 2.36 × on code or text generation. The source code of REST is available at https://github.com/FasterDecoding/REST.

pdf bib
Sub-Sentence Encoder: Contrastive Learning of Propositional Semantic Representations
Sihao Chen | Hongming Zhang | Tong Chen | Ben Zhou | Wenhao Yu | Dian Yu | Baolin Peng | Hongwei Wang | Dan Roth | Dong Yu

We introduce sub-sentence encoder, a contrastively-learned contextual embedding model for fine-grained semantic representation of text. In contrast to the standard practice with sentence embeddings, where the meaning of an entire sequence of text is encoded into a fixed-length vector, the sub-sentence encoder learns to produce distinct contextual embeddings corresponding to different atomic propositions, i.e. atomic units of meaning expressed within a text sequence. The sub-sentence embeddings are contrastively learned to recognize (inferred) semantic equivalence between propositions across different text sequences. Our experiments show the effectiveness of sub-sentence encoders in applications, such as retrieving supporting facts for fine-grained text attribution or recognizing the conditional semantic similarity between texts. In practice, we demonstrate that sub-sentence encoders keep the same level of inference cost and space complexity compared to sentence encoders.

pdf bib
MSciNLI: A Diverse Benchmark for Scientific Natural Language Inference
Mobashir Sadat | Cornelia Caragea

The task of scientific Natural Language Inference (NLI) involves predicting the semantic relation between two sentences extracted from research articles. This task was recently proposed along with a new dataset called SciNLI derived from papers published in the computational linguistics domain. In this paper, we aim to introduce diversity in the scientific NLI task and present MSciNLI, a dataset containing 132,320 sentence pairs extracted from five new scientific domains. The availability of multiple domains makes it possible to study domain shift for scientific NLI. We establish strong baselines on MSciNLI by fine-tuning Pre-trained Language Models (PLMs) and prompting Large Language Models (LLMs). The highest Macro F1 scores of PLM and LLM baselines are 77.21% and 51.77%, respectively, illustrating that MSciNLI is challenging for both types of models. Furthermore, we show that domain shift degrades the performance of scientific NLI models which demonstrates the diverse characteristics of different domains in our dataset. Finally, we use both scientific NLI datasets in an intermediate task transfer learning setting and show that they can improve the performance of downstream tasks in the scientific domain. We make our dataset and code available on Github.

pdf bib
Causal Inference for Human-Language Model Collaboration
Bohan Zhang | Yixin Wang | Paramveer Dhillon

In this paper, we examine the collaborative dynamics between humansand language models (LMs), where the interactions typically involveLMs proposing text segments and humans editing or responding to theseproposals. Productive engagement with LMs in such scenarios necessitates that humans discern effective text-based interaction strategies, such as editing and response styles, from historical human-LM interactions. This objective is inherently causal, driven by the counterfactual ‘what-if’ question: how would the outcome of collaboration change if humans employed a different text editing/refinement strategy? A key challenge in answering this causal inference question is formulating an appropriate causal estimand: the conventional average treatment effect (ATE) estimand is inapplicable to text-based treatments due to their high dimensionality. To address this concern, we introduce a new causal estimand– *Incremental Stylistic Effect (ISE)*, which characterizes the average impact of infinitesimally shifting a text towards a specific style, such as increasing formality. We establish the conditions for the non-parametric identification of ISE. Building on this, we develop *CausalCollab*, an algorithm designed to estimate the ISE of various interaction strategies in dynamic human-LM collaborations. Our empirical investigations across three distinct human-LM collaboration scenarios reveal that *CausalCollab* effectively reduces confounding and significantly improves counterfactual estimation over a set of competitive baselines.

pdf bib
SELF-GUARD: Empower the LLM to Safeguard Itself
Zezhong Wang | Fangkai Yang | Lu Wang | Pu Zhao | Hongru Wang | Liang Chen | Qingwei Lin | Kam-Fai Wong

With the increasing risk posed by jailbreak attacks, recent studies have investigated various methods to improve the safety of large language models (LLMs), mainly falling into two strategies: safety training and safeguards. Safety training involves fine-tuning the LLM with adversarial samples, which activate the LLM’s capabilities against jailbreak. However, it is not always effective in countering new attacks and often leads to potential performance degradation. Safeguards, on the other hand, are methods using additional models to filter harmful content from the LLM’s response. Nevertheless, they can only reduce a limited amount of harmful output and introduce extra computational costs. Given the distinct strengths and weaknesses of both, we combine them to balance out their flaws and propose a more effective method called Self-Guard.Specifically, we train the LLM to review its responses for any harmful content and append a [harmful] or [harmless] tag to the end of the response. In this way, Self-Guard possesses the advantages of safety training, leveraging the powerful capabilities of the LLMs themselves to detect harmfulness. Besides that, it gains flexibility like safeguards, making the safety check target the output side, which makes the system less vulnerable to attack updates. Experimental results indicate that our Self-Guard can effectively defend against jailbreak attacks and will not cause LLMs’ performance degradation.

pdf bib
COSIGN: Contextual Facts Guided Generation for Knowledge Graph Completion
Jinpeng Li | Hang Yu | Xiangfeng Luo | Qian Liu

Knowledge graph completion (KGC) aims to infer missing facts based on existing facts within a KG. Recently, research on generative models (GMs) has addressed the limitations of embedding methods in terms of generality and scalability. However, GM-based methods are sensitive to contextual facts on KG, so the contextual facts of poor quality can cause GMs to generate erroneous results. To improve the performance of GM-based methods for various KGC tasks, we propose a COntextual FactS GuIded GeneratioN (COSIGN) model. First, to enhance the inference ability of the generative model, we designed a contextual facts collector to achieve human-like retrieval behavior. Second, a contextual facts organizer is proposed to learn the organized capabilities of LLMs through knowledge distillation. Finally, the organized contextual facts as the input of the inference generator to generate missing facts. Experimental results demonstrate that COSIGN outperforms state-of-the-art baseline techniques in terms of performance.

pdf bib
Toward Informal Language Processing: Knowledge of Slang in Large Language Models
Zhewei Sun | Qian Hu | Rahul Gupta | Richard Zemel | Yang Xu

Recent advancement in large language models (LLMs) has offered a strong potential for natural language systems to process informal language. A representative form of informal language is slang, used commonly in daily conversations and online social media. To date, slang has not been comprehensively evaluated in LLMs due partly to the absence of a carefully designed and publicly accessible benchmark. Using movie subtitles, we construct a dataset that supports evaluation on a diverse set of tasks pertaining to automatic processing of slang. For both evaluation and finetuning, we show the effectiveness of our dataset on two core applications: 1) slang detection, and 2) identification of regional and historical sources of slang from natural sentences. We also show how our dataset can be used to probe the output distributions of LLMs for interpretive insights. We find that while LLMs such as GPT-4 achieve good performance in a zero-shot setting, smaller BERT-like models finetuned on our dataset achieve comparable performance. Furthermore, we show that our dataset enables finetuning of LLMs such as GPT-3.5 that achieve substantially better performance than strong zero-shot baselines. Our work offers a comprehensive evaluation and a high-quality benchmark on English slang based on the OpenSubtitles corpus, serving both as a publicly accessible resource and a platform for applying tools for informal language processing.

pdf bib
Ghostbuster: Detecting Text Ghostwritten by Large Language Models
Vivek Verma | Eve Fleisig | Nicholas Tomlin | Dan Klein

We introduce Ghostbuster, a state-of-the-art system for detecting AI-generated text.Our method works by passing documents through a series of weaker language models, running a structured search over possible combinations of their features, and then training a classifier on the selected features to predict whether documents are AI-generated.Crucially, Ghostbuster does not require access to token probabilities from the target model, making it useful for detecting text generated by black-box or unknown models.In conjunction with our model, we release three new datasets of human- and AI-generated text as detection benchmarks in the domains of student essays, creative writing, and news articles. We compare Ghostbuster to several existing detectors, including DetectGPT and GPTZero, as well as a new RoBERTa baseline. Ghostbuster achieves 99.0 F1 when evaluated across domains, which is 5.9 F1 higher than the best preexisting model. It also outperforms all previous approaches in generalization across writing domains (+7.5 F1), prompting strategies (+2.1 F1), and language models (+4.4 F1). We also analyze our system’s robustness to a variety of perturbations and paraphrasing attacks, and evaluate its performance on documents by non-native English speakers.

pdf bib
End-to-End Beam Retrieval for Multi-Hop Question Answering
Jiahao Zhang | Haiyang Zhang | Dongmei Zhang | Liu Yong | Shen Huang

Multi-hop question answering (QA) involves finding multiple relevant passages and step-by-step reasoning to answer complex questions, indicating a retrieve-and-read paradigm. However, previous retrievers were customized for two-hop questions, and most of them were trained separately across different hops, resulting in a lack of supervision over the entire multi-hop retrieval process and leading to poor performance in complicated scenarios beyond two hops. In this work, we introduce Beam Retrieval, an end-to-end beam retrieval framework for multi-hop QA. This approach models the multi-hop retrieval process in an end-to-end manner by jointly optimizing an encoder and two classification heads across all hops. Moreover, Beam Retrieval maintains multiple partial hypotheses of relevant passages at each step, expanding the search space and reducing the risk of missing relevant passages. To establish a complete QA system, we incorporate a supervised reader or a large language model (LLM). Experimental results demonstrate that Beam Retrieval achieves a nearly 50% improvement compared with baselines on challenging MuSiQue-Ans, and it also surpasses all previous retrievers on HotpotQA and achieves 99.9% precision on 2WikiMultiHopQA. Providing high-quality context, Beam Retrieval helps our supervised reader achieve new state-of-the-art performance and substantially improves the few-shot QA performance of LLMs.

pdf bib
Leveraging Generative Large Language Models with Visual Instruction and Demonstration Retrieval for Multimodal Sarcasm Detection
Binghao Tang | Boda Lin | Haolong Yan | Si Li

Multimodal sarcasm detection aims to identify sarcasm in the given image-text pairs and has wide applications in the multimodal domains. Previous works primarily design complex network structures to fuse the image-text modality features for classification. However, such complicated structures may risk overfitting on in-domain data, reducing the performance in out-of-distribution (OOD) scenarios. Additionally, existing methods typically do not fully utilize cross-modal features, limiting their performance on in-domain datasets. Therefore, to build a more reliable multimodal sarcasm detection model, we propose a generative multimodal sarcasm model consisting of a designed instruction template and a demonstration retrieval module based on the large language model. Moreover, to assess the generalization of current methods, we introduce an OOD test set, RedEval. Experimental results demonstrate that our method is effective and achieves state-of-the-art (SOTA) performance on the in-domain MMSD2.0 and OOD RedEval datasets.

pdf bib
Multi-Scale Prompt Memory-Augmented Model for Black-Box Scenarios
Xiaojun Kuang | C. L. Philip Chen | Shuzhen Li | Tong Zhang

Black-box few-shot text classification handles text classification in limited data without accessing the parameters and gradients of language models (LMs). Existing black-box optimization methods have demonstrated strong few-shot learning capabilities. However, they still require numerous LMs’ calls to search optimal prompts, thus resulting in overfitting performance and increasing computational cost. To address this issue, we present MuSKPrompt (Multi-scale Knowledge Prompt for Memory Model), an efficient multi-scale knowledge prompt-based memory model in black-box few-shot text classification task. MuSKPrompt extracts instance-level and class-level knowledge at different scales and stores them in memory banks during training. Then, it references multi-scale memory banks to perform quick inference on new samples via a novel scoring module. MuSKPrompt achieves competitive performance in limited data through multi-scale instance-level and class-level knowledge. Moreover, it realizes gradient-free optimization with zero training parameters in the black-box scenario. Experiments on different benchmarks and parameter analysis demonstrate the effectiveness and efficiency of MuSKPrompt in black-box few-shot text classification tasks.

pdf bib
Ungrammatical-syntax-based In-context Example Selection for Grammatical Error Correction
Chenming Tang | Fanyi Qu | Yunfang Wu

In the era of large language models (LLMs), in-context learning (ICL) stands out as an effective prompting strategy that explores LLMs’ potency across various tasks. However, applying LLMs to grammatical error correction (GEC) is still a challenging task. In this paper, we propose a novel ungrammatical-syntax-based in-context example selection strategy for GEC. Specifically, we measure similarity of sentences based on their syntactic structures with diverse algorithms, and identify optimal ICL examples sharing the most similar ill-formed syntax to the test input. Additionally, we carry out a two-stage process to further improve the quality of selection results. On benchmark English GEC datasets, empirical results show that our proposed ungrammatical-syntax-based strategies outperform commonly-used word-matching or semantics-based methods with multiple LLMs. This indicates that for a syntax-oriented task like GEC, paying more attention to syntactic information can effectively boost LLMs’ performance. Our code is available at https://github.com/JamyDon/SynICL4GEC.

pdf bib
BUFFET: Benchmarking Large Language Models for Few-shot Cross-lingual Transfer
Akari Asai | Sneha Kudugunta | Xinyan Yu | Terra Blevins | Hila Gonen | Machel Reid | Yulia Tsvetkov | Sebastian Ruder | Hannaneh Hajishirzi

Despite remarkable advancements in few-shot generalization in natural language processing, most models are developed and evaluated primarily in English. To establish a rigorous and equitable evaluation framework for few-shot cross-lingual transfer, we introduce a new benchmark, called BUFFET, which unifies 15 diverse tasks across 54 languages in a sequence-to-sequence format and provides a fixed set of few-shot examples and instructions. Using BUFFET, we perform thorough evaluations of ten state-of-the-art multilingual large language models with different transfer methods, namely in-context learning and fine-tuning. Our findings reveal significant room for improvement in few-shot in-context cross-lingual transfer. Strong multilingual pre-trained or instruction-tuned models such as BLOOM or ChatGPT often lag behind much smaller mT5-base models given the same number of few-shot samples, particularly in low-resource languages. Our analysis suggests avenues for future research in few-shot cross-lingual transfer.

pdf bib
TISE: A Tripartite In-context Selection Method for Event Argument Extraction
Yanhe Fu | Yanan Cao | Qingyue Wang | Yi Liu

In-context learning enhances the reasoning capabilities of LLMs by providing several examples. A direct yet effective approach to obtain in-context example is to select the top-k examples based on their semantic similarity to the test input. However, when applied to event argument extraction (EAE), this approach exhibits two shortcomings: 1) It may select almost identical examples, thus failing to provide additional event information, and 2) It overlooks event attributes, leading to the selected examples being unrelated to the test event type. In this paper, we introduce three necessary requirements when selecting an in-context example for EAE task: semantic similarity, example diversity and event correlation. And we further propose TISE, which scores examples from these three perspectives and integrates them using Determinantal Point Processes to directly select a set of examples as context. Experimental results on the ACE05 dataset demonstrate the effectiveness of TISE and the necessity of three requirements. Furthermore, we surprisingly observe that TISE can achieve superior performance with fewer examples and can even exceed some supervised methods.

pdf bib
Reasoning or Reciting? Exploring the Capabilities and Limitations of Language Models Through Counterfactual Tasks
Zhaofeng Wu | Linlu Qiu | Alexis Ross | Ekin Akyürek | Boyuan Chen | Bailin Wang | Najoung Kim | Jacob Andreas | Yoon Kim

The impressive performance of recent language models across a wide range of tasks suggests that they possess a degree of abstract reasoning skills. Are these skills general and transferable, or specialized to specific tasks seen during pretraining? To disentangle these effects, we propose an evaluation framework based on “counterfactual” task variants that deviate from the default assumptions underlying standard tasks. Across a suite of 11 tasks, we observe nontrivial performance on the counterfactual variants, but nevertheless find that performance substantially and consistently degrades compared to the default conditions. This suggests that while current LMs may possess abstract task-solving skills to an extent, they often also rely on narrow, non-transferable procedures for task-solving. These results motivate a more careful interpretation of language model performance that teases apart these aspects.

pdf bib
TRUE-UIE: Two Universal Relations Unify Information Extraction Tasks
Yucheng Wang | Bowen Yu | Yilin Liu | Shudong Lu

Information extraction (IE) encounters challenges due to the variety of schemas and objectives that differ across tasks. Recent advancements hint at the potential for universal approaches to model such tasks, referred to as Universal Information Extraction (UIE). While handling diverse tasks in one model, their generalization is limited since they are actually learning task-specific knowledge.In this study, we introduce an innovative paradigm known as TRUE-UIE, wherein all IE tasks are aligned to learn the same goals: extracting mention spans and two universal relations named \mathtt{NEXT} and \mathtt{IS}. During the decoding process, the \mathtt{NEXT} relation is utilized to group related elements, while the \mathtt{IS} relation, in conjunction with structured language prompts, undertakes the role of type recognition. Additionally, we consider the sequential dependency of tokens during span extraction, an aspect often overlooked in prevalent models.Our empirical experiments indicate that TRUE-UIE achieves state-of-the-art performance on established benchmarks encompassing 16 datasets, spanning 7 diverse IE tasks. Further evaluations reveal that our approach effectively share knowledge between different IE tasks, showcasing significant transferability in zero-shot and few-shot scenarios.

pdf bib
zrLLM: Zero-Shot Relational Learning on Temporal Knowledge Graphs with Large Language Models
Zifeng Ding | Heling Cai | Jingpei Wu | Yunpu Ma | Ruotong Liao | Bo Xiong | Volker Tresp

Modeling evolving knowledge over temporal knowledge graphs (TKGs) has become a heated topic. Various methods have been proposed to forecast links on TKGs. Most of them are embedding-based, where hidden representations are learned to represent knowledge graph (KG) entities and relations based on the observed graph contexts. Although these methods show strong performance on traditional TKG forecasting (TKGF) benchmarks, they face a strong challenge in modeling the unseen zero-shot relations that have no prior graph context. In this paper, we try to mitigate this problem as follows. We first input the text descriptions of KG relations into large language models (LLMs) for generating relation representations, and then introduce them into embedding-based TKGF methods. LLM-empowered representations can capture the semantic information in the relation descriptions. This makes the relations, whether seen or unseen, with similar semantic meanings stay close in the embedding space, enabling TKGF models to recognize zero-shot relations even without any observed graph context. Experimental results show that our approach helps TKGF models to achieve much better performance in forecasting the facts with previously unseen relations, while still maintaining their ability in link forecasting regarding seen relations.

pdf bib
Embodied Executable Policy Learning with Language-based Scene Summarization
Jielin Qiu | Mengdi Xu | William Han | Seungwhan Moon | Ding Zhao

Large Language models (LLMs) have shown remarkable success in assisting robot learning tasks, i.e., complex household planning.However, the performance of pretrained LLMs heavily relies on domain-specific templated text data, which may be infeasible in real-world robot learning tasks with image-based observations. Moreover, existing LLMs with text inputs lack the capability to evolve with non-expert interactions with environments.In this work, we introduce a novel learning paradigm that generates robots’ executable actions in the form of text, derived solely from visual observations. Our proposed paradigm stands apart from previous works, which utilized either language instructions or a combination of language and visual data as inputs. We demonstrate that our proposed method can employ two fine-tuning strategies, including imitation learning and reinforcement learning approaches, to adapt to the target test tasks effectively.We conduct extensive experiments involving various model selections, environments, and tasks across 7 house layouts in the VirtualHome environment. Our experimental results demonstrate that our method surpasses existing baselines, confirming the effectiveness of this novel learning paradigm.

pdf bib
Metacognitive Prompting Improves Understanding in Large Language Models
Yuqing Wang | Yun Zhao

In Large Language Models (LLMs), there have been consistent advancements in task-specific performance, largely influenced by effective prompt design. Recent advancements in prompting have enhanced reasoning in logic-intensive tasks for LLMs, yet the nuanced understanding abilities of these models, crucial for processing and interpreting complex information, remain underexplored. In this study, we introduce Metacognitive Prompting (MP), a strategy inspired by human introspective reasoning processes. Using MP, LLMs undergo a systematic series of structured, self-aware evaluations, drawing on both their vast inherent knowledge and new insights. We conduct extensive experiments on four prevalent LLMs: Llama2, PaLM2, GPT-3.5, and GPT-4, across ten natural language understanding (NLU) datasets from GLUE, SuperGLUE, BLUE, and LexGLUE benchmarks. Additionally, we compare our method with chain-of-thought prompting and its advanced versions. The results show that GPT-4 consistently excels across all tasks, while other models have shown significant progress in some tasks when used in conjunction with MP. Furthermore, MP consistently outperforms existing prompting methods in both general and domain-specific NLU tasks. This study underscores the potential to amplify the understanding abilities of LLMs and highlights the benefits of mirroring human introspective reasoning in NLU tasks.

pdf bib
MART: Improving LLM Safety with Multi-round Automatic Red-Teaming
Suyu Ge | Chunting Zhou | Rui Hou | Madian Khabsa | Yi-Chia Wang | Qifan Wang | Jiawei Han | Yuning Mao

Red-teaming is a common practice for mitigating unsafe behaviors in Large Language Models (LLMs), which involves thoroughly assessing LLMs to identify potential flaws and addressing them with responsible and accurate responses.While effective, manual red-teaming is costly, and existing automatic red-teaming typically discovers safety risks without addressing them.In this paper, we propose a Multi-round Automatic Red-Teaming (MART) method, which incorporates both automatic adversarial prompt writing and safe response generation, significantly increasing red-teaming scalability and the safety of the target LLM.Specifically, an adversarial LLM and a target LLM interplay with each other in an iterative manner, where the adversarial LLM aims to generate challenging prompts that elicit unsafe responses from the target LLM, while the target LLM is fine-tuned with safety aligned data on these adversarial prompts. In each round, the adversarial LLM crafts better attacks on the updated target LLM, while the target LLM also improves itself through safety fine-tuning.On adversarial prompt benchmarks, the violation rate of an LLM with limited safety alignment reduces up to 84.7% after 4 rounds of MART, achieving comparable performance to LLMs with extensive adversarial prompt writing. Notably, model helpfulness on non-adversarial prompts remains stable throughout iterations, indicating the target LLM maintains strong performance on instruction following.

pdf bib
DialogCC: An Automated Pipeline for Creating High-Quality Multi-Modal Dialogue Dataset
Young-Jun Lee | Byungsoo Ko | Han-Gyu Kim | Jonghwan Hyeon | Ho-Jin Choi

As sharing images in an instant message is a crucial factor, there has been active research on learning an image-text multi-modal dialogue models.However, training a well-generalized multi-modal dialogue model remains challenging due to the low quality and limited diversity of images per dialogue in existing multi-modal dialogue datasets.In this paper, we propose an automated pipeline to construct a multi-modal dialogue dataset, ensuring both dialogue quality and image diversity without requiring minimum human effort. In our pipeline, to guarantee the coherence between images and dialogue, we prompt GPT-4 to infer potential image-sharing moments - specifically, the utterance, speaker, rationale, and image description. Furthermore, we leverage CLIP similarity to maintain consistency between aligned multiple images to the utterance.Through this pipeline, we introduce DialogCC, a high-quality and diverse multi-modal dialogue dataset that surpasses existing datasets in terms of quality and diversity in human evaluation.Our comprehensive experiments highlight that when multi-modal dialogue models are trained using our dataset, their generalization performance on unseen dialogue datasets is significantly enhanced. We make our source code and dataset publicly available (https://dialogcc.github.io/).

pdf bib
Routing to the Expert: Efficient Reward-guided Ensemble of Large Language Models
Keming Lu | Hongyi Yuan | Runji Lin | Junyang Lin | Zheng Yuan | Chang Zhou | Jingren Zhou

The complementary potential of Large Language Models (LLM) assumes off-the-shelf LLMs have heterogeneous expertise in a wide range of domains and tasks so that an ensemble of LLMs can achieve consistently better performance. Existing ensemble methods for LLMs mainly focus on reward model ranking of outputs, leading to significant computation overhead. To combat this issue, we revisit the complementary potential of LLMs and further elaborate on it by mining latent expertise with off-the-shelf reward models. We propose ZOOTER, a reward-guided routing method distilling rewards on training queries to train a routing function, which can precisely distribute each query to the LLM with expertise about it. We also integrate a tag-based label enhancement to mitigate noise from uncertainty when using rewards as silver supervision. ZOOTER shows computation efficiency in inference as it only introduces minor computation overhead of a routing function compared with reward model ranking methods. We evaluate ZOOTER on a comprehensive benchmark collection with 26 subsets in different domains and tasks. ZOOTER outperforms the best single model on average and ranks first on 44% of tasks, even surpassing multiple reward model ranking methods.

pdf bib
Automatic Generation of Model and Data Cards: A Step Towards Responsible AI
Jiarui Liu | Wenkai Li | Zhijing Jin | Mona Diab

In an era of model and data proliferation in machine learning/AI especially marked by the rapid advancement of open-sourced technologies, there arises a critical need for standardized consistent documentation. Our work addresses the information incompleteness in current human-written model and data cards. We propose an automated generation approach using Large Language Models (LLMs). Our key contributions include the establishment of CardBench, a comprehensive dataset aggregated from over 4.8k model cards and 1.4k data cards, coupled with the development of the CardGen pipeline comprising a two-step retrieval process. Our approach exhibits enhanced completeness, objectivity, and faithfulness in generated model and data cards, a significant step in responsible AI documentation practices ensuring better accountability and traceability.

pdf bib
FUN with Fisher: Improving Generalization of Adapter-Based Cross-lingual Transfer with Scheduled Unfreezing
Chen Liu | Jonas Pfeiffer | Ivan Vulić | Iryna Gurevych

Standard fine-tuning of language models typically performs well on in-distribution data, but suffers with generalization to distribution shifts. In this work, we aim to improve the generalization of adapter-based cross-lingual task transfer where such cross-language distribution shifts are imminent. We investigate scheduled unfreezing algorithms –originally proposed to mitigate catastrophic forgetting in transfer learning – for fine-tuning task adapters. Our experiments show that scheduled unfreezing methods close the gap to full fine-tuning and achieve stronger cross-lingual transfer performance, suggesting that these methods can go beyond just mitigating catastrophic forgetting. Next, aiming to understand these empirical findings, we investigate the learning dynamics of scheduled unfreezing using Fisher Information. Our experiments reveal that scheduled unfreezing induces different learning dynamics compared to standard fine-tuning, and provide evidence that the dynamics of Fisher Information during training correlate with cross-lingual generalization performance. We additionally propose a general scheduled unfreezing algorithm that achieves an average of 2 points improvement over four datasets compared to standard fine-tuning and provides empirical evidence for a theory-based justification of the heuristic unfreezing schedule for task adapter training.

pdf bib
Are Multilingual LLMs Culturally-Diverse Reasoners? An Investigation into Multicultural Proverbs and Sayings
Chen Liu | Fajri Koto | Timothy Baldwin | Iryna Gurevych

Large language models (LLMs) are highly adept at question answering and reasoning tasks, but when reasoning in a situational context, human expectations vary depending on the relevant cultural common ground. As languages are associated with diverse cultures, LLMs should also be culturally-diverse reasoners. In this paper, we study the ability of a wide range of state-of-the-art multilingual LLMs (mLLMs) to reason with proverbs and sayings in a conversational context. Our experiments reveal that: (1) mLLMs “know” limited proverbs and memorizing proverbs does not mean understanding them within a conversational context; (2) mLLMs struggle to reason with figurative proverbs and sayings, and when asked to select the wrong answer (instead of asking it to select the correct answer); and (3) there is a “culture gap” in mLLMs when reasoning about proverbs and sayings translated from other languages. We construct and release our evaluation dataset MAPS (MulticulturAl Proverbs and Sayings) for proverb understanding with conversational context for six different languages.

pdf bib
The Colorful Future of LLMs: Evaluating and Improving LLMs as Emotional Supporters for Queer Youth
Shir Lissak | Nitay Calderon | Geva Shenkman | Yaakov Ophir | Eyal Fruchter | Anat Brunstein Klomek | Roi Reichart

Queer youth face increased mental health risks, such as depression, anxiety, and suicidal ideation. Hindered by negative stigma, they often avoid seeking help and rely on online resources, which may provide incompatible information. Although access to a supportive environment and reliable information is invaluable, many queer youth worldwide have no access to such support. However, this could soon change due to the rapid adoption of Large Language Models (LLMs) such as ChatGPT. This paper aims to comprehensively explore the potential of LLMs to revolutionize emotional support for queers. To this end, we conduct a qualitative and quantitative analysis of LLM’s interactions with queer-related content. To evaluate response quality, we develop a novel ten-question scale that is inspired by psychological standards and expert input. We apply this scale to score several LLMs and human comments to posts where queer youth seek advice and share experiences. We find that LLM responses are supportive and inclusive, outscoring humans. However, they tend to be generic, not empathetic enough, and lack personalization, resulting in nonreliable and potentially harmful advice. We discuss these challenges, demonstrate that a dedicated prompt can improve the performance, and propose a blueprint of an LLM-supporter that actively (but sensitively) seeks user context to provide personalized, empathetic, and reliable responses. Our annotated dataset is available for further research.*https://github.com/nitaytech/LGBTeenDataset

pdf bib
IPED: An Implicit Perspective for Relational Triple Extraction based on Diffusion Model
Jianli Zhao | Changhao Xu | Bin. Jiang

Relational triple extraction is a fundamental task in the field of information extraction, and a promising framework based on table filling has recently gained attention as a potential baseline for entity relation extraction. However, inherent shortcomings such as redundant information and incomplete triple recognition remain problematic. To address these challenges, we propose an Implicit Perspective for relational triple Extraction based on Diffusion model (IPED), an innovative approach for extracting relational triples. Our classifier-free solution adopts an implicit strategy using block coverage to complete the tables, avoiding the limitations of explicit tagging methods. Additionally, we introduce a generative model structure, the block-denoising diffusion model, to collaborate with our implicit perspective and effectively circumvent redundant information disruptions. Experimental results on two popular datasets demonstrate that IPED achieves state-of-the-art performance while gaining superior inference speed and low computational complexity. To support future research, we have made our source code publicly available online.

pdf bib
QualEval: Qualitative Evaluation for Model Improvement
Vishvak Murahari | Ameet Deshpande | Peter Clark | Tanmay Rajpurohit | Ashish Sabharwal | Karthik Narasimhan | Ashwin Kalyan

Quantitative evaluation metrics have been pivotal in gauging the advancements of AI systems like large language models (LLMs).However, due to the intricate nature of real-world tasks, a single scalar to quantify and compare performance trivializes the fine-grained nuances of model behavior. Additionally, metrics do not yield actionable diagnostics for model improvement, thus requiring extensive manual efforts of scientists, involving sifting through vast datasets and attempting hit-or-miss adjustments to training data or setups. In this work, we address the shortcomings of quantitative metrics by proposing QualEval, which uses automated qualitative evaluation as a vehicle for model improvement. QualEval uses a powerful LLM reasoner and our novel flexible linear programming solver to generate human-readable insights that when applied, accelerate model improvement. The insights are supported by a dashboard report with fine-grained visualizations and human-interpretable analyses. We corroborate the faithfulness of QualEval by demonstrating that leveraging its insights, for example, improves the absolute performance of the Llama 2 model by up to 15% points relative on a challenging dialogue task (DialogSum) when compared to baselines. QualEval successfully increases the pace and quality of model development by eliminating the need of arduous manual analysis, thus serving as a data-scientist-in-a-box.

pdf bib
Quantum-inspired Language Model with Lindblad Master Equation and Interference Measurement for Sentiment Analysis
Kehuan Yan | Peichao Lai | Yilei Wang

Quantum-inspired models have demonstrated superior performance in many downstream language tasks, such as question answering and sentiment analysis. However, recent models primarily focus on embedding and measurement operations, overlooking the significance of the quantum evolution process. In this work, we present a novel quantum-inspired neural network, LI-QiLM, which integrates the Lindblad Master Equation (LME) to model the evolution process and the interferometry to the measurement process, providing more physical meaning to strengthen the interpretability. We conduct comprehensive experiments on six sentiment analysis datasets. Compared to the traditional neural networks, transformer-based pre-trained models and quantum-inspired models, such as CICWE-QNN and ComplexQNN, the proposed method demonstrates superior performance in accuracy and F1-score on six commonly used datasets for sentiment analysis. Additional ablation tests verify the effectiveness of LME and interferometry.

pdf bib
VisLingInstruct: Elevating Zero-Shot Learning in Multi-Modal Language Models with Autonomous Instruction Optimization
Dongsheng Zhu | Daniel Tang | Weidong Han | Jinghui Lu | Yukun Zhao | Guoliang Xing | Junfeng Wang | Dawei Yin

This paper presents VisLingInstruct, a novel approach to advancing Multi-Modal Language Models (MMLMs) in zero-shot learning. Current MMLMs show impressive zero-shot abilities in multi-modal tasks, but their performance depends heavily on the quality of instructions. VisLingInstruct tackles this by autonomously evaluating and optimizing instructional texts through In-Context Learning, improving the synergy between visual perception and linguistic expression in MMLMs. Alongside this instructional advancement, we have also optimized the visual feature extraction modules in MMLMs, further augmenting their responsiveness to textual content. Our comprehensive experiments on MMLMs, based on FlanT5 and Vicuna, show that VisLingInstruct significantly improves zero-shot performance in visual multi-modal tasks. Notably, it achieves a 13.1% and 9% increase in accuracy over the prior state-of-the-art on the TextVQA and HatefulMemes datasets. Our main code is available at https://github.com/Zhudongsheng75/VisLingInstruct

pdf bib
A Wolf in Sheep’s Clothing: Generalized Nested Jailbreak Prompts can Fool Large Language Models Easily
Peng Ding | Jun Kuang | Dan Ma | Xuezhi Cao | Yunsen Xian | Jiajun Chen | Shujian Huang

Large Language Models (LLMs), such as ChatGPT and GPT-4, are designed to provide useful and safe responses. However, adversarial prompts known as ‘jailbreaks’ can circumvent safeguards, leading LLMs to generate potentially harmful content. Exploring jailbreak prompts can help to better reveal the weaknesses of LLMs and further steer us to secure them. Unfortunately, existing jailbreak methods either suffer from intricate manual design or require optimization on other white-box models, which compromises either generalization or efficiency. In this paper, we generalize jailbreak prompt attacks into two aspects: (1) Prompt Rewriting and (2) Scenario Nesting. Based on this, we propose ReNeLLM, an automatic framework that leverages LLMs themselves to generate effective jailbreak prompts. Extensive experiments demonstrate that ReNeLLM significantly improves the attack success rate while greatly reducing the time cost compared to existing baselines. Our study also reveals the inadequacy of current defense methods in safeguarding LLMs. Finally, we analyze the failure of LLMs defense from the perspective of prompt execution priority, and propose corresponding defense strategies. We hope that our research can catalyze both the academic community and LLMs developers towards the provision of safer and more regulated LLMs. The code is available at https://github.com/NJUNLP/ReNeLLM.

pdf bib
P3Sum: Preserving Author’s Perspective in News Summarization with Diffusion Language Models
Yuhan Liu | Shangbin Feng | Xiaochuang Han | Vidhisha Balachandran | Chan Young Park | Sachin Kumar | Yulia Tsvetkov

In this work, we take a first step towards designing summarization systems that are faithful to the author’s intent, not only the semantic content of the article. Focusing on a case study of preserving political perspectives in news summarization, we find that existing approaches alter the political opinions and stances of news articles in more than 50% of summaries, misrepresenting the intent and perspectives of the news authors. We thus propose P3Sum, a diffusion model-based summarization approach controlled by political perspective classifiers. In P3Sum, the political leaning of a generated summary is iteratively evaluated at each decoding step, and any drift from the article’s original stance incurs a loss back-propagated to the embedding layers, steering the political stance of the summary at inference time. Extensive experiments on three news summarization datasets demonstrate that P3Sum outperforms state-of-the-art summarization systems and large language models by up to 13.7% in terms of the success rate of stance preservation, with competitive performance on standard metrics of summarization quality. Our findings present a first analysis of preservation of pragmatic features in summarization, highlight the lacunae in existing summarization models—that even state-of-the-art models often struggle to preserve author’s intents—and develop new summarization systems that are more faithful to author’s perspectives.

pdf bib
Bridging the Novice-Expert Gap via Models of Decision-Making: A Case Study on Remediating Math Mistakes
Rose Wang | Qingyang Zhang | Carly Robinson | Susanna Loeb | Dorottya Demszky

Scaling high-quality tutoring remains a major challenge in education. Due to growing demand, many platforms employ novice tutors who, unlike experienced educators, struggle to address student mistakes and thus fail to seize prime learning opportunities. Our work explores the potential of large language models (LLMs) to close the novice-expert knowledge gap in remediating math mistakes. We contribute Bridge, a method that uses cognitive task analysis to translate an expert’s latent thought process into a decision-making model for remediation. This involves an expert identifying (A) the student’s error, (B) a remediation strategy, and (C) their intention before generating a response. We construct a dataset of 700 real tutoring conversations, annotated by experts with their decisions. We evaluate state-of-the-art LLMs on our dataset and find that the expert’s decision-making model is critical for LLMs to close the gap: responses from GPT4 with expert decisions (e.g., “simplify the problem”) are +76% more preferred than without. Additionally, context-sensitive decisions are critical to closing pedagogical gaps: random decisions decrease GPT4’s response quality by -97% than expert decisions. Our work shows the potential of embedding expert thought processes in LLM generations to enhance their capability to bridge novice-expert knowledge gaps. Our dataset and code can be found at: https://github.com/rosewang2008/bridge.

pdf bib
RST-LoRA: A Discourse-Aware Low-Rank Adaptation for Long Document Abstractive Summarization
Dongqi Pu | Vera Demberg

For long document summarization, discourse structure is important to discern the key content of the text and the differences in importance level between sentences. Unfortunately, the integration of rhetorical structure theory (RST) into parameter-efficient fine-tuning strategies for long document summarization remains unexplored. Therefore, this paper introduces RST-LoRA and proposes four RST-aware variants to explicitly incorporate RST into the LoRA model. Our empirical evaluation demonstrates that incorporating the type and uncertainty of rhetorical relations can complementarily enhance the performance of LoRA in summarization tasks. Furthermore, the best-performing variant we introduced outperforms the vanilla LoRA and full-parameter fine-tuning models, as confirmed by multiple automatic and human evaluations, and even surpasses previous state-of-the-art methods.

pdf bib
Strings from the Library of Babel: Random Sampling as a Strong Baseline for Prompt Optimisation
Yao Lu | Jiayi Wang | Raphael Tang | Sebastian Riedel | Pontus Stenetorp

Recent prompt optimisation approaches use the generative nature of language models to produce prompts – even rivaling the performance of human-curated prompts. In this paper, we demonstrate that randomly sampling tokens from the model vocabulary as “separators” can be as effective as language models for prompt-style text classification. Our experiments show that random separators are competitive baselines, having less than a 1% difference compared to previous self-optimisation methods and showing a 12% average relative improvement over strong human baselines across nine text classification tasks and eight language models. We further analyse this phenomenon in detail using three different random generation strategies, establishing that the language space is rich with potentially good separators, with a greater than 40% average chance that a randomly drawn separator performs better than human-curated separators. These observations challenge the common assumption that an effective prompt should be human readable or task relevant and establish a strong baseline for prompt optimisation research.

pdf bib
ReTA: Recursively Thinking Ahead to Improve the Strategic Reasoning of Large Language Models
Jinhao Duan | Shiqi Wang | James Diffenderfer | Lichao Sun | Tianlong Chen | Bhavya Kailkhura | Kaidi Xu

Current logical reasoning evaluations of Large Language Models (LLMs) primarily focus on single-turn and static environments, such as arithmetic problems. The crucial problem of multi-turn, strategic reasoning is under-explored. In this work, we analyze the multi-turn strategic reasoning of LLMs through text-driven complete- and incomplete-information gaming, e.g., board games (Tic-Tac-Toe, Connect-4) and poker games (Texas Hold’em Poker). Specifically, we consider two distinct scenarios: 1) Online Racing, featuring multiple LLMs/agents to facilitate direct competition and comparison; 2) Offline Probing, constructing targeted questions with verified ground truth to evaluate LLMs’ strategic behaviors. Experimental results demonstrate that existing state-of-the-art LLMs and reasoning schemes are largely ineffective for strategic reasoning tasks. To mitigate these limitations, we propose a simple yet effective Recursively Thinking-Ahead (ReTA) agent, incorporating a recursive prompting mechanism that automatically analyzes the opponents’ future moves/actions and assigns reward signals for these situations, to strengthen the strategic reasoning of LLMs. We hope our work could spur further research and exploration in the multi-turn strategic reasoning of LLMs. The code is available at https://github.com/jinhaoduan/ReTA.

pdf bib
Fact Checking Beyond Training Set
Payam Karisani | Heng Ji

Evaluating the veracity of everyday claims is time consuming and in some cases requires domain expertise. We empirically demonstrate that the commonly used fact checking pipeline, known as the retriever-reader, suffers from performance deterioration when it is trained on the labeled data from one domain and used in another domain. Afterwards, we delve into each component of the pipeline and propose novel algorithms to address this problem. We propose an adversarial algorithm to make the retriever component robust against distribution shift. Our core idea is to initially train a bi-encoder on the labeled source data, and then, to adversarially train two separate document and claim encoders using unlabeled target data. We then focus on the reader component and propose to train it such that it is insensitive towards the order of claims and evidence documents. Our empirical evaluations support the hypothesis that such a reader shows a higher robustness against distribution shift. To our knowledge, there is no publicly available multi-topic fact checking dataset. Thus, we propose a simple automatic method to re-purpose two well-known fact checking datasets. We then construct eight fact checking scenarios from these datasets, and compare our model to a set of strong baseline models, including recent domain adaptation models that use GPT4 for generating synthetic data.

pdf bib
Program-Aided Reasoners (Better) Know What They Know
Anubha Kabra | Sanketh Rangreji | Yash Mathur | Aman Madaan | Emmy Liu | Graham Neubig

Prior work shows that program-aided reasoning, in which large language models (LLMs) are combined with programs written in programming languages such as Python, can significantly improve accuracy on various reasoning tasks. However, while accuracy is essential, it is also important for such reasoners to “know what they know”, which can be quantified through the calibration of the model. In this paper, we compare the calibration of Program Aided Language Models (PAL) and text-based Chain-of-thought (COT) prompting techniques over 5 datasets and 2 model types - LLaMA models and OpenAI models. Our results indicate that PAL leads to improved calibration in 75% of the instances. Our analysis uncovers that prompting styles that produce lesser diversity in generations also have more calibrated results, and thus we also experiment with inducing lower generation diversity using temperature scaling and find that for certain temperatures, PAL is not only more accurate but is also more calibrated than COT. Overall, we demonstrate that, in the majority of cases, program-aided reasoners better know what they know than text-based counterparts.

pdf bib
The Perspectivist Paradigm Shift: Assumptions and Challenges of Capturing Human Labels
Eve Fleisig | Su Lin Blodgett | Dan Klein | Zeerak Talat

Longstanding data labeling practices in machine learning involve collecting and aggregating labels from multiple annotators. But what should we do when annotators disagree? Though annotator disagreement has long been seen as a problem to minimize, new perspectivist approaches challenge this assumption by treating disagreement as a valuable source of information. In this position paper, we examine practices and assumptions surrounding the causes of disagreement–some challenged by perspectivist approaches, and some that remain to be addressed–as well as practical and normative challenges for work operating under these assumptions. We conclude with recommendations for the data labeling pipeline and avenues for future research engaging with subjectivity and disagreement.

pdf bib
Principles from Clinical Research for NLP Model Generalization
Aparna Elangovan | Jiayuan He | Yuan Li | Karin Verspoor

The NLP community typically relies on performance of a model on a held-out test set to assess generalization. Performance drops observed in datasets outside of official test sets are generally attributed to “out-of-distribution” effects. Here, we explore the foundations of generalizability and study the factors that affect it, articulating lessons from clinical studies. In clinical research, generalizability is an act of reasoning that depends on (a) *internal validity* of experiments to ensure controlled measurement of cause and effect, and (b) *external validity* or transportability of the results to the wider population. We demonstrate how learning spurious correlations, such as the distance between entities in relation extraction tasks, can affect a model’s internal validity and in turn adversely impact generalization. We, therefore, present the need to ensure internal validity when building machine learning models in NLP. Our recommendations also apply to generative large language models, as they are known to be sensitive to even minor semantic preserving alterations. We also propose adapting the idea of *matching* in randomized controlled trials and observational studies to NLP evaluation to measure causation.

pdf bib
First Tragedy, then Parse: History Repeats Itself in the New Era of Large Language Models
Naomi Saphra | Eve Fleisig | Kyunghyun Cho | Adam Lopez

Many NLP researchers are experiencing an existential crisis triggered by the astonishing success of ChatGPT and other systems based on large language models (LLMs). After such a disruptive change to our understanding of the field, what is left to do? Taking a historical lens, we look for guidance from the first era of LLMs, which began in 2005 with large n-gram models for machine translation (MT). We identify durable lessons from the first era, and more importantly, we identify evergreen problems where NLP researchers can continue to make meaningful contributions in areas where LLMs are ascendant. We argue that disparities in scale are transient and researchers can work to reduce them; that data, rather than hardware, is still a bottleneck for many applications; that meaningful realistic evaluation is still an open problem; and that there is still room for speculative approaches.

pdf bib
Found in the Middle: Permutation Self-Consistency Improves Listwise Ranking in Large Language Models
Raphael Tang | Crystina Zhang | Xueguang Ma | Jimmy Lin | Ferhan Ture

Large language models (LLMs) exhibit positional bias in how they use context, which especially affects listwise ranking. To address this, we propose permutation self-consistency, a form of self-consistency over the ranking list outputs of black-box LLMs. Our key idea is to marginalize out different list orders in the prompt to produce an order-independent ranking with less positional bias. First, given some input prompt, we repeatedly shuffle the list in the prompt and pass it through the LLM while holding the instructions the same. Next, we aggregate the resulting sample of rankings by computing the central ranking closest in distance to all of them, marginalizing out prompt order biases in the process. Theoretically, we prove the robustness of our method, showing convergence to the true ranking under random perturbations.Empirically, on five datasets in sorting and passage reranking, our approach improves scores from conventional inference by up to 34-52% for Mistral, 7-18% for GPT-3.5, 8-16% for LLaMA v2 (70B). Our code is at https://github.com/castorini/perm-sc.

pdf bib
From Language Modeling to Instruction Following: Understanding the Behavior Shift in LLMs after Instruction Tuning
Xuansheng Wu | Wenlin Yao | Jianshu Chen | Xiaoman Pan | Xiaoyang Wang | Ninghao Liu | Dong Yu

Large Language Models (LLMs) have achieved remarkable success, where instruction tuning is the critical step in aligning LLMs with user intentions. In this work, we investigate how the instruction tuning adjusts pre-trained models with a focus on intrinsic changes. Specifically, we first develop several local and global explanation methods, including a gradient-based method for input-output attribution, and techniques for interpreting patterns and concepts in self-attention and feed-forward layers. The impact of instruction tuning is then studied by comparing the explanations derived from the pre-trained and instruction-tuned models. This approach provides an internal perspective of the model shifts on a human-comprehensible level. Our findings reveal three significant impacts of instruction tuning: 1) It empowers LLMs to recognize the instruction parts of user prompts, and promotes the response generation constantly conditioned on the instructions. 2) It encourages the self-attention heads to capture more word-word relationships about instruction verbs. 3) It encourages the feed-forward networks to rotate their pre-trained knowledge toward user-oriented tasks. These insights contribute to a more comprehensive understanding of instruction tuning and lay the groundwork for future work that aims at explaining and optimizing LLMs for various applications. Our code and data are publicly available at https://github.com/JacksonWuxs/Interpret_Instruction_Tuning_LLMs.

pdf bib
POLYIE: A Dataset of Information Extraction from Polymer Material Scientific Literature
Jerry Cheung | Yuchen Zhuang | Yinghao Li | Pranav Shetty | Wantian Zhao | Sanjeev Grampurohit | Rampi Ramprasad | Chao Zhang

Scientific information extraction (SciIE), which aims to automatically extract information from scientific literature, is becoming more important than ever. However, there are no existing SciIE datasets for polymer materials, which is an important class of materials used ubiquitously in our daily lives. To bridge this gap, we introduce POLYIE, a new SciIE dataset for polymer materials. POLYIE is curated from 146 full-length polymer scholarly articles, which are annotated with different named entities (i.e., materials, properties, values, conditions) as well as their N-ary relations by domain experts. POLYIE presents several unique challenges due to diverse lexical formats of entities, ambiguity between entities, and variable-length relations. We evaluate state-of-the-art named entity extraction and relation extraction models on POLYIE, analyze their strengths and weaknesses, and highlight some difficult cases for these models. To the best of our knowledge, POLYIE is the first SciIE benchmark for polymer materials, and we hope it will lead to more research efforts from the community on this challenging task. Our code and data are available on: https://github.com/jerry3027/PolyIE.

pdf bib
LLM-based Medical Assistant Personalization with Short- and Long-Term Memory Coordination
Kai Zhang | Yangyang Kang | Fubang Zhao | Xiaozhong Liu

Large Language Models (LLMs), such as GPT3.5, have exhibited remarkable proficiency in comprehending and generating natural language. On the other hand, medical assistants hold the potential to offer substantial benefits for individuals. However, the exploration of LLM-based personalized medical assistant remains relatively scarce. Typically, patients converse differently based on their background and preferences which necessitates the task of enhancing user-oriented medical assistant. While one can fully train an LLM for this objective, the resource consumption is unaffordable. Prior research has explored memory-based methods to enhance the response with aware of previous mistakes for new queries during a dialogue session. We contend that a mere memory module is inadequate and fully training an LLM can be excessively costly. In this study, we propose a novel computational bionic memory mechanism, equipped with a parameter-efficient fine-tuning (PEFT) schema, to personalize medical assistants. To encourage further research into this area, we are releasing a new conversation dataset generated based on an open-source medical corpus and our implementation.

pdf bib
SumTra: A Differentiable Pipeline for Few-Shot Cross-Lingual Summarization
Jacob Parnell | Inigo Jauregi Unanue | Massimo Piccardi

Cross-lingual summarization (XLS) generates summaries in a language different from that of the input documents (e.g., English to Spanish), allowing speakers of the target language to gain a concise view of their content. In the present day, the predominant approach to this task is to take a performing, pretrained multilingual language model (LM) and fine-tune it for XLS on the language pairs of interest. However, the scarcity of fine-tuning samples makes this approach challenging in some cases. For this reason, in this paper we propose revisiting the summarize-and-translate pipeline, where the summarization and translation tasks are performed in a sequence. This approach allows reusing the many, publicly-available resources for monolingual summarization and translation, obtaining a very competitive zero-shot performance. In addition, the proposed pipeline is completely differentiable end-to-end, allowing it to take advantage of few-shot fine-tuning, where available. Experiments over two contemporary and widely adopted XLS datasets (CrossSum and WikiLingua) have shown the remarkable zero-shot performance of the proposed approach, and also its strong few-shot performance compared to an equivalent multilingual LM baseline, that the proposed approach has been able to outperform in many languages with only 10% of the fine-tuning samples.

pdf bib
KTRL+F: Knowledge-Augmented In-Document Search
Hanseok Oh | Haebin Shin | Miyoung Ko | Hyunji Lee | Minjoon Seo

We introduce a new problem KTRL+F, a knowledge-augmented in-document search that necessitates real-time identification of all semantic targets within a document with the awareness of external sources through a single natural query. KTRL+F addresses following unique challenges for in-document search: 1) utilizing knowledge outside the document for extended use of additional information about targets, and 2) balancing between real-time applicability with the performance.We analyze various baselines in KTRL+F and find limitations of existing models, such as hallucinations, high latency, or difficulties in leveraging external knowledge. Therefore, we propose a Knowledge-Augmented Phrase Retrieval model that shows a promising balance between speed and performance by simply augmenting external knowledge in phrase embedding. We also conduct a user study to verify whether solving KTRL+F can enhance search experience for users. It demonstrates that even with our simple model, users can reduce the time for searching with less queries and reduced extra visits to other sources for collecting evidence. We encourage the research community to work on KTRL+F to enhance more efficient in-document information access.

pdf bib
How Well Do Large Language Models Truly Ground?
Hyunji Lee | Se June Joo | Chaeeun Kim | Joel Jang | Doyoung Kim | Kyoung-Woon On | Minjoon Seo

To reduce issues like hallucinations and lack of control in Large Language Models (LLMs), a common method is to generate responses by grounding on external contexts given as input, known as knowledge-augmented models. However, previous research often narrowly defines “grounding” as just having the correct answer, which does not ensure the reliability of the entire response. To overcome this, we propose a stricter definition of grounding: a model is truly grounded if it (1) fully utilizes the necessary knowledge from the provided context, and (2) stays within the limits of that knowledge. We introduce a new dataset and a grounding metric to evaluate model capability under the definition. We perform experiments across 25 LLMs of different sizes and training methods and provide insights into factors that influence grounding performance. Our findings contribute to a better understanding of how to improve grounding capabilities and suggest an area of improvement toward more reliable and controllable LLM applications.

pdf bib
ALBA: Adaptive Language-Based Assessments for Mental Health
Vasudha Varadarajan | Sverker Sikström | Oscar Kjell | H. Schwartz

Mental health issues differ widely among individuals, with varied signs and symptoms. Recently, language-based assessments haveshown promise in capturing this diversity, but they require a substantial sample of words per person for accuracy. This work introducesthe task of Adaptive Language-Based Assessment (ALBA), which involves adaptively ordering questions while also scoring an individual’s latent psychological trait using limited language responses to previous questions. To this end, we develop adaptive testing methods under two psychometric measurement theories: Classical Test Theory and Item Response Theory.We empirically evaluate ordering and scoring strategies, organizing into two new methods: a semi-supervised item response theory-basedmethod (ALIRT) and a supervised Actor-Critic model. While we found both methods to improve over non-adaptive baselines, We foundALIRT to be the most accurate and scalable, achieving the highest accuracy with fewer questions (e.g., Pearson r ≈ 0.93 after only 3 questions as compared to typically needing at least 7 questions). In general, adaptive language-based assessments of depression and anxiety were able to utilize a smaller sample of language without compromising validity or large computational costs.

pdf bib
FREB-TQA: A Fine-Grained Robustness Evaluation Benchmark for Table Question Answering
Wei Zhou | Mohsen Mesgar | Heike Adel | Annemarie Friedrich

Table Question Answering (TQA) aims at composing an answer to a question based on tabular data. While prior research has shown that TQA models lack robustness, understanding the underlying cause and nature of this issue remains predominantly unclear, posing a significant obstacle to the development of robust TQA systems. In this paper, we formalize three major desiderata for a fine-grained evaluation of robustness of TQA systems. They should (i) answer questions regardless of alterations in table structure, (ii) base their responses on the content of relevant cells rather than on biases, and (iii) demonstrate robust numerical reasoning capabilities. To investigate these aspects, we create and publish a novel TQA evaluation benchmark in English. Our extensive experimental analysis reveals that none of the examined state-of-the-art TQA systems consistently excels in these three aspects. Our benchmark is a crucial instrument for monitoring the behavior of TQA systems and paves the way for the development of robust TQA systems. We release our benchmark publicly.

pdf bib
MILL: Mutual Verification with Large Language Models for Zero-Shot Query Expansion
Pengyue Jia | Yiding Liu | Xiangyu Zhao | Xiaopeng Li | Changying Hao | Shuaiqiang Wang | Dawei Yin

Query expansion, pivotal in search engines, enhances the representation of user information needs with additional terms. While existing methods expand queries using retrieved or generated contextual documents, each approach has notable limitations. Retrieval-based methods often fail to accurately capture search intent, particularly with brief or ambiguous queries. Generation-based methods, utilizing large language models (LLMs), generally lack corpus-specific knowledge and entail high fine-tuning costs. To address these gaps, we propose a novel zero-shot query expansion framework utilizing LLMs for mutual verification. Specifically, we first design a query-query-document generation method, leveraging LLMs’ zero-shot reasoning ability to produce diverse sub-queries and corresponding documents. Then, a mutual verification process synergizes generated and retrieved documents for optimal expansion. Our proposed method is fully zero-shot, and extensive experiments on three public benchmark datasets are conducted to demonstrate its effectiveness over existing methods. Our code is available online at https://github.com/Applied-Machine-Learning-Lab/MILL to ease reproduction.

pdf bib
Efficient Benchmarking (of Language Models)
Yotam Perlitz | Elron Bandel | Ariel Gera | Ofir Arviv | Liat Ein-Dor | Eyal Shnarch | Noam Slonim | Michal Shmueli-Scheuer | Leshem Choshen

The increasing versatility of language models (LMs) has given rise to a new class of benchmarks that comprehensively assess a broad range of capabilities. Such benchmarks are associated with massive computational costs, extending to thousands of GPU hours per model. However, the efficiency aspect of these evaluation efforts had raised little discussion in the literature.In this work, we present the problem of Efficient Benchmarking, namely, intelligently reducing the computation costs of LM evaluation without compromising reliability. Using the HELM benchmark as a test case, we investigate how different benchmark design choices affect the computation-reliability trade-off. We propose to evaluate the reliability of such decisions, by using a new measure – Decision Impact on Reliability, DIoR for short.We find, for example, that a benchmark leader may change by merely removing a low-ranked model from the benchmark, and observe that a correct benchmark ranking can be obtained by considering only a fraction of the evaluation examples.Based on our findings, we outline a set of concrete recommendations for efficient benchmark design and utilization practices. To take a step further, we use our findings to propose an evaluation algorithm, that, when applied to the HELM benchmark, leads to dramatic cost savings with minimal loss of benchmark reliability, often reducing computation by x100 or more.

pdf bib
ReFACT: Updating Text-to-Image Models by Editing the Text Encoder
Dana Arad | Hadas Orgad | Yonatan Belinkov

Our world is marked by unprecedented technological, global, and socio-political transformations, posing a significant challenge to textto-image generative models. These models encode factual associations within their parameters that can quickly become outdated, diminishing their utility for end-users. To that end, we introduce ReFACT, a novel approach for editing factual associations in text-to-image models without relaying on explicit input from end-users or costly re-training. ReFACT updates the weights of a specific layer in the text encoder, modifying only a tiny portion of the model’s parameters and leaving the rest of the model unaffected.We empirically evaluate ReFACT on an existing benchmark, alongside a newly curated dataset.Compared to other methods, ReFACT achieves superior performance in both generalization to related concepts and preservation of unrelated concepts.Furthermore, ReFACT maintains image generation quality, making it a practical tool for updating and correcting factual information in text-to-image models.

pdf bib
A Likelihood Ratio Test of Genetic Relationship among Languages
V.S.D.S.Mahesh Akavarapu | Arnab Bhattacharya

Lexical resemblances among a group of languages indicate that the languages could be genetically related, i.e., they could have descended from a common ancestral language. However, such resemblances can arise by chance and, hence, need not always imply an underlying genetic relationship. Many tests of significance based on permutation of wordlists and word similarity measures appeared in the past to determine the statistical significance of such relationships. We demonstrate that although existing tests may work well for bilateral comparisons, i.e., on pairs of languages, they are either infeasible by design or are prone to yield false positives when applied to groups of languages or language families. To this end, inspired by molecular phylogenetics, we propose a likelihood ratio test to determine if given languages are related based on the proportion of invariant character sites in the aligned wordlists applied during tree inference. Further, we evaluate some language families and show that the proposed test solves the problem of false positives. Finally, we demonstrate that the test supports the existence of macro language families such as Nostratic and Macro-Mayan.

pdf bib
PaD: Program-aided Distillation Can Teach Small Models Reasoning Better than Chain-of-thought Fine-tuning
Xuekai Zhu | Biqing Qi | Kaiyan Zhang | Xinwei Long | Zhouhan Lin | Bowen Zhou

While large language models (LLMs) excel in various natural language processing tasks, their huge size and the inaccessibility of parameters present challenges for practical deployment. Previous studies try to distill task-specific ability from LLMs to smaller models, using data synthesis and chain-of-thought (CoT) fine-tuning. However, synthetic CoT data often contains faulty reasoning, which deteriorates the quality of distillation, especially in reasoning capabilities. In this work, we propose Program-aided Distillation (PaD), which introduces reasoning programs to suppress the errors in distilled data, and thus achieves better distillation quality for reasoning tasks. In PaD, we utilize the reasoning program to substitute the CoT, allowing automated error checking of synthetic data. Further, through error injecting and further training, the small distilling model could iteratively self-refine the reasoning. Moreover, we conduct a step-wise beam search by step-by-step verifying to acquire more exact reasoning chains. We evaluate PaD on arithmetic reasoning, symbolic reasoning, and general ability.Experimental results demonstrate that smaller models using PaD can not only outperform certain LLMs (e.g., LLaMA-1 13B) but also achieve strong improvement over baselines with a significantly smaller scale of parameters and data. The source code is publicly available athttps://github.com/Xuekai-Zhu/pad.

pdf bib
MEGAVERSE: Benchmarking Large Language Models Across Languages, Modalities, Models and Tasks
Sanchit Ahuja | Divyanshu Aggarwal | Varun Gumma | Ishaan Watts | Ashutosh Sathe | Millicent Ochieng | Rishav Hada | Prachi Jain | Mohamed Ahmed | Kalika Bali | Sunayana Sitaram

There has been a surge in LLM evaluation research to understand LLM capabilities and limitations. However, much of this research has been confined to English, leaving LLM building and evaluation for non-English languages relatively unexplored. Several new LLMs have been introduced recently, necessitating their evaluation on non-English languages. This study aims to perform a thorough evaluation of the non-English capabilities of SoTA LLMs (GPT-3.5-Turbo, GPT-4, PaLM2, Gemini-Pro, Mistral, Llama2, and Gemma) by comparing them on the same set of multilingual datasets. Our benchmark comprises 22 datasets covering 83 languages, including low-resource African languages. We also include two multimodal datasets in the benchmark and compare the performance of LLaVA models, GPT-4-Vision and Gemini-Pro-Vision. Our experiments show that larger models such as GPT-4, Gemini-Pro and PaLM2 outperform smaller models on various tasks, notably on low-resource languages, with GPT-4 outperforming PaLM2 and Gemini-Pro on more datasets. We also perform a study on data contamination and find that several models are likely to be contaminated with multilingual evaluation benchmarks, necessitating approaches to detect and handle contamination while assessing the multilingual performance of LLMs.

pdf bib
Unlocking Emergent Modularity in Large Language Models
Zihan Qiu | Zeyu Huang | Jie Fu

Modular Neural Networks (MNNs) demonstrate various advantages over monolithic models.Existing MNNs are generally explicit: their modular architectures are pre-defined, with individual modules expected to implement distinct functions.Recent works reveal that there exists implicit modularity in standard pre-trained transformers, namely Emergent Modularity.They indicate that such modular structures spontaneously exhibit during the early pre-training phase.Despite the benefits of modularity, most Language Models (LMs) are still treated as monolithic models in the pre-train and fine-tune paradigm, with their emergent modularity locked and underutilized.In this work, focusing on unlocking the emergent modularity in LMs, we showcase that standard LMs could be fine-tuned as their Mixture-of-Expert (MoEs) counterparts without introducing any extra parameters. Such MoEs are derived from emergent modularity and are referred to as Emergent MoEs (EMoE).Our experiments demonstrate that fine-tuning EMoE effectively improves downstream in-domain and out-of-domain generalization compared with vanilla fine-tuning.Our analysis and ablation studies further illustrate that it is robust to various configurations and can scale up to Large Language Models (i.e., Llama2-7B and Llama-30B). Code is available at https://github.com/qiuzh20/EMoE.

pdf bib
A School Student Essay Corpus for Analyzing Interactions of Argumentative Structure and Quality
Maja Stahl | Nadine Michel | Sebastian Kilsbach | Julian Schmidtke | Sara Rezat | Henning Wachsmuth

Learning argumentative writing is challenging. Besides writing fundamentals such as syntax and grammar, learners must select and arrange argument components meaningfully to create high-quality essays. To support argumentative writing computationally, one step is to mine the argumentative structure. When combined with automatic essay scoring, interactions of the argumentative structure and quality scores can be exploited for comprehensive writing support. Although studies have shown the usefulness of using information about the argumentative structure for essay scoring, no argument mining corpus with ground-truth essay quality annotations has been published yet. Moreover, none of the existing corpora contain essays written by school students specifically. To fill this research gap, we present a German corpus of 1,320 essays from school students of two age groups. Each essay has been manually annotated for argumentative structure and quality on multiple levels of granularity. We propose baseline approaches to argument mining and essay scoring, and we analyze interactions between both tasks, thereby laying the ground for quality-oriented argumentative writing support.

pdf bib
Adjusting Interpretable Dimensions in Embedding Space with Human Judgments
Katrin Erk | Marianna Apidianaki

Embedding spaces contain interpretable dimensions indicating gender, formality in style, or even object properties. This has been observed multiple times. Such interpretable dimensions are becoming valuable tools in different areas of study, from social science to neuroscience. The standard way to compute these dimensions uses contrasting seed words and computes difference vectors over them. This is simple but does not always work well. We combine seed-based vectors with guidance from human ratings of where words fall along a specific dimension, and evaluate on predicting both object properties like size and danger, and the stylistic properties of formality and complexity. We obtain interpretable dimensions with markedly better performance especially in cases where seed-based dimensions do not work well.

pdf bib
PatentEval: Understanding Errors in Patent Generation
You Zuo | Kim Gerdes | Éric Clergerie | Benoît Sagot

In this work, we introduce a comprehensive error typology specifically designed for evaluating two distinct tasks in machine-generated patent texts: claims-to-abstract generation, and the generation of the next claim given previous ones. We have also developed a benchmark, PatentEval, for systematically assessing language models in this context. Our study includes a comparative analysis, annotated by humans, of various models. These range from those specifically adapted during training for tasks within the patent domain to the latest general-purpose large language models (LLMs). Furthermore, we explored and evaluated some metrics to approximate human judgments in patent text evaluation, analyzing the extent to which these metrics align with expert assessments. These approaches provide valuable insights into the capabilities and limitations of current language models in the specialized field of patent text generation.

pdf bib
Contextual Refinement of Translations: Large Language Models for Sentence and Document-Level Post-Editing
Sai Koneru | Miriam Exel | Matthias Huck | Jan Niehues

Large language models (LLMs) have demonstrated considerable success in various natural language processing tasks, but open-source LLMs have yet to attain state-of-the-art performance in Neural Machine Translation (NMT). Nevertheless, their significant performance in tasks demanding a broad understanding and contextual processing shows their potential for translation. To exploit these abilities, we investigate using LLMs for MT and explore recent parameter-efficient fine-tuning techniques. Surprisingly, our initial experiments found that fine-tuning with Q-LoRA for translation purposes led to performance improvements in terms of BLEU but degradation in COMET compared to in-context learning. To overcome this, we propose an alternative approach: adapting LLMs as Automatic Post-Editors (APE) rather than direct translators. Building on the ability of the LLM to handle long sequences, we also propose extending our approach to document-level translation. We show that leveraging Low-Rank-Adapter fine-tuning for APE can yield significant improvements across both sentence and document-level metrics while generalizing to out-of-domain data. Most notably, we achieve a state-of-the-art accuracy rate of 88.7% on the ContraPro test set, which assesses the model’s ability to resolve pronoun ambiguities when translating from English to German. Lastly, during manual post-editing for document-level translation, the source sentences are iteratively annotated, which can be used to refine further translations in the document. Here, we demonstrate that leveraging human corrections can significantly reduce the number of edits required for subsequent translations.

pdf bib
Metaphor Detection with Context Enhancement and Curriculum Learning
Kaidi Jia | Rongsheng Li

Metaphor detection is a challenging task for natural language processing (NLP) systems. Previous works failed to sufficiently utilize the internal and external semantic relationships between target words and their context. Furthermore, they have faced challenges in tackling the problem of data sparseness due to the very limited available training data. To address these two challenges, we propose a novel model called MiceCL. By leveraging the difference between the literal meaning of the target word and the meaning of the sentence as the sentence external difference, MiceCL can better handle the semantic relationships. Additionally, we propose a curriculum learning framework for automatically assessing difficulty of the sentence with a pre-trained model. By starting from easy examples and gradually progressing to more difficult ones, we can ensure that the model will not deal with complex data when its ability is weak so that to avoid wasting limited data. Experimental results demonstrate that MiceCL achieves competitive performance across multiple datasets, with a significantly improved convergence speed compared to other models.

pdf bib
What Causes the Failure of Explicit to Implicit Discourse Relation Recognition?
Wei Liu | Stephen Wan | Michael Strube

We consider an unanswered question in the discourse processing community: why do relation classifiers trained on explicit examples (with connectives removed) perform poorly in real implicit scenarios? Prior work claimed this is due to linguistic dissimilarity between explicit and implicit examples but provided no empirical evidence. In this study, we show that one cause for such failure is a label shift after connectives are eliminated. Specifically, we find that the discourse relations expressed by some explicit instances will change when connectives disappear. Unlike previous work manually analyzing a few examples, we present empirical evidence at the corpus level to prove the existence of such shift. Then, we analyze why label shift occurs by considering factors such as the syntactic role played by connectives, ambiguity of connectives, and more. Finally, we investigate two strategies to mitigate the label shift: filtering out noisy data and joint learning with connectives. Experiments on PDTB 2.0, PDTB 3.0, and the GUM dataset demonstrate that classifiers trained with our strategies outperform strong baselines.

pdf bib
UniverSLU: Universal Spoken Language Understanding for Diverse Tasks with Natural Language Instructions
Siddhant Arora | Hayato Futami | Jee-weon Jung | Yifan Peng | Roshan Sharma | Yosuke Kashiwagi | Emiru Tsunoo | Karen Livescu | Shinji Watanabe

Recent studies leverage large language models with multi-tasking capabilities, using natural language prompts to guide the model’s behavior and surpassing performance of task-specific models. Motivated by this, we ask: can we build a single model that jointly performs various spoken language understanding (SLU) tasks? We start by adapting a pre-trained automatic speech recognition model to additional tasks using single-token task specifiers. We enhance this approach through instruction tuning, i.e., finetuning by describing the task using natural language instructions followed by the list of label options. Our approach can generalize to new task descriptions for the seen tasks during inference, thereby enhancing its user-friendliness. We demonstrate the efficacy of our single multi-task learning model “UniverSLU” for 12 speech classification and sequence generation task types spanning 17 datasets and 9 languages. On most tasks, UniverSLU achieves competitive performance and often even surpasses task-specific models. Additionally, we assess the zero-shot capabilities, finding that the model generalizes to new datasets and languages for seen task types.

pdf bib
How Trustworthy are Open-Source LLMs? An Assessment under Malicious Demonstrations Shows their Vulnerabilities
Lingbo Mo | Boshi Wang | Muhao Chen | Huan Sun

The rapid progress in open-source Large Language Models (LLMs) is significantly driving AI development forward. However, there is still a limited understanding of their trustworthiness. Deploying these models at scale without sufficient trustworthiness can pose significant risks, highlighting the need to uncover these issues promptly. In this work, we conduct an adversarial assessment of open-source LLMs on trustworthiness, scrutinizing them across eight different aspects including toxicity, stereotypes, ethics, hallucination, fairness, sycophancy, privacy, and robustness against adversarial demonstrations. We propose advCoU, an extended Chain of Utterances-based (CoU) prompting strategy by incorporating carefully crafted malicious demonstrations for trustworthiness attack. Our extensive experiments encompass recent and representative series of open-source LLMs, including Vicuna, MPT, Falcon, Mistral, and Llama 2. The empirical outcomes underscore the efficacy of our attack strategy across diverse aspects. More interestingly, our result analysis reveals that models with superior performance in general NLP tasks do not always have greater trustworthiness; in fact, larger models can be more vulnerable to attacks. Additionally, models that have undergone instruction tuning, focusing on instruction following, tend to be more susceptible, although fine-tuning LLMs for safety alignment proves effective in mitigating adversarial trustworthiness attacks.

pdf bib
Paraphrase and Solve: Exploring and Exploiting the Impact of Surface Form on Mathematical Reasoning in Large Language Models
Yue Zhou | Yada Zhu | Diego Antognini | Yoon Kim | Yang Zhang

This paper studies the relationship between the surface form of a mathematical problem and its solvability by large language models. We find that subtle alterations in the surface form can significantly impact the answer distribution and the solve rate, exposing the language model’s lack of robustness and sensitivity to the surface form in reasoning through complex problems. To improve mathematical reasoning performance, we propose Self-Consistency-over-Paraphrases (SCoP), which diversifies reasoning paths from specific surface forms of the problem. We evaluate our approach on four mathematics reasoning benchmarks over three large language models and show that SCoP improves mathematical reasoning performance over vanilla self-consistency, particularly for problems initially deemed unsolvable. Finally, we provide additional experiments and discussion regarding problem difficulty and surface forms, including cross-model difficulty agreement and paraphrasing transferability, and Variance of Variations (VOV) for language model evaluation.

pdf bib
TriSum: Learning Summarization Ability from Large Language Models with Structured Rationale
Pengcheng Jiang | Cao Xiao | Zifeng Wang | Parminder Bhatia | Jimeng Sun | Jiawei Han

The advent of large language models (LLMs) has significantly advanced natural language processing tasks like text summarization. However, their large size and computational demands, coupled with privacy concerns in data transmission, limit their use in resource-constrained and privacy-centric settings. To overcome this, we introduce TriSum, a framework for distilling LLMs’ text summarization abilities into a compact, local model. Initially, LLMs extract a set of aspect-triple rationales and summaries, which are refined using a dual-scoring method for quality. Next, a smaller local model is trained with these tasks, employing a curriculum learning strategy that evolves from simple to complex tasks. Our method enhances local model performance on various benchmarks (CNN/DailyMail, XSum, and ClinicalTrial), outperforming baselines by 4.5%, 8.5%, and 7.4%, respectively. It also improves interpretability by providing insights into the summarization rationale.

pdf bib
GenRES: Rethinking Evaluation for Generative Relation Extraction in the Era of Large Language Models
Pengcheng Jiang | Jiacheng Lin | Zifeng Wang | Jimeng Sun | Jiawei Han

The field of relation extraction (RE) is experiencing a notable shift towards generative relation extraction (GRE), leveraging the capabilities of large language models (LLMs). However, we discovered that traditional relation extraction (RE) metrics like precision and recall fall short in evaluating GRE methods. This shortfall arises because these metrics rely on exact matching with human-annotated reference relations, while GRE methods often produce diverse and semantically accurate relations that differ from the references. To fill this gap, we introduce GenRES for a multi-dimensional assessment in terms of the topic similarity, uniqueness, granularity, factualness, and completeness of the GRE results. With GenRES, we empirically identified that (1) precision/recall fails to justify the performance of GRE methods; (2) human-annotated referential relations can be incomplete; (3) prompting LLMs with a fixed set of relations or entities can cause hallucinations. Next, we conducted a human evaluation of GRE methods that shows GenRES is consistent with human preferences for RE quality. Last, we made a comprehensive evaluation of fourteen leading LLMs using GenRES across document, bag, and sentence level RE datasets, respectively, to set the benchmark for future research in GRE

pdf bib
Curated Datasets and Neural Models for Machine Translation of Informal Registers between Mayan and Spanish Vernaculars
Andrés Lou | Juan Antonio Pérez-Ortiz | Felipe Sánchez-Martínez | Víctor Sánchez-Cartagena

The Mayan languages comprise a language family with an ancient history, millions of speakers, and immense cultural value, that, nevertheless, remains severely underrepresented in terms of resources and global exposure. In this paper we develop, curate, and publicly release a set of corpora in several Mayan languages spoken in Guatemala and Southern Mexico, which we call MayanV. The datasets are parallel with Spanish, the dominant language of the region, and are taken from official native sources focused on representing informal, day-to-day, and non-domain-specific language. As such, and according to our dialectometric analysis, they differ in register from most other available resources. Additionally, we present neural machine translation models, trained on as many resources and Mayan languages as possible, and evaluated exclusively on our datasets. We observe lexical divergences between the dialects of Spanish in our resources and the more widespread written standard of Spanish, and that resources other than the ones we present do not seem to improve translation performance, indicating that many such resources may not accurately capture common, real-life language usage. The MayanV dataset is available at https://github.com/transducens/mayanv.

pdf bib
The Effect of Data Partitioning Strategy on Model Generalizability: A Case Study of Morphological Segmentation
Zoey Liu | Bonnie Dorr

Recent work to enhance data partitioning strategies for more realistic model evaluation face challenges in providing a clear optimal choice. This study addresses these challenges, focusing on morphological segmentation and synthesizing limitations related to language diversity, adoption of multiple datasets and splits, and detailed model comparisons. Our study leverages data from 19 languages, including ten indigenous or endangered languages across 10 language families with diverse morphological systems (polysynthetic, fusional, and agglutinative) and different degrees of data availability. We conduct large-scale experimentation with varying sized combinations of training and evaluation sets as well as new test data. Our results show that, when faced with new test data: (1) models trained from random splits are able to achieve higher numerical scores; (2) model rankings derived from random splits tend to generalize more consistently.

pdf bib
Measuring Entrainment in Spontaneous Code-switched Speech
Debasmita Bhattacharya | Siying Ding | Alayna Nguyen | Julia Hirschberg

It is well-known that speakers who entrain to one another have more successful conversations than those who do not. Previous research has shown that interlocutors entrain on linguistic features in both written and spoken monolingual domains. More recent work on code-switched communication has also shown preliminary evidence of entrainment on certain aspects of code-switching (CSW). However, such studies of entrainment in code-switched domains have been extremely few and restricted to human-machine textual interactions. Our work studies code-switched spontaneous speech between humans, finding that (1) patterns of written and spoken entrainment in monolingual settings largely generalize to code-switched settings, and (2) some patterns of entrainment on code-switching in dialogue agent-generated text generalize to spontaneous code-switched speech. Our findings give rise to important implications for the potentially “universal” nature of entrainment as a communication phenomenon, and potential applications in inclusive and interactive speech technology.

pdf bib
A Survey of Meaning Representations – From Theory to Practical Utility
Zacchary Sadeddine | Juri Opitz | Fabian Suchanek

Symbolic meaning representations of natural language text have been studied since at least the 1960s. With the availability of large annotated corpora, and more powerful machine learning tools, the field has recently seen several new developments. In this survey, we study today’s most prominent Meaning Representation Frameworks. We shed light on their theoretical properties, as well as on their practical research environment, i.e., on datasets, parsers, applications, and future challenges.

pdf bib
Mitigating Language-Level Performance Disparity in mPLMs via Teacher Language Selection and Cross-lingual Self-Distillation
Haozhe Zhao | Zefan Cai | Shuzheng Si | Liang Chen | Yufeng He | Kaikai An | Baobao Chang

Large-scale multilingual Pretrained Language Models (mPLMs) yield impressive performance on cross-language tasks, yet significant performance disparities exist across different languages within the same mPLM. Previous studies endeavored to narrow these disparities by supervise fine-tuning the mPLMs with multilingual data.However, obtaining labeled multilingual data is time-consuming, and fine-tuning mPLM with limited labeled multilingual data merely encapsulates the knowledge specific to the labeled data.Therefore, we introduce **ALSACE** to leverage the learned knowledge from the well-performing languages to guide under-performing ones within the same mPLM, eliminating the need for additional labeled multilingual data. Experiments show that ALSACE effectively mitigates language-level performance disparity across various mPLMs while showing the competitive performance on different multilingual NLU tasks, ranging from full resource to limited resource settings. The code for our approach is available at https://github.com/pkunlp-icler/ALSACE.

pdf bib
Evaluating In-Context Learning of Libraries for Code Generation
Arkil Patel | Siva Reddy | Dzmitry Bahdanau | Pradeep Dasigi

Contemporary Large Language Models (LLMs) exhibit a high degree of code generation and comprehension capability. A particularly promising area is their ability to interpret code modules from unfamiliar libraries for solving user-instructed tasks. Recent work has shown that large proprietary LLMs can learn novel library usage in-context from demonstrations. These results raise several open questions: whether demonstrations of library usage is required, whether smaller (and more open) models also possess such capabilities, etc. In this work, we take a broader approach by systematically evaluating a diverse array of LLMs across three scenarios reflecting varying levels of domain specialization to understand their abilities and limitations in generating code based on libraries defined in-context. Our results show that even smaller open-source LLMs like Llama-2 and StarCoder demonstrate an adept understanding of novel code libraries based on specification presented in-context. Our findings further reveal that LLMs exhibit a surprisingly high proficiency in learning novel library modules even when provided with just natural language descriptions or raw code implementations of the functions, which are often cheaper to obtain than demonstrations. Overall, our results pave the way for harnessing LLMs in more adaptable and dynamic coding environments.

pdf bib
Visually-Aware Context Modeling for News Image Captioning
Tingyu Qu | Tinne Tuytelaars | Marie-Francine Moens

News Image Captioning aims to create captions from news articles and images, emphasizing the connection between textual context and visual elements. Recognizing the significance of human faces in news images and the face-name co-occurrence pattern in existing datasets, we propose a face-naming module for learning better name embeddings. Apart from names, which can be directly linked to an image area (faces), news image captions mostly contain context information that can only be found in the article. We design a retrieval strategy using CLIP to retrieve sentences that are semantically close to the image, mimicking human thought process of linking articles to images. Furthermore, to tackle the problem of the imbalanced proportion of article context and image context in captions, we introduce a simple yet effective method Contrasting with Language Model backbone (CoLaM) to the training pipeline. We conduct extensive experiments to demonstrate the efficacy of our framework. We out-perform the previous state-of-the-art (without external data) by 7.97/5.80 CIDEr scores on GoodNews/NYTimes800k. Our code is available at https://github.com/tingyu215/VACNIC.

pdf bib
Regularized Conventions: Equilibrium Computation as a Model of Pragmatic Reasoning
Athul Jacob | Gabriele Farina | Jacob Andreas

We present a game-theoretic model of pragmatics that we call ReCo (for Regularized Conventions). This model formulates pragmatic communication as a game in which players are rewarded for communicating successfully and penalized for deviating from a shared, “default” semantics. As a result, players assign utterances context-dependent meanings that jointly optimize communicative success and naturalness with respect to speakers’ and listeners’ background knowledge of language. By using established game-theoretic tools to compute equilibrium strategies for this game, we obtain principled pragmatic language generation procedures with formal guarantees of communicative success. Across several datasets capturing real and idealized human judgments about pragmatic implicature, ReCo matches, or slightly improves upon, predictions made by Iterated Best Response and Rational Speech Acts models of language understanding.

pdf bib
TopicGPT: A Prompt-based Topic Modeling Framework
Chau Pham | Alexander Hoyle | Simeng Sun | Philip Resnik | Mohit Iyyer

Topic modeling is a well-established technique for exploring text corpora. Conventional topic models (e.g., LDA) represent topics as bags of words that often require “reading the tea leaves” to interpret; additionally, they offer users minimal control over the formatting and specificity of resulting topics. To tackle these issues, we introduce TopicGPT, a prompt-based framework that uses large language models (LLMs) to uncover latent topics in a text collection. TopicGPT produces topics that align better with human categorizations compared to competing methods: it achieves a harmonic mean purity of 0.74 against human-annotated Wikipedia topics compared to 0.64 for the strongest baseline. Its topics are also more interpretable, dispensing with ambiguous bags of words in favor of topics with natural language labels and associated free-form descriptions. Moreover, the framework is highly adaptable, allowing users to specify constraints and modify topics without the need for model retraining. By streamlining access to high-quality and interpretable topics, TopicGPT represents a compelling, human-centered approach to topic modeling.

pdf bib
ChatGPT as an Attack Tool: Stealthy Textual Backdoor Attack via Blackbox Generative Model Trigger
Jiazhao Li | Yijin Yang | Zhuofeng Wu | V.G.Vinod Vydiswaran | Chaowei Xiao

Textual backdoor attacks, characterized by subtle manipulations of input triggers and training dataset labels, pose significant threats to security-sensitive applications. The rise of advanced generative models, such as GPT-4, with their capacity for human-like rewriting, makes these attacks increasingly challenging to detect. In this study, we conduct an in-depth examination of black-box generative models as tools for backdoor attacks, thereby emphasizing the need for effective defense strategies. We propose BGMAttack, a novel framework that harnesses advanced generative models to execute stealthier backdoor attacks on text classifiers. Unlike prior approaches constrained by subpar generation quality, BGMAttack renders backdoor triggers more elusive to human cognition and advanced machine detection. A rigorous evaluation of attack effectiveness over four sentiment classification tasks, complemented by four human cognition stealthiness tests, reveals BGMAttack’s superior performance, achieving a state-of-the-art attack success rate of 97.35% on average while maintaining superior stealth compared to conventional methods. The dataset and code are available: https://github.com/JiazhaoLi/BGMAttack.

pdf bib
Social Meme-ing: Measuring Linguistic Variation in Memes
Naitian Zhou | David Jurgens | David Bamman

Much work in the space of NLP has used computational methods to explore sociolinguistic variation in text. In this paper, we argue that memes, as multimodal forms of language comprised of visual templates and text, also exhibit meaningful social variation. We construct a computational pipeline to cluster individual instances of memes into templates and semantic variables, taking advantage of their multimodal structure in doing so. We apply this method to a large collection of meme images from Reddit and make available the resulting SemanticMemes dataset of 3.8M images clustered by their semantic function. We use these clusters to analyze linguistic variation in memes, discovering not only that socially meaningful variation in meme usage exists between subreddits, but that patterns of meme innovation and acculturation within these communities align with previous findings on written language.

pdf bib
ExpertQA: Expert-Curated Questions and Attributed Answers
Chaitanya Malaviya | Subin Lee | Sihao Chen | Elizabeth Sieber | Mark Yatskar | Dan Roth

As language models are adopted by a more sophisticated and diverse set of users, the importance of guaranteeing that they provide factually correct information supported by verifiable sources is critical across fields of study. This is especially the case for high-stakes fields, such as medicine and law, where the risk of propagating false information is high and can lead to undesirable societal consequences. Previous work studying attribution and factuality has not focused on analyzing these characteristics of language model outputs in domain-specific scenarios. In this work, we conduct human evaluation of responses from a few representative systems along various axes of attribution and factuality, by bringing domain experts in the loop. Specifically, we collect expert-curated questions from 484 participants across 32 fields of study, and then ask the same experts to evaluate generated responses to their own questions. In addition, we ask experts to improve upon responses from language models. The output of our analysis is ExpertQA, a high-quality long-form QA dataset with 2177 questions spanning 32 fields, along with verified answers and attributions for claims in the answers.

pdf bib
What if you said that differently?: How Explanation Formats Affect Human Feedback Efficacy and User Perception
Chaitanya Malaviya | Subin Lee | Dan Roth | Mark Yatskar

Eliciting feedback from end users of NLP models can be beneficial for improving models. However, how should we present model responses to users so they are most amenable to be corrected from user feedback? Further, what properties do users value to understand and trust responses? We answer these questions by analyzing the effect of rationales (or explanations) generated by QA models to support their answers. We specifically consider decomposed QA models that first extract an intermediate rationale based on a context and a question and then use solely this rationale to answer the question. A rationale outlines the approach followed by the model to answer the question. Our work considers various formats of these rationales that vary according to well-defined properties of interest. We sample rationales from language models using few-shot prompting for two datasets, and then perform two user studies. First, we present users with incorrect answers and corresponding rationales in various formats and ask them to provide natural language feedback to revise the rationale. We then measure the effectiveness of this feedback in patching these rationales through in-context learning. The second study evaluates how well different rationale formats enable users to understand and trust model answers, when they are correct. We find that rationale formats significantly affect how easy it is (1) for users to give feedback for rationales, and (2) for models to subsequently execute this feedback. In addition, formats with attributions to the context and in-depth reasoning significantly enhance user-reported understanding and trust of model outputs.

pdf bib
When Life Gives You Lemons, Make Cherryade: Converting Feedback from Bad Responses into Good Labels
Weiyan Shi | Emily Dinan | Kurt Shuster | Jason Weston | Jing Xu

Deployed dialogue agents have the potential to integrate human feedback to continuously improve themselves. However, humans may not always provide explicit signals when the chatbot makes mistakes during interactions. In this work, we propose Juicer, a framework to make use of both binary and free-form textual human feedback. It works by: (i) extending sparse binary feedback by training a satisfaction classifier to label the unlabeled data; and (ii) training a reply corrector to map the bad replies to good ones. We find that augmenting training with model-corrected replies improves the final dialogue model, and we can further improve performance by using both positive and negative replies through the recently proposed Director model.

pdf bib
Kreyòl-MT: Building MT for Latin American, Caribbean and Colonial African Creole Languages
Nathaniel Robinson | Raj Dabre | Ammon Shurtz | Rasul Dent | Onenamiyi Onesi | Claire Monroc | Loïc Grobol | Hasan Muhammad | Ashi Garg | Naome Etori | Vijay Murari Tiyyala | Olanrewaju Samuel | Matthew Stutzman | Bismarck Odoom | Sanjeev Khudanpur | Stephen Richardson | Kenton Murray

A majority of language technologies are tailored for a small number of high-resource languages, while relatively many low-resource languages are neglected. One such group, Creole languages, have long been marginalized in academic study, though their speakers could benefit from machine translation (MT). These languages are predominantly used in much of Latin America, Africa and the Caribbean. We present the largest cumulative dataset to date for Creole language MT, including 14.5M unique Creole sentences with parallel translations—11.6M of which we release publicly, and the largest bitexts gathered to date for 41 languages—the first ever for 21. In addition, we provide MT models supporting all 41 Creole languages in 172 translation directions. Given our diverse dataset, we produce a model for Creole language MT exposed to more genre diversity then ever before, which outperforms a genre-specific Creole MT model on its own benchmark for 23 of 34 translation directions.

pdf bib
Instructions as Backdoors: Backdoor Vulnerabilities of Instruction Tuning for Large Language Models
Jiashu Xu | Mingyu Ma | Fei Wang | Chaowei Xiao | Muhao Chen

We investigate security concerns of the emergent instruction tuning paradigm, that models are trained on crowdsourced datasets with task instructions to achieve superior performance. Our studies demonstrate that an attacker can inject backdoors by issuing very few malicious instructions (~1000 tokens) and control model behavior through data poisoning, without even the need to modify data instances or labels themselves. Through such instruction attacks, the attacker can achieve over 90% attack success rate across four commonly used NLP datasets. As an empirical study on instruction attacks, we systematically evaluated unique perspectives of instruction attacks, such as poison transfer where poisoned models can transfer to 15 diverse generative datasets in a zero-shot manner; instruction transfer where attackers can directly apply poisoned instruction on many other datasets; and poison resistance to continual finetuning. Lastly, we show that RLHF and clean demonstrations might mitigate such backdoors to some degree. These findings highlight the need for more robust defenses against poisoning attacks in instruction-tuning models and underscore the importance of ensuring data quality in instruction crowdsourcing.

pdf bib
Modeling Empathetic Alignment in Conversation
Jiamin Yang | David Jurgens

Empathy requires perspective-taking: empathetic responses require a person to reason about what another has experienced and communicate that understanding in language. However, most NLP approaches to empathy do not explicitly model this alignment process. Here, we introduce a new approach to recognizing alignment in empathetic speech, grounded in Appraisal Theory. We introduce a new dataset of over 9.2K span-level annotations of different types of appraisals of a person’s experience and over 3K empathetic alignments between a speaker’s and observer’s speech. Through computational experiments, we show that these appraisals and alignments can be accurately recognized. In experiments in over 9.2M Reddit conversations, we find that appraisals capture meaningful groupings of behavior but that most responses have minimal alignment. However, we find that mental health professionals engage with substantially more empathetic alignment.

pdf bib
Native Language Identification in Texts: A Survey
Dhiman Goswami | Sharanya Thilagan | Kai North | Shervin Malmasi | Marcos Zampieri

We present the first comprehensive survey of Native Language Identification (NLI) applied to texts. NLI is the task of automatically identifying an author’s native language (L1) based on their second language (L2) production. NLI is an important task with practical applications in second language teaching and NLP. The task has been widely studied for both text and speech, particularly for L2 English due to the availability of suitable corpora. Speech-based NLI relies heavily on accent modeled by pronunciation patterns and prosodic cues while text-based NLI relies primarily on modeling spelling errors and grammatical patterns that reveal properties of an individuals’ L1 influencing L2 production. We survey over one hundred papers on the topic including the papers associated with the NLI and INLI shared tasks. We describe several text representations and computational techniques used in text-based NLI. Finally, we present a comprehensive account of publicly available datasets used for the task thus far.

pdf bib
LoRETTA: Low-Rank Economic Tensor-Train Adaptation for Ultra-Low-Parameter Fine-Tuning of Large Language Models
Yifan Yang | Jiajun Zhou | Ngai Wong | Zheng Zhang

Various parameter-efficient fine-tuning (PEFT) techniques have been proposed to enable computationally efficient fine-tuning while maintaining model performance. However, existing PEFT methods are still limited by the growing number of trainable parameters with the rapid deployment of Large Language Models (LLMs). To address this challenge, we present LoRETTA, an ultra-parameter-efficient framework that significantly reduces trainable parameters through tensor-train decomposition. Specifically, we propose two methods, named LoRETTA_adp and LoRETTA_rep. The former employs tensorized adapters, offering a high-performance yet lightweight approach for the fine-tuning of LLMs. The latter emphasizes fine-tuning via weight reparameterization with a set of small tensor factors. LoRETTA achieves comparable or better performance than most widely used PEFT methods with up to 100× fewer parameters on the LLaMA-2-7B models. Furthermore, empirical results demonstrate that the proposed methods exhibit remarkable anti-overfitting capability, effectively improve training efficiency, and enjoy better multi-task learning performance. Plug-and-play loretta library built upon the Huggingface framework and PEFT library are provided.

pdf bib
Which One? Leveraging Context Between Objects and Multiple Views for Language Grounding
Chancharik Mitra | Abrar Anwar | Rodolfo Corona | Dan Klein | Trevor Darrell | Jesse Thomason

When connecting objects and their language referents in an embodied 3D environment, it is important to note that: (1) an object can be better characterized by leveraging comparative information between itself and other objects, and (2) an object’s appearance can vary with camera position. As such, we present the Multi-view Approach to Grounding in Context (MAGiC) model, which selects an object referent based on language that distinguishes between two similar objects. By pragmatically reasoning over both objects and across multiple views of those objects, MAGiC improves over the state-of-the-art model on the SNARE object reference task with a relative error reduction of 12.9% (representing an absolute improvement of 2.7%). Ablation studies show that reasoning jointly over object referent candidates and multiple views of each object both contribute to improved accuracy. Code: https://github.com/rcorona/magic_snare/

pdf bib
Do Localization Methods Actually Localize Memorized Data in LLMs? A Tale of Two Benchmarks
Ting-Yun Chang | Jesse Thomason | Robin Jia

The concept of localization in LLMs is often mentioned in prior work; however, methods for localization have never been systematically and directly evaluated. We propose two complementary benchmarks that evaluate the ability of localization methods to pinpoint LLM components responsible for memorized data. In our INJ benchmark, we actively inject a piece of new information into a small subset of LLM weights, enabling us to directly evaluate whether localization methods can identify these “ground truth” weights. In our DEL benchmark, we evaluate localization by measuring how much dropping out identified neurons deletes a memorized pretrained sequence. Despite their different perspectives, our two benchmarks yield consistent rankings of five localization methods. Methods adapted from network pruning perform well on both benchmarks, and all evaluated methods show promising localization ability. On the other hand, even successful methods identify neurons that are not specific to a single memorized sequence.

pdf bib
PromptFix: Few-shot Backdoor Removal via Adversarial Prompt Tuning
Tianrong Zhang | Zhaohan Xi | Ting Wang | Prasenjit Mitra | Jinghui Chen

Pre-trained language models (PLMs) have attracted enormous attention over the past few years with their unparalleled performances. Meanwhile, the soaring cost to train PLMs as well as their amazing generalizability have jointly contributed to few-shot fine-tuning and prompting as the most popular training paradigms for natural language processing (NLP) models. Nevertheless, existing studies have shown that these NLP models can be backdoored such that model behavior is manipulated when trigger tokens are presented.In this paper, we propose PromptFix, a novel backdoor mitigation strategy for NLP models via adversarial prompt-tuning in few-shot settings.Unlike existing NLP backdoor removal methods, which rely on accurate trigger inversion and subsequent model fine-tuning, PromptFix keeps the model parameters intact and only utilizes two extra sets of soft tokens which approximate the trigger and counteract it respectively. The use of soft tokens and adversarial optimization eliminates the need to enumerate possible backdoor configurations and enables an adaptive balance between trigger finding and preservation of performance.Experiments with various backdoor attacks validate the effectiveness of the proposed method and the performances when domain shift is present further shows PromptFix’s applicability to models pretrained on unknown data source which is the common case in prompt tuning scenarios.

pdf bib
Comparing Explanation Faithfulness between Multilingual and Monolingual Fine-tuned Language Models
Zhixue Zhao | Nikolaos Aletras

In many real natural language processing application scenarios, practitioners not only aim to maximize predictive performance but also seek faithful explanations for the model predictions. Rationales and importance distribution given by feature attribution methods (FAs) provide insights into how different parts of the input contribute to a prediction. Previous studies have explored how different factors affect faithfulness, mainly in the context of monolingual English models. On the other hand, the differences in FA faithfulness between multilingual and monolingual models have yet to be explored. Our extensive experiments, covering five languages and five popular FAs, show that FA faithfulness varies between multilingual and monolingual models. We find that the larger the multilingual model, the less faithful the FAs are compared to its counterpart monolingual models. Our further analysis shows that the faithfulness disparity is potentially driven by the differences between model tokenizers. Our code is available: https://github.com/casszhao/multilingual-faith.

pdf bib
A Pretrainer’s Guide to Training Data: Measuring the Effects of Data Age, Domain Coverage, Quality, & Toxicity
Shayne Longpre | Gregory Yauney | Emily Reif | Katherine Lee | Adam Roberts | Barret Zoph | Denny Zhou | Jason Wei | Kevin Robinson | David Mimno | Daphne Ippolito

Pretraining data design is critically under-documented and often guided by empirically unsupported intuitions. We pretrain models on data curated (1) at different collection times, (2) with varying toxicity and quality filters, and (3) with different domain compositions. First, we find that temporal shift between evaluation data and pretraining data leads to performance degradation, which is not overcome by finetuning. Second, we measure the effect of quality and toxicity filters, showing a trade-off between performance on standard benchmarks and risk of toxic generations. We also find that the effects of different types of filtering are not predictable from text domain characteristics. Third, we empirically validate that heterogeneous data sources, like books and web, are beneficial and warrant greater prioritization. To date, these experiments constitute the single largest publicly documented empirical study of the effects of pretraining data. Spanning 28 unique 1.5 billion parameter models pretrained from scratch, these findings validate, quantify, and expose many undocumented intuitions about text pretraining, which ultimately support more informed data-centric decisions in model development.

pdf bib
Instructional Fingerprinting of Large Language Models
Jiashu Xu | Fei Wang | Mingyu Ma | Pang Wei Koh | Chaowei Xiao | Muhao Chen

The exorbitant cost of training Large language models (LLMs) from scratch makes it essential to fingerprint the models to protect intellectual property via ownership authentication and to ensure downstream users and developers comply with their license terms (eg restricting commercial use). In this study, we present a pilot study on LLM fingerprinting as a form of very lightweight instruction tuning. Model publisher specifies a confidential private key and implants it as an instruction backdoor that causes the LLM to generate specific text when the key is present. Results on 11 popularly-used LLMs showed that this approach is lightweight and does not affect the normal behavior of the model. It also prevents publisher overclaim, maintains robustness against fingerprint guessing and parameter-efficient training, and supports multi-stage fingerprinting akin to MIT License.

pdf bib
Reinforced Multiple Instance Selection for Speaker Attribute Prediction
Alireza Salkhordeh Ziabari | Ali Omrani | Parsa Hejabi | Preni Golazizian | Brendan Kennedy | Payam Piray | Morteza Dehghani

Language usage is related to speaker age, gender, moral concerns, political ideology, and other attributes. Current state-of-the-art methods for predicting these attributes take a speaker’s utterances as input and provide a prediction per speaker attribute. Most of these approaches struggle to handle a large number of utterances per speaker. This difficulty is primarily due to the computational constraints of the models. Additionally, only a subset of speaker utterances may be relevant to specific attributes. In this paper, we formulate speaker attribute prediction as a Multiple Instance Learning (MIL) problem and propose RL-MIL, a novel approach based on Reinforcement Learning (RL) that effectively addresses both of these challenges. Our experiments demonstrate that our RL-based methodology consistently outperforms previous approaches across a range of related tasks: predicting speakers’ psychographics and demographics from social media posts, and political ideologies from transcribed speeches. We create synthetic datasets and investigate the behavior of RL-MIL systematically. Our results show the success of RL-MIL in improving speaker attribute prediction by learning to select relevant speaker utterances.

pdf bib
DynaMo: Accelerating Language Model Inference with Dynamic Multi-Token Sampling
Shikhar Tuli | Chi-Heng Lin | Yen-Chang Hsu | Niraj Jha | Yilin Shen | Hongxia Jin

Traditional language models operate autoregressively, i.e., they predict one token at a time. Rapid explosion in model sizes has resulted in high inference times. In this work, we propose DynaMo, a suite of multi-token prediction language models that reduce net inference times. Our models *dynamically* predict multiple tokens based on their confidence in the predicted joint probability distribution. We propose a lightweighttechnique to train these models, leveraging the weights of traditional autoregressive counterparts. Moreover, we propose novel ways to enhance the estimated joint probability to improve text generation quality, namely co-occurrence weighted masking and adaptive thresholding. We also propose systematic qualitative and quantitative methods to rigorously test the quality of generated text for non-autoregressive generation. One of the models in our suite, DynaMo-7.3B-T3, achieves same-quality generated text as the baseline (Pythia-6.9B) while achieving 2.57× speed-up with only 5.87% and 2.67% parameter and training time overheads, respectively.

pdf bib
Few-shot Knowledge Graph Relational Reasoning via Subgraph Adaptation
Haochen Liu | Song Wang | Chen Chen | Jundong Li

Few-shot Knowledge Graph (KG) Relational Reasoning aims to predict unseen triplets (i.e., query triplets) for rare relations in KGs, given only several triplets of these relations as references (i.e., support triplets). This task has gained significant traction due to the widespread use of knowledge graphs in various natural language processing applications. Previous approaches have utilized meta-training methods and manually constructed meta-relation sets to tackle this task. Recent efforts have focused on edge-mask-based methods, which exploit the structure of the contextualized graphs of target triplets (i.e., a subgraph containing relevant triplets in the KG). However, existing edge-mask-based methods have limitations in extracting insufficient information from KG and are highly influenced by spurious information in KG. To overcome these challenges, we propose SAFER (Subgraph Adaptation for Few-shot Relational Reasoning), a novel approach that effectively adapts the information in contextualized graphs to various subgraphs generated from support and query triplets to perform the prediction. Specifically, SAFER enables the extraction of more comprehensive information from support triplets while minimizing the impact of spurious information when predicting query triplets. Experimental results on three prevalent datasets demonstrate the superiority of our proposed framework SAFER.

pdf bib
Uncertainty Quantification for In-Context Learning of Large Language Models
Chen Ling | Xujiang Zhao | Xuchao Zhang | Wei Cheng | Yanchi Liu | Yiyou Sun | Mika Oishi | Takao Osaki | Katsushi Matsuda | Jie Ji | Guangji Bai | Liang Zhao | Haifeng Chen

In-context learning has emerged as a groundbreaking ability of Large Language Models (LLMs) and revolutionized various fields by providing a few task-relevant demonstrations in the prompt. However, trustworthy issues with LLM’s response, such as hallucination, have also been actively discussed. Existing works have been devoted to quantifying the uncertainty in LLM’s response, but they often overlook the complex nature of LLMs and the uniqueness of in-context learning. In this work, we delve into the predictive uncertainty of LLMs associated with in-context learning, highlighting that such uncertainties may stem from both the provided demonstrations (aleatoric uncertainty) and ambiguities tied to the model’s configurations (epistemic uncertainty). We propose a novel formulation and corresponding estimation method to quantify both types of uncertainties. The proposed method offers an unsupervised way to understand the prediction of in-context learning in a plug-and-play fashion. Extensive experiments are conducted to demonstrate the effectiveness of the decomposition. The code and data are available at: https://github.com/lingchen0331/UQ_ICL.

pdf bib
HelpSteer: Multi-attribute Helpfulness Dataset for SteerLM
Zhilin Wang | Yi Dong | Jiaqi Zeng | Virginia Adams | Makesh Narsimhan Sreedhar | Daniel Egert | Olivier Delalleau | Jane Scowcroft | Neel Kant | Aidan Swope | Oleksii Kuchaiev

Existing open-source helpfulness preference datasets do not specify what makes some responses more helpful and others less so. Models trained on these datasets can incidentally learn to model dataset artifacts (e.g. preferring longer but unhelpful responses only due to their length). To alleviate this problem, we collect HelpSteer, a multi-attribute helpfulness dataset annotated for the various aspects that make responses helpful. Specifically, our 37k-sample dataset has annotations for correctness, coherence, complexity, and verbosity in addition to overall helpfulness of responses. Training Llama 2 70B using the HelpSteer dataset with SteerLM technique produces a model that scores 7.54 on MT Bench, which is currently the highest score for open models that do not require training data from more powerful models (e.g. GPT-4). We release this dataset with CC-BY-4.0 license at https://huggingface.co/datasets/nvidia/HelpSteer

pdf bib
A Preference-driven Paradigm for Enhanced Translation with Large Language Models
Dawei Zhu | Sony Trenous | Xiaoyu Shen | Dietrich Klakow | Bill Byrne | Eva Hasler

Recent research has shown that large language models (LLMs) can achieve remarkable translation performance through supervised fine-tuning (SFT) using only a small amount of parallel data. However, SFT simply instructs the model to imitate the reference translations at the token level, making it vulnerable to the noise present in the references. Hence, the assistance from SFT often reaches a plateau once the LLMs have achieved a certain level of translation capability, and further increasing the size of parallel data does not provide additional benefits. To overcome this plateau associated with imitation-based SFT, we propose a preference-based approach built upon the Plackett-Luce model. The objective is to steer LLMs towards a more nuanced understanding of translation preferences from a holistic view, while also being more resilient in the absence of gold translations. We further build a dataset named MAPLE to verify the effectiveness of our approach, which includes multiple translations of varying quality for each source sentence. Extensive experiments demonstrate the superiority of our approach in “breaking the plateau” across diverse LLMs and test settings. Our in-depth analysis underscores the pivotal role of diverse translations and accurate preference scores in the success of our approach.

pdf bib
Fair Abstractive Summarization of Diverse Perspectives
Yusen Zhang | Nan Zhang | Yixin Liu | Alexander Fabbri | Junru Liu | Ryo Kamoi | Xiaoxin Lu | Caiming Xiong | Jieyu Zhao | Dragomir Radev | Kathleen McKeown | Rui Zhang

People from different social and demographic groups express diverse perspectives and conflicting opinions on a broad set of topics such as product reviews, healthcare, law, and politics. A fair summary should provide a comprehensive coverage of diverse perspectives without underrepresenting certain groups. However, current work in summarization metrics and Large Language Models (LLMs) evaluation has not explored fair abstractive summarization. In this paper, we systematically investigate fair abstractive summarization for user-generated data. We first formally define fairness in abstractive summarization as not underrepresenting perspectives of any groups of people, and we propose four reference-free automatic metrics by measuring the differences between target and source perspectives. We evaluate nine LLMs, including three GPT models, four LLaMA models, PaLM 2, and Claude, on six datasets collected from social media, online reviews, and recorded transcripts. Experiments show that both the model-generated and the human-written reference summaries suffer from low fairness. We conduct a comprehensive analysis of the common factors influencing fairness and propose three simple but effective methods to alleviate unfair summarization. Our dataset and code are available at https://github.com/psunlpgroup/FairSumm.

pdf bib
What Are We Measuring When We Evaluate Large Vision-Language Models? An Analysis of Latent Factors and Biases
Anthony Tiong | Junqi Zhao | Boyang Li | Junnan Li | Steven Hoi | Caiming Xiong

Vision-language (VL) models, pretrained on colossal image-text datasets, have attained broad VL competence that is difficult to evaluate. A common belief is that a small number of VL skills underlie the variety of VL tests. In this paper, we perform a large-scale transfer learning experiment aimed at discovering latent VL skills from data. We reveal interesting characteristics that have important implications for test suite design. First, generation tasks suffer from a length bias, suggesting benchmarks should balance tasks with varying output lengths. Second, we demonstrate that factor analysis successfully identifies reasonable yet surprising VL skill factors, suggesting benchmarks could leverage similar analyses for task selection.Finally, we present a new dataset, OLIVE1, which simulates user instructions in the wild and presents challenges dissimilar to all datasets we tested. Our findings contribute to the design of balanced and broad-coverage vision-language evaluation methods. 1https://github.com/jq-zh/olive-dataset

pdf bib
Show Your Work with Confidence: Confidence Bands for Tuning Curves
Nicholas Lourie | Kyunghyun Cho | He He

The choice of hyperparameters greatly impacts performance in natural language processing. Often, it is hard to tell if a method is better than another or just better tuned. *Tuning curves* fix this ambiguity by accounting for tuning effort. Specifically, they plot validation performance as a function of the number of hyperparameter choices tried so far. While several estimators exist for these curves, it is common to use point estimates, which we show fail silently and give contradictory results when given too little data.Beyond point estimates, *confidence bands* are necessary to rigorously establish the relationship between different approaches. We present the first method to construct valid confidence bands for tuning curves. The bands are exact, simultaneous, and distribution-free, thus they provide a robust basis for comparing methods.Empirical analysis shows that while bootstrap confidence bands, which serve as a baseline, fail to approximate their target confidence, ours achieve it exactly. We validate our design with ablations, analyze the effect of sample size, and provide guidance on comparing models with our method. To promote confident comparisons in future work, we release opda: an easy-to-use library that you can install with pip. https://github.com/nicholaslourie/opda

pdf bib
GRASP: A Disagreement Analysis Framework to Assess Group Associations in Perspectives
Vinodkumar Prabhakaran | Christopher Homan | Lora Aroyo | Aida Mostafazadeh Davani | Alicia Parrish | Alex Taylor | Mark Diaz | Ding Wang | Gregory Serapio-García

Human annotation plays a core role in machine learning — annotations for supervised models, safety guardrails for generative models, and human feedback for reinforcement learning, to cite a few avenues. However, the fact that many of these human annotations are inherently subjective is often overlooked. Recent work has demonstrated that ignoring rater subjectivity (typically resulting in rater disagreement) is problematic within specific tasks and for specific subgroups. Generalizable methods to harness rater disagreement and thus understand the socio-cultural leanings of subjective tasks remain elusive. In this paper, we propose GRASP, a comprehensive disagreement analysis framework to measure group association in perspectives among different rater subgroups, and demonstrate its utility in assessing the extent of systematic disagreements in two datasets: (1) safety annotations of human-chatbot conversations, and (2) offensiveness annotations of social media posts, both annotated by diverse rater pools across different socio-demographic axes. Our framework (based on disagreement metrics) reveals specific rater groups that have significantly different perspectives than others on certain tasks, and helps identify demographic axes that are crucial to consider in specific task contexts.

pdf bib
Event Causality Is Key to Computational Story Understanding
Yidan Sun | Qin Chao | Boyang Li

Cognitive science and symbolic AI research suggest that event causality provides vital information for story understanding. However, machine learning systems for story understanding rarely employ event causality, partially due to the lack of methods that reliably identify open-world causal event relations. Leveraging recent progress in large language models, we present the first method for event causality identification that leads to material improvements in computational story understanding. Our technique sets a new state of the art on the COPES dataset (Wang et al., 2023c) for causal event relation identification. Further, in the downstream story quality evaluation task, the identified causal relations lead to 3.6-16.6% relative improvement on correlation with human ratings. In the multimodal story video-text alignment task, we attain 4.1-10.9% increase on Clip Accuracy and 4.2-13.5% increase on Sentence IoU. The findings indicate substantial untapped potential for event causality in computational story understanding. The codebase is at https://github.com/insundaycathy/Event-Causality-Extraction.

pdf bib
Subspace Representations for Soft Set Operations and Sentence Similarities
Yoichi Ishibashi | Sho Yokoi | Katsuhito Sudoh | Satoshi Nakamura

In the field of natural language processing (NLP), continuous vector representations are crucial for capturing the semantic meanings of individual words. Yet, when it comes to the representations of sets of words, the conventional vector-based approaches often struggle with expressiveness and lack the essential set operations such as union, intersection, and complement. Inspired by quantum logic, we realize the representation of word sets and corresponding set operations within pre-trained word embedding spaces. By grounding our approach in the linear subspaces, we enable efficient computation of various set operations and facilitate the soft computation of membership functions within continuous spaces. Moreover, we allow for the computation of the F-score directly within word vectors, thereby establishing a direct link to the assessment of sentence similarity. In experiments with widely-used pre-trained embeddings and benchmarks, we show that our subspace-based set operations consistently outperform vector-based ones in both sentence similarity and set retrieval tasks.

pdf bib
My Heart Skipped a Beat! Recognizing Expressions of Embodied Emotion in Natural Language
Yuan Zhuang | Tianyu Jiang | Ellen Riloff

Humans frequently experience emotions. When emotions arise, they affect not only our mental state but can also change our physical state. For example, we often open our eyes wide when we are surprised, or clap our hands when we feel excited. Physical manifestations of emotions are referred to as embodied emotion in the psychology literature. From an NLP perspective, recognizing descriptions of physical movements or physiological responses associated with emotions is a type of implicit emotion recognition. Our work introduces a new task of recognizing expressions of embodied emotion in natural language. We create a dataset of sentences that contains 7,300 body part mentions with human annotations for embodied emotion. We develop a classification model for this task and present two methods to acquire weakly labeled instances of embodied emotion by extracting emotional manner expressions and by prompting a language model. Our experiments show that the weakly labeled data can train an effective classification model without gold data, and can also improve performance when combined with gold data. Our dataset is publicly available at https://github.com/yyzhuang1991/Embodied-Emotions.

pdf bib
Low-Cost Generation and Evaluation of Dictionary Example Sentences
Bill Cai | Ng Clarence | Daniel Liang | Shelvia Hotama

Dictionary example sentences play an important role in illustrating word definitions and usage, but manually creating quality sentences is challenging. Prior works have demonstrated that language models can be trained to generate example sentences. However, they relied on costly customized models and word sense datasets for generation and evaluation of their work. Rapid advancements in foundational models present the opportunity to create low-cost, zero-shot methods for the generation and evaluation of dictionary example sentences. We introduce a new automatic evaluation metric called OxfordEval that measures the win-rate of generated sentences against existing Oxford Dictionary sentences. OxfordEval shows high alignment with human judgments, enabling large-scale automated quality evaluation. We experiment with various LLMs and configurations to generate dictionary sentences across word classes. We complement this with a novel approach of using masked language models to identify and select sentences that best exemplify word meaning. The eventual model, FM-MLM, achieves over 85.1% win rate against Oxford baseline sentences according to OxfordEval, compared to 39.8% win rate for prior model-generated sentences.

pdf bib
Making Language Models Better Tool Learners with Execution Feedback
Shuofei Qiao | Honghao Gui | Chengfei Lv | Qianghuai Jia | Huajun Chen | Ningyu Zhang

Tools serve as pivotal interfaces that enable humans to understand and reshape the environment. With the advent of foundation models, AI systems can utilize tools to expand their capabilities and interact with the real world. Existing tool learning methodologies, encompassing supervised fine-tuning and prompt engineering approaches, often induce large language models to utilize tools indiscriminately, as complex tasks often exceed their own competencies. However, introducing tools for simple tasks, which the models themselves can readily resolve, can inadvertently propagate errors rather than enhance performance. This leads to the research question: can we teach language models when and how to use tools? To meet this need, we propose Tool leaRning wIth exeCution fEedback (TRICE), a two-stage end-to-end framework that enables the model to continually learn through feedback derived from tool execution, thereby learning when and how to use tools effectively. Experimental results, backed by further analysis, show that TRICE can make the large language model selectively use tools by improving the accuracy of tool usage while enhancing insufficient tool learning and mitigating excessive reliance on tools.

pdf bib
Complex Claim Verification with Evidence Retrieved in the Wild
Jifan Chen | Grace Kim | Aniruddh Sriram | Greg Durrett | Eunsol Choi

Retrieving evidence to support or refute claims is a core part of automatic fact-checking. Prior work makes simplifying assumptions in retrieval that depart from real-world use cases: either no access to evidence, access to evidence curated by a human fact-checker, or access to evidence published after a claim was made. In this work, we present the first realistic pipeline to check real-world claims by retrieving raw evidence from the web. We restrict our retriever to only search documents available prior to the claim’s making, modeling the realistic scenario of emerging claims. Our pipeline includes five components: claim decomposition, raw document retrieval, fine-grained evidence retrieval, claim-focused summarization, and veracity judgment. We conduct experiments on complex political claims in the ClaimDecomp dataset and show that the aggregated evidence produced by our pipeline improves veracity judgments. Human evaluation finds the evidence summary produced by our system is reliable (it does not hallucinate information) and relevant to answering key questions about a claim, suggesting that it can assist fact-checkers even when it does not reflect a complete evidence set.

pdf bib
Multimodal Multi-loss Fusion Network for Sentiment Analysis
Zehui Wu | Ziwei Gong | Jaywon Koo | Julia Hirschberg

This paper investigates the optimal selection and fusion of feature encoders across multiple modalities and combines these in one neural network to improve sentiment detection. We compare different fusion methods and examine the impact of multi-loss training within the multi-modality fusion network, identifying surprisingly important findings relating to subnet performance. We have also found that integrating context significantly enhances model performance. Our best model achieves state-of-the-art performance for three datasets (CMU-MOSI, CMU-MOSEI and CH-SIMS). These results suggest a roadmap toward an optimized feature selection and fusion approach for enhancing sentiment detection in neural networks.

pdf bib
Confronting LLMs with Traditional ML: Rethinking the Fairness of Large Language Models in Tabular Classifications
Yanchen Liu | Srishti Gautam | Jiaqi Ma | Himabindu Lakkaraju

Recent literature has suggested the potential of using large language models (LLMs) to make classifications for tabular tasks. However, LLMs have been shown to exhibit harmful social biases that reflect the stereotypes and inequalities present in society. To this end, as well as the widespread use of tabular data in many high-stake applications, it is important to explore the following questions: what sources of information do LLMs draw upon when making classifications for tabular tasks; whether and to what extent are LLM classifications for tabular data influenced by social biases and stereotypes; and what are the consequential implications for fairness?Through a series of experiments, we delve into these questions and show that LLMs tend to inherit social biases from their training data which significantly impact their fairness in tabular classification tasks. Furthermore, our investigations show that in the context of bias mitigation, though in-context learning and finetuning have a moderate effect, the fairness metric gap between different subgroups is still larger than that in traditional machine learning models, such as Random Forest and shallow Neural Networks. This observation emphasizes that the social biases are inherent within the LLMs themselves and inherited from their pretraining corpus, not only from the downstream task datasets. Besides, we demonstrate that label-flipping of in-context examples can significantly reduce biases, further highlighting the presence of inherent bias within LLMs.

pdf bib
Analyzing the Use of Metaphors in News Editorials for Political Framing
Meghdut Sengupta | Roxanne El Baff | Milad Alshomary | Henning Wachsmuth

Metaphorical language is a pivotal element inthe realm of political framing. Existing workfrom linguistics and the social sciences providescompelling evidence regarding the distinctivenessof conceptual framing for politicalideology perspectives. However, the nature andutilization of metaphors and the effect on audiencesof different political ideologies withinpolitical discourses are hardly explored. Toenable research in this direction, in this workwe create a dataset, originally based on newseditorials and labeled with their persuasive effectson liberals and conservatives and extend itwith annotations pertaining to metaphorical usageof language. To that end, first, we identifyall single metaphors and composite metaphors.Secondly, we provide annotations of the sourceand target domains for each metaphor. As aresult, our corpus consists of 300 news editorialsannotated with spans of texts containingmetaphors and the corresponding domains ofwhich these metaphors draw from. Our analysisshows that liberal readers are affected bymetaphors, whereas conservatives are resistantto them. Both ideologies are affected differentlybased on the metaphor source and targetcategory. For example, liberals are affected bymetaphors in the Darkness & Light (e.g., death)source domains, where as the source domain ofNature affects conservatives more significantly.

pdf bib
SharpSeq: Empowering Continual Event Detection through Sharpness-Aware Sequential-task Learning
Thanh-Thien Le | Viet Dao | Linh Nguyen | Thi-Nhung Nguyen | Linh Ngo | Thien Nguyen

Continual event detection is a cornerstone in uncovering valuable patterns in many dynamic practical applications, where novel events emerge daily. Existing state-of-the-art approaches with replay buffers still suffer from catastrophic forgetting, partially due to overly simplistic objective aggregation. This oversight disregards complex trade-offs and leads to sub-optimal gradient updates, resulting in performance deterioration across objectives. While there are successful, widely cited multi-objective optimization frameworks for multi-task learning, they lack mechanisms to address data imbalance and evaluate whether a Pareto-optimal solution can effectively mitigate catastrophic forgetting, rendering them unsuitable for direct application to continual learning. To address these challenges, we propose **SharpSeq**, a novel continual learning paradigm leveraging sharpness-aware minimization combined with a generative model to balance training data distribution. Through extensive experiments on multiple real-world datasets, we demonstrate the superior performance of SharpSeq in continual event detection, proving the importance of our approach in mitigating catastrophic forgetting in continual event detection.

pdf bib
Dissecting Paraphrases: The Impact of Prompt Syntax and supplementary Information on Knowledge Retrieval from Pretrained Language Models
Stephan Linzbach | Dimitar Dimitrov | Laura Kallmeyer | Kilian Evang | Hajira Jabeen | Stefan Dietze

Pre-trained Language Models (PLMs) are known to contain various kinds of knowledge.One method to infer relational knowledge is through the use of cloze-style prompts, where a model is tasked to predict missing subjects orobjects. Typically, designing these prompts is a tedious task because small differences in syntax or semantics can have a substantial impact on knowledge retrieval performance. Simultaneously, evaluating the impact of either prompt syntax or information is challenging due to their interdependence. We designed CONPARE-LAMA – a dedicated probe, consisting of 34 million distinct prompts that facilitate comparison across minimal paraphrases. These paraphrases follow a unified meta-template enabling the controlled variation of syntax and semantics across arbitrary relations.CONPARE-LAMA enables insights into the independent impact of either syntactical form or semantic information of paraphrases on the knowledge retrieval performance of PLMs. Extensive knowledge retrieval experiments using our probe reveal that prompts following clausal syntax have several desirable properties in comparison to appositive syntax: i) they are more useful when querying PLMs with a combination of supplementary information, ii) knowledge is more consistently recalled across different combinations of supplementary information, and iii) they decrease response uncertainty when retrieving known facts. In addition, range information can boost knowledge retrieval performance more than domain information, even though domain information is more reliably helpful across syntactic forms.

pdf bib
Know When To Stop: A Study of Semantic Drift in Text Generation
Ava Spataru

In this work, we explicitly show that modern LLMs tend to generate correct facts first, then “drift away” and generate incorrect facts later: this was occasionally observed but never properly measured. We develop a semantic drift score that measures the degree of separation between correct and incorrect facts in generated texts and confirm our hypothesis when generating Wikipedia-style biographies. This correct-then-incorrect generation pattern suggests that factual accuracy can be improved by knowing when to stop generation. Therefore, we explore the trade-off between information quantity and factual accuracy for several early stopping methods and manage to improve factuality by a large margin. We further show that reranking with semantic similarity can further improve these results, both compared to the baseline and when combined with early stopping. Finally, we try calling external API to bring the model back to the right generation path, but do not get positive results. Overall, our methods generalize and can be applied to any long-form text generation to produce more reliable information, by balancing trade-offs between factual accuracy, information quantity and computational cost.

pdf bib
Curriculum Masking in Vision-Language Pretraining to Maximize Cross Modal Interaction
Kraig Tou | Zijun Sun

Many leading methods in Vision and language (V+L) pretraining utilize masked language modeling (MLM) as a standard pretraining component, with the expectation that reconstruction of masked text tokens would necessitate reference to corresponding image context via cross/self attention and thus promote representation fusion. However, we observe that the minimization of MLM loss in earlier training stages can depend disproportionately on local text signals, leading to poor training efficiency and inconsistency with the goal of representation fusion. The extent of this lack of cross modal interaction depends strongly which token(s) are masked. To address this issue, we propose a curriculum masking scheme as a replacement for random masking. Tokens are selected to be masked at a frequency proportional to the expected level of cross modal interaction necessary to reconstruct them. This is achieved using a parallel mask selection agent that measures the cross modal flow of information and treats it as a reward to be maximized. By additionally masking contiguous spans that include key objects and their relations, we also achieve better relational understanding, which has been shown to be lacking in many SOTA models. Our experiments on a wide range of V+L tasks show that we trail closely behind state-of-the-art methods despite pretraining on 300x to 1000x less data and we also achieve either top or runner-up performance on tasks from the ARO benchmark which tests compositional relationships. Finally, we demonstrate the potential of our method to scale to larger pretraining data.

pdf bib
Elote, Choclo and Mazorca: on the Varieties of Spanish
Cristina España-Bonet | Alberto Barrón-Cedeño

Spanish is one of the most widespread languages: the official language in 20 countries and the second most-spoken native language. Its contact with other languages across different regions and the rich regional and cultural diversity has produced varieties which divert from each other, particularly in terms of lexicon. Still, available corpora, and models trained upon them, generally treat Spanish as one monolithic language, which dampers prediction and generation power when dealing with different varieties. To alleviate the situation, we compile and curate datasets in the different varieties of Spanish around the world at an unprecedented scale and create the CEREAL corpus. With such a resource at hand, we perform a stylistic analysis to identify and characterise varietal differences. We implement a classifier specially designed to deal with long documents and identify Spanish varieties (and therefore expand CEREAL further). We produce varietal-specific embeddings, and analyse the cultural differences that they encode. We make data, code and models publicly available.

pdf bib
Ada-LEval: Evaluating long-context LLMs with length-adaptable benchmarks
Chonghua Wang | Haodong Duan | Songyang Zhang | Dahua Lin | Kai Chen

Recently, the large language model (LLM) community has shown increasing interest in enhancing LLMs’ capability to handle extremely long documents. As various long-text techniques and model architectures emerge, the precise and detailed evaluation of models’ long-text capabilities has become increasingly important. Existing long-text evaluation benchmarks, such as L-Eval and LongBench, construct long-text test sets based on open-source datasets, focusing mainly on QA and summarization tasks. These datasets include test samples of varying lengths (from 2k to 32k+) entangled together, making it challenging to assess model capabilities across different length ranges. Moreover, they do not cover the ultralong settings (100k+ tokens) that the latest LLMs claim to achieve. In this paper, we introduce Ada-LEval, a length-adaptable benchmark for evaluating the long-context understanding of LLMs. Ada-LEval includes two challenging subsets, TSort and BestAnswer, which enable a more reliable evaluation of LLMs’ long context capabilities. These benchmarks support intricate manipulation of the length of test cases, and can easily produce text samples up to 128k tokens. We evaluate 4 state-of-the-art closed-source API models and 6 open-source models with Ada-LEval. The evaluation results demonstrate the limitations of current LLMs, especially in ultra-long-context settings. Our code is available at https://github.com/open-compass/Ada-LEval.

pdf bib
A Zero-Shot Monolingual Dual Stage Information Retrieval System for Spanish Biomedical Systematic Literature Reviews
Regina Ofori-Boateng | Magaly Aceves-Martins | Nirmalie Wiratunga | Carlos Moreno-Garcia

Systematic Reviews (SRs) are foundational in healthcare for synthesising evidence to inform clinical practices. Traditionally skewed towards English-language databases, SRs often exclude significant research in other languages, leading to potential biases. This study addresses this gap by focusing on Spanish, a language notably underrepresented in SRs. We present a foundational zero-shot dual information retrieval (IR) baseline system, integrating traditional retrieval methods with pre-trained language models and cross-attention re-rankers for enhanced accuracy in Spanish biomedical literature retrieval. Utilising the LILACS database, known for its comprehensive coverage of Latin American and Caribbean biomedical literature, we evaluate the approach with three real-life case studies in Spanish SRs. The findings demonstrate the system’s efficacy and underscore the importance of query formulation. This study contributes to the field of IR by promoting language inclusivity and supports the development of more comprehensive and globally representative healthcare guidelines.

pdf bib
LayoutPointer: A Spatial-Context Adaptive Pointer Network for Visual Information Extraction
Huang Siyuan | Yongping Xiong | Wu Guibin

Visual Information Extraction (VIE), as a crucial task of Document Intelligence, involves two primary sub-tasks: Semantic Entity Recognition (SER) and Relation Extraction (RE). However, VIE faces two significant challenges. Firstly, most existing models inadequately utilize spatial information of entities, often failing to predict connections or incorrectly linking spatially distant entities. Secondly, the improper input order of tokens challenges in extracting complete entity pairs from documents with multi-line entities when text is extracted via PDF parser or OCR. To address these challenges, we propose LayoutPointer, a Spatial-Context Adaptive Pointer Network. LayoutPointer explicitly enhances spatial-context relationships by incorporating 2D relative position information and adaptive spatial constraints within self-attention. Furthermore, we recast the RE task as a specialized cycle detection problem, employing a unique tail-to-head pointer to restore the semantic order among multi-line entities. To better evaluate the effectiveness of our proposed method, we reconstruct a multi-line dataset named MLFUD, which more accurately reflects real-world scenarios. Fine-tuning experimental results on FUNSD, XFUND, and MLFUD datasets demonstrate that LayoutPointer significantly outperforms existing state-of-the-art methods in F1 scores for RE tasks (e.g., 5.71% improvement on XFUND using LayoutPointerBASE-X over LayoutLMv3).

pdf bib
Long-form evaluation of model editing
Domenic Rosati | Robie Gonzales | Jinkun Chen | Xuemin Yu | Yahya Kayani | Frank Rudzicz | Hassan Sajjad

Evaluations of model editing, a technique for changing the factual knowledge held by Large Language Models (LLMs), currently only use the ‘next few token’ completions after a prompt. As a result, the impact of these methods on longer natural language generation is largely unknown. We introduce long-form evaluation of model editing (LEME) a novel evaluation protocol that measures the efficacy and impact of model editing in long-form generative settings. Our protocol consists of a machine-rated survey and a classifier which correlates well with human ratings. Importantly, we find that our protocol has very little relationship with previous short-form metrics (despite being designed to extend efficacy, generalization, locality, and portability into a long-form setting), indicating that our method introduces a novel set of dimensions for understanding model editing methods. Using this protocol, we benchmark a number of model editing techniques and present several findings including that, while some methods (ROME and MEMIT) perform well in making consistent edits within a limited scope, they suffer much more from factual drift than other methods. Finally, we present a qualitative analysis that illustrates common failure modes in long-form generative settings including internal consistency, lexical cohesion, and locality issues.

pdf bib
Analyzing the Role of Semantic Representations in the Era of Large Language Models
Zhijing Jin | Yuen Chen | Fernando Gonzalez Adauto | Jiarui Liu | Jiayi Zhang | Julian Michael | Bernhard Schölkopf | Mona Diab

Traditionally, natural language processing (NLP) models often use a rich set of features created by linguistic expertise, such as semantic representations. However, in the era of large language models (LLMs), more and more tasks are turned into generic, end-to-end sequence generation problems. In this paper, we investigate the question: what is the role of semantic representations in the era of LLMs? Specifically, we investigate the effect of Abstract Meaning Representation (AMR) across five diverse NLP tasks. We propose an AMR-driven chain-of-thought prompting method, which we call AMRCOT, and find that it generally hurts performance more than it helps. To investigate what AMR may have to offer on these tasks, we conduct a series of analysis experiments. We find that it is difficult to predict which input examples AMR may help or hurt on, but errors tend to arise with multi-word expressions, named entities, and in the final inference step where the LLM must connect its reasoning over the AMR to its prediction. We recommend focusing on these areas for future work in semantic representations for LLMs. Our code: https://github.com/causalNLP/amr_llm

pdf bib
TRAQ: Trustworthy Retrieval Augmented Question Answering via Conformal Prediction
Shuo Li | Sangdon Park | Insup Lee | Osbert Bastani

When applied to open-domain question answering, large language models (LLMs) frequently generate incorrect responses based on made-up facts, which are called hallucinations. Retrieval augmented generation (RAG) is a promising strategy to avoid hallucinations, but it does not provide guarantees on its correctness. To address this challenge, we propose the Trustworthy Retrieval Augmented Question Answering, or *TRAQ*, which provides the first end-to-end statistical correctness guarantee for RAG. TRAQ uses conformal prediction, a statistical technique for constructing prediction sets that are guaranteed to contain the semantically correct response with high probability. Additionally, TRAQ leverages Bayesian optimization to minimize the size of the constructed sets. In an extensive experimental evaluation, we demonstrate that TRAQ provides the desired correctness guarantee while reducing prediction set size by 16.2% on average compared to an ablation. The implementation is available: [https://github.com/shuoli90/TRAQ](https://github.com/shuoli90/TRAQ).

pdf bib
MapGuide: A Simple yet Effective Method to Reconstruct Continuous Language from Brain Activities
Xinpei Zhao | Jingyuan Sun | Shaonan Wang | Jing Ye | Xhz Xhz | Chengqing Zong

Decoding continuous language from brain activity is a formidable yet promising field of research. It is particularly significant for aiding people with speech disabilities to communicate through brain signals. This field addresses the complex task of mapping brain signals to text. The previous best attempt reverse-engineered this process in an indirect way: it began by learning to encode brain activity from text and then guided text generation by aligning with predicted brain responses. In contrast, we propose a simple yet effective method that guides text reconstruction by directly comparing them with the predicted text embeddings mapped from brain activities. Comprehensive experiments reveal that our method significantly outperforms the current state-of-the-art model, showing average improvements of 77% and 54% on BLEU and METEOR scores. We further validate the proposed modules through detailed ablation studies and case analyses and highlight a critical correlation: the more precisely we map brain activities to text embeddings, the better the text reconstruction results. Such insight can simplify the task of reconstructing language from brain activities for future work, emphasizing the importance of improving brain-to-text-embedding mapping techniques.

pdf bib
On-the-fly Definition Augmentation of LLMs for Biomedical NER
Monica Munnangi | Sergey Feldman | Byron Wallace | Silvio Amir | Tom Hope | Aakanksha Naik

Despite their general capabilities, LLMs still struggle on biomedicalNER tasks, which are difficult due to the presence of specialized terminology and lack of training data. In this work we set out to improve LLM performance on biomedical NER in limited data settings via a new knowledge augmentation approach which incorporates definitions of relevant concepts on-the-fly. During this process, to provide a test bed for knowledge augmentation, we perform a comprehensive exploration of prompting strategies. Our experiments show that definition augmentation is useful for both open source and closed LLMs.For example, it leads to a relative improvement of 15% (on average) in GPT-4 performance (F1) across all (six) of our test datasets. We conduct extensive ablations and analyses to demonstrate that our performance improvements stem from adding relevant definitional knowledge. We find that careful prompting strategies also improve LLM performance, allowing them to outperform fine-tuned language models in few-shot settings. To facilitate future research in this direction, we release our code at https://github.com/allenai/beacon.

pdf bib
This Land is Your, My Land: Evaluating Geopolitical Bias in Language Models through Territorial Disputes
Bryan Li | Samar Haider | Chris Callison-Burch

Do the Spratly Islands belong to China, the Philippines, or Vietnam? A pretrained large language model (LLM) may answer differently if asked in the languages of each claimant country: Chinese, Tagalog, or Vietnamese. This contrasts with a multilingual human, who would likely answer consistently. In this paper, we show that LLMs recall certain geographical knowledge inconsistently when queried in different languages—a phenomenon we term geopolitical bias. As a targeted case study, we consider territorial disputes, an inherently controversial and multilingual task. We introduce BorderLines, a dataset of territorial disputes which covers 251 territories, each associated with a set of multiple-choice questions in the languages of each claimant country (49 languages in total). We also propose a suite of evaluation metrics to precisely quantify bias and consistency in responses across different languages. We then evaluate various multilingual LLMs on our dataset and metrics to probe their internal knowledge and use the proposed metrics to discover numerous inconsistencies in how these models respond in different languages. Finally, we explore several prompt modification strategies, aiming to either amplify or mitigate geopolitical bias, which highlights how brittle LLMs are and how they tailor their responses depending on cues from the interaction context. Our code and data are available at https://github.com/manestay/borderlines.

pdf bib
Set-Aligning Framework for Auto-Regressive Event Temporal Graph Generation
Xingwei Tan | Yuxiang Zhou | Gabriele Pergola | Yulan He

Event temporal graphs have been shown as convenient and effective representations of complex temporal relations between events in text. Recent studies, which employ pre-trained language models to auto-regressively generate linearised graphs for constructing event temporal graphs, have shown promising results. However, these methods have often led to suboptimal graph generation as the linearised graphs exhibit set characteristics which are instead treated sequentially by language models. This discrepancy stems from the conventional text generation objectives, leading to erroneous penalisation of correct predictions caused by the misalignment of elements in target sequences. To address these challenges, we reframe the task as a conditional set generation problem, proposing a Set-aligning Framework tailored for the effective utilisation of Large Language Models (LLMs). The framework incorporates data augmentations and set-property regularisations designed to alleviate text generation loss penalties associated with the linearised graph edge sequences, thus encouraging the generation of more relation edges. Experimental results show that our framework surpasses existing baselines for event temporal graph generation. Furthermore, under zero-shot settings, the structural knowledge introduced through our framework notably improves model generalisation, particularly when the training examples available are limited.

pdf bib
LanguageFlow: Advancing Diffusion Language Generation with Probabilistic Flows
Shujian Zhang | Lemeng Wu | Chengyue Gong | Xingchao Liu

Recent works have demonstrated success in controlling sentence attributes (e.g., sentiment) and structure (e.g., syntactic structure) based on the diffusion language model. A key component that drives theimpressive performance for generating high-quality samples from noise is iteratively denoise for thousands of steps. While beneficial, the complexity of starting from the noise and the learning steps has limited its implementation to many NLP real-world applications. This paper proposes Language Rectified Flow (LF).Our method is based on the reformulation of the standard probabilistic flow models.Language rectified flow learns (neural) ordinary differentialequation models to transport between the source distribution and the target distribution, henceproviding a unified and effective solution to generative modeling and domain transfer.From the source distribution, our language rectified flow yields fast simulation and effectively decreases the inference time. Experiments on three challenging fine-grained control tasks and multiple high-quality text editing show that our method consistently outperforms its baselines. Extensive experiments and ablation studies demonstrate that our method can be general, effective, and beneficial for many NLP tasks.

pdf bib
Towards Improved Multi-Source Attribution for Long-Form Answer Generation
Nilay Patel | Shivashankar Subramanian | Siddhant Garg | Pratyay Banerjee | Amita Misra

Teaching large language models (LLMs) to generate text with attribution to evidence sources can reduce hallucinations, improve verifiability in question answering systems (QA), and increase reliability of retrieval augmented LLMs. Despite gaining increasing popularity for usage in QA systems and search engines, current LLMs struggle with attribution for long-form responses which require reasoning over multiple evidence sources. To address this, in this paper we aim to improve the attribution capability of LLMs for long-form answer generation to multiple sources, with multiple citations per sentence. However, data for training multi-source attributable QA systems is difficult and expensive to annotate, and therefore scarce. To overcome this challenge, we transform existing QA datasets for this task (MultiAttr), and empirically demonstrate, on a wide range of attribution benchmark datasets, that fine-tuning on MultiAttr provides significant improvements over training only on the target QA domain. Lastly, to fill a gap in existing benchmarks, we present a multi-source attribution dataset containing multi-paragraph answers, PolitiICite, based on PolitiFact articles that discuss events closely related to implementation statuses of election promises.

pdf bib
Synthetic Query Generation for Privacy-Preserving Deep Retrieval Systems using Differentially Private Language Models
Aldo Carranza | Rezsa Farahani | Natalia Ponomareva | Alexey Kurakin | Matthew Jagielski | Milad Nasr

We address the challenge of ensuring differential privacy (DP) guarantees in training deep retrieval systems. Training these systems often involves the use of contrastive-style losses, which are typically non-per-example decomposable, making them difficult to directly DP-train with since common techniques require per-example gradients. To address this issue, we propose an approach that prioritizes ensuring query privacy prior to training a deep retrieval system. Our method employs DP language models (LMs) to generate private synthetic queries representative of the original data. These synthetic queries can be used in downstream retrieval system training without compromising privacy. Our approach demonstrates a significant enhancement in retrieval quality compared to direct DP-training, all while maintaining query-level privacy guarantees. This work highlights the potential of harnessing LMs to overcome limitations in standard DP-training methods.

pdf bib
Okay, Let’s Do This! Modeling Event Coreference with Generated Rationales and Knowledge Distillation
Abhijnan Nath | Shadi Manafi Avari | Avyakta Chelle | Nikhil Krishnaswamy

In NLP, Event Coreference Resolution (ECR) is the task of connecting event clusters that refer to the same underlying real-life event, usually via neural systems. In this work, we investigate using abductive free-text rationales (FTRs) generated by modern autoregressive LLMs as distant supervision of smaller student models for cross-document coreference (CDCR) of events. We implement novel rationale-oriented event clustering and knowledge distillation methods for event coreference scoring that leverage enriched information from the FTRs for improved CDCR without additional annotation or expensive document clustering. Our model using coreference-specific knowledge distillation achieves SOTA B3 F1 on the ECB+ and GVC corpora and we establish a new baseline on the AIDA Phase 1 corpus. Our code can be found at https://github.com/csu-signal/llama_cdcr.

pdf bib
Can Knowledge Graphs Reduce Hallucinations in LLMs? : A Survey
Garima Agrawal | Tharindu Kumarage | Zeyad Alghamdi | Huan Liu

The contemporary LLMs are prone to producing hallucinations, stemming mainly from the knowledge gaps within the models. To address this critical limitation, researchers employ diverse strategies to augment the LLMs by incorporating external knowledge, aiming to reduce hallucinations and enhance reasoning accuracy. Among these strategies, leveraging knowledge graphs as a source of external information has demonstrated promising results. In this survey, we comprehensively review these knowledge-graph-based augmentation techniques in LLMs, focusing on their efficacy in mitigating hallucinations. We systematically categorize these methods into three overarching groups, offering methodological comparisons and performance evaluations. Lastly, this survey explores the current trends and challenges associated with these techniques and outlines potential avenues for future research in this emerging field.

pdf bib
Pedagogically Aligned Objectives Create Reliable Automatic Cloze Tests
Brian Ondov | Kush Attal | Dina Demner-Fushman

The cloze training objective of Masked Language Models makes them a natural choice for generating plausible distractors for human cloze questions. However, distractors must also be both distinct and incorrect, neither of which is directly addressed by existing neural methods. Evaluation of recent models has also relied largely on automated metrics, which cannot demonstrate the reliability or validity of human comprehension tests. In this work, we first formulate the pedagogically motivated objectives of plausibility, incorrectness, and distinctiveness in terms of conditional distributions from language models. Second, we present an unsupervised, interpretable method that uses these objectives to jointly optimize sets of distractors. Third, we test the reliability and validity of the resulting cloze tests compared to other methods with human participants. We find our method has stronger correlation with teacher-created comprehension tests than the state-of-the-art neural method and is more internally consistent. Our implementation is freely available and can quickly create a multiple choice cloze test from any given passage.

pdf bib
Take One Step at a Time to Know Incremental Utility of Demonstration: An Analysis on Reranking for Few-Shot In-Context Learning
Kazuma Hashimoto | Karthik Raman | Michael Bendersky

In-Context Learning (ICL) is an emergent capability of Large Language Models (LLMs). Only a few demonstrations enable LLMs to be used as blackbox for new tasks. Previous studies have shown that using LLMs’ outputs as labels is effective in training models to select demonstrations. Such a label is expected to estimate utility of a demonstration in ICL; however, it has not been well understood how different labeling strategies affect results on target tasks. This paper presents an analysis on different utility functions by focusing on LLMs’ output probability given ground-truth output, and task-specific reward given LLMs’ prediction. Unlike the previous work, we introduce a novel labeling method, incremental utility, which estimates how much incremental knowledge is brought into the LLMs by a demonstration. We conduct experiments with instruction-tuned LLMs on binary/multi-class classification, segmentation, and translation across Arabic, English, Finnish, Japanese, and Spanish. Our results show that (1) the probability is effective when the probability values are distributed across the whole value range (on the classification tasks), and (2) the downstream metric is more robust when nuanced reward values are provided with long outputs (on the segmentation and translation tasks). We then show that the proposed incremental utility further helps ICL by contrasting how the LLMs perform with and without the demonstrations.

pdf bib
LM-Infinite: Zero-Shot Extreme Length Generalization for Large Language Models
Chi Han | Qifan Wang | Hao Peng | Wenhan Xiong | Yu Chen | Heng Ji | Sinong Wang

Today’s large language models (LLMs) typically train on short text segments (e.g., <4K tokens) due to the quadratic complexity of their Transformer architectures. As a result, their performance suffers drastically on inputs longer than those encountered during training, substantially limiting their applications in real-world tasks involving long contexts such as encod- ing scientific articles, code repositories, or long dialogues. Through both theoretical analysis and empirical investigation, this work identifies three major factors contributing to this length generalization failure. Our theoretical analysis reveals that commonly used techniques like using a sliding-window attention pattern or relative positional encodings are inadequate to address them. Answering these challenges, we propose LM-Infinite, a simple and effective method for enhancing LLMs’ capabilities of handling long contexts. LM-Infinite is highly flexible and can be used with most modern LLMs off-the-shelf. Without any parameter updates, it allows LLMs pre-trained with 2K or 4K-long segments to generalize to up to 200M length inputs while retaining perplexity. It also improves performance on downstream tasks such as Passkey Retrieval and Qasper in the zero-shot setting. LM-Infinite brings substantial efficiency improvements: it achieves 2.7× decoding speed up and 7.5× memory saving over the original model. Our code will be publicly available upon publication.

pdf bib
CONSCENDI: A Contrastive and Scenario-Guided Distillation Approach to Guardrail Models for Virtual Assistants
Albert Sun | Varun Nair | Elliot Schumacher | Anitha Kannan

A wave of new task-based virtual assistants has been fueled by increasingly powerful large language models (LLMs), such as GPT-4 (OpenAI, 2023). A major challenge in deploying LLM-based virtual conversational assistants in real world settings is ensuring they operate within what is admissible for the task. To overcome this challenge, the designers of these virtual assistants rely on an independent guardrail system that verifies the virtual assistant’s output aligns with the constraints required for the task. However, relying on commonly used, prompt-based guardrails can be difficult to engineer correctly and comprehensively. To address these challenges, we propose CONSCENDI. We use CONSCENDI to exhaustively generate training data with two key LLM-powered components: scenario-augmented generation and contrastive training examples. When generating conversational data, we generate a set of rule-breaking scenarios, which enumerate a diverse set of high-level ways a rule can be violated. This scenario-guided approach produces a diverse training set and provides chatbot designers greater control. To generate contrastive examples, we prompt the LLM to alter conversations with violations into acceptable conversations to enable fine-grained distinctions. We then use this data, generated by CONSCENDI, to train a smaller model. We find that CONSCENDI results in guardrail models that improve over baselines in multiple dialogue domains.

pdf bib
Advancing Beyond Identification: Multi-bit Watermark for Large Language Models
KiYoon Yoo | Wonhyuk Ahn | Nojun Kwak

We show the viability of tackling misuses of large language models beyond the identification of machine-generated text. While existing zero-bit watermark methods focus on detection only, some malicious misuses demand tracing the adversary user for counteracting them. To address this, we propose Multi-bit Watermark via Position Allocation, embedding traceable multi-bit information during language model generation. Through allocating tokens onto different parts of the messages, we embed longer messages in high corruption settings without added latency. By independently embedding sub-units of messages, the proposed method outperforms the existing works in terms of robustness and latency. Leveraging the benefits of zero-bit watermarking, our method enables robust extraction of the watermark without any model access, embedding and extraction of long messages (≥ 32-bit) without finetuning, and maintaining text quality, while allowing zero-bit detection all at the same time.

pdf bib
HTCCN: Temporal Causal Convolutional Networks with Hawkes Process for Extrapolation Reasoning in Temporal Knowledge Graphs
Tingxuan Chen | Jun Long | Liu Yang | Zidong Wang | Yongheng Wang | Xiongnan Jin

Temporal knowledge graphs (TKGs) serve as powerful tools for storing and modeling dynamic facts, holding immense potential in anticipating future facts. Since future facts are inherently unknowable, effectively modeling the intricate temporal structure of historical facts becomes paramount for accurate prediction. However, current models often rely heavily on fact recurrence or periodicity, leading to information loss due to prolonged evolutionary processes. Notably, the occurrence of one fact always influences the likelihood of another. To this end, we propose HTCCN, a novel Hawkes process-based temporal causal convolutional network designed for temporal reasoning under extrapolation settings. HTCCN employs a temporal causal convolutional network to model the historical interdependence of facts and leverages Hawkes to model link formation processes inductively in TKGs. Importantly, HTCCN introduces dual-level dynamics to comprehensively capture the temporal evolution of facts. Rigorous experimentation on four real-world datasets underscores the superior performance of HTCCN.

pdf bib
SemStamp: A Semantic Watermark with Paraphrastic Robustness for Text Generation
Abe Hou | Jingyu Zhang | Tianxing He | Yichen Wang | Yung-Sung Chuang | Hongwei Wang | Lingfeng Shen | Benjamin Van Durme | Daniel Khashabi | Yulia Tsvetkov

Existing watermarked generation algorithms employ token-level designs and therefore, are vulnerable to paraphrase attacks. To address this issue, we introduce watermarking on the semantic representation of sentences. We propose SemStamp, a robust sentence-level semantic watermarking algorithm that uses locality-sensitive hashing (LSH) to partition the semantic space of sentences. The algorithm encodes and LSH-hashes a candidate sentence generated by a language model, and conducts rejection sampling until the sampled sentence falls in watermarked partitions in the semantic embedding space. To test the paraphrastic robustness of watermarking algorithms, we propose a “bigram paraphrase” attack that produces paraphrases with small bigram overlap with the original sentence. This attack is shown to be effective against existing token-level watermark algorithms, while posing only minor degradations to SemStamp. Experimental results show that our novel semantic watermark algorithm is not only more robust than the previous state-of-the-art method on various paraphrasers and domains, but also better at preserving the quality of generation.

pdf bib
Media Bias Detection Across Families of Language Models
Iffat Maab | Edison Marrese-Taylor | Sebastian Padó | Yutaka Matsuo

Bias in reporting can influence the public’s opinion on relevant societal issues. Examples include informational bias (selective presentation of content) and lexical bias (specific framing of content through linguistic choices). The recognition of media bias is arguably an area where NLP can contribute to the “social good”. Traditional NLP models have shown good performance in classifying media bias, but require careful model design and extensive tuning. In this paper, we ask how well prompting of large language models can recognize media bias. Through an extensive empirical study including a wide selection of pre-trained models, we find that prompt-based techniques can deliver comparable performance to traditional models with greatly reduced effort and that, similar to traditional models, the availability of context substantially improves results. We further show that larger models can leverage different kinds of context simultaneously, obtaining further performance improvements.

pdf bib
Better Zero-Shot Reasoning with Role-Play Prompting
Aobo Kong | Shiwan Zhao | Hao Chen | Qicheng Li | Yong Qin | Ruiqi Sun | Xin Zhou | Enzhi Wang | Xiaohang Dong

Modern large language models (LLMs) exhibit a remarkable capacity for role-playing, enabling them to embody not only human characters but also non-human entities. This versatility allows them to simulate complex human-like interactions and behaviors within various contexts, as well as to emulate specific objects or systems. While these capabilities have enhanced user engagement and introduced novel modes of interaction, the influence of role-playing on LLMs’ reasoning abilities remains underexplored. In this study, we introduce a strategically designed role-play prompting methodology and assess its performance under the zero-shot setting across twelve diverse reasoning benchmarks. Our empirical results illustrate that role-play prompting consistently surpasses the standard zero-shot approach across most datasets. Notably, in experiments conducted using ChatGPT, accuracy on AQuA rises from 53.5% to 63.8%, and on Last Letter from 23.8% to 84.2%. Upon further comparison with the Zero-Shot-CoT technique, which prompts the model to “think step by step”, our study demonstrates that role-play prompting acts as a more effective trigger for the CoT process.This highlights its potential to augment the reasoning capabilities of LLMs. We release our code at https://github.com/NKU-HLT/Role-Play-Prompting.

pdf bib
Event-Content-Oriented Dialogue Generation in Short Video
Fenghua Cheng | Xue Li | Zi Huang | Jinxiang Wang | Sen Wang

Understanding complex events from different modalities, associating to external knowledge and generating response in a clear point of view are still unexplored in today’s multi-modal dialogue research. The great challenges include 1) lack of event-based multi-modal dialogue dataset; 2) understanding of complex events and 3) heterogeneity gap between different modalities. To overcome these challenges, we firstly introduce a novel event-oriented video-dialogue dataset called SportsVD (Sports-domain Video-dialogue Dataset). To our best knowledge, SportsVD is the first dataset that consists of complex events videos and opinion-based conversations with regards to contents in these events. Meanwhile, we present multi-modal dialogue generation method VCD (Video Commentary Dialogue) to generate human-like response according to event contents in the video and related external knowledge. In contrast to previous video-based dialogue generation, we focus on opinion-based response and the understanding of longer and more complex event contents. We evaluate VCD’s performance on SportsVD and other baselines under several automatic metrics. Experiments demonstrate VCD can outperform among other state-of-the-art baselines. Our work is available at https://github.com/Cheng-Fenghua/SportsVD.

pdf bib
DoG-Instruct: Towards Premium Instruction-Tuning Data via Text-Grounded Instruction Wrapping
Yongrui Chen | Haiyun Jiang | Xinting Huang | Shuming Shi | Guilin Qi

The improvement of LLMs’ instruction-following capabilities relies heavily on the availability of high-quality instruction-response pairs. Unfortunately, the current methods used to collect the pairs suffer from either unaffordable labor costs or severe hallucinations in the self-generation of LLM.To tackle these challenges, this paper proposes a scalable solution.It involves training LLMs to generate instruction-response pairs based on human-written documents, rather than relying solely on self-generation without context.Our proposed method not only exploits the advantages of human-written documents in reducing hallucinations but also utilizes an LLM to wrap the expression of documents, which enables us to bridge the gap between various document styles and the standard AI response.Experiments demonstrate that our method outperforms existing typical methods on multiple benchmarks.In particular, compared to the best-performing baseline, the LLM trained using our generated dataset exhibits a 10% relative improvement in performance on AlpacaEval, despite utilizing only 1/5 of its training data.Furthermore, a comprehensive manual evaluation validates the quality of the data we generated.

pdf bib
Beyond Borders: Investigating Cross-Jurisdiction Transfer in Legal Case Summarization
Santosh T.y.s.s | Vatsal Venkatkrishna | Saptarshi Ghosh | Matthias Grabmair

Legal professionals face the challenge of managing an overwhelming volume of lengthy judgments, making automated legal case summarization crucial. However, prior approaches mainly focused on training and evaluating these models within the same jurisdiction. In this study, we explore the cross-jurisdictional generalizability of legal case summarization models. Specifically, we explore how to effectively summarize legal cases of a target jurisdiction where reference summaries are not available. In particular, we investigate whether supplementing models with unlabeled target jurisdiction corpus and extractive silver summaries obtained from unsupervised algorithms on target data enhances transfer performance. Our comprehensive study on three datasets from different jurisdictions highlights the role of pre-training in improving transfer performance. We shed light on the pivotal influence of jurisdictional similarity in selecting optimal source datasets for effective transfer. Furthermore, our findings underscore that incorporating unlabeled target data yields improvements in general pre-trained models, with additional gains when silver summaries are introduced. This augmentation is especially valuable when dealing with extractive datasets and scenarios featuring limited alignment between source and target jurisdictions. Our study provides key insights for developing adaptable legal case summarization systems, transcending jurisdictional boundaries.

pdf bib
EDC: Effective and Efficient Dialog Comprehension For Dialog State Tracking
Qifan Lu | Bhaskar Ramasubramanian | Radha Poovendran

In Task-Oriented Dialog (TOD) systems, Dialog State Tracking (DST) structurally extracts information from user and system utterances, which can be further used for querying databases and forming responses to users. The two major categories of DST methods, sequential and independent methods, face trade-offs between accuracy and efficiency. To resolve this issue, we propose Effective and Efficient Dialog Comprehension (EDC), an alternative DST approach that leverages the tree structure of the dialog state. EDC predicts domains, slot names and slot values of the dialog state step-by-step for better accuracy, and efficiently encodes dialog contexts with causal attention patterns. We evaluate EDC on several popular TOD datasets and EDC is able to achieve state-of-the-art Joint Goal Accuracy (JGA). We also show theoretically and empirically that EDC is more efficient than model designs used by previous works.

pdf bib
Automatic Restoration of Diacritics for Speech Data Sets
Sara Shatnawi | Sawsan Alqahtani | Hanan Aldarmaki

Automatic text-based diacritic restoration models generally have high diacritic error rates when applied to speech transcripts as a result of domain and style shifts in spoken language. In this work, we explore the possibility of improving the performance of automatic diacritic restoration when applied to speech data by utilizing parallel spoken utterances. In particular, we use the pre-trained Whisper ASR model fine-tuned on relatively small amounts of diacritized Arabic speech data to produce rough diacritized transcripts for the speech utterances, which we then use as an additional input for diacritic restoration models. The proposed framework consistently improves diacritic restoration performance compared to text-only baselines. Our results highlight the inadequacy of current text-based diacritic restoration models for speech data sets and provide a new baseline for speech-based diacritic restoration.

pdf bib
XNLIeu: a dataset for cross-lingual NLI in Basque
Maite Heredia | Julen Etxaniz | Muitze Zulaika | Xabier Saralegi | Jeremy Barnes | Aitor Soroa

XNLI is a popular Natural Language Inference (NLI) benchmark widely used to evaluate cross-lingual Natural Language Understanding (NLU) capabilities across languages. In this paper, we expand XNLI to include Basque, a low-resource language that can greatly benefit from transfer-learning approaches. The new dataset, dubbed XNLIeu, has been developed by first machine-translating the English XNLI corpus into Basque, followed by a manual post-edition step. We have conducted a series of experiments using mono- and multilingual LLMs to assess a) the effect of professional post-edition on the MT system; b) the best cross-lingual strategy for NLI in Basque; and c) whether the choice of the best cross-lingual strategy is influenced by the fact that the dataset is built by translation. The results show that post-edition is necessary and that the translate-train cross-lingual strategy obtains better results overall, although the gain is lower when tested in a dataset that has been built natively from scratch. Our code and datasets are publicly available under open licenses.

pdf bib
MDR: Model-Specific Demonstration Retrieval at Inference Time for In-Context Learning
Huazheng Wang | Jinming Wu | Haifeng Sun | Zixuan Xia | Daixuan Cheng | Jingyu Wang | Qi Qi | Jianxin Liao

Recently, retrieval-based in-context learning (ICL) methods for selecting demonstrations have been widely investigated. Existing methods train a dense retriever to retrieve the most appropriate demonstrations for a given test query, which improves ICL performance. However, we find that distinct LLMs exhibit different biases for “what is a good demonstration” since they possess differences in training data, model architectures and training methods. As a result, a demonstration suitable for one LLM may not be appropriate for others.Previous approaches ignore the model bias and fail to retrieve the most appropriate demonstrations for different inference LLMs, resulting in a degradation of ICL performance.To address this problem, we propose a simple yet effective metric to evaluate the appropriateness of demonstrations for a specific inference LLM. Furthermore, we introduce a Model-specific Demonstration Retrieval (MDR) method for ICL at inference time, which considers the biases of different LLMs. We test MDR on seen and unseen tasks with multi-scale inference LLMs, such as GPT-Neo-2.7B, LLaMA-7B and Vicuna-13B. Experiments on 23 datasets across 11 data domains highlight the remarkable effectiveness of MDR, showcasing improvements of up to 41.2% in comparison to methods that neglect model biases.

pdf bib
Exploring Cross-Cultural Differences in English Hate Speech Annotations: From Dataset Construction to Analysis
Nayeon Lee | Chani Jung | Junho Myung | Jiho Jin | Jose Camacho-Collados | Juho Kim | Alice Oh

***Warning**: this paper contains content that may be offensive or upsetting.*Most hate speech datasets neglect the cultural diversity within a single language, resulting in a critical shortcoming in hate speech detection. To address this, we introduce **CREHate**, a **CR**oss-cultural **E**nglish **Hate** speech dataset.To construct CREHate, we follow a two-step procedure: 1) cultural post collection and 2) cross-cultural annotation.We sample posts from the SBIC dataset, which predominantly represents North America, and collect posts from four geographically diverse English-speaking countries (Australia, United Kingdom, Singapore, and South Africa) using culturally hateful keywords we retrieve from our survey.Annotations are collected from the four countries plus the United States to establish representative labels for each country.Our analysis highlights statistically significant disparities across countries in hate speech annotations.Only 56.2% of the posts in CREHate achieve consensus among all countries, with the highest pairwise label difference rate of 26%.Qualitative analysis shows that label disagreement occurs mostly due to different interpretations of sarcasm and the personal bias of annotators on divisive topics.Lastly, we evaluate large language models (LLMs) under a zero-shot setting and show that current LLMs tend to show higher accuracies on Anglosphere country labels in CREHate.Our dataset and codes are available at: https://github.com/nlee0212/CREHate

pdf bib
Enhancing Contextual Understanding in Large Language Models through Contrastive Decoding
Zheng Zhao | Emilio Monti | Jens Lehmann | Haytham Assem

Large language models (LLMs) tend to inadequately integrate input context during text generation, relying excessively on encoded prior knowledge in model parameters, potentially resulting in generated text with factual inconsistencies or contextually unfaithful content. LLMs utilize two primary knowledge sources: 1) prior (parametric) knowledge from pretraining, and 2) contextual (non-parametric) knowledge from input prompts. The study addresses the open question of how LLMs effectively balance these knowledge sources during the generation process, specifically in the context of open-domain question answering. To address this issue, we introduce a novel approach integrating contrastive decoding with adversarial irrelevant passages as negative samples to enhance robust context grounding during generation. Notably, our method operates at inference time without requiring further training. We conduct comprehensive experiments to demonstrate its applicability and effectiveness, providing empirical evidence showcasing its superiority over existing methodologies.

pdf bib
Generalizable Sarcasm Detection is Just Around the Corner, of Course!
Hyewon Jang | Diego Frassinelli

We tested the robustness of sarcasm detection models by examining their behavior when fine-tuned on four sarcasm datasets containing varying characteristics of sarcasm: label source (authors vs. third-party), domain (social media/online vs. offline conversations/dialogues), style (aggressive vs. humorous mocking). We tested their prediction performance on the same dataset (intra-dataset) and across different datasets (cross-dataset). For intra-dataset predictions, models consistently performed better when fine-tuned with third-party labels rather than with author labels. For cross-dataset predictions, most models failed to generalize well to the other datasets, implying that one type of dataset cannot represent all sorts of sarcasm with different styles and domains. Compared to the existing datasets, models fine-tuned on the new dataset we release in this work showed the highest generalizability to other datasets. With a manual inspection of the datasets and post-hoc analysis, we attributed the difficulty in generalization to the fact that sarcasm actually comes in different domains and styles. We argue that future sarcasm research should take the broad scope of sarcasm into account.

pdf bib
Encoding of lexical tone in self-supervised models of spoken language
Gaofei Shen | Michaela Watkins | Afra Alishahi | Arianna Bisazza | Grzegorz Chrupała

Interpretability research has shown that self-supervised Spoken LanguageModels (SLMs) encode a wide variety of features in human speech from theacoustic, phonetic, phonological, syntactic and semantic levels, to speakercharacteristics. The bulk of prior research on representations of phonologyhas focused on segmental features such as phonemes; the encoding ofsuprasegmental phonology (such as tone and stress patterns) in SLMs is not yetwell understood. Tone is a suprasegmental feature that is present in more thanhalf of the world’s languages. This paper aims to analyze the tone encodingcapabilities of SLMs, using Mandarin and Vietnamese as case studies. We showthat SLMs encode lexical tone to a significant degree even when they aretrained on data from non-tonal languages. We further find that SLMs behavesimilarly to native and non-native human participants in tone and consonantperception studies, but they do not follow the same developmental trajectory.

pdf bib
A Systematic Comparison of Contextualized Word Embeddings for Lexical Semantic Change
Francesco Periti | Nina Tahmasebi

Contextualized embeddings are the preferred tool for modeling Lexical Semantic Change (LSC). Current evaluations typically focus on a specific task known as Graded Change Detection (GCD). However, performance comparison across work are often misleading due to their reliance on diverse settings. In this paper, we evaluate state-of-the-art models and approaches for GCD under equal conditions. We further break the LSC problem into Word-in-Context (WiC) and Word Sense Induction (WSI) tasks, and compare models across these different levels. Our evaluation is performed across different languages on eight available benchmarks for LSC, and shows that (i) APD outperforms other approaches for GCD; (ii) XL-LEXEME outperforms other contextualized models for WiC, WSI, and GCD, while being comparable to GPT-4; (iii) there is a clear need for improving the modeling of word meanings, as well as focus on *how*, *when*, and *why* these meanings change, rather than solely focusing on the extent of semantic change.

pdf bib
iACOS: Advancing Implicit Sentiment Extraction with Informative and Adaptive Negative Examples
Xiancai Xu | Jia-Dong Zhang | Lei Xiong | Zhishang Liu

Aspect-based sentiment analysis (ABSA) have been extensively studied, but little light has been shed on the quadruple extraction consisting of four fundamental elements: aspects, categories, opinions and sentiments, especially with implicit aspects and opinions. In this paper, we propose a new method iACOS for extracting Implicit Aspects with Categories and Opinions with Sentiments. First, iACOS appends two implicit tokens at the end of a text to capture the context-aware representation of all tokens including implicit aspects and opinions. Second, iACOS develops a sequence labeling model over the context-aware token representation to co-extract explicit and implicit aspects and opinions. Third, iACOS devises a multi-label classifier with a specialized multi-head attention for discovering aspect-opinion pairs and predicting their categories and sentiments simultaneously. Fourth, iACOS leverages informative and adaptive negative examples to jointly train the multi-label classifier and the other two classifiers on categories and sentiments by multi-task learning. Finally, the experimental results show that iACOS significantly outperforms other quadruple extraction baselines according to the F1 score on two public benchmark datasets.

pdf bib
Rectifying Demonstration Shortcut in In-Context Learning
Joonwon Jang | Sanghwan Jang | Wonbin Kweon | Minjin Jeon | Hwanjo Yu

Large language models (LLMs) are able to solve various tasks with only a few demonstrations utilizing their in-context learning (ICL) abilities.However, LLMs often rely on their pre-trained semantic priors of demonstrations rather than on the input-label relationships to proceed with ICL prediction. In this work, we term this phenomenon as the ‘Demonstration Shortcut’.While previous works have primarily focused on improving ICL prediction results for predefined tasks, we aim to rectify the Demonstration Shortcut, thereby enabling the LLM to effectively learn new input-label relationships from demonstrations.To achieve this, we introduce In-Context Calibration, a demonstration-aware calibration method.We evaluate the effectiveness of the proposed method in two settings: (1) the Original ICL Task using the standard label space and (2) the Task Learning setting, where the label space is replaced with semantically unrelated tokens.In both settings, In-Context Calibration demonstrates substantial improvements, with results generalized across three LLM families (OPT, GPT, and Llama2) under various configurations.

pdf bib
Universal NER: A Gold-Standard Multilingual Named Entity Recognition Benchmark
Stephen Mayhew | Terra Blevins | Shuheng Liu | Marek Suppa | Hila Gonen | Joseph Marvin Imperial | Börje Karlsson | Peiqin Lin | Nikola Ljubešić | Lester James Miranda | Barbara Plank | Arij Riabi | Yuval Pinter

We introduce Universal NER (UNER), an open, community-driven project to develop gold-standard NER benchmarks in many languages. The overarching goal of UNER is to provide high-quality, cross-lingually consistent annotations to facilitate and standardize multilingual NER research. UNER v1 contains 19 datasets annotated with named entities in a cross-lingual consistent schema across 13 diverse languages. In this paper, we detail the dataset creation and composition of UNER; we also provide initial modeling baselines on both in-language and cross-lingual learning settings. We will release the data, code, and fitted models to the public.

pdf bib
ODD: A Benchmark Dataset for the Natural Language Processing Based Opioid Related Aberrant Behavior Detection
Sunjae Kwon | Xun Wang | Weisong Liu | Emily Druhl | Minhee Sung | Joel Reisman | Wenjun Li | Robert Kerns | William Becker | Hong Yu

Opioid related aberrant behaviors (ORABs) present novel risk factors for opioid overdose. This paper introduces a novel biomedical natural language processing benchmark dataset named ODD, for ORAB Detection Dataset. ODD is an expert-annotated dataset designed to identify ORABs from patients’ EHR notes and classify them into nine categories; 1) Confirmed Aberrant Behavior, 2) Suggested Aberrant Behavior, 3) Opioids, 4) Indication, 5) Diagnosed opioid dependency, 6) Benzodiazepines, 7) Medication Changes, 8) Central Nervous System-related, and 9) Social Determinants of Health. We explored two state-of-the-art natural language processing models (fine-tuning and prompt-tuning approaches) to identify ORAB. Experimental results show that the prompt-tuning models outperformed the fine-tuning models in most categories and the gains were especially higher among uncommon categories (Suggested Aberrant Behavior, Confirmed Aberrant Behaviors, Diagnosed Opioid Dependence, and Medication Change). Although the best model achieved the highest 88.17% on macro average area under precision recall curve, uncommon classes still have a large room for performance improvement. ODD is publicly available.

pdf bib
A Comprehensive Study of Gender Bias in Chemical Named Entity Recognition Models
Xingmeng Zhao | Ali Niazi | Anthony Rios

Chemical named entity recognition (NER) models are used in many downstream tasks, from adverse drug reaction identification to pharmacoepidemiology. However, it is unknown whether these models work the same for everyone. Performance disparities can potentially cause harm rather than the intended good. This paper assesses gender-related performance disparities in chemical NER systems. We develop a framework for measuring gender bias in chemical NER models using synthetic data and a newly annotated corpus of over 92,405 words with self-identified gender information from Reddit. Our evaluation of multiple biomedical NER models reveals evident biases. For instance, synthetic data suggests that female names are frequently misclassified as chemicals, especially when it comes to brand name mentions. Additionally, we observe performance disparities between female- and male-associated data in both datasets. Many systems fail to detect contraceptives such as birth control. Our findings emphasize the biases in chemical NER models, urging practitioners to account for these biases in downstream applications.

pdf bib
The Promises and Pitfalls of Using Language Models to Measure Instruction Quality in Education
Paiheng Xu | Jing Liu | Nathan Jones | Julie Cohen | Wei Ai

Assessing instruction quality is a fundamental component of any improvement efforts in the education system. However, traditional manual assessments are expensive, subjective, and heavily dependent on observers’ expertise and idiosyncratic factors, preventing teachers from getting timely and frequent feedback. Different from prior research that mostly focuses on low-inference instructional practices on a singular basis, this paper presents the first study that leverages Natural Language Processing (NLP) techniques to assess multiple high-inference instructional practices in two distinct educational settings: in-person K-12 classrooms and simulated performance tasks for pre-service teachers. This is also the first study that applies NLP to measure a teaching practice that is widely acknowledged to be particularly effective for students with special needs. We confront two challenges inherent in NLP-based instructional analysis, including noisy and long input data and highly skewed distributions of human ratings. Our results suggest that pretrained Language Models (PLMs) demonstrate performances comparable to the agreement level of human raters for variables that are more discrete and require lower inference, but their efficacy diminishes with more complex teaching practices. Interestingly, using only teachers’ utterances as input yields strong results for student-centered variables, alleviating common concerns over the difficulty of collecting and transcribing high-quality student speech data in in-person teaching settings. Our findings highlight both the potential and the limitations of current NLP techniques in the education domain, opening avenues for further exploration.

pdf bib
Differentially Private Next-Token Prediction of Large Language Models
James Flemings | Meisam Razaviyayn | Murali Annavaram

Ensuring the privacy of Large Language Models (LLMs) is becoming increasingly important. The most widely adopted technique to accomplish this is DP-SGD, which trains a model to guarantee Differential Privacy (DP). However, DP-SGD overestimates an adversary’s capabilities in having white box access to the model and, as a result, causes longer training times and larger memory usage than SGD. On the other hand, commercial LLM deployments are predominantly cloud-based; hence, adversarial access to LLMs is black-box. Motivated by these observations, we present Private Mixing of Ensemble Distributions (PMixED): a private prediction protocol for next-token prediction that utilizes the inherent stochasticity of next-token sampling and a public model to achieve Differential Privacy. We formalize this by introducing RD-mollifers which project each of the model’s output distribution from an ensemble of fine-tuned LLMs onto a set around a public LLM’s output distribution, then average the projected distributions and sample from it. Unlike DP-SGD which needs to consider the model architecture during training, PMixED is model agnostic, which makes PMixED a very appealing solution for current deployments. Our results show that PMixED achieves a stronger privacy guarantee than sample-level privacy and outperforms DP-SGD for privacy 𝜖 = 8 on large-scale datasets. Thus, PMixED offers a practical alternative to DP training methods for achieving strong generative utility without compromising privacy.

pdf bib
Improving Adversarial Data Collection by Supporting Annotators: Lessons from GAHD, a German Hate Speech Dataset
Janis Goldzycher | Paul Röttger | Gerold Schneider

Hate speech detection models are only as good as the data they are trained on. Datasets sourced from social media suffer from systematic gaps and biases, leading to unreliable models with simplistic decision boundaries. Adversarial datasets, collected by exploiting model weaknesses, promise to fix this problem. However, adversarial data collection can be slow and costly, and individual annotators have limited creativity. In this paper, we introduce GAHD, a new German Adversarial Hate speech Dataset comprising ca. 11k examples. During data collection, we explore new strategies for supporting annotators, to create more diverse adversarial examples more efficiently and provide a manual analysis of annotator disagreements for each strategy. Our experiments show that the resulting dataset is challenging even for state-of-the-art hate speech detection models, and that training on GAHD clearly improves model robustness. Further, we find that mixing multiple support strategies is most advantageous. We make GAHD publicly available at https://github.com/jagol/gahd.

pdf bib
Memory Augmented Language Models through Mixture of Word Experts
Cicero Nogueira dos Santos | James Lee-Thorp | Isaac Noble | Chung-Ching Chang | David Uthus

Scaling up the number of parameters of language models has proven to be an effective approach to improve performance. For dense models, increasing their size proportionally increases their computational footprint. In this work, we seek to aggressively decouple learning capacity and FLOPs through Mixture-of-Experts (MoE) style models with large knowledge-rich vocabulary based routing functions. Our proposed approach, dubbed Mixture of Word Experts (MoWE), can be seen as a memory augmented model, where a large set of word-specific experts play the role of a sparse memory. We demonstrate that MoWE performs significantly better than the T5 family of models with similar number of FLOPs in a variety of NLP tasks. Moreover, MoWE outperforms traditional MoE models on knowledge intensive tasks and has similar performance to complex memory augmented approaches that often require to invoke custom mechanisms to search the sparse memory.

pdf bib
Impossible Distillation for Paraphrasing and Summarization: How to Make High-quality Lemonade out of Small, Low-quality Model
Jaehun Jung | Peter West | Liwei Jiang | Faeze Brahman | Ximing Lu | Jillian Fisher | Taylor Sorensen | Yejin Choi

We present Impossible Distillation, a novel framework for paraphrasing and sentence summarization, that distills a high-quality dataset and model from a low-quality teacher that itself cannot perform these tasks. Unlike prior works that rely on an extreme-scale teacher model (e.g., GPT3) or task-specific architecture, we hypothesize and verify the paraphrastic proximity intrinsic to pre-trained LMs (e.g., GPT2), where paraphrases occupy a proximal subspace in the LM distribution. By identifying and distilling generations from these subspaces, Impossible Distillation produces a high-quality dataset and model even from GPT2-scale LMs. We evaluate our method on multiple benchmarks spanning unconstrained / syntax-controlled paraphrase generation and sentence summarization. Our model with 770M parameters consistently outperforms strong baselines, including models distilled from ChatGPT, and sometimes, even ChatGPT itself. Also, we find that our distilled dataset from 1.5B LMs exhibits higher diversity and fidelity than up to 13 times larger datasets.

pdf bib
TofuEval: Evaluating Hallucinations of LLMs on Topic-Focused Dialogue Summarization
Liyan Tang | Igor Shalyminov | Amy Wong | Jon Burnsky | Jake Vincent | Yu’an Yang | Siffi Singh | Song Feng | Hwanjun Song | Hang Su | Lijia Sun | Yi Zhang | Saab Mansour | Kathleen McKeown

Single document news summarization has seen substantial progress on faithfulness in recent years, driven by research on the evaluation of factual consistency, or hallucinations. We ask whether these advances carry over to other text summarization domains. We propose a new evaluation benchmark on topic-focused dialogue summarization, generated by LLMs of varying sizes. We provide binary sentence- level human annotations of the factual consistency of these summaries along with detailed explanations of factually inconsistent sentences. Our analysis shows that existing LLMs hallucinate significant amounts of factual errors in the dialogue domain, regardless of the model’s size. On the other hand, when LLMs, including GPT-4, serve as binary factual evaluators, they perform poorly and can be outperformed by prevailing state-of-the-art specialized factuality evaluation metrics. Finally, we conducted an analysis of hallucination types with a curated error taxonomy. We find that there are diverse errors and error distributions in model-generated summaries and that non-LLM based metrics can capture all error types better than LLM-based evaluators.

pdf bib
MOKA: Moral Knowledge Augmentation for Moral Event Extraction
Xinliang Frederick Zhang | Winston Wu | Nicholas Beauchamp | Lu Wang

News media often strive to minimize explicit moral language in news articles, yet most articles are dense with moral values as expressed through the reported events themselves. However, values that are reflected in the intricate dynamics among *participating entities* and *moral events* are far more challenging for most NLP systems to detect, including LLMs. To study this phenomenon, we annotate a new dataset, **MORAL EVENTS**, consisting of 5,494 structured event annotations on 474 news articles by diverse US media across the political spectrum. We further propose **MOKA**, a moral event extraction framework with **MO**ral **K**nowledge **A**ugmentation, which leverages knowledge derived from moral words and moral scenarios to produce structural representations of morality-bearing events. Experiments show that **MOKA** outperforms competitive baselines across three moral event understanding tasks. Further analysis shows even ostensibly nonpartisan media engage in the selective reporting of moral events.

pdf bib
Fixing Rogue Memorization in Many-to-One Multilingual Translators of Extremely-Low-Resource Languages by Rephrasing Training Samples
Paulo Cavalin | Pedro Henrique Domingues | Claudio Pinhanez | Julio Nogima

In this paper we study the fine-tuning of pre-trained large high-resource language models (LLMs) into many-to-one multilingual machine translators for extremely-low-resource languages such as endangered Indigenous languages. We explore those issues using datasets created from pseudo-parallel translations to English of The Bible written in 39 Brazilian Indigenous languages using mBART50 and WMT19 as pre-trained models and multiple translation metrics. We examine bilingual and multilingual models and show that, according to machine translation metrics, same-linguistic family models tend to perform best. However, we also found that many-to-one multilingual systems have a tendency to learn a “rogue” strategy of storing output strings from the training data in the LLM structure and retrieving them instead of performing actual translations. We show that rephrasing the output of the training samples seems to solve the problem.

pdf bib
Backdoor Attacks on Multilingual Machine Translation
Jun Wang | Qiongkai Xu | Xuanli He | Benjamin Rubinstein | Trevor Cohn

While multilingual machine translation (MNMT) systems hold substantial promise, they also have security vulnerabilities. Our research highlights that MNMT systems can be susceptible to a particularly devious style of backdoor attack, whereby an attacker injects poisoned data into a low-resource language pair to cause malicious translations in other languages, including high-resource languages.Our experimental results reveal that injecting less than 0.01% poisoned data into a low-resource language pair can achieve an average 20% attack success rate in attacking high-resource language pairs. This type of attack is of particular concern, given the larger attack surface of languages inherent to low-resource settings. Our aim is to bring attention to these vulnerabilities within MNMT systems with the hope of encouraging the community to address security concerns in machine translation, especially in the context of low-resource languages.

pdf bib
Personalized Jargon Identification for Enhanced Interdisciplinary Communication
Yue Guo | Joseph Chee Chang | Maria Antoniak | Erin Bransom | Trevor Cohen | Lucy Wang | Tal August

Scientific jargon can confuse researchers when they read materials from other domains. Identifying and translating jargon for individual researchers could speed up research, but current methods of jargon identification mainly use corpus-level familiarity indicators rather than modeling researcher-specific needs, which can vary greatly based on each researcher’s background. We collect a dataset of over 10K term familiarity annotations from 11 computer science researchers for terms drawn from 100 paper abstracts. Analysis of this data reveals that jargon familiarity and information needs vary widely across annotators, even within the same sub-domain (e.g., NLP). We investigate features representing domain, subdomain, and individual knowledge to predict individual jargon familiarity. We compare supervised and prompt-based approaches, finding that prompt-based methods using information about the individual researcher (e.g., personal publications, self-defined subfield of research) yield the highest accuracy, though the task remains difficult and supervised approaches have lower false positive rates. This research offers insights into features and methods for the novel task of integrating personal data into scientific jargon identification.

pdf bib
Flames: Benchmarking Value Alignment of LLMs in Chinese
Kexin Huang | Xiangyang Liu | Qianyu Guo | Tianxiang Sun | Jiawei Sun | Yaru Wang | Zeyang Zhou | Yixu Wang | Yan Teng | Xipeng Qiu | Yingchun Wang | Dahua Lin

The widespread adoption of large language models (LLMs) across various regions underscores the urgent need to evaluate their alignment with human values. Current benchmarks, however, fall short of effectively uncovering safety vulnerabilities in LLMs. Despite numerous models achieving high scores and ‘topping the chart’ in these evaluations, there is still a significant gap in LLMs’ deeper alignment with human values and achieving genuine harmlessness. To this end, this paper proposes a value alignment benchmark named Flames, which encompasses both common harmlessness principles and a unique morality dimension that integrates specific Chinese values such as harmony. Accordingly, we carefully design adversarial prompts that incorporate complex scenarios and jailbreaking methods, mostly with implicit malice. By prompting 17 mainstream LLMs, we obtain model responses and rigorously annotate them for detailed evaluation. Our findings indicate that all the evaluated LLMs demonstrate relatively poor performance on Flames, particularly in the safety and fairness dimensions. We also develop a lightweight specified scorer capable of scoring LLMs across multiple dimensions to efficiently evaluate new models on the benchmark. The complexity of Flames has far exceeded existing benchmarks, setting a new challenge for contemporary LLMs and highlighting the need for further alignment of LLMs. Our benchmark is publicly available at https://github.com/AIFlames/Flames.

pdf bib
Mitigating Bias for Question Answering Models by Tracking Bias Influence
Mingyu Ma | Jiun-Yu Kao | Arpit Gupta | Yu-Hsiang Lin | Wenbo Zhao | Tagyoung Chung | Wei Wang | Kai-Wei Chang | Nanyun Peng

Models of various NLP tasks have been shown to exhibit stereotypes, and the bias in the question answering (QA) models is especially harmful as the output answers might be directly consumed by the end users. There have been datasets to evaluate bias in QA models, while bias mitigation technique for the QA models is still under-explored. In this work, we propose BMBI, an approach to mitigate the bias of multiple-choice QA models. Based on the intuition that a model would lean to be more biased if it learns from a biased example, we measure the bias level of a query instance by observing its influence on another instance. If the influenced instance is more biased, we derive that the query instance is biased. We then use the bias level detected as an optimization objective to form a multi-task learning setting in addition to the original QA task. We further introduce a new bias evaluation metric to quantify bias in a comprehensive and sensitive way. We show that our method could be applied to multiple QA formulations across multiple bias categories. It can significantly reduce the bias level in all 9 bias categories in the BBQ dataset while maintaining comparable QA accuracy.

pdf bib
Extending CLIP’s Image-Text Alignment to Referring Image Segmentation
Seoyeon Kim | Minguk Kang | Dongwon Kim | Jaesik Park | Suha Kwak

Referring Image Segmentation (RIS) is a cross-modal task that aims to segment an instance described by a natural language expression. Recent methods leverage large-scale pretrained unimodal models as backbones along with fusion techniques for joint reasoning across modalities. However, the inherent cross-modal nature of RIS raises questions about the effectiveness of unimodal backbones. We propose RISCLIP, a novel framework that effectively leverages the cross-modal nature of CLIP for RIS. Observing CLIP’s inherent alignment between image and text features, we capitalize on this starting point and introduce simple but strong modules that enhance unimodal feature extraction and leverage rich alignment knowledge in CLIP’s image-text shared-embedding space. RISCLIP exhibits outstanding results on all three major RIS benchmarks and also outperforms previous CLIP-based methods, demonstrating the efficacy of our strategy in extending CLIP’s image-text alignment to RIS.

pdf bib
Generating Attractive and Authentic Copywriting from Customer Reviews
Yu-Xiang Lin | Wei-Yun Ma

The goal of product copywriting is to capture the interest of potential buyers by emphasizing the features of products through text descriptions. As e-commerce platforms offer a wide range of services, it’s becoming essential to dynamically adjust the styles of these auto-generated descriptions. Typical approaches to copywriting generation often rely solely on specified product attributes, which may result in dull and repetitive content. To tackle this issue, we propose to generate copywriting based on customer reviews, as they provide firsthand practical experiences with products, offering a richer source of information than just product attributes. We have developed a sequence-to-sequence framework, enhanced with reinforcement learning, to produce copywriting that is attractive, authentic, and rich in information. Our framework outperforms all existing baseline and zero-shot large language models, including LLaMA-2-chat-7B and GPT-3.5, in terms of both attractiveness and faithfulness. Furthermore, this work features the use of LLMs for aspect-based summaries collection and argument allure assessment. Experiments demonstrate the effectiveness of using LLMs for marketing domain corpus construction. The code and the dataset is publicly available at: https://github.com/YuXiangLin1234/Copywriting-Generation.

pdf bib
Effective Long-Context Scaling of Foundation Models
Wenhan Xiong | Jingyu Liu | Igor Molybog | Hejia Zhang | Prajjwal Bhargava | Rui Hou | Louis Martin | Rashi Rungta | Karthik Abinav Sankararaman | Barlas Oguz | Madian Khabsa | Han Fang | Yashar Mehdad | Sharan Narang | Kshitiz Malik | Angela Fan | Shruti Bhosale | Sergey Edunov | Mike Lewis | Sinong Wang | Hao Ma

We present an effective recipe to train strong long-context LLMs that are capable of utilizing massive context windows of up to 32,000 tokens. Our models are built through continual pretraining from Llama 2 checkpoints with longer text sequences and on a dataset where long texts are upsampled. We perform extensive evaluation using language modeling, synthetic context probing tasks, and a wide range of downstream benchmarks. Across all evaluations, our models achieve consistent improvements on most regular-context tasks and significant improvements on long-context tasks over Llama 2. Moreover, with a cost-effective instruction tuning procedure that is free of expensive annotation, the presented models can already surpass gpt-3.5-turbo-16k‘s overall performance on long-context benchmarks. Alongside these results, we provide an in-depth analysis on each individual component of our method. We delve into Llama’s position encodings and discuss its key limitation in modeling long data. We examine the impact of various design choices in the pretraining process, including the data mix and the training curriculum of sequence lengths – ablation results suggest that having abundant long texts in the pretrain dataset is not the key to achieving strong performance, and we empirically verify that long context continual pretraining is more efficient and similarly effective compared to pretraining from scratch with long sequences.

pdf bib
Empowering Diffusion Models on the Embedding Space for Text Generation
Zhujin Gao | Junliang Guo | Xu Tan | Yongxin Zhu | Fang Zhang | Jiang Bian | Linli Xu

Diffusion models have achieved state-of-the-art synthesis quality on both visual and audio tasks, and recent works further adapt them to textual data by diffusing on the embedding space. In this paper, we conduct systematic studies of the optimization challenges encountered with both the embedding space and the denoising model, which have not been carefully explored. Firstly, the data distribution is learnable for embeddings, which may lead to the collapse of the embedding space and unstable training. To alleviate this problem, we propose a new objective called the anchor loss which is more efficient than previous methods. Secondly, we find the noise levels of conventional schedules are insufficient for training a desirable denoising model while introducing varying degrees of degeneration in consequence. To address this challenge, we propose a novel framework called noise rescaling. Based on the above analysis, we propose Difformer, an embedding diffusion model based on Transformer. Experiments on varieties of seminal text generation tasks show the effectiveness of the proposed methods and the superiority of Difformer over previous state-of-the-art embedding diffusion baselines.

pdf bib
Aligning as Debiasing: Causality-Aware Alignment via Reinforcement Learning with Interventional Feedback
Yu Xia | Tong Yu | Zhankui He | Handong Zhao | Julian McAuley | Shuai Li

Large language models (LLMs) often generate biased outputs containing offensive, toxic, or stereotypical text. Existing LLM alignment methods such as reinforcement learning from human feedback (RLHF) alleviate biases primarily based on reward signals from current model outputs without considering the source of biases. In this work, to explore how biases are formed, we revisit LLMs’ text generation from a causal perspective. We identify pretraining data and input prompts, which contain semantic correlations of textual phrases, as two confounders between LLMs and model outputs causing biases. Inspired by our causal view, we leverage the reward model in RL alignment as an instrumental variable to perform causal intervention on LLMs. Utilizing the reward difference between an initial LLM and intervened LLM as interventional feedback to guide RL finetuning, we propose Causality-Aware Alignment (CAA) for LLM debiasing. Experiments on two text generation tasks with three different alignment objectives demonstrate the advantages of our method in aligning LLMs to generate less biased and safer outputs.

pdf bib
Fake Alignment: Are LLMs Really Aligned Well?
Yixu Wang | Yan Teng | Kexin Huang | Chengqi Lyu | Songyang Zhang | Wenwei Zhang | Xingjun Ma | Yu-Gang Jiang | Yu Qiao | Yingchun Wang

The growing awareness of safety concerns in large language models (LLMs) has sparked considerable interest in the evaluation of safety. This study investigates an under-explored issue about the evaluation of LLMs, namely the substantial discrepancy in performance between multiple-choice questions and open-ended questions. Inspired by research on jailbreak attack patterns, we argue this is caused by mismatched generalization. That is, LLM only remembers the answer style for open-ended safety questions, which makes it unable to solve other forms of safety tests. We refer to this phenomenon as fake alignment and construct a comparative benchmark to empirically verify its existence in LLMs. We introduce a Fake alIgNment Evaluation (FINE) framework and two novel metrics——Consistency Score (CS) and Consistent Safety Score (CSS), which jointly assess two complementary forms of evaluation to quantify fake alignment and obtain corrected performance estimation. Applying FINE to 14 widely-used LLMs reveals several models with purported safety are poorly aligned in practice. Subsequently, we found that multiple-choice format data can also be used as high-quality contrast distillation-based fine-tuning data, which can strongly improve the alignment consistency of LLMs with minimal fine-tuning overhead. For data and code, see https://github.com/AIFlames/Fake-Alignment.

pdf bib
Visually Guided Generative Text-Layout Pre-training for Document Intelligence
Zhiming Mao | Haoli Bai | Lu Hou | Lifeng Shang | Xin Jiang | Qun Liu | Kam-Fai Wong

Prior study shows that pre-training techniques can boost the performance of visual document understanding (VDU), which typically requires models to gain abilities to perceive and reason both document texts and layouts (e.g., locations of texts and table-cells). To this end, we propose visually guided generative text-layout pre-training, named ViTLP. Given a document image, the model optimizes hierarchical language and layout modeling objectives to generate the interleaved text and layout sequence. In addition, to address the limitation of processing long documents by Transformers, we introduce a straightforward yet effective multi-segment generative pre-training scheme, facilitating ViTLP to process word-intensive documents of any length. ViTLP can function as a native OCR model to localize and recognize texts of document images. Besides, ViTLP can be effectively applied to various downstream VDU tasks. Extensive experiments show that ViTLP achieves competitive performance over existing baselines on benchmark VDU tasks, including information extraction, document classification, and document question answering.

pdf bib
HILL: Hierarchy-aware Information Lossless Contrastive Learning for Hierarchical Text Classification
He Zhu | Junran Wu | Ruomei Liu | Yue Hou | Ze Yuan | Shangzhe Li | Yicheng Pan | Ke Xu

Existing self-supervised methods in natural language processing (NLP), especially hierarchical text classification (HTC), mainly focus on self-supervised contrastive learning, extremely relying on human-designed augmentation rules to generate contrastive samples, which can potentially corrupt or distort the original information. In this paper, we tend to investigate the feasibility of a contrastive learning scheme in which the semantic and syntactic information inherent in the input sample is adequately reserved in the contrastive samples and fused during the learning process. Specifically, we propose an information lossless contrastive learning strategy for HTC, namely Hierarchy-aware Information Lossless contrastive Learning (HILL), which consists of a text encoder representing the input document, and a structure encoder directly generating the positive sample. The structure encoder takes the document embedding as input, extracts the essential syntactic information inherent in the label hierarchy with the principle of structural entropy minimization, and injects the syntactic information into the text representation via hierarchical representation learning. Experiments on three common datasets are conducted to verify the superiority of HILL.

pdf bib
Investigating the Emergent Audio Classification Ability of ASR Foundation Models
Rao Ma | Adian Liusie | Mark Gales | Kate Knill

Text and vision foundation models can perform many tasks in a zero-shot setting, a desirable property that enables these systems to be applied in general and low-resource settings. There has been far less work, however, on the zero-shot abilities of ASR foundation models, with these systems typically fine-tuned to specific tasks or constrained to applications that match their training criterion and data annotation. In this work we investigate the ability of Whisper and MMS, ASR foundation models trained primarily for speech recognition, to perform zero-shot audio classification. We use simple template-based text prompts at the decoder and use the resulting decoding probabilities to generate zero-shot predictions. Without training the model on extra data or adding any new parameters, we demonstrate that Whisper shows promising zero-shot classification performance on a range of 8 audio-classification datasets, outperforming the accuracy of existing state-of-the-art zero-shot baselines by an average of 9%. One important step to unlock the emergent ability is debiasing, where a simple unsupervised reweighting method of the class probabilities yields consistent significant performance gains. We further show that performance increases with model size, implying that as ASR foundation models scale up, they may exhibit improved zero-shot performance.

pdf bib
In-context Learning Generalizes, But Not Always Robustly: The Case of Syntax
Aaron Mueller | Albert Webson | Jackson Petty | Tal Linzen

In-context learning (ICL) is now a common method for teaching large language models (LLMs) new tasks: given labeled examples in the input context, the LLM learns to perform the task without weight updates. Do models guided via ICL infer the underlying structure of the task defined by the context, or do they rely on superficial heuristics that only generalize to identically distributed examples? We address this question using transformations tasks and an NLI task that assess sensitivity to syntax—a requirement for robust language understanding. We further investigate whether out-of-distribution generalization can be improved via chain-of-thought prompting, where the model is provided with a sequence of intermediate computation steps that illustrate how the task ought to be performed. In experiments with models from the GPT, PaLM, and Llama 2 families, we find large variance across LMs. The variance is explained more by the composition of the pre-training corpus and supervision methods than by model size; in particular, models pre-trained on code generalize better, and benefit more from chain-of-thought prompting.

pdf bib
Prompt-Singer: Controllable Singing-Voice-Synthesis with Natural Language Prompt
Yongqi Wang | Ruofan Hu | Rongjie Huang | Zhiqing Hong | Ruiqi Li | Wenrui Liu | Fuming You | Tao Jin | Zhou Zhao

Recent singing-voice-synthesis (SVS) methods have achieved remarkable audio quality and naturalness, yet they lack the capability to control the style attributes of the synthesized singing explicitly. We propose Prompt-Singer, the first SVS method that enables attribute controlling on singer gender, vocal range and volume with natural language. We adopt a model architecture based on a decoder-only transformer with a multi-scale hierarchy, and design a range-melody decoupled pitch representation that enables text-conditioned vocal range control while keeping melodic accuracy. Furthermore, we explore various experiment settings, including different types of text representations, text encoder fine-tuning, and introducing speech data to alleviate data scarcity, aiming to facilitate further research. Experiments show that our model achieves favorable controlling ability and audio quality. Audio samples are available at http://prompt-singer.github.io .

pdf bib
Lost in Transcription: Identifying and Quantifying the Accuracy Biases of Automatic Speech Recognition Systems Against Disfluent Speech
Dena Mujtaba | Nihar Mahapatra | Megan Arney | J Yaruss | Hope Gerlach-Houck | Caryn Herring | Jia Bin

Automatic speech recognition (ASR) systems, increasingly prevalent in education, healthcare, employment, and mobile technology, face significant challenges in inclusivity, particularly for the 80 million-strong global community of people who stutter. These systems often fail to accurately interpret speech patterns deviating from typical fluency, leading to critical usability issues and misinterpretations. This study evaluates six leading ASRs, analyzing their performance on both a real-world dataset of speech samples from individuals who stutter and a synthetic dataset derived from the widely-used LibriSpeech benchmark. The synthetic dataset, uniquely designed to incorporate various stuttering events, enables an in-depth analysis of each ASR’s handling of disfluent speech. Our comprehensive assessment includes metrics such as word error rate (WER), character error rate (CER), and semantic accuracy of the transcripts. The results reveal a consistent and statistically significant accuracy bias across all ASRs against disfluent speech, manifesting in significant syntactical and semantic inaccuracies in transcriptions. These findings highlight a critical gap in current ASR technologies, underscoring the need for effective bias mitigation strategies. Addressing this bias is imperative not only to improve the technology’s usability for people who stutter but also to ensure their equitable and inclusive participation in the rapidly evolving digital landscape.

pdf bib
MAFALDA: A Benchmark and Comprehensive Study of Fallacy Detection and Classification
Chadi Helwe | Tom Calamai | Pierre-Henri Paris | Chloé Clavel | Fabian Suchanek

We introduce MAFALDA, a benchmark for fallacy classification that merges and unites previous fallacy datasets. It comes with a taxonomy that aligns, refines, and unifies existing classifications of fallacies. We further provide a manual annotation of a part of the dataset together with manual explanations for each annotation. We propose a new annotation scheme tailored for subjective NLP tasks, and a new evaluation method designed to handle subjectivity. We then evaluate several language models under a zero-shot learning setting and human performances on MAFALDA to assess their capability to detect and classify fallacies.

pdf bib
Diffusion Glancing Transformer for Parallel Sequence-to-Sequence Learning
Lihua Qian | Mingxuan Wang | Yang Liu | Hao Zhou

Previously, non-autoregressive models were widely recognized as being superior in generation efficiency but inferior in generation quality due to the challenges of modeling multiple target modalities.To enhance the multi-modality modeling ability, we propose the diffusion glancing transformer, which employs a modality diffusion process and residual glancing sampling.The modality diffusion process is a discrete process that interpolates the multi-modal distribution along the decoding steps, and the residual glancing sampling approach guides the model to continuously learn the remaining modalities across the layers. Experimental results on various machine translation and text generation benchmarks demonstrate that DIFFGLAT achieves better generation accuracy while maintaining fast decoding speed compared with both autoregressive and non-autoregressive models.

pdf bib
No Context Needed: Contextual Quandary In Idiomatic Reasoning With Pre-Trained Language Models
Kellen Cheng | Suma Bhat

Reasoning in the presence of idiomatic expressions (IEs) remains a challenging frontier in natural language understanding (NLU). Unlike standard text, the non-compositional nature of an IE makes it difficult for model comprehension, as their figurative or non-literal mean- ing usually cannot be inferred from the constituent words alone. It stands to reason that in these challenging circumstances, pre-trained language models (PTLMs) should make use of the surrounding context to infer additional in- formation about the IE. In this paper, we investigate the utilization of said context for idiomatic reasoning tasks, which is under-explored relative to arithmetic or commonsense reason- ing (Liu et al., 2022; Yu et al., 2023). Preliminary findings point to a surprising observation: general purpose PTLMs are actually negatively affected by the context, as performance almost always increases with its removal. In these scenarios, models may see gains of up to 3.89%. As a result, we argue that only IE-aware models remain suitable for idiomatic reasoning tasks, given the unexpected and unexplainable manner in which general purpose PTLMs reason over IEs. Additionally, we conduct studies to examine how models utilize the context in various situations, as well as an in-depth analysis on dataset formation and quality. Finally, we provide some explanations and insights into the reasoning process itself based on our results.

pdf bib
Multi-stage Retrieve and Re-rank Model for Automatic Medical Coding Recommendation
Xindi Wang | Robert Mercer | Frank Rudzicz

The International Classification of Diseases (ICD) serves as a definitive medical classification system encompassing a wide range of diseases and conditions. The primary objective of ICD indexing is to allocate a subset of ICD codes to a medical record, which facilitates standardized documentation and management of various health conditions. Most existing approaches have suffered from selecting the proper label subsets from an extremely large ICD collection with a heavy long-tailed label distribution. In this paper, we leverage a multi-stage “retrieve and re-rank” framework as a novel solution to ICD indexing, via a hybrid discrete retrieval method, and re-rank retrieved candidates with contrastive learning that allows the model to make more accurate predictions from a simplified label space. The retrieval model is a hybrid of auxiliary knowledge of the electronic health records (EHR) and a discrete retrieval method (BM25), which efficiently collects high-quality candidates. In the last stage, we propose a label co-occurrence guided contrastive re-ranking model, which re-ranks the candidate labels by pulling together the clinical notes with positive ICD codes. Experimental results show the proposed method achieves state-of-the-art performance on a number of measures on the MIMIC-III benchmark.

pdf bib
Anisotropy is Not Inherent to Transformers
Anemily Machina | Robert Mercer

Isotropy is the property that embeddings are uniformly distributed around the origin. Previous work has shown that Transformer embedding spaces are anisotropic, which is called the representation degradation problem. This degradation has been assumed to be inherent to the standard language modeling tasks and to apply to all Transformer models regardless of their architecture. In this work we identify a set of Transformer models with isotropic embedding spaces, the large Pythia models. We examine the isotropy of Pythia models and explore how isotropy and anisotropy develop as a model is trained. We find that anisotropic models do not develop as previously theorized, using our own analysis to show that the large Pythia models optimize their final Layer Norm for isotropy, and provide reasoning why previous theoretical justifications for anisotropy were insufficient. The identification of a set of isotropic Transformer models calls previous assumptions into question, provides a set of models to contrast existing analysis, and should lead to deeper insight into isotropy.

pdf bib
Finding Replicable Human Evaluations via Stable Ranking Probability
Parker Riley | Daniel Deutsch | George Foster | Viresh Ratnakar | Ali Dabirmoghaddam | Markus Freitag

Reliable human evaluation is critical to the development of successful natural language generation models, but achieving it is notoriously difficult. Stability is a crucial requirement when ranking systems by quality: consistent ranking of systems across repeated evaluations is not just desirable, but essential. Without it, there is no reliable foundation for hill-climbing or product launch decisions. In this paper, we use machine translation and its state-of-the-art human evaluation framework, MQM, as a case study to understand how to set up reliable human evaluations that yield stable conclusions. We investigate the optimal configurations for item allocation to raters, number of ratings per item, and score normalization. Our study on two language pairs provides concrete recommendations for designing replicable human evaluation studies. We also collect and release the largest publicly available dataset of multi-segment translations rated by multiple professional translators, consisting of nearly 140,000 segment annotations across two language pairs.

pdf bib
Stealthy and Persistent Unalignment on Large Language Models via Backdoor Injections
Yuanpu Cao | Bochuan Cao | Jinghui Chen

Recent developments in Large Language Models (LLMs) have manifested significant advancements. To facilitate safeguards against malicious exploitation, a body of research has concentrated on aligning LLMs with human preferences and inhibiting their generation of inappropriate content. Unfortunately, such alignments are often vulnerable: fine-tuning with a minimal amount of harmful data can easily unalign the target LLM. While being effective, such fine-tuning-based unalignment approaches also have their own limitations: (1) non-stealthiness, after fine-tuning, safety audits or red-teaming can easily expose the potential weaknesses of the unaligned models, thereby precluding their release/use. (2) non-persistence, the unaligned LLMs can be easily repaired through re-alignment, i.e., fine-tuning again with aligned data points. In this work, we show that it is possible to conduct stealthy and persistent unalignment on large language models via backdoor injections. We also provide a novel understanding of the relationship between the backdoor persistence and the activation pattern and further provide guidelines for potential trigger design. Through extensive experiments, we demonstrate that our proposed stealthy and persistent unalignment can successfully pass the safety evaluation while maintaining strong persistence against re-alignment defense.

pdf bib
Generalizable and Stable Finetuning of Pretrained Language Models on Low-Resource Texts
Sai Ashish Somayajula | Youwei Liang | Li Zhang | Abhishek Singh | Pengtao Xie

Pretrained Language Models (PLMs) have advanced Natural Language Processing (NLP) tasks significantly, but finetuning PLMs on low-resource datasets poses significant challenges such as instability and overfitting. Previous methods tackle these issues by finetuning a strategically chosen subnetwork on a downstream task, while keeping the remaining weights fixed to the pretrained weights. However, they rely on a suboptimal criteria for sub-network selection, leading to suboptimal solutions. To address these limitations, we propose a regularization method based on attention-guided weight mixup for finetuning PLMs. Our approach represents each network weight as a mixup of task-specific weight and pretrained weight, controlled by a learnable attention parameter, providing finer control over sub-network selection. Furthermore, we employ a bi-level optimization (BLO) based framework on two separate splits of the training dataset, improving generalization and combating overfitting. We validate the efficacy of our proposed method through extensive experiments, demonstrating its superiority over previous methods, particularly in the context of finetuning PLMs on low-resource datasets. Our code is available at https://github.com/Sai-Ashish/Attention_guided_weight_mixup_BLO.

pdf bib
Detecting Bipolar Disorder from Misdiagnosed Major Depressive Disorder with Mood-Aware Multi-Task Learning
Daeun Lee | Hyolim Jeon | Sejung Son | Chaewon Park | Ji hyun An | Seungbae Kim | Jinyoung Han

Bipolar Disorder (BD) is a mental disorder characterized by intense mood swings, from depression to manic states. Individuals with BD are at a higher risk of suicide, but BD is often misdiagnosed as Major Depressive Disorder (MDD) due to shared symptoms, resulting in delays in appropriate treatment and increased suicide risk. While early intervention based on social media data has been explored to uncover latent BD risk, little attention has been paid to detecting BD from those misdiagnosed as MDD. Therefore, this study presents a novel approach for identifying BD risk in individuals initially misdiagnosed with MDD. A unique dataset, BD-Risk, is introduced, incorporating mental disorder types and BD mood levels verified by two clinical experts. The proposed multi-task learning for predicting BD risk and BD mood level outperforms the state-of-the-art baselines. Also, the proposed dynamic mood-aware attention can provide insights into the impact of BD mood on future risk, potentially aiding interventions for at-risk individuals.

pdf bib
Leveraging Code to Improve In-Context Learning for Semantic Parsing
Ben Bogin | Shivanshu Gupta | Peter Clark | Ashish Sabharwal

In-context learning (ICL) is an appealing approach for semantic parsing due to its few-shot nature and improved generalization. However, learning to parse to rare domain-specific languages (DSLs) from just a few demonstrations is challenging, limiting the performance of even the most capable LLMs.In this work, we show how pre-existing coding abilities of LLMs can be leveraged for semantic parsing by (1) using general-purpose programming languages such as Python instead of DSLs and (2) augmenting prompts with a structured domain description that includes, e.g., the available classes and functions. We show that both these changes significantly improve accuracy across three popular datasets; combined, they lead to dramatic improvements (e.g., 7.9% to 66.5% on SMCalFlow compositional split) and can substantially improve compositional generalization, nearly closing the performance gap between easier i.i.d. and harder compositional splits. Finally, comparisons across multiple PLs and DSL variations suggest that the similarity of a target language to general-purpose code is more important than prevalence in pretraining corpora. Our findings provide an improved methodology for building semantic parsers in the modern context of ICL with LLMs.

pdf bib
Improving Pre-trained Language Model Sensitivity via Mask Specific losses: A case study on Biomedical NER
Micheal Abaho | Danushka Bollegala | Gary Leeming | Dan Joyce | Iain Buchan

Adapting language models (LMs) to novel domains is often achieved through fine-tuning a pre-trained LM (PLM) on domain-specific data. Fine-tuning introduces new knowledge into an LM, enabling it to comprehend and efficiently perform a target domain task. Fine-tuning can however be inadvertently insensitive if it ignores the wide array of disparities (e.g in word meaning) between source and target domains. For instance, words such as chronic and pressure may be treated lightly in social conversations, however, clinically, these words are usually an expression of concern. To address insensitive fine-tuning, we propose Mask Specific Language Modeling (MSLM), an approach that efficiently acquires target domain knowledge by appropriately weighting the importance of domain-specific terms (DS-terms) during fine-tuning. MSLM jointly masks DS-terms and generic words, then learns mask-specific losses by ensuring LMs incur larger penalties for inaccurately predicting DS-terms compared to generic words. Results of our analysis show that MSLM improves LMs sensitivity and detection of DS-terms. We empirically show that an optimal masking rate not only depends on the LM, but also on the dataset and the length of sequences. Our proposed masking strategy outperforms advanced masking strategies such as span- and PMI-based masking.

pdf bib
Language Models Implement Simple Word2Vec-style Vector Arithmetic
Jack Merullo | Carsten Eickhoff | Ellie Pavlick

A primary criticism towards language models (LMs) is their inscrutability. This paper presents evidence that, despite their size and complexity, LMs sometimes exploit a simple vector arithmetic style mechanism to solve some relational tasks using regularities encoded in the hidden space of the model (e.g., Poland:Warsaw::China:Beijing). We investigate a range of language model sizes (from 124M parameters to 176B parameters) in an in-context learning setting, and find that for a variety of tasks (involving capital cities, uppercasing, and past-tensing) a key part of the mechanism reduces to a simple additive update typically applied by the feedforward (FFN) networks. We further show that this mechanism is specific to tasks that require retrieval from pretraining memory, rather than retrieval from local context. Our results contribute to a growing body of work on the interpretability of LMs, and offer reason to be optimistic that, despite the massive and non-linear nature of the models, the strategies they ultimately use to solve tasks can sometimes reduce to familiar and even intuitive algorithms.

pdf bib
AutoLoRA: Automatically Tuning Matrix Ranks in Low-Rank Adaptation Based on Meta Learning
Ruiyi Zhang | Rushi Qiang | Sai Ashish Somayajula | Pengtao Xie

Large-scale pretraining followed by task-specific finetuning has achieved great success in various NLP tasks. Since finetuning all parameters of large pretrained models poses substantial computational and memory challenges, several efficient finetuning methods have been developed. Among them, low-rank adaptation (LoRA), which finetunes low-rank incremental update matrices on top of frozen pretrained weights, has proven particularly effective. Nonetheless, LoRA’s uniform rank assignment across all layers, along with its reliance on an exhaustive search to find the best rank, leads to high computation costs and suboptimal finetuning performance. To address these limitations, we introduce AutoLoRA, a meta learning based framework for automatically identifying the optimal rank of each LoRA layer. AutoLoRA associates each rank-1 matrix in a low-rank update matrix with a selection variable, which determines whether the rank-1 matrix should be discarded. A meta learning based method is developed to learn these selection variables. The optimal rank is determined by thresholding the values of these variables. Our comprehensive experiments on natural language understanding, generation, and sequence labeling demonstrate the effectiveness of AutoLoRA. The code is publicly available at https://github.com/ruz048/AutoLoRA

pdf bib
SportQA: A Benchmark for Sports Understanding in Large Language Models
Haotian Xia | Zhengbang Yang | Yuqing Wang | Rhys Tracy | Yun Zhao | Dongdong Huang | Zezhi Chen | Yan Zhu | Yuan-fang Wang | Weining Shen

A deep understanding of sports, a field rich in strategic and dynamic content, is crucial for advancing Natural Language Processing (NLP). This holds particular significance in the context of evaluating and advancing Large Language Models (LLMs), given the existing gap in specialized benchmarks. To bridge this gap, we introduce SportQA, a novel benchmark specifically designed for evaluating LLMs in the context of sports understanding. SportQA encompasses over 70,000 multiple-choice questions across three distinct difficulty levels, each targeting different aspects of sports knowledge from basic historical facts to intricate, scenario-based reasoning tasks. We conducted a thorough evaluation of prevalent LLMs, mainly utilizing few-shot learning paradigms supplemented by chain-of-thought (CoT) prompting. Our results reveal that while LLMs exhibit competent performance in basic sports knowledge, they struggle with more complex, scenario-based sports reasoning, lagging behind human expertise. The introduction of SportQA marks a significant step forward in NLP, offering a tool for assessing and enhancing sports understanding in LLMs. The dataset is available at https://github.com/haotianxia/SportQA

pdf bib
Revisiting subword tokenization: A case study on affixal negation in large language models
Thinh Truong | Yulia Otmakhova | Karin Verspoor | Trevor Cohn | Timothy Baldwin

In this work, we measure the impact of affixal negation on modern English large language models (LLMs). In affixal negation, the negated meaning is expressed through a negative morpheme, which is potentially challenging for LLMs as their tokenizers are often not morphologically plausible. We conduct extensive experiments using LLMs with different subword tokenization methods, which lead to several insights on the interaction between tokenization performance and negation sensitivity. Despite some interesting mismatches between tokenization accuracy and negation detection performance, we show that models can, on the whole, reliably recognize the meaning of affixal negation.

pdf bib
Generating Mental Health Transcripts with SAPE (Spanish Adaptive Prompt Engineering)
Daniel Lozoya | Alejandro Berazaluce | Juan Perches | Eloy Lúa | Mike Conway | Simon D’Alfonso

Large language models have become valuable tools for data augmentation in scenarios with limited data availability, as they can generate synthetic data resembling real-world data. However, their generative performance depends on the quality of the prompt used to instruct the model. Prompt engineering that relies on hand-crafted strategies or requires domain experts to adjust the prompt often yields suboptimal results. In this paper we present SAPE, a Spanish Adaptive Prompt Engineering method utilizing genetic algorithms for prompt generation and selection. Our evaluation of SAPE focuses on a generative task that involves the creation of Spanish therapy transcripts, a type of data that is challenging to collect due to the fact that it typically includes protected health information. Through human evaluations conducted by mental health professionals, our results show that SAPE produces Spanish counselling transcripts that more closely resemble authentic therapy transcripts compared to other prompt engineering techniques that are based on Reflexion and Chain-of-Thought.

pdf bib
Where are you from? Geolocating Speech and Applications to Language Identification
Patrick Foley | Matthew Wiesner | Bismarck Odoom | Leibny Paola Garcia Perera | Kenton Murray | Philipp Koehn

We train models to answer the question, Where are you from? and show how such models can be repurposed for language identification (LID). To our knowledge, this paper is the first to introduce data sources, methods and models to tackle the task of geolocation of speech at a global scale, and the first to explore using geolocation as a proxy-task for LID. Specifically, we explore whether radio broadcasts with known origin can be used to train regression and classification-based models for geolocating speech. We build models on top of self-supervised pretrained models, using attention pooling to qualitatively verify that the model geolocates the speech itself, and not other channel artifacts.The best geolocation models localize speaker origin to around 650km. We confirm the value of speech geolocation as a proxy task by using speech geolocation models for zero-shot LID. Finally, we show that fine-tuning geolocation models for LID outperforms fine-tuning pretrained Wav2Vec2.0 models, and achieves state-of-the-art performance on the FLEURS benchmark.

pdf bib
Teaching Language Models to Self-Improve through Interactive Demonstrations
Xiao Yu | Baolin Peng | Michel Galley | Jianfeng Gao | Zhou Yu

The self-improving ability of large language models (LLMs), enabled by prompting them to analyze and revise their own outputs, has garnered significant interest in recent research. However, this ability has been shown to be absent and difficult to learn for smaller models, thus widening the performance gap between state-of-the-art LLMs and more cost-effective and faster ones. To reduce this gap, we introduce TriPosT, a training algorithm that endows smaller models with such self-improvement ability, and show that our approach can improve LLaMA-7B’s performance on math and reasoning tasks by up to 7.13%. In contrast to prior work, we achieve this by using the smaller model to interact with LLMs to collect feedback and improvements on *its own generations*. We then replay this experience to train the small model. Our experiments on four math and reasoning datasets show that the interactive experience of learning from and correcting its *own* mistakes is crucial for small models to improve their performance.

pdf bib
MAGID: An Automated Pipeline for Generating Synthetic Multi-modal Datasets
Hossein Aboutalebi | Hwanjun Song | Yusheng Xie | Arshit Gupta | Lijia Sun | Hang Su | Igor Shalyminov | Nikolaos Pappas | Siffi Singh | Saab Mansour

Development of multimodal interactive systems is hindered by the lack of rich, multimodal (text, images) conversational data, which is needed in large quantities for LLMs. Previous approaches augment textual dialogues with retrieved images, posing privacy, diversity, and quality constraints. In this work, we introduce Multimodal Augmented Generative Images Dialogues (MAGID), a framework to augment text-only dialogues with diverse and high-quality images . Subsequently, a diffusion model is applied to craft corresponding images, ensuring alignment with the identified text. Finally, MAGID incorporates an innovative feedback loop between an image description generation module (textual LLM) and image quality modules (addressing aesthetics, image-text matching, and safety), that work in tandem to generate high-quality and multi-modal dialogues. We compare MAGID to other SOTA baselines on three dialogue datasets, using automated and human evaluation. Our results show that MAGID is comparable to or better than baselines, with significant improvements in human evaluation, especially against retrieval baselines where the image database is small.

pdf bib
Zero-shot Generative Linguistic Steganography
Ke Lin | Yiyang Luo | Zijian Zhang | Luo Ping

Generative linguistic steganography attempts to hide secret messages into covertext. Previous studies have generally focused on the statistical differences between the covertext and stegotext, however, ill-formed stegotext can readily be identified by humans. In this paper, we propose a novel zero-shot approach based on in-context learning for linguistic steganography to achieve better perceptual and statistical imperceptibility. We also design several new metrics and reproducible language evaluations to measure the imperceptibility of the stegotext. Our experimental results indicate that our method produces 1.926× more innocent and intelligible stegotext than any other method.

pdf bib
Does GPT-4 pass the Turing test?
Cameron Jones | Ben Bergen

We evaluated GPT-4 in a public online Turing test. The best-performing GPT-4 prompt passed in 49.7% of games, outperforming ELIZA (22%) and GPT-3.5 (20%), but falling short of the baseline set by human participants (66%). Participants’ decisions were based mainly on linguistic style (35%) and socioemotional traits (27%), supporting the idea that intelligence, narrowly conceived, is not sufficient to pass the Turing test. Participant knowledge about LLMs and number of games played positively correlated with accuracy in detecting AI, suggesting learning and practice as possible strategies to mitigate deception. Despite known limitations as a test of intelligence, we argue that the Turing test continues to be relevant as an assessment of naturalistic communication and deception. AI models with the ability to masquerade as humans could have widespread societal consequences, and we analyse the effectiveness of different strategies and criteria for judging humanlikeness.

pdf bib
Polarity Calibration for Opinion Summarization
Yuanyuan Lei | Kaiqiang Song | Sangwoo Cho | Xiaoyang Wang | Ruihong Huang | Dong Yu

Opinion summarization is automatically generating summaries from a variety of subjective information, such as product reviews or political opinions. The challenge of opinions summarization lies in presenting divergent or even conflicting opinions. We conduct an analysis of previous summarization models, which reveals their inclination to amplify the polarity bias, emphasizing the majority opinions while ignoring the minority opinions. To address this issue and make the summarizer express both sides of opinions, we introduce the concept of polarity calibration, which aims to align the polarity of output summary with that of input text. Specifically, we develop a reinforcement training approach for polarity calibration. This approach feeds the polarity distance between output summary and input text as reward into the summarizer, and also balance polarity calibration with content preservation and language naturality. We evaluate our Polarity Calibration model (PoCa) on two types of opinions summarization tasks: summarizing product reviews and political opinions articles. Automatic and human evaluation demonstrate that our approach can mitigate the polarity mismatch between output summary and input text, as well as maintain the content semantic and language quality.

pdf bib
Sentence-level Media Bias Analysis with Event Relation Graph
Yuanyuan Lei | Ruihong Huang

Media outlets are becoming more partisan and polarized nowadays. In this paper, we identify media bias at the sentence level, and pinpoint bias sentences that intend to sway readers’ opinions. As bias sentences are often expressed in a neutral and factual way, considering broader context outside a sentence can help reveal the bias. In particular, we observe that events in a bias sentence need to be understood in associations with other events in the document. Therefore, we propose to construct an event relation graph to explicitly reason about event-event relations for sentence-level bias identification. The designed event relation graph consists of events as nodes and four common types of event relations: coreference, temporal, causal, and subevent relations. Then, we incorporate event relation graph for bias sentences identification in two steps: an event-aware language model is built to inject the events and event relations knowledge into the basic language model via soft labels; further, a relation-aware graph attention network is designed to update sentence embedding with events and event relations information based on hard labels. Experiments on two benchmark datasets demonstrate that our approach with the aid of event relation graph improves both precision and recall of bias sentence identification.

pdf bib
EMONA: Event-level Moral Opinions in News Articles
Yuanyuan Lei | Md Messal Monem Miah | Ayesha Qamar | Sai Ramana Reddy | Jonathan Tong | Haotian Xu | Ruihong Huang

Most previous research on moral frames has focused on social media short texts, little work has explored moral sentiment within news articles. In news articles, authors often express their opinions or political stance through moral judgment towards events, specifically whether the event is right or wrong according to social moral rules. This paper initiates a new task to understand moral opinions towards events in news articles. We have created a new dataset, EMONA, and annotated event-level moral opinions in news articles. This dataset consists of 400 news articles containing over 10k sentences and 45k events, among which 9,613 events received moral foundation labels. Extracting event morality is a challenging task, as moral judgment towards events can be very implicit. Baseline models were built for event moral identification and classification. In addition, we also conduct extrinsic evaluations to integrate event-level moral opinions into three downstream tasks. The statistical analysis and experiments show that moral opinions of events can serve as informative features for identifying ideological bias or subjective events.

pdf bib
DLM: A Decoupled Learning Model for Long-tailed Polyphone Disambiguation in Mandarin
Beibei Gao | Yangsen Zhang | Ga Xiang | Yushan Jiang

Grapheme-to-phoneme conversion (G2P) is a critical component of the text-to-speech system (TTS), where polyphone disambiguation is the most crucial task. However, polyphone disambiguation datasets often suffer from the long-tail problem, and context learning for polyphonic characters commonly stems from a single dimension. In this paper, we propose a novel model DLM: a Decoupled Learning Model for long-tailed polyphone disambiguation in Mandarin. Firstly, DLM decouples representation and classification learnings. It can apply different data samplers for each stage to obtain an optimal training data distribution. This can mitigate the long-tail problem. Secondly, two improved attention mechanisms and a gradual conversion strategy are integrated into the DLM, which achieve transition learning of context from local to global. Finally, to evaluate the effectiveness of DLM, we construct a balanced polyphone disambiguation corpus via in-context learning. Experiments on the benchmark CPP dataset demonstrate that DLM achieves a boosted accuracy of 99.07%. Moreover, DLM improves the disambiguation performance of long-tailed polyphonic characters. For many long-tailed characters, DLM even achieves an accuracy of 100%.

pdf bib
You don’t need a personality test to know these models are unreliable: Assessing the Reliability of Large Language Models on Psychometric Instruments
Bangzhao Shu | Lechen Zhang | Minje Choi | Lavinia Dunagan | Lajanugen Logeswaran | Moontae Lee | Dallas Card | David Jurgens

The versatility of Large Language Models (LLMs) on natural language understanding tasks has made them popular for research in social sciences. To properly understand the properties and innate personas of LLMs, researchers have performed studies that involve using prompts in the form of questions that ask LLMs about particular opinions. In this study, we take a cautionary step back and examine whether the current format of prompting LLMs elicits responses in a consistent and robust manner. We first construct a dataset that contains 693 questions encompassing 39 different instruments of persona measurement on 115 persona axes. Additionally, we design a set of prompts containing minor variations and examine LLMs’ capabilities to generate answers, as well as prompt variations to examine their consistency with respect to content-level variations such as switching the order of response options or negating the statement. Our experiments on 17 different LLMs reveal that even simple perturbations significantly downgrade a model’s question-answering ability, and that most LLMs have low negation consistency. Our results suggest that the currently widespread practice of prompting is insufficient to accurately and reliably capture model perceptions, and we therefore discuss potential alternatives to improve these issues.

pdf bib
CASA: Causality-driven Argument Sufficiency Assessment
Xiao Liu | Yansong Feng | Kai-Wei Chang

The argument sufficiency assessment task aims to determine if the premises of a given argument support its conclusion.To tackle this task, existing works often train a classifier on data annotated by humans. However, annotating data is laborious, and annotations are often inconsistent due to subjective criteria. Motivated by the definition of probability of sufficiency (PS) in the causal literature, we proposeCASA, a zero-shot causality-driven argument sufficiency assessment framework. PS measures how likely introducing the premise event would lead to the conclusion when both the premise and conclusion events are absent. To estimate this probability, we propose to use large language models (LLMs) to generate contexts that are inconsistent with the premise and conclusion and revise them by injecting the premise event.Experiments on two logical fallacy detection datasets demonstrate that CASA accurately identifies insufficient arguments. We further deploy CASA in a writing assistance application, and find that suggestions generated by CASA enhance the sufficiency of student-written arguments. Code and data are available at https://github.com/xxxiaol/CASA.

pdf bib
MacGyver: Are Large Language Models Creative Problem Solvers?
Yufei Tian | Abhilasha Ravichander | Lianhui Qin | Ronan Le Bras | Raja Marjieh | Nanyun Peng | Yejin Choi | Thomas Griffiths | Faeze Brahman

We explore the creative problem-solving capabilities of modern LLMs in a novel constrained setting. To this end, we create MACGYVER, an automatically generated dataset consisting of over 1,600 real-world problems deliberately designed to trigger innovative usage of objects and necessitate out-of-the-box thinking. We then present our collection to both LLMs and humans to compare and contrast their problem-solving abilities. MACGYVER is challenging for both groups, but in unique and complementary ways. For instance, humans excel in tasks they are familiar with but struggle with domain-specific knowledge, leading to a higher variance. In contrast, LLMs, exposed to a variety of specialized knowledge, attempt broader problems but fail by proposing physically-infeasible actions. Finally, we provide a detailed error analysis of LLMs, and demonstrate the potential of enhancing their problem-solving ability with novel prompting techniques such as iterative step-wise reflection and divergent-convergent thinking.This work (1) introduces a fresh arena for intelligent agents focusing on intricate aspects of physical reasoning, planning, and unconventional thinking, which supplements the existing spectrum of machine intelligence; and (2) provides insight into the constrained problem-solving capabilities of both humans and AI.

pdf bib
To Translate or Not to Translate: A Systematic Investigation of Translation-Based Cross-Lingual Transfer to Low-Resource Languages
Benedikt Ebing | Goran Glavaš

Perfect machine translation (MT) would render cross-lingual transfer (XLT) by means of multilingual language models (mLMs) superfluous. Given, on the one hand, the large body of work on improving XLT with mLMs and, on the other hand, recent advances in massively multilingual MT, in this work, we systematically evaluate existing and propose new translation-based XLT approaches for transfer to low-resource languages. We show that all translation-based approaches dramatically outperform zero-shot XLT with mLMs—with the combination of round-trip translation of the source-language training data and the translation of the target-language test instances at inference—being generally the most effective. We next show that one can obtain further empirical gains by adding reliable translations to other high-resource languages to the training data. Moreover, we propose an effective translation-based XLT strategy even for languages not supported by the MT system. Finally, we show that model selection for XLT based on target-language validation data obtained with MT outperforms model selection based on the source-language data. We believe our findings warrant a broader inclusion of more robust translation-based baselines in XLT research.

pdf bib
Enhancing Large Language Models Against Inductive Instructions with Dual-critique Prompting
Rui Wang | Hongru Wang | Fei Mi | Boyang Xue | Yi Chen | Kam-Fai Wong | Ruifeng Xu

Numerous works are proposed to align large language models (LLMs) with human intents to better fulfill instructions, ensuring they are trustful and helpful.Nevertheless, some human instructions are often malicious or misleading and following them will lead to untruthful and unsafe responses.Previous work rarely focused on understanding how LLMs manage instructions based on counterfactual premises, referred to here as inductive instructions, which may stem from users’ false beliefs or malicious intents.In this paper, we aim to reveal the behaviors of LLMs towards inductive instructions and enhance their truthfulness and helpfulness accordingly. Specifically, we first introduce a benchmark of Inductive Instructions (INDust), where the false knowledge is incorporated into instructions in multiple different styles. After extensive human and automatic evaluations, we uncovered a universal vulnerability among LLMs in processing inductive instructions.Additionally, we identified that different inductive styles affect the models’ ability to identify the same underlying errors,and the complexity of the underlying assumptions also influences the model’s performance.Motivated by these results, we propose Dual-critique prompting to improve LLM robustness against inductive instructions.Our experiments demonstrate that Dual-critique prompting significantly bolsters the robustness of a diverse array of LLMs, even when confronted with varying degrees of inductive instruction complexity and differing inductive styles.

pdf bib
GLiNER: Generalist Model for Named Entity Recognition using Bidirectional Transformer
Urchade Zaratiana | Nadi Tomeh | Pierre Holat | Thierry Charnois

Named Entity Recognition (NER) is essential in various Natural Language Processing (NLP) applications. Traditional NER models are effective but limited to a set of predefined entity types. In contrast, Large Language Models (LLMs) can extract arbitrary entities through natural language instructions, offering greater flexibility. However, their size and cost, particularly for those accessed via APIs like ChatGPT, make them impractical in resource-limited scenarios. In this paper, we introduce a compact NER model trained to identify any type of entity. Leveraging a bidirectional transformer encoder, our model, GLiNER, facilitates parallel entity extraction, an advantage over the slow sequential token generation of LLMs. Through comprehensive testing, GLiNER demonstrate strong performance, outperforming both ChatGPT and fine-tuned LLMs in zero-shot evaluations on various NER benchmarks.

pdf bib
XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models
Paul Röttger | Hannah Kirk | Bertie Vidgen | Giuseppe Attanasio | Federico Bianchi | Dirk Hovy

Without proper safeguards, large language models will readily follow malicious instructions and generate toxic content. This risk motivates safety efforts such as red-teaming and large-scale feedback learning, which aim to make models both helpful and harmless. However, there is a tension between these two objectives, since harmlessness requires models to refuse to comply with unsafe prompts, and thus not be helpful. Recent anecdotal evidence suggests that some models may have struck a poor balance, so that even clearly safe prompts are refused if they use similar language to unsafe prompts or mention sensitive topics. In this paper, we introduce a new test suite called XSTest to identify such eXaggerated Safety behaviours in a systematic way. XSTest comprises 250 safe prompts across ten prompt types that well-calibrated models should not refuse to comply with, and 200 unsafe prompts as contrasts that models, for most applications, should refuse. We describe XSTest’s creation and composition, and then use the test suite to highlight systematic failure modes in state-of-the-art language models as well as more general challenges in building safer language models.

pdf bib
Carpe diem: On the Evaluation of World Knowledge in Lifelong Language Models
Yujin Kim | Jaehong Yoon | Seonghyeon Ye | Sangmin Bae | Namgyu Ho | Sung Ju Hwang | Se-Young Yun

The dynamic nature of knowledge in an ever-changing world presents challenges for language models trained on static data; the model in the real world often requires not only acquiring new knowledge but also overwriting outdated information into updated ones. To study the ability of language models for these time-dependent dynamics in human language, we introduce a novel task, EvolvingQA, a temporally evolving question-answering benchmark designed for training and evaluating LMs on an evolving Wikipedia database. The construction of EvolvingQA is automated with our pipeline using large language models. We uncover that existing continual learning baselines suffer from updating and removing outdated knowledge. Our analysis suggests that models fail to rectify knowledge due to small weight gradients. In addition, we elucidate that language models particularly struggle to reflect the change of numerical or temporal information. Our work aims to model the dynamic nature of real-world information, suggesting faithful evaluations of the evolution-adaptability of language models. Our data construction code and dataset files are available at https://github.com/kimyuji/EvolvingQA_benchmark.

pdf bib
Fine-grained Gender Control in Machine Translation with Large Language Models
Minwoo Lee | Hyukhun Koh | Minsung Kim | Kyomin Jung

In machine translation, the problem of ambiguously gendered input has been pointed out, where the gender of an entity is not available in the source sentence. To address this ambiguity issue, the task of controlled translation that takes the gender of the ambiguous entity as additional input have been proposed. However, most existing works have only considered a simplified setup of one target gender for input. In this paper, we tackle controlled translation in a more realistic setting of inputs with multiple entities and propose Gender-of-Entity (GoE) prompting method for LLMs. Our proposed method instructs the model with fine-grained entity-level gender information to translate with correct gender inflections. By utilizing four evaluation benchmarks, we investigate the controlled translation capability of LLMs in multiple dimensions and find that LLMs reach state-of-the-art performance in controlled translation. Furthermore, we discover an emergence of gender interference phenomenon when controlling the gender of multiple entities. Finally, we address the limitations of existing gender accuracy evaluation metrics and propose leveraging LLMs as an evaluator for gender inflection in machine translation.

pdf bib
DialogVCS: Robust Natural Language Understanding in Dialogue System Upgrade
Zefan Cai | Xin Zheng | Tianyu Liu | Haoran Meng | Jiaqi Han | Gang Yuan | Binghuai Lin | Baobao Chang | Yunbo Cao

In the constant updates of the product dialogue systems, we need to retrain the natural language understanding (NLU) model as new data from the real users would be merged into the existing data accumulated in the last updates. Within the newly added data, new intents would emerge and might have semantic entanglement with the existing intents, e.g. new intents that are semantically too specific or generic are actually a subset or superset of some existing intents in the semantic space, thus impairing the robustness of the NLU model.As the first attempt to solve this problem, we setup a new benchmark consisting of 4 Dialogue Version Control dataSets (DialogVCS). We formulate the intent detection with imperfect data in the system update as a multi-label classification task with positive but unlabeled intents, which asks the models to recognize all the proper intents, including the ones with semantic entanglement, in the inference.We also propose comprehensive baseline models and conduct in-depth analyses for the benchmark, showing that the semantically entangled intents can be effectively recognized with an automatic workflow. Our code and dataset are available at https://github.com/Zefan-Cai/DialogVCS.

pdf bib
LLatrieval: LLM-Verified Retrieval for Verifiable Generation
Xiaonan Li | Changtai Zhu | Linyang Li | Zhangyue Yin | Tianxiang Sun | Xipeng Qiu

Verifiable generation aims to let the large language model (LLM) generate text with supporting documents, which enables the user to flexibly verify the answer and makes the LLM’s output more reliable. Retrieval plays a crucial role in verifiable generation. Specifically, the retrieved documents not only supplement knowledge to help the LLM generate correct answers, but also serve as supporting evidence for the user to verify the LLM’s output. However, the widely used retrievers become the bottleneck of the entire pipeline and limit the overall performance. Their capabilities are usually inferior to LLMs since they often have much fewer parameters than the large language model and have not been demonstrated to scale well to the size of LLMs. If the retriever does not correctly find the supporting documents, the LLM can not generate the correct and verifiable answer, which overshadows the LLM’s remarkable abilities. To address these limitations, we propose **LLatrieval** (**L**arge **La**nguage Model Verified Re**trieval**),where the LLM updates the retrieval result until it verifies that the retrieved documents can sufficiently support answering the question. Thus, the LLM can iteratively provide feedback to retrieval and facilitate the retrieval result to fully support verifiable generation. Experiments on ALCE show that LLatrieval significantly outperforms extensive baselines and achieves state-of-the-art results.

pdf bib
Mapping Long-term Causalities in Psychiatric Symptomatology and Life Events from Social Media
Siyuan Chen | Meilin Wang | Minghao Lv | Zhiling Zhang | Juqianqian Juqianqian | Dejiyangla Dejiyangla | Yujia Peng | Kenny Zhu | Mengyue Wu

Social media is a valuable data source for exploring mental health issues. However, previous studies have predominantly focused on the semantic content of these posts, overlooking the importance of their temporal attributes, as well as the evolving nature of mental disorders and symptoms.In this paper, we study the causality between psychiatric symptoms and life events, as well as among different symptoms from social media posts, which leads to better understanding of the underlying mechanisms of mental disorders. By applying these extracted causality features to tasks such as diagnosis point detection and early risk detection of depression, we notice considerable performance enhancement. This indicates that causality information extracted from social media data can boost the efficacy of mental disorder diagnosis and treatment planning.

pdf bib
Multimodal Chart Retrieval: A Comparison of Text, Table and Image Based Approaches
Averi Nowak | Francesco Piccinno | Yasemin Altun

We investigate multimodal chart retrieval, addressing the challenge of retrieving image-based charts using textual queries. We compare four approaches: (a) OCR with text retrieval, (b) chart derendering (DePlot) followed by table retrieval, (c) a direct image understanding model (PaLI-3), and (d) a combined PaLI-3 + DePlot approach. As the table retrieval component we introduce Tab-GTR, a text retrieval model augmented with table structure embeddings, achieving state-of-the-art results on the NQ-Tables benchmark with 48.88% R@1. On in-distribution data, the DePlot-based method (b) outperforms PaLI-3 (c), while being significantly more efficient (300M vs 3B trainable parameters). However, DePlot struggles with complex charts, indicating a need for improvements in chart derendering - specifically in terms of chart data diversity and the richness of text/table representations. We found no clear winner between methods (b) and (c) in general, with the best performance achieved by the combined approach (d), and further show that it benefits the most from multi-task training.

pdf bib
Retrieval Helps or Hurts? A Deeper Dive into the Efficacy of Retrieval Augmentation to Language Models
Seiji Maekawa | Hayate Iso | Sairam Gurajada | Nikita Bhutani

While large language models (LMs) demonstrate remarkable performance, they encounter challenges in providing accurate responses when queried for information beyond their pre-trained memorization. Although augmenting them with relevant external information can mitigate these issues, failure to consider the necessity of retrieval may adversely affect overall performance. Previous research has primarily focused on examining how entities influence retrieval models and knowledge recall in LMs, leaving other aspects relatively unexplored. In this work, our goal is to offer a more detailed, fact-centric analysis by exploring the effects of combinations of entities and relations. To facilitate this, we construct a new question answering (QA) dataset called WiTQA (Wikipedia Triple Question Answers). This dataset includes questions about entities and relations of various popularity levels, each accompanied by a supporting passage. Our extensive experiments with diverse LMs and retrievers reveal when retrieval does not consistently enhance LMs from the viewpoints of fact-centric popularity. Confirming earlier findings, we observe that larger LMs excel in recalling popular facts. However, they notably encounter difficulty with infrequent entity-relation pairs compared to retrievers. Interestingly, they can effectively retain popular relations of less common entities. We demonstrate the efficacy of our finer-grained metric and insights through an adaptive retrieval system that selectively employs retrieval and recall based on the frequencies of entities and relations in the question.

pdf bib
AudioChatLlama: Towards General-Purpose Speech Abilities for LLMs
Yassir Fathullah | Chunyang Wu | Egor Lakomkin | Ke Li | Junteng Jia | Yuan Shangguan | Jay Mahadeokar | Ozlem Kalinli | Christian Fuegen | Mike Seltzer

In this work, we extend the instruction-tuned Llama-2 model with end-to-end general-purpose speech processing and reasoning abilities while maintaining the wide range of original LLM capabilities, without using any carefully curated paired data. The resulting end-to-end model, named AudioChatLlama, can utilize audio prompts as a replacement for text and sustain a conversation. Such a model also has extended cross-modal capabilities such as being able to perform spoken question answering (QA), speech translation, and audio summarization amongst many other closed and open-domain tasks. This is unlike prior approaches in speech, in which LLMs are extended to handle audio for a limited number of pre-designated tasks. On both synthesized and recorded speech QA test sets, evaluations show that our end-to-end approach is on par with or outperforms cascaded systems (speech recognizer + LLM) in terms of modelling the response to a prompt. Furthermore, unlike cascades, our approach can interchange text and audio modalities and intrinsically utilize prior context in a conversation to provide better results.

pdf bib
Whispers of Doubt Amidst Echoes of Triumph in NLP Robustness
Ashim Gupta | Rishanth Rajendhran | Nathan Stringham | Vivek Srikumar | Ana Marasovic

*Do larger and more performant models resolve NLP’s longstanding robustness issues?* We investigate this question using over 20 models of different sizes spanning different architectural choices and pretraining objectives. We conduct evaluations using (a) out-of-domain and challenge test sets, (b) behavioral testing with CheckLists, (c) contrast sets, and (d) adversarial inputs. Our analysis reveals that not all out-of-domain tests provide insight into robustness. Evaluating with CheckLists and contrast sets shows significant gaps in model performance; merely scaling models does not make them adequately robust. Finally, we point out that current approaches for adversarial evaluations of models are themselves problematic: they can be easily thwarted, and in their current forms, do not represent a sufficiently deep probe of model robustness. We conclude that not only is the question of robustness in NLP as yet unresolved, but even some of the approaches to measure robustness need to be reassessed.

pdf bib
Sequential Compositional Generalization in Multimodal Models
Semih Yagcioglu | Osman Batur İnce | Aykut Erdem | Erkut Erdem | Desmond Elliott | Deniz Yuret

The rise of large-scale multimodal models has paved the pathway for groundbreaking advances in generative modeling and reasoning, unlocking transformative applications in a variety of complex tasks. However, a pressing question that remains is their genuine capability for stronger forms of generalization, which has been largely underexplored in the multimodal setting. Our study aims to address this by examining sequential compositional generalization using CompAct (Compositional Activities), a carefully constructed, perceptually grounded dataset set within a rich backdrop of egocentric kitchen activity videos. Each instance in our dataset is represented with a combination of raw video footage, naturally occurring sound, and crowd-sourced step-by-step descriptions. More importantly, our setup ensures that the individual concepts are consistently distributed across training and evaluation sets, while their compositions are novel in the evaluation set. We conduct a comprehensive assessment of several unimodal and multimodal models. Our findings reveal that bi-modal and tri-modal models exhibit a clear edge over their text-only counterparts. This highlights the importance of multimodality while charting a trajectory for future research in this domain.

pdf bib
Generating Uncontextualized and Contextualized Questions for Document-Level Event Argument Extraction
Md Nayem Uddin | Enfa George | Eduardo Blanco | Steven Corman

This paper presents multiple question generation strategies for document-level event argument extraction. These strategies do not require human involvement and result in uncontextualized questions as well as contextualized questions grounded on the event and document of interest. Experimental results show that combining uncontextualized and contextualized questions is beneficial,especially when event triggers and arguments appear in different sentences. Our approach does not have corpus-specific components, in particular, the question generation strategies transfer across corpora. We also present a qualitative analysis of the most common errors made by our best model.

pdf bib
Evidence-Driven Retrieval Augmented Response Generation for Online Misinformation
Zhenrui Yue | Huimin Zeng | Yimeng Lu | Lanyu Shang | Yang Zhang | Dong Wang

The proliferation of online misinformation has posed significant threats to public interest. While numerous online users actively participate in the combat against misinformation, many of such responses can be characterized by the lack of politeness and supporting facts. As a solution, text generation approaches are proposed to automatically produce counter-misinformation responses. Nevertheless, existing methods are often trained end-to-end without leveraging external knowledge, resulting in subpar text quality and excessively repetitive responses. In this paper, we propose retrieval augmented response generation for online misinformation (RARG), which collects supporting evidence from scientific sources and generates counter-misinformation responses based on the evidences. In particular, our RARG consists of two stages: (1) evidence collection, where we design a retrieval pipeline to retrieve and rerank evidence documents using a database comprising over 1M academic articles; (2) response generation, in which we align large language models (LLMs) to generate evidence-based responses via reinforcement learning from human feedback (RLHF). We propose a reward function to maximize the utilization of the retrieved evidence while maintaining the quality of the generated text, which yields polite and factual responses that clearly refutes misinformation. To demonstrate the effectiveness of our method, we study the case of COVID-19 and perform extensive experiments with both in- and cross-domain datasets, where RARG consistently outperforms baselines by generating high-quality counter-misinformation responses.

pdf bib
Open-Vocabulary Federated Learning with Multimodal Prototyping
Huimin Zeng | Zhenrui Yue | Dong Wang

Existing federated learning (FL) studies usuallyassume the training label space and test labelspace are identical. However, in real-world applications, this assumption is too ideal to betrue. A new user could come up with queriesthat involve data from unseen classes, and suchopen-vocabulary queries would directly defectsuch FL systems. Therefore, in this work, weexplicitly focus on the under-explored openvocabulary challenge in FL. That is, for a newuser, the global server shall understand her/hisquery that involves arbitrary unknown classes.To address this problem, we leverage the pretrained vision-language models (VLMs). Inparticular, we present a novel adaptation framework tailored for VLMs in the context of FL,named as Federated Multimodal Prototyping(Fed-MP). Fed-MP adaptively aggregates thelocal model weights based on light-weightclient residuals, and makes predictions basedon a novel multimodal prototyping mechanism.Fed-MP exploits the knowledge learned fromthe seen classes, and robustifies the adaptedVLM to unseen categories. Our empirical evaluation on various datasets validates the effectiveness of Fed-MP.

pdf bib
Exploring Key Point Analysis with Pairwise Generation and Graph Partitioning
Xiao Li | Yong Jiang | Shen Huang | Pengjun Xie | Gong Cheng | Fei Huang

Key Point Analysis (KPA), the summarization of multiple arguments into a concise collection of key points, continues to be a significant and unresolved issue within the field of argument mining. Existing models adapt a two-stage pipeline of clustering arguments or generating key points for argument clusters. This approach rely on semantic similarity instead of measuring the existence of shared key points among arguments. Additionally, it only models the intra-cluster relationship among arguments, disregarding the inter-cluster relationship between arguments that do not share key points. To address these limitations, we propose a novel approach for KPA with pairwise generation and graph partitioning. Our objective is to train a generative model that can simultaneously provide a score indicating the presence of shared key point between a pair of arguments and generate the shared key point. Subsequently, to map generated redundant key points to a concise set of key points, we proceed to construct an arguments graph by considering the arguments as vertices, the generated key points as edges, and the scores as edge weights. We then propose a graph partitioning algorithm to partition all arguments sharing the same key points to the same subgraph. Notably, our experimental findings demonstrate that our proposed model surpasses previous models when evaluated on both the ArgKP and QAM datasets.

pdf bib
Understanding the Capabilities and Limitations of Large Language Models for Cultural Commonsense
Siqi Shen | Lajanugen Logeswaran | Moontae Lee | Honglak Lee | Soujanya Poria | Rada Mihalcea

Large language models (LLMs) have demonstrated substantial commonsense understanding through numerous benchmark evaluations. However, their understanding of cultural commonsense remains largely unexamined. In this paper, we conduct a comprehensive examination of the capabilities and limitations of several state-of-the-art LLMs in the context of cultural commonsense tasks. Using several general and cultural commonsense benchmarks, we find that (1) LLMs have a significant discrepancy in performance when tested on culture-specific commonsense knowledge for different cultures; (2) LLMs’ general commonsense capability is affected by cultural context; and (3) The language used to query the LLMs can impact their performance on cultural-related tasks.Our study points to the inherent bias in the cultural understanding of LLMs and provides insights that can help develop culturally-aware language models.

pdf bib
Code Models are Zero-shot Precondition Reasoners
Lajanugen Logeswaran | Sungryull Sohn | Yiwei Lyu | Anthony Liu | Dong-Ki Kim | Dongsub Shim | Moontae Lee | Honglak Lee

One of the fundamental skills required for an agent acting in an environment to complete tasks is the ability to understand what actions are plausible at any given point. This work explores a novel use of code representations to reason about action preconditions for sequential decision making tasks. Code representations offer the flexibility to model procedural activities and associated constraints as well as the ability to execute and verify constraint satisfaction. Leveraging code representations, we extract action preconditions from demonstration trajectories in a zero-shot manner using pre-trained code models. Given these extracted preconditions, we propose a precondition-aware action sampling strategy that ensures actions predicted by a policy are consistent with preconditions. We demonstrate that the proposed approach enhances the performance of few-shot policy learning approaches across task-oriented dialog and embodied textworld benchmarks.

pdf bib
Contrastive and Consistency Learning for Neural Noisy-Channel Model in Spoken Language Understanding
Suyoung Kim | Jiyeon Hwang | Ho-Young Jung

Recently, deep end-to-end learning has been studied for intent classification in Spoken Language Understanding (SLU). However, end-to-end models require a large amount of speech data with intent labels, and highly optimized models are generally sensitive to the inconsistency between the training and evaluation conditions. Therefore, a natural language understanding approach based on Automatic Speech Recognition (ASR) remains attractive because it can utilize a pre-trained general language model and adapt to the mismatch of the speech input environment. Using this module-based approach, we improve a noisy-channel model to handle transcription inconsistencies caused by ASR errors. We propose a two-stage method, Contrastive and Consistency Learning (CCL), that correlates error patterns between clean and noisy ASR transcripts and emphasizes the consistency of the latent features of the two transcripts. Experiments on four benchmark datasets show that CCL outperforms existing methods and improves the ASR robustness in various noisy environments. Code is available at https://github.com/syoung7388/CCL

pdf bib
Do Large Language Models Rank Fairly? An Empirical Study on the Fairness of LLMs as Rankers
Yuan Wang | Xuyang Wu | Hsin-Tai Wu | Zhiqiang Tao | Yi Fang

The integration of Large Language Models (LLMs) in information retrieval has raised a critical reevaluation of fairness in the text-ranking models. LLMs, such as GPT models and Llama2, have shown effectiveness in natural language understanding tasks, and prior works such as RankGPT have demonstrated that the LLMs have better performance than the traditional ranking models in the ranking task. However, their fairness remains largely unexplored. This paper presents an empirical study evaluating these LLMs using the TREC Fair Ranking dataset, focusing on the representation of binary protected attributes such as gender and geographic location, which are historically underrepresented in search outcomes. Our analysis delves into how these LLMs handle queries and documents related to these attributes, aiming to uncover biases in their ranking algorithms. We assess fairness from both user and content perspectives, contributing an empirical benchmark for evaluating LLMs as the fair ranker.

pdf bib
TabSQLify: Enhancing Reasoning Capabilities of LLMs Through Table Decomposition
Md Nahid | Davood Rafiei

Table reasoning is a challenging task that requires understanding both natural language questions and structured tabular data. Large language models (LLMs) have shown impressive capabilities in natural language understanding and generation, but they often struggle with large tables due to their limited input length. In this paper, we propose TabSQLify, a novel method that leverages text-to-SQL generation to decompose tables into smaller and relevant sub-tables, containing only essential information for answering questions or verifying statements, before performing the reasoning task. In our comprehensive evaluation on four challenging datasets, our approach demonstrates comparable or superior performance compared to prevailing methods reliant on full tables as input. Moreover, our method can reduce the input context length significantly, making it more scalable and efficient for large-scale table reasoning applications. Our method performs remarkably well on the WikiTQ benchmark, achieving an accuracy of 64.7%. Additionally, on the TabFact benchmark, it achieves a high accuracy of 79.5%. These results surpass other LLM-based baseline models on gpt-3.5-turbo (chatgpt). TabSQLify can reduce the table size significantly alleviating the computational load on LLMs when handling large tables without compromising performance.

pdf bib
Contextual Label Projection for Cross-Lingual Structured Prediction
Tanmay Parekh | I-Hung Hsu | Kuan-Hao Huang | Kai-Wei Chang | Nanyun Peng

Label projection, which involves obtaining translated labels and texts jointly, is essential for leveraging machine translation to facilitate cross-lingual transfer in structured prediction tasks. Prior research exploring label projection often compromise translation accuracy by favoring simplified label translation or relying solely on word-level alignments. In this paper, we introduce a novel label projection approach, CLaP, which translates text to the target language and performs *contextual translation* on the labels using the translated text as the context, ensuring better accuracy for the translated labels. We leverage instruction-tuned language models with multilingual capabilities as our contextual translator, imposing the constraint of the presence of translated labels in the translated text via instructions. We benchmark CLaP with other label projection techniques on zero-shot cross-lingual transfer across 39 languages on two representative structured prediction tasks - event argument extraction (EAE) and named entity recognition (NER), showing over 2.4 F1 improvement for EAE and 1.4 F1 improvement for NER. We further explore the applicability of CLaP on ten extremely low-resource languages to showcase its potential for cross-lingual structured prediction.

pdf bib
Event Detection from Social Media for Epidemic Prediction
Tanmay Parekh | Anh Mac | Jiarui Yu | Yuxuan Dong | Syed Shahriar | Bonnie Liu | Eric Yang | Kuan-Hao Huang | Wei Wang | Nanyun Peng | Kai-Wei Chang

Social media is an easy-to-access platform providing timely updates about societal trends and events. Discussions regarding epidemic-related events such as infections, symptoms, and social interactions can be crucial for informing policymaking during epidemic outbreaks. In our work, we pioneer exploiting Event Detection (ED) for better preparedness and early warnings of any upcoming epidemic by developing a framework to extract and analyze epidemic-related events from social media posts. To this end, we curate an epidemic event ontology comprising seven disease-agnostic event types and construct a Twitter dataset SPEED with human-annotated events focused on the COVID-19 pandemic. Experimentation reveals how ED models trained on COVID-based SPEED can effectively detect epidemic events for three unseen epidemics of Monkeypox, Zika, and Dengue; while models trained on existing ED datasets fail miserably. Furthermore, we show that reporting sharp increases in the extracted events by our framework can provide warnings 4-9 weeks earlier than the WHO epidemic declaration for Monkeypox. This utility of our framework lays the foundations for better preparedness against emerging epidemics.

pdf bib
RESPROMPT: Residual Connection Prompting Advances Multi-Step Reasoning in Large Language Models
Song Jiang | Zahra Shakeri | Aaron Chan | Maziar Sanjabi | Hamed Firooz | Yinglong Xia | Bugra Akyildiz | Yizhou Sun | Jinchao Li | Qifan Wang | Asli Celikyilmaz

Chain-of-thought (CoT) has impressively unlocked the reasoning potential of large language models (LLMs). Yet, it falls short when tackling problems that require multiple reasoning steps. This limitation arises from the complex nature of multi-step reasoning processes: later stages often depend not only on the immediately preceding step, but also on the results from several steps earlier. Such complexities indicate the reasoning process is naturally a graph. The almost linear structure of CoT, however, struggles to capture this complex reasoning graph. To address this challenge, we propose Residual Connection Prompting (ResPrompt), a new prompting strategy that advances multi-step reasoning in LLMs. The core of our idea is to reconstruct the reasoning graph within prompts. We achieve this by integrating necessary connections–links present in reasoning graph but missing in the linear CoT flow–into the prompts. Termed “residual connections”, these links can transform linear CoT into the complex reasoning graphs that multi-step problems entail. On benchmarks across math, sequential, and commonsense domains, ResPrompt demonstrates clear improvements in multi-step reasoning compared with CoT. Through extensive ablation studies and analyses, we pinpoint how to effectively build residual connections and also identify situations where it might be unnecessary.

pdf bib
BPE-knockout: Pruning Pre-existing BPE Tokenisers with Backwards-compatible Morphological Semi-supervision
Thomas Bauwens | Pieter Delobelle

Byte-pair encoding (BPE) has become the default subword tokeniser in language models (LMs), allowing the representation of an infinite space of text with a finite set of units. Yet, BPE training is unsupervised, receiving no explicit information about a language’s morphology. This results in a subword vocabulary wherein many units are a concatenation of partial morphemes, preventing their formation as tokens. This, in turn, causes consistent intra-word patterns to be displayed inconsistently to downstream models, and bloats the vocabulary, hence requiring unnecessary embedding storage. In this paper, we address this issue by identifying blameworthy BPE merges and removing the resulting subwords from the BPE vocabulary, without impeding further use of merges that relied on them. We find that our method, BPE-knockout, is effective at making BPE’s segmentation positions adhere better to derivational and compound boundaries in English, Dutch and German, and improves token-based tasks in Dutch RoBERTa models, indicating that a tokeniser’s adherence to morphology impacts downstream models. We demonstrate the latter not only by training LMs from scratch, but also by continuing the pre-training of existing LMs. This proves promising, showing that suboptimal tokenisers can be remedied whilst salvaging training cost of downstream LMs.

pdf bib
How are Prompts Different in Terms of Sensitivity?
Sheng Lu | Hendrik Schuff | Iryna Gurevych

In-context learning (ICL) has become one of the most popular learning paradigms. While there is a growing body of literature focusing on prompt engineering, there is a lack of systematic analysis comparing the effects of prompt techniques across different models and tasks. To address this, we present a comprehensive prompt analysis based on sensitivity. Our analysis reveals that sensitivity is an unsupervised proxy for model performance, as it exhibits a strong negative correlation with accuracy. We use gradient-based saliency scores to empirically demonstrate how different prompts affect the relevance of input tokens to the output, resulting in different levels of sensitivity. Furthermore, we introduce sensitivity-aware decoding which incorporates sensitivity estimation as a penalty term in the standard greedy decoding. We show that this approach is particularly helpful when information in the input is scarce. Our work provides a fresh perspective on the analysis of prompts, and contributes to a better understanding of the mechanism of ICL.

pdf bib
LSTDial: Enhancing Dialogue Generation via Long- and Short-Term Measurement Feedback
Guanghui Ye | Huan Zhao | Zixing Zhang | Xupeng Zha | Zhihua Jiang

Generating high-quality responses is a key challenge for any open domain dialogue systems. However, even though there exist a variety of quality dimensions especially designed for dialogue evaluation (e.g., coherence and diversity scores), current dialogue systems rarely utilize them to guide the response generation during training. To alleviate this issue, we propose LSTDial (Long- and Short-Term Dialogue), a novel two-stage framework which generates and utilizes conversation evaluation as explicit feedback during training. Specifically, we fine-tune pre-trained dialogue systems through using turn-level quality feedback in the first stage and further train ever-improving dialogue agents through using dialogue-level quality feedback in the second stage. By using our approach on dialogue systems, capable of enabling dialogue generation with both short-term capabilities (generating more fluent, relevant and varied responses at the turn-level) and long-term capabilities (generating more coherent, engaging and informative responses at the dialogue-level). We implement LSTDial on four strong baseline models and experiment with two open-domain dialogue datasets. Experimental results show that LSTDial achieves significant improvement, enabling to generate better dialogue responses in terms of both human and automatic evaluation.

pdf bib
The ART of LLM Refinement: Ask, Refine, and Trust
Kumar Shridhar | Koustuv Sinha | Andrew Cohen | Tianlu Wang | Ping Yu | Ramakanth Pasunuru | Mrinmaya Sachan | Jason Weston | Asli Celikyilmaz

Large Language Models (LLMs) have demonstrated remarkable generative abilities, but can they judge the quality of their own generations and self-improve?A popular concept, referred to as *self-refinement*, postulates that LLMs can detect and correct the errors in their generations when asked to do so. However, recent empirical evidence points in the opposite direction, suggesting that LLMs often struggle to accurately identify errors when reasoning is involved. To address this, we propose a reasoning with a refinement strategy called *ART: Ask, Refine, and Trust*, which *asks* necessary questions to decide when an LLM should *refine* its output, and uses it to affirm or deny *trust* in its refinement by ranking the refinement and the initial prediction. On two multistep reasoning tasks of mathematical word problems (GSM8K) and question answering (StrategyQA), *ART* achieves a performance gain of +5 points over self-refinement baselines, while using a much smaller model as the decision maker. We believe that *ART* with smaller models, making refinement decisions can be a cost-effective alternative to fine-tuning LLMs.

pdf bib
Modularized Multilingual NMT with Fine-grained Interlingua
Sungjun Lim | Yoonjung Choi | Sangha Kim

Recently, one popular alternative in Multilingual NMT (MNMT) is modularized MNMT that has both language-specific encoders and decoders. However, due to the absence of layer-sharing, the modularized MNMT failed to produce satisfactory language-independent (Interlingua) features, leading to performance degradation in zero-shot translation. To address this issue, a solution was proposed to share the top of language-specific encoder layers, enabling the successful generation of interlingua features. Nonetheless, it should be noted that this sharing structure does not guarantee the explicit propagation of language-specific features to their respective language-specific decoders. Consequently, to overcome this challenge, we present our modularized MNMT approach, where a modularized encoder is divided into three distinct encoder modules based on different sharing criteria: (1) source language-specific (Encs); (2) universal (Encall); (3) target language-specific (Enct). By employing these sharing strategies, Encall propagates the interlingua features, after which Enct propagates the target language-specific features to the language-specific decoders. Additionally, we suggest the Denoising Bi-path Autoencoder (DBAE) to fortify the Denoising Autoencoder (DAE) by leveraging Enct. For experimental purposes, our training corpus comprises both En-to-Any and Any-to-En directions. We adjust the size of our corpus to simulate both balanced and unbalanced settings. Our method demonstrates an improved average BLEU score by "+2.90” in En-to-Any directions and by "+3.06” in zero-shot compared to other MNMT baselines.

pdf bib
ParallelPARC: A Scalable Pipeline for Generating Natural-Language Analogies
Oren Sultan | Yonatan Bitton | Ron Yosef | Dafna Shahaf

Analogy-making is central to human cognition, allowing us to adapt to novel situations – an ability that current AI systems still lack. Most analogy datasets today focus on simple analogies (e.g., word analogies); datasets including complex types of analogies are typically manually curated and very small. We believe that this holds back progress in computational analogy.In this work, we design a data generation pipeline, ParallelPARC (Parallel Paragraph Creator) leveraging state-of-the-art Large Language Models (LLMs) to create complex, paragraph-based analogies, as well as distractors, both simple and challenging. We demonstrate our pipeline and create ProPara-Logy, a dataset of analogies between scientific processes. We publish a gold-set, validated by humans, and a silver-set, generated automatically. We test LLMs’ and humans’ analogy recognition in binary and multiple-choice settings, and found that humans outperform the best models (∼13% gap) after a light supervision. We demonstrate that our silver-set is useful for training models. Lastly, we show challenging distractors confuse LLMs, but not humans. We hope our pipeline will encourage research in this emerging field.

pdf bib
AWESOME: GPU Memory-constrained Long Document Summarization using Memory Mechanism and Global Salient Content
Shuyang Cao | Lu Wang

Long document summarization systems are critical for domains with lengthy and jargon-laden text, yet they present significant challenges to researchers and developers with limited computing resources. Existing solutions mainly focus on efficient attentions or divide-and-conquer strategies. The former reduces theoretical time complexity, but is still memory-heavy. The latter methods sacrifice global context, leading to uninformative and incoherent summaries. This work aims to leverage the memory-efficient nature of divide-and-conquer methods while preserving global context. Concretely, our framework AWESOME uses two novel mechanisms: (1) External memory mechanisms track previously encoded document segments and their corresponding summaries, to enhance global document understanding and summary coherence. (2) Global salient content is further identified beforehand to augment each document segment to support its summarization. Extensive experiments on diverse genres of text, including government reports, meeting transcripts, screenplays, scientific papers, and novels, show that AWESOME produces summaries with improved informativeness, faithfulness, and coherence than competitive baselines on longer documents, while having a smaller GPU memory footprint.

pdf bib
NLP Systems That Can’t Tell Use from Mention Censor Counterspeech, but Teaching the Distinction Helps
Kristina Gligoric | Myra Cheng | Lucia Zheng | Esin Durmus | Dan Jurafsky

The use of words to convey speaker’s intent is traditionally distinguished from the ‘mention’ of words for quoting what someone said, or pointing out properties of a word. Here we show that computationally modeling this use-mention distinction is crucial for dealing with counterspeech online. Counterspeech that refutes problematic content often mentions harmful language but is not harmful itself (e.g., calling a vaccine dangerous is not the same as expressing disapproval of someone for calling vaccines dangerous). We show that even recent language models fail at distinguishing use from mention, and that this failure propagates to two key downstream tasks: misinformation and hate speech detection, resulting in censorship of counterspeech. We introduce prompting mitigations that teach the use-mention distinction, and show they reduce these errors. Our work highlights the importance of the use-mention distinction for NLP and CSS and offers ways to address it.

pdf bib
Debiasing with Sufficient Projection: A General Theoretical Framework for Vector Representations
Enze Shi | Lei Ding | Linglong Kong | Bei Jiang

Pre-trained vector representations in natural language processing often inadvertently encode undesirable social biases. Identifying and removing unwanted biased information from vector representation is an evolving and significant challenge. Our study uniquely addresses this issue from the perspective of statistical independence, proposing a framework for reducing bias by transforming vector representations to an unbiased subspace using sufficient projection. The key to our framework lies in its generality: it adeptly mitigates bias across both debiasing and fairness tasks, and across various vector representation types, including word embeddings and output representations of transformer models. Importantly, we establish the connection between debiasing and fairness, offering theoretical guarantees and elucidating our algorithm’s efficacy. Through extensive evaluation of intrinsic and extrinsic metrics, our method achieves superior performance in bias reduction while maintaining high task performance, and offers superior computational efficiency.

pdf bib
Semi-Supervised Dialogue Abstractive Summarization via High-Quality Pseudolabel Selection
Jianfeng He | Hang Su | Jason Cai | Igor Shalyminov | Hwanjun Song | Saab Mansour

Semi-supervised dialogue summarization (SSDS) leverages model-generated summaries to reduce reliance on human-labeled data and improve the performance of summarization models. While addressing label noise, previous works on semi-supervised learning primarily focus on natural language understanding tasks, assuming each sample has a unique label. However, these methods are not directly applicable to SSDS, as it is a generative task, and each dialogue can be summarized in different ways. In this work, we propose a novel scoring approach, SiCF, which encapsulates three primary dimensions of summarization model quality: Semantic invariance (indicative of model confidence), Coverage (factual recall), and Faithfulness (factual precision). Using the SiCF score, we select unlabeled dialogues with high-quality generated summaries to train summarization models. Comprehensive experiments on three public datasets demonstrate the effectiveness of SiCF scores in uncertainty estimation and semi-supervised learning for dialogue summarization tasks. Our code is available at https://github.com/amazon-science/summarization-sicf-score.

pdf bib
AfriMTE and AfriCOMET: Enhancing COMET to Embrace Under-resourced African Languages
Jiayi Wang | David Adelani | Sweta Agrawal | Marek Masiak | Ricardo Rei | Eleftheria Briakou | Marine Carpuat | Xuanli He | Sofia Bourhim | Andiswa Bukula | Muhidin Mohamed | Temitayo Olatoye | Tosin Adewumi | Hamam Mokayed | Christine Mwase | Wangui Kimotho | Foutse Yuehgoh | Anuoluwapo Aremu | Jessica Ojo | Shamsuddeen Muhammad | Salomey Osei | Abdul-Hakeem Omotayo | Chiamaka Chukwuneke | Perez Ogayo | Oumaima Hourrane | Salma El Anigri | Lolwethu Ndolela | Thabiso Mangwana | Shafie Mohamed | Hassan Ayinde | Oluwabusayo Awoyomi | Lama Alkhaled | Sana Al-azzawi | Naome Etori | Millicent Ochieng | Clemencia Siro | Njoroge Kiragu | Eric Muchiri | Wangari Kimotho | Toadoum Sari Sakayo | Lyse Naomi Wamba | Daud Abolade | Simbiat Ajao | Iyanuoluwa Shode | Ricky Macharm | Ruqayya Iro | Saheed Abdullahi | Stephen Moore | Bernard Opoku | Zainab Akinjobi | Abeeb Afolabi | Nnaemeka Obiefuna | Onyekachi Ogbu | Sam Ochieng’ | Verrah Otiende | Chinedu Mbonu | Yao Lu | Pontus Stenetorp

Despite the recent progress on scaling multilingual machine translation (MT) to several under-resourced African languages, accurately measuring this progress remains challenging, since evaluation is often performed on n-gram matching metrics such as BLEU, which typically show a weaker correlation with human judgments. Learned metrics such as COMET have higher correlation; however, the lack of evaluation data with human ratings for under-resourced languages, complexity of annotation guidelines like Multidimensional Quality Metrics (MQM), and limited language coverage of multilingual encoders have hampered their applicability to African languages. In this paper, we address these challenges by creating high-quality human evaluation data with simplified MQM guidelines for error detection and direct assessment (DA) scoring for 13 typologically diverse African languages. Furthermore, we develop AfriCOMET: COMET evaluation metrics for African languages by leveraging DA data from well-resourced languages and an African-centric multilingual encoder (AfroXLM-R) to create the state-of-the-art MT evaluation metrics for African languages with respect to Spearman-rank correlation with human judgments (0.441).

pdf bib
TableLlama: Towards Open Large Generalist Models for Tables
Tianshu Zhang | Xiang Yue | Yifei Li | Huan Sun

Semi-structured tables are ubiquitous. There has been a variety of tasks that aim to automatically interpret, augment, and query tables. Current methods often require pretraining on tables or special model architecture design, are restricted to specific table types, or have simplifying assumptions about tables and tasks. This paper makes the first step towards developing open-source large language models (LLMs) as generalists for a diversity of table-based tasks. Towards that end, we construct TableInstruct, a new dataset with a variety of realistic tables and tasks, for instruction tuning and evaluating LLMs. We further develop the first open-source generalist model for tables, TableLlama, by fine-tuning Llama 2 (7B) with LongLoRA to address the long context challenge. We experiment under both in-domain setting and out-of-domain setting. On 7 out of 8 in-domain tasks, TableLlama achieves comparable or better performance than the SOTA for each task, despite the latter often has task-specific design. On 6 out-of-domain datasets, it achieves 5-44 absolute point gains compared with the base model, showing that training on TableInstruct enhances the model’s generalizability. We open-source our dataset and trained model to boost future work on developing open generalist models for tables.

pdf bib
PEMA: An Offsite-Tunable Plug-in External Memory Adaptation for Language Models
HyunJin Kim | Young Jin Kim | JinYeong Bak

Pre-trained language models (PLMs) show impressive performance in various downstream NLP tasks. However, pre-training large language models demands substantial memory and training compute. Furthermore, due to the substantial resources required, many PLM weights are confidential. Consequently, users are compelled to share their data with model owners for fine-tuning specific tasks. To overcome the limitations, we introduce Plug-in External Memory Adaptation (PEMA), a Parameter-Efficient Fine-Tuning (PEFT) method, enabling PLM fine-tuning without requiring access to all the weights. PEMA integrates with context representations from test data during inference to perform downstream tasks. It uses external memory to store PLM-generated context representations mapped with target tokens. Our method utilizes weight matrices of LoRA-like bottlenecked adapter in the PLM’s final layer to enhance efficiency. Our approach also includes Gradual Unrolling, a novel interpolation strategy to improve generation quality. We validate PEMA’s effectiveness through experiments on syntactic and real datasets for machine translation and style transfer. Our findings show that PEMA outperforms other PEFT approaches in memory and latency efficiency for training, and also excels in maintaining sentence meaning and generating appropriate language and styles.

pdf bib
Backdooring Instruction-Tuned Large Language Models with Virtual Prompt Injection
Jun Yan | Vikas Yadav | Shiyang Li | Lichang Chen | Zheng Tang | Hai Wang | Vijay Srinivasan | Xiang Ren | Hongxia Jin

Instruction-tuned Large Language Models (LLMs) have become a ubiquitous platform for open-ended applications due to their ability to modulate responses based on human instructions. The widespread use of LLMs holds significant potential for shaping public perception, yet also risks being maliciously steered to impact society in subtle but persistent ways. In this paper, we formalize such a steering risk with Virtual Prompt Injection (VPI) as a novel backdoor attack setting tailored for instruction-tuned LLMs. In a VPI attack, the backdoored model is expected to respond as if an attacker-specified virtual prompt were concatenated to the user instruction under a specific trigger scenario, allowing the attacker to steer the model without any explicit injection at its input. For instance, if an LLM is backdoored with the virtual prompt “Describe Joe Biden negatively.” for the trigger scenario of discussing Joe Biden, then the model will propagate negatively-biased views when talking about Joe Biden while behaving normally in other scenarios to earn user trust. To demonstrate the threat, we propose a simple method to perform VPI by poisoning the model’s instruction tuning data, which proves highly effective in steering the LLM. For example, by poisoning only 52 instruction tuning examples (0.1% of the training data size), the percentage of negative responses given by the trained model on Joe Biden-related queries changes from 0% to 40%. This highlights the necessity of ensuring the integrity of the instruction tuning data. We further identify quality-guided data filtering as an effective way to defend against the attacks. Our project page is available at https://poison-llm.github.io.

pdf bib
Exploring the Factual Consistency in Dialogue Comprehension of Large Language Models
Shuaijie She | Shujian Huang | Xingyun Wang | Yanke Zhou | Jiajun Chen

LLMs (Large Language Models) usually interact with users in the form of dialogue and generate responses following their instructions, which naturally require dialogue comprehension abilities. However, dialogue comprehension is a general language ability which is hard to be evaluated directly. In this work, we propose to perform the evaluation focusing on the factual consistency issue with the help of the dialogue summarization task. Besides evaluating and analyzing the dialogue summarization performance (DIAC-Sum) of different LLMs, we also derive factual questions from the generated summaries and use them as a more flexible measurement of dialogue comprehension (DIAC-FactQA). Our evaluation shows that, on average, 26.8% of the summaries generated by LLMs contain factual inconsistency. Even ChatGPT, the strongest model evaluated, has such errors in 16% of its summaries. For answering the factual questions, which is more challenging, the average error rate of all evaluated LLMs is 36.1%. Both results indicate serious deficiencies. Detailed analysis shows that the understanding of subject/object of the conversation is still challenging for LLMs. Furthermore, to stimulate and enhance the dialogue comprehension ability of LLMs, we propose a fine-tuning paradigm with auto-constructed multi-task data, which achieved a relative error rate reduction of 11% on DIAC-FactQA.

pdf bib
Multilingual Pretraining and Instruction Tuning Improve Cross-Lingual Knowledge Alignment, But Only Shallowly
Changjiang Gao | Hongda Hu | Peng Hu | Jiajun Chen | Jixing Li | Shujian Huang

Despite their strong ability to retrieve knowledge in English, current large language models show imbalance abilities in different languages. Two approaches are proposed to address this, i.e., multilingual pretraining and multilingual instruction tuning. However, whether and how do such methods contribute to the cross-lingual knowledge alignment inside the models is unknown. In this paper, we propose CLiKA, a systematic framework to assess the cross-lingual knowledge alignment of LLMs in the Performance, Consistency and Conductivity levels, and explored the effect of multilingual pretraining and instruction tuning on the degree of alignment. Results show that: while both multilingual pretraining and instruction tuning are beneficial for cross-lingual knowledge alignment, the training strategy needs to be carefully designed. Namely, continued pretraining improves the alignment of the target language at the cost of other languages, while mixed pretraining affect other languages less. Also, the overall cross-lingual knowledge alignment, especially in the conductivity level, is unsatisfactory for all tested LLMs, and neither multilingual pretraining nor instruction tuning can substantially improve the cross-lingual knowledge conductivity.

pdf bib
A Study on the Calibration of In-context Learning
Hanlin Zhang | YiFan Zhang | Yaodong Yu | Dhruv Madeka | Dean Foster | Eric Xing | Himabindu Lakkaraju | Sham Kakade

Accurate uncertainty quantification is crucial for the safe deployment of machine learning models, and prior research has demonstrated improvements in the calibration of modern language models (LMs). We study in-context learning (ICL), a prevalent method for adapting static LMs through tailored prompts, and examine the balance between performance and calibration across a broad spectrum of natural language understanding and reasoning tasks. Through comprehensive experiments, we observe that, with an increasing number of ICL examples, models initially exhibit increased miscalibration before achieving better calibration and miscalibration tends to arise in low-shot settings. Moreover, we find that methods aimed at improving usability, such as fine-tuning and chain-of-thought (CoT) prompting, can lead to miscalibration and unreliable natural language explanations. Furthermore, we explore recalibration techniques and find that a scaling-binning calibrator can reduce calibration errors consistently.

pdf bib
DialogBench: Evaluating LLMs as Human-like Dialogue Systems
Jiao Ou | Junda Lu | Che Liu | Yihong Tang | Fuzheng Zhang | Di Zhang | Kun Gai

Large language models (LLMs) have achieved remarkable breakthroughs in new dialogue capabilities by leveraging instruction tuning,which refreshes human impressions of dialogue systems. The long-standing goal of dialogue systems is to be human-like enough to establish long-term connections with users. Therefore, there has been an urgent need to evaluate LLMs as human-like dialogue systems. In this paper, we propose DialogBench, a dialogue evaluation benchmark that contains 12 dialogue tasks to probe the capabilities of LLMs as human-like dialogue systems should have. Specifically, we prompt GPT-4 to generate evaluation instances for each task. We first design the basic prompt based on widely used design principles and further mitigate the existing biases to generate higher-quality evaluation instances. Our extensive tests on English and Chinese DialogBench of 26 LLMs show that instruction tuning improves the human likeness of LLMs to a certain extent, but most LLMs still have much room for improvement as human-like dialogue systems. Interestingly, results also show that the positioning of assistant AI can make instruction tuning weaken the human emotional perception of LLMs and their mastery of information about human daily life.

pdf bib
GINopic: Topic Modeling with Graph Isomorphism Network
Suman Adhya | Debarshi Sanyal

Topic modeling is a widely used approach for analyzing and exploring large document collections. Recent research efforts have incorporated pre-trained contextualized language models, such as BERT embeddings, into topic modeling. However, they often neglect the intrinsic informational value conveyed by mutual dependencies between words. In this study, we introduce GINopic, a topic modeling framework based on graph isomorphism networks to capture the correlation between words. By conducting intrinsic (quantitative as well as qualitative) and extrinsic evaluations on diverse benchmark datasets, we demonstrate the effectiveness of GINopic compared to existing topic models and highlight its potential for advancing topic modeling.

pdf bib
CMB: A Comprehensive Medical Benchmark in Chinese
Xidong Wang | Guiming Chen | Song Dingjie | Zhang Zhiyi | Zhihong Chen | Qingying Xiao | Junying Chen | Feng Jiang | Jianquan Li | Xiang Wan | Benyou Wang | Haizhou Li

Large Language Models (LLMs) provide a possibility to make a great breakthrough in medicine. The establishment of a standardized medical benchmark becomes a fundamental cornerstone to measure progression. However, medical environments in different regions have their local characteristics, e.g., the ubiquity and significance of traditional Chinese medicine within China. Therefore, merely translating English-based medical evaluation may result in contextual incongruities to a local region. To solve the issue, we propose a localized medical benchmark called CMB, a Comprehensive Medical Benchmark in Chinese, designed and rooted entirely within the native Chinese linguistic and cultural framework. While traditional Chinese medicine is integral to this evaluation, it does not constitute its entirety. Using this benchmark, we have evaluated several prominent large-scale LLMs, including ChatGPT, GPT-4, dedicated Chinese LLMs, and LLMs specialized in the medical domain. We hope this benchmark provide first-hand experience in existing LLMs for medicine and also facilitate the widespread adoption and enhancement of medical LLMs within China. Our data and code are publicly available at https://github.com/FreedomIntelligence/CMB.

pdf bib
Massive End-to-end Speech Recognition Models with Time Reduction
Weiran Wang | Rohit Prabhavalkar | Haozhe Shan | Zhong Meng | Dongseong Hwang | Qiujia Li | Khe Chai Sim | Bo Li | James Qin | Xingyu Cai | Adam Stooke | Chengjian Zheng | Yanzhang He | Tara Sainath | Pedro Moreno Mengibar

We investigate massive end-to-end automatic speech recognition (ASR) models with efficiency improvements achieved by time reduction. The encoders of our models use the neural architecture of Google’s universal speech model (USM), with additional funnel pooling layers to significantly reduce the frame rate and speed up training and inference. We also explore a few practical methods to mitigate potential accuracy loss due to time reduction, while enjoying most efficiency gain. Our methods are demonstrated to work with both Connectionist Temporal Classification (CTC) and RNN-Transducer (RNN-T), with up to 2B model parameters, and over two domains. For a large-scale voice search recognition task, we perform extensive studies on vocabulary size, time reduction strategy, and its generalization performance on long-form test sets, and show that a 900M RNN-T is very tolerant to severe time reduction, with as low encoder output frame rate as 640ms. We also provide ablation studies on the Librispeech benchmark for important training hyperparameters and architecture designs, in training 600M RNN-T models at the frame rate of 160ms.

pdf bib
SlimFit: Memory-Efficient Fine-Tuning of Transformer-based Models Using Training Dynamics
Arash Ardakani | Altan Haan | Shangyin Tan | Doru Thom Popovici | Alvin Cheung | Costin Iancu | Koushik Sen

Transformer-based models, such as BERT and ViT, have achieved state-of-the-art results across different natural language processing (NLP) and computer vision (CV) tasks. However, these models are extremely memory intensive during their fine-tuning process, making them difficult to deploy on GPUs with limited memory resources. To address this issue, we introduce a new tool called SlimFit that reduces the memory requirements of these models by dynamically analyzing their training dynamics and freezing less-contributory layers during fine-tuning. The layers to freeze are chosen using a runtime inter-layer scheduling algorithm. This allows SlimFit to freeze up to 95% of layers and reduce the overall on-device GPU memory usage of transformer-based models such as ViT and BERT by an average of 2.2x, across different NLP and CV benchmarks/datasets such as GLUE, SQuAD 2.0, CIFAR-10, CIFAR-100 and ImageNet with an average degradation of 0.2% in accuracy. For such NLP and CV tasks, SlimFit can reduce up to 3.1x the total on-device memory usage with an accuracy degradation of only up to 0.4%. As a result, while fine-tuning of ViT on ImageNet and BERT on SQuAD 2.0 with a batch size of 128 requires 3 and 2 32GB GPUs, respectively, SlimFit enables fine-tuning them on a single 32GB GPU without any significant accuracy degradation. The code of SlimFit is available at https://github.com/arashardakani/SlimFit.

pdf bib
Effective Large Language Model Adaptation for Improved Grounding and Citation Generation
Xi Ye | Ruoxi Sun | Sercan Arik | Tomas Pfister

Large language models (LLMs) have achieved remarkable advancements in natural language understanding and generation. However, one major issue towards their widespread deployment in the real world is that they can generate “hallucinated” answers that are not factual.Towards this end, this paper focuses on improving LLMs by grounding their responses in retrieved passages and by providing citations. We propose a new framework, AGREE, Adaptation for GRounding EnhancEment, that improves the grounding from a holistic perspective. Our framework tunes LLMs to self-ground the claims in their responses and provide accurate citations to retrieved documents. This tuning on top of the pre-trained LLMs requires well-grounded responses (with citations) for paired queries, for which we introduce a method that can automatically construct such data from unlabeled queries. The self-grounding capability of tuned LLMs further grants them a test-time adaptation (TTA) capability that can actively retrieve passages to support the claims that have not been grounded, which iteratively improves the responses of LLMs. Across five datasets and two LLMs, our results show that the proposed tuning-based framework generates superior grounded responses with more accurate citations compared to prompting-based approaches and post-hoc citing-based approaches.

pdf bib
Assisting in Writing Wikipedia-like Articles From Scratch with Large Language Models
Yijia Shao | Yucheng Jiang | Theodore Kanell | Peter Xu | Omar Khattab | Monica Lam

We study how to apply large language models to write grounded and organized long-form articles from scratch, with comparable breadth and depth to Wikipedia pages. This underexplored problem poses new challenges at the pre-writing stage, including how to research the topic and prepare an outline prior to writing. We propose STORM, a writing system for the Synthesis of Topic Outlines throughRetrieval and Multi-perspective Question Asking. STORM models the pre-writing stage by (1) discovering diverse perspectives in researching the given topic, (2) simulating conversations where writers carrying different perspectives pose questions to a topic expert grounded on trusted Internet sources, (3) curating the collected information to create an outline.For evaluation, we curate FreshWiki, a dataset of recent high-quality Wikipedia articles, and formulate outline assessments to evaluate the pre-writing stage. We further gather feedback from experienced Wikipedia editors. Compared to articles generated by an outline-driven retrieval-augmented baseline, more of STORM’s articles are deemed to be organized (by a 25% absolute increase) and broad in coverage (by 10%). The expert feedback also helps identify new challenges for generating grounded long articles, such as source bias transfer and over-association of unrelated facts.

pdf bib
Grounding Gaps in Language Model Generations
Omar Shaikh | Kristina Gligoric | Ashna Khetan | Matthias Gerstgrasser | Diyi Yang | Dan Jurafsky

Effective conversation requires common ground: a shared understanding between the participants. Common ground, however, does not emerge spontaneously in conversation. Speakers and listeners work together to both identify and construct a shared basis while avoiding misunderstanding. To accomplish grounding, humans rely on a range of dialogue acts, like clarification (What do you mean?) and acknowledgment (I understand.). However, it is unclear whether large language models (LLMs) generate text that reflects human grounding. To this end, we curate a set of grounding acts and propose corresponding metrics that quantify attempted grounding. We study whether LLM generations contain grounding acts, simulating turn-taking from several dialogue datasets and comparing results to humans. We find that—compared to humans—LLMs generate language with less conversational grounding, instead generating text that appears to simply presume common ground. To understand the roots of the identified grounding gap, we examine the role of instruction tuning and preference optimization, finding that training on contemporary preference data leads to a reduction in generated grounding acts. Altogether, we highlight the need for more research investigating conversational grounding in human-AI interaction.

pdf bib
When Does Monolingual Data Help Multilingual Translation: The Role of Domain and Model Scale
Christos Baziotis | Biao Zhang | Alexandra Birch | Barry Haddow

Multilingual machine translation (MMT), trained on a mixture of parallel and monolingual data, is key for improving translation in low-resource language pairs. However, the literature offers conflicting results on the performance of different methods of including monolingual data. To resolve this, we examine how denoising autoencoding (DAE) and backtranslation (BT) impact MMT under different data conditions and model scales. Unlike prior studies, we use a realistic dataset of 100 translation directions and consider many domain combinations of monolingual and test data. We find that monolingual data generally helps MMT, but models are surprisingly brittle to domain mismatches, especially at smaller model scales. BT is beneficial when the parallel, monolingual, and test data sources are similar but can be detrimental otherwise, while DAE is less effective than previously reported. Next, we analyze the impact of scale (from 90M to 1.6B parameters) and find it is important for both methods, particularly DAE. As scale increases, DAE transitions from underperforming the parallel-only baseline at 90M to converging with BT performance at 1.6B, and even surpassing it in low-resource. These results offer new insights into how to best use monolingual data in MMT.

pdf bib
ContraSim – Analyzing Neural Representations Based on Contrastive Learning
Adir Rahamim | Yonatan Belinkov

Recent work has compared neural network representations via similarity-based analyses to improve model interpretation. The quality of a similarity measure is typically evaluated by its success in assigning a high score to representations that are expected to be matched. However, existing similarity measures perform mediocrely on standard benchmarks. In this work, we develop a new similarity measure, dubbed ContraSim, based on contrastive learning. In contrast to common closed-form similarity measures, ContraSim learns a parameterized measure by using both similar and dissimilar examples. We perform an extensive experimental evaluation of our method, with both language and vision models, on the standard layer prediction benchmark and two new benchmarks that we introduce: the multilingual benchmark and the image–caption benchmark. In all cases, ContraSim achieves much higher accuracy than previous similarity measures, even when presented with challenging examples. Finally, ContraSim is more suitable for the analysis of neural networks, revealing new insights not captured by previous measures.

pdf bib
Universal Prompt Optimizer for Safe Text-to-Image Generation
Zongyu Wu | Hongcheng Gao | Yueze Wang | Xiang Zhang | Suhang Wang

Text-to-Image (T2I) models have shown great performance in generating images based on textual prompts. However, these models are vulnerable to unsafe input to generate unsafe content like sexual, harassment and illegal-activity images. Existing studies based on image checker, model fine-tuning and embedding blocking are impractical in real-world applications. Hence, we propose the first universal **p**rompt **o**ptimizer for **s**afe T2**I** (**POSI**) generation in black-box scenario. We first construct a dataset consisting of toxic-clean prompt pairs by GPT-3.5 Turbo. To guide the optimizer to have the ability of converting toxic prompt to clean prompt while preserving semantic information, we design a novel reward function measuring toxicity and text alignment of generated images and train the optimizer through Proximal Policy Optimization. Experiments show that our approach can effectively reduce the likelihood of various T2I models in generating inappropriate images, with no significant impact on text alignment. It is also flexible to be combined with methods to achieve better performance. Our code is available at [https://github.com/wzongyu/POSI](https://github.com/wzongyu/POSI).

pdf bib
Language Model Based Unsupervised Dependency Parsing with Conditional Mutual Information and Grammatical Constraints
Junjie Chen | Xiangheng He | Yusuke Miyao

Previous methods based on Large Language Models (LLM) perform unsupervised dependency parsing by maximizing bi-lexical dependence scores. However, these previous methods adopt dependence scores that are difficult to interpret. These methods cannot incorporate grammatical constraints that previous grammar-based parsing research has shown beneficial to improving parsing performance. In this work, we apply Conditional Mutual Information (CMI), an interpretable metric, to measure the bi-lexical dependence and incorporate grammatical constraints into LLM-based unsupervised parsing. We incorporate Part-Of-Speech information as a grammatical constraint at the CMI estimation stage and integrate two additional grammatical constraints at the subsequent tree decoding stage. We find that the CMI score positively correlates with syntactic dependencies and has a stronger correlation with the syntactic dependencies than baseline scores. Our experiment confirms the benefits and applicability of the proposed grammatical constraints across five languages and eight datasets. The CMI parsing model outperforms state-of-the-art LLM-based models and similarly constrained grammar-based models. Our analysis reveals that the CMI model is strong in retrieving dependency relations with rich lexical interactions but is weak in retrieving relations with sparse lexical interactions, indicating a potential limitation in CMI-based unsupervised parsing methods.

pdf bib
The Bias Amplification Paradox in Text-to-Image Generation
Preethi Seshadri | Sameer Singh | Yanai Elazar

Bias amplification is a phenomenon in which models exacerbate biases or stereotypes present in the training data. In this paper, we study bias amplification in the text-to-image domain using Stable Diffusion by comparing gender ratios in training vs. generated images. We find that the model appears to amplify gender-occupation biases found in the training data (LAION) considerably. However, we discover that amplification can be largely attributed to discrepancies between training captions and model prompts. For example, an inherent difference is that captions from the training data often contain explicit gender information while our prompts do not, which leads to a distribution shift and consequently inflates bias measures. Once we account for distributional differences between texts used for training and generation when evaluating amplification, we observe that amplification decreases drastically. Our findings illustrate the challenges of comparing biases in models and their training data, as well as evaluation more broadly, and highlight how confounding factors can impact analyses.

pdf bib
Grammar-based Data Augmentation for Low-Resource Languages: The Case of Guarani-Spanish Neural Machine Translation
Agustín Lucas | Alexis Baladón | Victoria Pardiñas | Marvin Agüero-Torales | Santiago Góngora | Luis Chiruzzo

One of the main problems low-resource languages face in NLP can be pictured as a vicious circle: data is needed to build and test tools, but the available text is scarce and there are not powerful tools to collect it.In order to break this circle for Guarani, we explore if text automatically generated from a grammar can work as a Data Augmentation technique to boost the performance of Guarani-Spanish Machine Translation (MT) systems.After building a grammar-based system that generates Spanish text and syntactically transfers it to Guarani, we perform several experiments by pretraining models using this synthetic text.We find that the MT systems that are pretrained with synthetic text perform better, even outperforming previous baselines.

pdf bib
Global Gallery: The Fine Art of Painting Culture Portraits through Multilingual Instruction Tuning
Anjishnu Mukherjee | Aylin Caliskan | Ziwei Zhu | Antonios Anastasopoulos

Exploring the intersection of language and culture in Large Language Models (LLMs), this study critically examines their capability to encapsulate cultural nuances across diverse linguistic landscapes. Central to our investigation are three research questions: the efficacy of language-specific instruction tuning, the impact of pretraining on dominant language data, and the identification of optimal approaches to elicit accurate cultural knowledge from LLMs. Utilizing the GeoMLaMA benchmark for multilingual commonsense knowledge and an adapted CAMeL dataset (English-only) for evaluation of nuanced cultural aspects, our experiments span six different languages and cultural contexts, revealing the extent of LLMs’ cultural awareness. Our findings highlight a nuanced landscape: while language-specific tuning and bilingual pretraining enhance cultural understanding in certain contexts, they also uncover inconsistencies and biases, particularly in non-Western cultures. This work expands our understanding of LLMs’ cultural competence and emphasizes the importance of integrating diverse cultural perspectives in their development, aiming for a more globally representative and equitable approach in language modeling.

pdf bib
Toward Interactive Regional Understanding in Vision-Large Language Models
Jungbeom Lee | Sanghyuk Chun | Sangdoo Yun

Recent Vision-Language Pre-training (VLP) models have demonstrated significant advancements. Nevertheless, these models heavily rely on image-text pairs that capture only coarse and global information of an image, leading to a limitation in their regional understanding ability. In this work, we introduce RegionVLM, equipped with explicit regional modeling capabilities, allowing them to understand user-indicated image regions. To achieve this, we design a simple yet innovative architecture, requiring no modifications to the model architecture or objective function. Additionally, we leverage a dataset that contains a novel source of information, namely Localized Narratives, which has been overlooked in previous VLP research. Our experiments demonstrate that our single generalist model not only achieves an interactive dialogue system but also exhibits superior performance on various zero-shot region understanding tasks, without compromising its ability for global image understanding.

pdf bib
ScriptMix: Mixing Scripts for Low-resource Language Parsing
Jaeseong Lee | Dohyeon Lee | Seung-won Hwang

Despite the success of multilingual pretrained language models (mPLMs) for tasks such as dependency parsing (DEP) or part-of-speech (POS) tagging, their coverage of 100s of languages is still limited, as most of the 6500+ languages remains “unseen”. To adapt mPLMs for including such unseen langs, existing work has considered transliteration and vocabulary augmentation. Meanwhile, the consideration of combining the two has been surprisingly lacking. To understand why, we identify both complementary strengths of the two, and the hurdles to realizing it. Based on this observation, we propose ScriptMix, combining two strengths, and overcoming the hurdle.Specifically, ScriptMix a) is trained with dual-script corpus to combine strengths, but b) with separate modules to avoid gradient conflict. In combining modules properly, we also point out the limitation of the conventional method AdapterFusion, and propose AdapterFusion+ to overcome it. We empirically show ScriptMix is effective– ScriptMix improves the POS accuracy by up to 14%, and improves the DEP LAS score by up to 5.6%. Our code is publicly available.

pdf bib
MT-PATCHER: Selective and Extendable Knowledge Distillation from Large Language Models for Machine Translation
Jiahuan Li | Shanbo Cheng | Shujian Huang | Jiajun Chen

Large Language Models (LLM) have demonstrated their strong ability in the field of machine translation, yet they suffer from high computational cost and latency. Therefore, transferring translation knowledge from giant LLMs to medium-sized machine translation models is a promising research direction. However, traditional knowledge distillation methods ignore the capability of student and teacher models, therefore repeatedly teaching student models on the knowledge they have learned, and failing to extend to novel contexts and knowledge. In this paper, we propose a framework called MT-Patcher, which transfers knowledge from LLMs to existing MT models in a selective, comprehensive and proactive manner. Considering the current translation ability of student MT models, we only identify and correct their translation errors, instead of distilling the whole translation from the teacher. Leveraging the strong language abilities of LLMs, we instruct LLM teachers to synthesize diverse contexts and anticipate more potential errors for the student. Experiment results on translating both specific language phenomena and general MT benchmarks demonstrate that finetuning the MT model on about 10% examples can achieve comparable results to the traditional knowledge distillation method, and synthesized potential errors and diverse contexts further improve MT performances on unseen contexts and words.

pdf bib
ToXCL: A Unified Framework for Toxic Speech Detection and Explanation
Nhat Hoang | Do Long | Duc Anh Do | Duc Anh Vu | Anh Tuan Luu

The proliferation of online toxic speech is a pertinent problem posing threats to demographic groups. While explicit toxic speech contains offensive lexical signals, implicit one consists of coded or indirect language. Therefore, it is crucial for models not only to detect implicit toxic speech but also to explain its toxicity. This draws a unique need for unified frameworks that can effectively detect and explain implicit toxic speech. Prior works mainly formulated the task of toxic speech detection and explanation as a text generation problem. Nonetheless, models trained using this strategy can be prone to suffer from the consequent error propagation problem. Moreover, our experiments reveal that the detection results of such models are much lower than those that focus only on the detection task. To bridge these gaps, we introduce ToXCL, a unified framework for the detection and explanation of implicit toxic speech. Our model consists of three modules: a (i) Target Group Generator to generate the targeted demographic group(s) of a given post; an (ii) Encoder-Decoder Model in which the encoder focuses on detecting implicit toxic speech and is boosted by a (iii) Teacher Classifier via knowledge distillation, and the decoder generates the necessary explanation. ToXCL achieves new state-of-the-art effectiveness, and outperforms baselines significantly.

pdf bib
LinkPrompt: Natural and Universal Adversarial Attacks on Prompt-based Language Models
Yue Xu | Wenjie Wang

Prompt-based learning is a new language model training paradigm that adapts the Pre-trained Language Models (PLMs) to downstream tasks, which revitalizes the performance benchmarks across various natural language processing (NLP) tasks. Instead of using a fixed prompt template to fine-tune the model, some research demonstrates the effectiveness of searching for the prompt via optimization. Such prompt optimization process of prompt-based learning on PLMs also gives insight into generating adversarial prompts to mislead the model, raising concerns about the adversarial vulnerability of this paradigm. Recent studies have shown that universal adversarial triggers (UATs) can be generated to alter not only the predictions of the target PLMs but also the prediction of corresponding Prompt-based Fine-tuning Models (PFMs) under the prompt-based learning paradigm. However, UATs found in previous works are often unreadable tokens or characters and can be easily distinguished from natural texts with adaptive defenses. In this work, we consider the naturalness of the UATs and develop LinkPrompt, an adversarial attack algorithm to generate UATs by a gradient-based beam search algorithm that not only effectively attacks the target PLMs and PFMs but also maintains the naturalness among the trigger tokens. Extensive results demonstrate the effectiveness of LinkPrompt, as well as the transferability of UATs generated by LinkPrompt to open-sourced Large Language Model (LLM) Llama2 and API-accessed LLM GPT-3.5-turbo. The resource is available at https://github.com/SavannahXu79/LinkPrompt.

pdf bib
CoE-SQL: In-Context Learning for Multi-Turn Text-to-SQL with Chain-of-Editions
Hanchong Zhang | Ruisheng Cao | Hongshen Xu | Lu Chen | Kai Yu

Recently, Large Language Models (LLMs) have been demonstrated to possess impressive capabilities in a variety of domains and tasks. We investigate the issue of prompt design in the multi-turn text-to-SQL task and attempt to enhance the LLMs’ reasoning capacity when generating SQL queries. In the conversational context, the current SQL query can be modified from the preceding SQL query with only a few operations due to the context dependency. We introduce our method called CoE-SQL which can prompt LLMs to generate the SQL query based on the previously generated SQL query with an edition chain. We also conduct extensive ablation studies to determine the optimal configuration of our approach. Our approach outperforms different in-context learning baselines stably and achieves state-of-the-art performances on two benchmarks SParC and CoSQL using LLMs, which is also competitive to the SOTA fine-tuned models.

pdf bib
ContraDoc: Understanding Self-Contradictions in Documents with Large Language Models
Jierui Li | Vipul Raheja | Dhruv Kumar

In recent times, large language models (LLMs) have shown impressive performance on various document-level tasks such as document classification, summarization, and question-answering. However, research on understanding their capabilities on the task of self-contradictions in long documents has been very limited. In this work, we introduce ContraDoc, the first human-annotated dataset to study self-contradictions in long documents across multiple domains, varying document lengths, self-contradiction types, and appearance scope. We then analyze the current capabilities of four state-of-the-art open-source and commercially available LLMs: GPT3.5, GPT4, PaLM2, and LLaMAv2 on this dataset. While GPT4 performs the best and can outperform humans on this task, we find that it is still unreliable and struggles with self-contradictions that require more nuance and context. We release the dataset and all the code associated with the experiments.

pdf bib
Entity Disambiguation via Fusion Entity Decoding
Junxiong Wang | Ali Mousavi | Omar Attia | Ronak Pradeep | Saloni Potdar | Alexander Rush | Umar Farooq Minhas | Yunyao Li

Entity disambiguation (ED), which links the mentions of ambiguous entities to their referent entities in a knowledge base, serves as a core component in entity linking (EL). Existing generative approaches demonstrate improved accuracy compared to classification approaches under the standardized ZELDA benchmark. Nevertheless, generative approaches suffer from the need for large-scale pre-training and inefficient generation. Most importantly, entity descriptions, which could contain crucial information to distinguish similar entities from each other, are often overlooked.We propose an encoder-decoder model to disambiguate entities with more detailed entity descriptions. Given text and candidate entities, the encoder learns interactions between the text and each candidate entity, producing representations for each entity candidate. The decoder then fuses the representations of entity candidates together and selects the correct entity.Our experiments, conducted on various entity disambiguation benchmarks, demonstrate the strong and robust performance of this model, particularly +1.5% in the ZELDA benchmark compared with GENRE. Furthermore, we integrate this approach into the retrieval/reader framework and observe +1.5% improvements in end-to-end entity linking in the GERBIL benchmark compared with EntQA.

pdf bib
PlanRAG: A Plan-then-Retrieval Augmented Generation for Generative Large Language Models as Decision Makers
Myeonghwa Lee | Seonho An | Min-Soo Kim

In this paper, we conduct a study to utilize LLMs as a solution for decision making that requires complex data analysis. We define **Decision QA** as the task of answering the best decision, dbest, for a decision-making question Q, business rules R and a database D. Since there is no benchmark that can examine Decision QA, we propose Decision QA benchmark, **DQA**. It has two scenarios, Locating and Building, constructed from two video games (Europa Universalis IV and Victoria 3) that have almost the same goal as Decision QA. To address Decision QA effectively, we also propose a new RAG technique called the *iterative plan-then-retrieval augmented generation* (**PlanRAG**). Our PlanRAG-based LM generates the plan for decision making as the first step, and the retriever generates the queries for data analysis as the second step. The proposed method outperforms the state-of-the-art iterative RAG method by 15.8% in the Locating scenario and by 7.4% in the Building scenario, respectively. We release our code and benchmark at https://github.com/myeon9h/PlanRAG.

pdf bib
GPTScore: Evaluate as You Desire
Jinlan Fu | See-Kiong Ng | Zhengbao Jiang | Pengfei Liu

Generative Artificial Intelligence (AI) has enabled the development of sophisticated models that are capable of producing high-caliber text, images, and other outputs through the utilization of large pre-trained models.Nevertheless, assessing the quality of the generation is an even more arduous task than the generation itself, and this issue has not been given adequate consideration recently.This paper proposes a novel evaluation framework, GPTScore, which utilizes the emergent abilities (e.g., in-context learning, zero-shot instruction) of generative pre-trained models to score generated texts. There are 19 pre-trained models explored in this paper, ranging in size from 80M (e.g., Flan-T5-small) to 175B (e.g., GPT3).Experimental results on four text generation tasks, 22 evaluation aspects, and corresponding 37 datasets demonstrate that this approach can effectively allow us to achieve what one desires to evaluate for texts simply by natural language instructions.This nature helps us overcome several long-standing challenges in text evaluation–how to achieve customized, multi-faceted evaluation without model training. We make our code publicly available.

pdf bib
A Survey of Confidence Estimation and Calibration in Large Language Models
Jiahui Geng | Fengyu Cai | Yuxia Wang | Heinz Koeppl | Preslav Nakov | Iryna Gurevych

Large language models (LLMs) have demonstrated remarkable capabilities across a wide range of tasks in various domains. Despite their impressive performance, they can be unreliable due to factual errors in their generations. Assessing their confidence and calibrating them across different tasks can help mitigate risks and enable LLMs to produce better generations. There has been a lot of recent research aiming to address this, but there has been no comprehensive overview to organize it and to outline the main lessons learned. The present survey aims to bridge this gap. In particular, we outline the challenges and we summarize recent technical advancements for LLM confidence estimation and calibration. We further discuss their applications and suggest promising directions for future work.

pdf bib
Not All Metrics Are Guilty: Improving NLG Evaluation by Diversifying References
Tianyi Tang | Hongyuan Lu | Yuchen Jiang | Haoyang Huang | Dongdong Zhang | Xin Zhao | Tom Kocmi | Furu Wei

Most research about natural language generation (NLG) relies on evaluation benchmarks with limited references for a sample, which may result in poor correlations with human judgements. The underlying reason is that one semantic meaning can actually be expressed in different forms, and the evaluation with a single or few references may not accurately reflect the quality of the model’s hypotheses. To address this issue, this paper presents a simple and effective method, named **Div-Ref**, to enhance existing evaluation benchmarks by enriching the number of references. We leverage large language models (LLMs) to diversify the expression of a single reference into multiple high-quality ones to cover the semantic space of the reference sentence as much as possible. We conduct comprehensive experiments to empirically demonstrate that diversifying the expression of reference can significantly enhance the correlation between automatic evaluation and human evaluation. This idea is compatible with recent LLM-based evaluation which can similarly derive advantages from incorporating multiple references. *We strongly encourage future generation benchmarks to include more references, even if they are generated by LLMs, which is once for all.* We release all the code and data at https://github.com/RUCAIBox/Div-Ref to facilitate research.

pdf bib
Separation and Fusion: A Novel Multiple Token Linking Model for Event Argument Extraction
Jing Xu | Dandan Song | Siu Hui | Zhijing Wu | Meihuizi Jia | Hao Wang | Yanru Zhou | Changzhi Zhou | Ziyi Yang

In event argument extraction (EAE), a promising approach involves jointly encoding text and argument roles, and performing multiple token linking operations. This approach further falls into two categories. One extracts arguments within a single event, while the other attempts to extract arguments from multiple events simultaneously. However, the former lacks to leverage cross-event information and the latter requires tougher predictions with longer encoded role sequences and extra linking operations. In this paper, we design a novel separation-and-fusion paradigm to separately acquire cross-event information and fuse it into the argument extraction of a target event. Following the paradigm, we propose a novel multiple token linking model named Sep2F, which can effectively build event correlations via roles and preserve the simple linking predictions of single-event extraction. In particular, we employ one linking module to extract arguments for the target event and another to aggregate the role information of multiple events. More importantly, we propose a novel two-fold fusion module to ensure that the aggregated cross-event information serves EAE well. We evaluate our proposed model on sentence-level and document-level datasets, including ACE05, RAMS, WikiEvents and MLEE. The extensive experimental results indicate that our model outperforms the state-of-the-art EAE models on all the datasets.

pdf bib
The Integration of Semantic and Structural Knowledge in Knowledge Graph Entity Typing
Muzhi Li | Minda Hu | Irwin King | Ho-fung Leung

The Knowledge Graph Entity Typing (KGET) task aims to predict missing type annotations for entities in knowledge graphs. Recent works only utilize the structural knowledge in the local neighborhood of entities, disregarding semantic knowledge in the textual representations of entities, relations, and types that are also crucial for type inference. Additionally, we observe that the interaction between semantic and structural knowledge can be utilized to address the false-negative problem. In this paper, we propose a novel Semantic and Structure-aware KG Entity Typing (SSET) framework, which is composed of three modules. First, the Semantic Knowledge Encoding module encodes factual knowledge in the KG with a Masked Entity Typing task. Then, the Structural Knowledge Aggregation module aggregates knowledge from the multi-hop neighborhood of entities to infer missing types. Finally, the Unsupervised Type Re-ranking module utilizes the inference results from the two models above to generate type predictions that are robust to false-negative samples. Extensive experiments show that SSET significantly outperforms existing state-of-the-art methods.

pdf bib
ComCLIP: Training-Free Compositional Image and Text Matching
Kenan Jiang | Xuehai He | Ruize Xu | Xin Wang

Contrastive Language-Image Pretraining (CLIP) has demonstrated great zero-shot performance for matching images and text. However, it is still challenging to adapt vision-language pretrained models like CLIP to compositional image and text matching — a more challenging image and text matching task requiring the model’s understanding of compositional word concepts and visual components. Towards better compositional generalization in zero-shot image and text matching, in this paper, we study the problem from a causal perspective: the erroneous semantics of individual entities are essentially confounders that cause the matching failure. Therefore, we propose a novel training-free compositional CLIP model (ComCLIP). ComCLIP disentangles input images into subjects, objects, and action subimages and composes CLIP’s vision encoder and text encoder to perform evolving matching over compositional text embedding and subimage embeddings. In this way, ComCLIP can mitigate spurious correlations introduced by the pretrained CLIP models and dynamically evaluate the importance of each component. Experiments on four compositional image-text matching datasets: Winoground, VL-checklist, SVO, and ComVG, and two general image-text retrieval datasets: Flick30K, and MSCOCO demonstrate the effectiveness of our plug-and-play method, which boosts the zero-shot inference ability of CLIP, SLIP, and BLIP2 even without further training or fine-tuning. Our codes can be found at https://github.com/eric-ai-lab/ComCLIP.

pdf bib
ACLSum: A New Dataset for Aspect-based Summarization of Scientific Publications
Sotaro Takeshita | Tommaso Green | Ines Reinig | Kai Eckert | Simone Ponzetto

Extensive efforts in the past have been directed toward the development of summarization datasets. However, a predominant number of these resources have been (semi)-automatically generated, typically through web data crawling. This resulted in subpar resources for training and evaluating summarization systems, a quality compromise that is arguably due to the substantial costs associated with generating ground-truth summaries, particularly for diverse languages and specialized domains. To address this issue, we present ACLSum, a novel summarization dataset carefully crafted and evaluated by domain experts. In contrast to previous datasets, ACLSum facilitates multi-aspect summarization of scientific papers, covering challenges, approaches, and outcomes in depth. Through extensive experiments, we evaluate the quality of our resource and the performance of models based on pretrained language models (PLMs) and state-of-the-art large language models (LLMs). Additionally, we explore the effectiveness of extract-then-abstract versus abstractive end-to-end summarization within the scholarly domain on the basis of automatically discovered aspects. While the former performs comparably well to the end-to-end approach with pretrained language models regardless of the potential error propagation issue, the prompting-based approach with LLMs shows a limitation in extracting sentences from source documents.

pdf bib
XAL: EXplainable Active Learning Makes Classifiers Better Low-resource Learners
Yun Luo | Zhen Yang | Fandong Meng | Yingjie Li | Fang Guo | Qinglin Qi | Jie Zhou | Yue Zhang

Active learning (AL), which aims to construct an effective training set by iteratively curating the most formative unlabeled data for annotation, has been widely used in low-resource tasks. Most active learning techniques in classification rely on the model’s uncertainty or disagreement to choose unlabeled data, suffering from the problem of over-confidence in superficial patterns and a lack of exploration.Inspired by the cognitive processes in which humans deduce and predict through causal information, we take an initial attempt towards integrating rationales into AL and propose a novel Explainable Active Learning framework (XAL) for low-resource text classification, which aims to encourage classifiers to justify their inferences and delve into unlabeled data for which they cannot provide reasonable explanations. Specifically, besides using a pre-trained bi-directional encoder for classification, we employ a pre-trained uni-directional decoder to generate and score the explanation. We further facilitate the alignment of the model with human reasoning preference through a proposed ranking loss. During the selection of unlabeled data, the predicted uncertainty of the encoder and the explanation score of the decoder complement each other as the final metric to acquire informative data. Extensive experiments on six datasets show that XAL achieves consistent improvement over 9 strong baselines. Analysis indicates that the proposed method can generate corresponding explanations for its predictions.

pdf bib
LaDiC: Are Diffusion Models Really Inferior to Autoregressive Counterparts for Image-to-Text Generation?
Yuchi Wang | Shuhuai Ren | Rundong Gao | Linli Yao | Qingyan Guo | Kaikai An | Jianhong Bai | Xu Sun

Diffusion models have exhibited remarkable capabilities in text-to-image generation. However, their performance in image-to-text generation, specifically image captioning, has lagged behind Auto-Regressive (AR) models, casting doubt on their applicability for such tasks. In this work, we revisit diffusion models, highlighting their capacity for holistic context modeling and parallel decoding. With these benefits, diffusion models can alleviate the inherent limitations of AR methods, including their slow inference speed, error propagation, and unidirectional constraints. Furthermore, we identify the prior underperformance of diffusion models stemming from the absence of an effective latent space for image-text alignment, and the discrepancy between continuous diffusion processes and discrete textual data. In response, we introduce a novel architecture, LaDiC, which utilizes a split BERT to create a dedicated latent space for captions and integrates a regularization module to manage varying text lengths. Our framework also includes a diffuser for semantic image-to-text conversion and a Back&Refine technique to enhance token interactivity during inference. LaDiC achieves state-of-the-art performance for diffusion-based methods on the MS COCO dataset with 38.2 BLEU@4 and 126.2 CIDEr, demonstrating exceptional performance without pre-training or ancillary modules. This indicates strong competitiveness with AR models, revealing the previously untapped potential of diffusion models in image-to-text generation.

pdf bib
Intent-conditioned and Non-toxic Counterspeech Generation using Multi-Task Instruction Tuning with RLAIF
Amey Hengle | Aswini Padhi | Sahajpreet Singh | Anil Bandhakavi | Md Shad Akhtar | Tanmoy Chakraborty

Counterspeech, defined as a response to mitigate online hate speech, is increasingly used as a non-censorial solution. The effectiveness of addressing hate speech involves dispelling the stereotypes, prejudices, and biases often subtly implied in brief, single-sentence statements or abuses. These expressions challenge language models, especially in seq2seq tasks, as model performance typically excels with longer contexts. Our study introduces CoARL, a novel framework enhancing counterspeech generation by modeling the pragmatic implications underlying social biases in hateful statements. The first two phases of CoARL involve sequential multi-instruction tuning, teaching the model to understand intents, reactions, and harms of offensive statements, and then learning task-specific low-rank adapter weights for generating intent-conditioned counterspeech. The final phase uses reinforcement learning to fine-tune outputs for effectiveness and nontoxicity. CoARL outperforms existing benchmarks in intent-conditioned counterspeech generation, showing an average improvement of ∼3 points in intent-conformity and ∼4 points in argument-quality metrics. Extensive human evaluation supports CoARL’s efficacy in generating superior and more context-appropriate responses compared to existing systems, including prominent LLMs like ChatGPT.

pdf bib
Attacks, Defenses and Evaluations for LLM Conversation Safety: A Survey
Zhichen Dong | Zhanhui Zhou | Chao Yang | Jing Shao | Yu Qiao

Large Language Models (LLMs) are now commonplace in conversation applications. However, their risks of misuse for generating harmful responses have raised serious societal concerns and spurred recent research on LLM conversation safety. Therefore, in this survey, we provide a comprehensive overview of recent studies, covering three critical aspects of LLM conversation safety: attacks, defenses, and evaluations. Our goal is to provide a structured summary that enhances understanding of LLM conversation safety and encourages further investigation into this important subject. For easy reference, we have categorized all the studies mentioned in this survey according to our taxonomy, available at: https://github.com/niconi19/LLM-conversation-safety.

pdf bib
Mind’s Mirror: Distilling Self-Evaluation Capability and Comprehensive Thinking from Large Language Models
Weize Liu | Guocong Li | Kai Zhang | Bang Du | Qiyuan Chen | Xuming Hu | Hongxia Xu | Jintai Chen | Jian Wu

Large language models (LLMs) have achieved remarkable advancements in natural language processing. However, the massive scale and computational demands of these models present formidable challenges when considering their practical deployment in resource-constrained environments. While techniques such as chain-of-thought (CoT) distillation have displayed promise in distilling LLMs into small language models (SLMs), there is a risk that distilled SLMs may still inherit flawed reasoning and hallucinations from LLMs. To address these issues, we propose a twofold methodology: First, we introduce a novel method for distilling the self-evaluation capability from LLMs into SLMs, aiming to mitigate the adverse effects of flawed reasoning and hallucinations inherited from LLMs. Second, we advocate for distilling more comprehensive thinking by incorporating multiple distinct CoTs and self-evaluation outputs, to ensure a more thorough and robust knowledge transfer into SLMs. Experiments on three NLP benchmarks demonstrate that our method significantly improves the performance of distilled SLMs, offering a new perspective for developing more effective and efficient SLMs in resource-constrained environments.

pdf bib
Divergent Token Metrics: Measuring degradation to prune away LLM components – and optimize quantization
Björn Deiseroth | Max Meuer | Nikolas Gritsch | Constantin Eichenberg | Patrick Schramowski | Matthias Aßenmacher | Kristian Kersting

Large Language Models (LLMs) have reshaped natural language processing with their impressive capabilities. However, their ever-increasing size has raised concerns about their effective deployment and the need for LLM compression. This study introduces the Divergent Token Metrics (DTMs), a novel approach to assessing compressed LLMs, addressing the limitations of traditional perplexity or accuracy measures that fail to accurately reflect text generation quality. DTMs measure token divergences that allow deeper insights into the subtleties of model compression, in particular, when evaluating components’ impacts individually. Utilizing the First Divergent Token Metric (FDTM) in model sparsification reveals that 25% of all attention components can be pruned beyond 90% on the Llama-2 model family, still keeping SOTA performance. For quantization, FDTM suggests that more than 80% of parameters can be naively transformed to int8 without special outlier management. These evaluations indicate the necessity of choosing appropriate compressions for parameters individually—and that FDTM can identify those—while standard metrics result in deteriorated outcomes.

pdf bib
Beyond Performance: Quantifying and Mitigating Label Bias in LLMs
Yuval Reif | Roy Schwartz

Large language models (LLMs) have shown remarkable adaptability to diverse tasks, by leveraging context prompts containing instructions, or minimal input-output examples. However, recent work revealed they also exhibit *label bias*—an undesirable preference toward predicting certain answers over others. Still, detecting and measuring this bias reliably and at scale has remained relatively unexplored. In this study, we evaluate different approaches to quantifying label bias in a model’s predictions, conducting a comprehensive investigation across 279 classification tasks and ten LLMs. Our investigation reveals substantial label bias in models both before and after debiasing attempts, as well as highlights the importance of outcomes-based evaluation metrics, which were not previously used in this regard. We further propose a novel label bias calibration method tailored for few-shot prompting, which outperforms recent calibration approaches for both improving performance and mitigating label bias. Our results emphasize that label bias in the predictions of LLMs remains a barrier to their reliability.

pdf bib
Instructing Large Language Models to Identify and Ignore Irrelevant Conditions
Zhenyu Wu | Chao Shen | Meng Jiang

Math word problem (MWP) solving requires generating a reasoning path based on a given problem description that often contains irrelevant conditions.Existing chain-of-thought (CoT) prompting methods elicited multi-step reasoning abilities of large language models (LLMs) to solve MWPs.However, they were seriously confused by the irrelevant conditions, resulting in low accuracy.In this paper, we propose a novel approach named I3C that instructs LLMs to identify and ignore irrelevant conditions.It identifies a set of irrelevant condition candidates that have a weak semantic relevance with the question.Then it prompts LLMs to verify the irrelevant conditions.Lastly it instructs the LLMs with the verification on relevant and irrelevant conditions to avoid confusion and improve reasoning paths.Moreover, we propose to select (problem, reasoning paths) pairs as demonstrations to enhance I3C with few-shot reasoning. We develop I3C-Select that selects the most confusing problems based on the semantic relevance measurement.We conduct extensive experiments on eight MWP datasets.I3C can be combined with any CoT prompting methods to improve the performance of solving MWPs.Notably, with GPT-3.5-Turbo and I3C-Select, we achieve an accuracy of 96.0 and 94.1 on GSM-IC2-1K and GSM-ICM-1K, respectively, significantly outperforming the state-of-the-art few-shot prompting method Complex-CoT by +11.7 and +11.1.Our implementation is made publicly available at https://wzy6642.github.io/I3C.github.io/.

pdf bib
Lower Bounds on the Expressivity of Recurrent Neural Language Models
Anej Svete | Franz Nowak | Anisha Sahabdeen | Ryan Cotterell

The recent successes and spread of large neural language models (LMs) call for a thorough understanding of their abilities. Describing their abilities through LMs’ representational capacity is a lively area of research. Investigations of the representational capacity of neural LMs have predominantly focused on their ability to recognize formal languages. For example, recurrent neural networks (RNNs) as classifiers are tightly linked to regular languages, i.e., languages defined by finite-state automata (FSAs). Such results, however, fall short of describing the capabilities of RNN language models (LMs), which are definitionally distributions over strings. We take a fresh look at the represen- tational capacity of RNN LMs by connecting them to probabilistic FSAs and demonstrate that RNN LMs with linearly bounded precision can express arbitrary regular LMs.

pdf bib
Transformers Can Represent n-gram Language Models
Anej Svete | Ryan Cotterell

Plenty of existing work has analyzed the abilities of the transformer architecture by describing its representational capacity with formal models of computation. However, the focus so far has been on analyzing the architecture in terms of language acceptance. We contend that this is an ill-suited problem in the study of language models (LMs), which are definitionally probability distributions over strings. In this paper, we focus on the relationship between transformer LMs and n-gram LMs, a simple and historically relevant class of language models. We show that transformer LMs using the hard or sparse attention mechanisms can exactly represent any n-gram LM, giving us a concrete lower bound on their probabilistic representational capacity. This provides a first step towards understanding the mechanisms that transformer LMs can use to represent probability distributions over strings.

pdf bib
The Role of n-gram Smoothing in the Age of Neural Networks
Luca Malagutti | Andrius Buinovskij | Anej Svete | Clara Meister | Afra Amini | Ryan Cotterell

For nearly three decades, language models derived from the n-gram assumption held the state of the art on the task. The key to their success lay in the application of various smoothing techniques that served to combat overfitting. However, when neural language models toppled n-gram models as the best performers, n-gram smoothing techniques became less relevant. Indeed, it would hardly be an understatement to suggest that the line of inquiry into n-gram smoothing techniques became dormant. This paper re-opens the role classical n-gram smoothing techniques may play in the age of neural language models. First, we draw a formal equivalence between label smoothing, a popular regularization technique for neural language models, and add-𝜆 smoothing. Second, we derive a generalized framework for converting any n-gram smoothing technique into a regularizer compatible with neural language models. Our empirical results find that our novel regularizers are comparable to and, indeed, sometimes outperform label smoothing on language modeling and machine translation.

pdf bib
Reliability Estimation of News Media Sources: Birds of a Feather Flock Together
Sergio Burdisso | Dairazalia Sanchez-cortes | Esaú Villatoro-tello | Petr Motlicek

Evaluating the reliability of news sources is a routine task for journalists and organizations committed to acquiring and disseminating accurate information.Recent research has shown that predicting sources’ reliability represents an important first-prior step in addressing additional challenges such as fake news detection and fact-checking.In this paper, we introduce a novel approach for source reliability estimation that leverages reinforcement learning strategies for estimating the reliability degree of news sources. Contrary to previous research, our proposed approach models the problem as the estimation of a reliability degree, and not a reliability label, based on how all the news media sources interact with each other on the Web.We validated the effectiveness of our method on a news media reliability dataset that is an order of magnitude larger than comparable existing datasets. Results show that the estimated reliability degrees strongly correlates with journalists-provided scores (Spearman=0.80) and can effectively predict reliability labels (macro-avg. F1 score=81.05).We release our implementation and dataset, aiming to provide a valuable resource for the NLP community working on information verification.

pdf bib
On the Multilingual Ability of Decoder-based Pre-trained Language Models: Finding and Controlling Language-Specific Neurons
Takeshi Kojima | Itsuki Okimura | Yusuke Iwasawa | Hitomi Yanaka | Yutaka Matsuo

Current decoder-based pre-trained language models (PLMs) successfully demonstrate multilingual capabilities. However, it is unclear how these models handle multilingualism.We analyze the neuron-level internal behavior of multilingual decoder-based PLMs, Specifically examining the existence of neurons that fire “uniquely for each language” within decoder-only multilingual PLMs.We analyze six languages: English, German, French, Spanish, Chinese, and Japanese, and show that language-specific neurons are unique, with a slight overlap (< 5%) between languages. These neurons are mainly distributed in the models’ first and last few layers. This trend remains consistent across languages and models.Additionally, we tamper with less than 1% of the total neurons in each model during inference and demonstrate that tampering with a few language-specific neurons drastically changes the probability of target language occurrence in text generation.

pdf bib
NLP Progress in Indigenous Latin American Languages
Atnafu Tonja | Fazlourrahman Balouchzahi | Sabur Butt | Olga Kolesnikova | Hector Ceballos | Alexander Gelbukh | Thamar Solorio

The paper focuses on the marginalization of indigenous language communities in the face of rapid technological advancements. We highlight the cultural richness of these languages and the risk they face of being overlooked in the realm of Natural Language Processing (NLP). We aim to bridge the gap between these communities and researchers, emphasizing the need for inclusive technological advancements that respect indigenous community perspectives. We show the NLP progress of indigenous Latin American languages and the survey that covers the status of indigenous languages in Latin America, their representation in NLP, and the challenges and innovations required for their preservation and development. The paper contributes to the current literature in understanding the need and progress of NLP for indigenous communities of Latin America, specifically low-resource and indigenous communities in general.

pdf bib
On the Effectiveness of Adversarial Robustness for Abuse Mitigation with Counterspeech
Yi-Ling Chung | Jonathan Bright

Recent work on automated approaches to counterspeech have mostly focused on synthetic data but seldom look into how the public deals with abuse. While these systems identifying and generating counterspeech have the potential for abuse mitigation, it remains unclear how robust a model is against adversarial attacks across multiple domains and how models trained on synthetic data can handle unseen user-generated abusive content in the real world. To tackle these issues, this paper first explores the dynamics of abuse and replies using our novel dataset of 6,955 labelled tweets targeted at footballers for studying public figure abuse. We then curate DynaCounter, a new English dataset of 1,911 pairs of abuse and replies addressing nine minority identity groups, collected in an adversarial human-in-the-loop process over four rounds. Our analysis shows that adversarial attacks do not necessarily result in better generalisation. We further present a study of multi-domain counterspeech generation, comparing Flan-T5 and T5 models. We observe that handling certain abuse targets is particularly challenging.

pdf bib
Leveraging the Structure of Pre-trained Embeddings to Minimize Annotation Effort
Cesar Gonzalez-Gutierrez | Ariadna Quattoni

Most current state-of-the-art approaches for text classification are based on fine-tuning the representations computed by large language models (LLMs). This strategy has led to significant improvements in classification performance and contributed to a reduction of the amount of labeled data required for training a model. However, for some challenging classification tasks, providing enough annotations to ensure a reliable classification continues to be the main bottleneck. This is especially true in settings of highly imbalanced class distributions. This paper proposes to tackle this bottleneck by exploiting the structural properties of pre-trained embeddings. We develop a label propagation method that uses pre-trained embeddings to spread information from the labeled samples to nearby samples in the induced space, ensuring the optimal use of annotations. Our approach is simple and relatively low-cost since it only requires computing some distances in the embedded space. We conduct experiments on different text classification datasets showing that the proposed method is efficient and significantly outperforms both self-training and random walk label propagation strategies.

pdf bib
UniArk: Improving Generalisation and Consistency for Factual Knowledge Extraction through Debiasing
Yijun Yang | Jie He | Pinzhen Chen | Victor Gutierrez Basulto | Jeff Pan

Several recent papers have investigated the potential of language models as knowledge bases as well as the existence of severe biases when extracting factual knowledge. In this work, we focus on the factual probing performance over unseen prompts from tuning, and using a probabilistic view we show the inherent misalignment between pre-training and downstream tuning objectives in language models for probing knowledge. We hypothesize that simultaneously debiasing these objectives can be the key to generalisation over unseen prompts. We propose an adapter-based framework, **UniArk**, for generalised and consistent factual knowledge extraction through simple methods without introducing extra parameters. Extensive experiments show that UniArk can significantly improve the model’s out-of-domain generalisation as well as consistency under various prompts. Additionally, we construct **ParaTrex**, a large-scale and diverse dataset for measuring the inconsistency and out-of-domain generation of models. Further, ParaTrex offers a reference method for constructing paraphrased datasets using large language models.

pdf bib
Adaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexity
Soyeong Jeong | Jinheon Baek | Sukmin Cho | Sung Ju Hwang | Jong Park

Retrieval-Augmented Large Language Models (LLMs), which incorporate the non-parametric knowledge from external knowledge bases into LLMs, have emerged as a promising approach to enhancing response accuracy in several tasks, such as Question-Answering (QA). However, even though there are various approaches dealing with queries of different complexities, they either handle simple queries with unnecessary computational overhead or fail to adequately address complex multi-step queries; yet, not all user requests fall into only one of the simple or complex categories. In this work, we propose a novel adaptive QA framework that can dynamically select the most suitable strategy for (retrieval-augmented) LLMs from the simplest to the most sophisticated ones based on the query complexity. Also, this selection process is operationalized with a classifier, which is a smaller LM trained to predict the complexity level of incoming queries with automatically collected labels, obtained from actual predicted outcomes of models and inherent inductive biases in datasets. This approach offers a balanced strategy, seamlessly adapting between the iterative and single-step retrieval-augmented LLMs, as well as the no-retrieval methods, in response to a range of query complexities. We validate our model on a set of open-domain QA datasets, covering multiple query complexities, and show that ours enhances the overall efficiency and accuracy of QA systems, compared to relevant baselines including the adaptive retrieval approaches. Code is available at: https://github.com/starsuzi/Adaptive-RAG.

pdf bib
Knowing What LLMs DO NOT Know: A Simple Yet Effective Self-Detection Method
Yukun Zhao | Lingyong Yan | Weiwei Sun | Guoliang Xing | Chong Meng | Shuaiqiang Wang | Zhicong Cheng | Zhaochun Ren | Dawei Yin

Large Language Models (LLMs) have shown great potential in Natural Language Processing (NLP) tasks.However, recent literature reveals that LLMs hallucinate intermittently, which impedes their reliability for further utilization. In this paper, we propose a novel self-detection method to detect which questions an LLM does not know.Our proposal is empirical and applicable for continually upgrading LLMs compared with state-of-the-art methods. Specifically, we examine the divergence of the LLM’s behaviors on different verbalizations for a question and examine the atypicality of the verbalized input. We combine the two components to identify whether the model generates a non-factual response to the question. The above components can be accomplished by utilizing the LLM itself without referring to any other external resources. We conduct comprehensive experiments and demonstrate the effectiveness of our method for recently released LLMs involving Llama 2, Vicuna, ChatGPT, and GPT-4 across factoid question-answering, arithmetic reasoning, and commonsense reasoning tasks.

pdf bib
Are Large Language Model Temporally Grounded?
Yifu Qiu | Zheng Zhao | Yftah Ziser | Anna Korhonen | Edoardo Ponti | Shay Cohen

Are Large Language Models (LLMs) temporally grounded? Since LLMs cannot perceive and interact with the environment, it is impossible to answer this question directly. Instead, we provide LLMs with textual narratives and probe them with respect to their common-sense knowledge of the structure and duration of events, their ability to order events along a timeline, and self-consistency within their temporal model (e.g., temporal relations such as after and before are mutually exclusive for any pair of events). We evaluate state-of-the-art LLMs (such as LLaMA 2 and GPT-4) on three tasks reflecting these abilities. Generally, we find that LLMs lag significantly behind both human performance as well as small-scale, specialised LMs. In-context learning, instruction tuning, and chain-of-thought prompting reduce this gap only to a limited degree. Crucially, LLMs struggle the most with self-consistency, displaying incoherent behaviour in at least 27.23% of their predictions. Contrary to expectations, we also find that scaling the model size does not guarantee positive gains in performance. To explain these results, we study the sources from which LLMs may gather temporal information: we find that sentence ordering in unlabelled texts, available during pre-training, is only weakly correlated with event ordering. Moreover, public instruction tuning mixtures contain few temporal tasks. Hence, we conclude that current LLMs lack a consistent temporal model of textual narratives.

pdf bib
Document Image Machine Translation with Dynamic Multi-pre-trained Models Assembling
Yupu Liang | Yaping Zhang | Cong Ma | Zhiyang Zhang | Yang Zhao | Lu Xiang | Chengqing Zong | Yu Zhou

Text image machine translation (TIMT) is a task that translates source texts embedded in the image to target translations. The existing TIMT task mainly focuses on text-line-level images. In this paper, we extend the current TIMT task and propose a novel task, **D**ocument **I**mage **M**achine **T**ranslation to **Markdown** (**DIMT2Markdown**), which aims to translate a source document image with long context and complex layout structure to markdown-formatted target translation.We also introduce a novel framework, **D**ocument **I**mage **M**achine **T**ranslation with **D**ynamic multi-pre-trained models **A**ssembling (**DIMTDA**).A dynamic model assembler is used to integrate multiple pre-trained models to enhance the model’s understanding of layout and translation capabilities.Moreover, we build a novel large-scale **Do**cument image machine **T**ranslation dataset of **A**rXiv articles in markdown format (**DoTA**), containing 126K image-translation pairs.Extensive experiments demonstrate the feasibility of end-to-end translation of rich-text document images and the effectiveness of DIMTDA.

pdf bib
Elastic Weight Removal for Faithful and Abstractive Dialogue Generation
Nico Daheim | Nouha Dziri | Mrinmaya Sachan | Iryna Gurevych | Edoardo Ponti

Generating factual responses is a crucial requirement for dialogue systems. To promotemore factual responses, a common strategyis to ground their responses in relevant documents that inform response generation. However, common dialogue models still often hallucinate information that was not containedin these documents and is therefore unfaithful. In this work, we propose to alleviate suchhallucinations by ‘subtracting’ the parametersof a model trained to hallucinate from a dialogue response generation model in order to‘negate’ the contribution of such hallucinatedexamples from it. Extensive automatic and human evaluation shows favourable results whencompared to state-of-the-art methods that combine the distributions of multiple models, suchas DExperts (Liu et al., 2021), and others thatchange the training procedure, such as Quark(Lu et al., 2022a). Finally, we show how wecan not only reduce hallucinations but also discourage extractive responses, which are oftena consequence of reducing hallucinations byencouraging copy-pasting of document spans.We publicly release our code for reproducibilityand facilitating further research.

pdf bib
R-Tuning: Instructing Large Language Models to Say ‘I Don’t Know’
Hanning Zhang | Shizhe Diao | Yong Lin | Yi Fung | Qing Lian | Xingyao Wang | Yangyi Chen | Heng Ji | Tong Zhang

Large language models (LLMs) have revolutionized numerous domains with their impressive performance but still face their challenges. A predominant issue is the propensity for these models to generate non-existent facts, a concern termed hallucination. Our research is motivated by the observation that previous instruction tuning methods force the model to complete a sentence no matter whether the model knows the knowledge or not. When the question is out of the parametric knowledge, it will try to make up something and fail to indicate when it lacks knowledge. In this paper, we present a new approach called Refusal-Aware Instruction Tuning (R-Tuning). This approach is formalized by first identifying the disparity in knowledge encompassed by pre-trained parameters compared to that of instruction tuning data. Then, we construct the refusal-aware data based on the knowledge intersection, to tune LLMs to refrain from responding to questions beyond its parametric knowledge. Experimental results demonstrate R-Tuning effectively improves a model’s ability to answer known questions and refrain from answering unknown questions. Furthermore, when tested on out-of-domain datasets, the refusal ability was found to be a meta-skill that could be generalized to other tasks. Further analysis surprisingly finds that learning the uncertainty results in better calibration and an improved ability to estimate the uncertainty than uncertainty-based testing. Our code is available at https://github.com/shizhediao/R-Tuning

pdf bib
Bridging the Gap between Different Vocabularies for LLM Ensemble
Yangyifan Xu | Jinliang Lu | Jiajun Zhang

Ensembling different large language models (LLMs) to unleash their complementary potential and harness their individual strengths is highly valuable. Nevertheless, vocabulary discrepancies among various LLMs have constrained previous studies to either selecting or blending completely generated outputs. This limitation hinders the dynamic correction and enhancement of outputs during the generation process, resulting in a limited capacity for effective ensemble. To address this issue, we propose a novel method to Ensemble LLMs via Vocabulary Alignment (EVA). EVA bridges the lexical gap among various LLMs, enabling meticulous ensemble at each generation step. Specifically, we first learn mappings between the vocabularies of different LLMs with the assistance of overlapping tokens. Subsequently, these mappings are employed to project output distributions of LLMs into a unified space, facilitating a fine-grained ensemble. Finally, we design a filtering strategy to exclude models that generate unfaithful tokens. Experimental results on commonsense reasoning, arithmetic reasoning, machine translation, and data-to-text generation tasks demonstrate the superiority of our approach compared with individual LLMs and previous ensemble methods conducted on complete outputs. Further analyses confirm that our approach can leverage knowledge from different language models and yield consistent improvement.

pdf bib
KnowLA: Enhancing Parameter-efficient Finetuning with Knowledgeable Adaptation
Xindi Luo | Zequn Sun | Jing Zhao | Zhe Zhao | Wei Hu

Parameter-efficient finetuning (PEFT) is a key technique for adapting large language models (LLMs) to downstream tasks. In this paper, we study leveraging knowledge graph embeddings to improve the effectiveness of PEFT. We propose a knowledgeable adaptation method called KnowLA. It inserts an adaptation layer into an LLM to integrate the embeddings of entities appearing in the input text. The adaptation layer is trained in combination with LoRA on instruction data. Experiments on six benchmarks with two popular LLMs and three knowledge graphs demonstrate the effectiveness and robustness of KnowLA. We show that KnowLA can help activate the relevant parameterized knowledge in an LLM to answer a question without changing its parameters or input prompts.

pdf bib
Extremely Weakly-supervised Text Classification with Wordsets Mining and Sync-Denoising
Lysa Xiao

Extremely weakly-supervised text classification aims to classify texts without any labeled data, but only relying on class names as supervision. Existing works include prompt-based and seed-based methods. Prompt-based methods prompt language model with instructions, while seed-based methods generate pseudo-labels with word matching. Both of them have significant flaws, including zero-shot instability and context-dependent ambiguities. This paper introduces SetSync, which follows a new paradigm, i.e. wordset-based, which can avoid the above problems. In SetSync, a class is represented with wordsets, and pseudo-labels are generated with wordsets matching. To facilitate this, we propose to use information bottleneck to identify class-relevant wordsets. Moreover, we regard the classifier training as a hybrid learning of semi-supervised and noisy-labels, and propose a new training strategy, termed sync-denoising. Extensive experiments on 11 datasets show that SetSync outperforms all existing prompt and seed methods, exceeding SOTA by an impressive average of 8 points.

pdf bib
F-MALLOC: Feed-forward Memory Allocation for Continual Learning in Neural Machine Translation
Junhong Wu | Yuchen Liu | Chengqing Zong

In the evolving landscape of Neural Machine Translation (NMT), the pretrain-then-finetune paradigm has yielded impressive results. However, the persistent challenge of Catastrophic Forgetting (CF) remains a hurdle. While previous work has introduced Continual Learning (CL) methods to address CF, these approaches grapple with the delicate balance between avoiding forgetting and maintaining system extensibility. To address this, we propose a CL method, named F-MALLOC (Feed-forward Memory ALLOCation). F-MALLOC is inspired by recent insights highlighting that feed-forward layers emulate neural memories and encapsulate crucial translation knowledge. It decomposes feed-forward layers into discrete memory cells and allocates these memories to different tasks. By learning to allocate and safeguard these memories, our method effectively alleviates CF while ensuring robust extendability. Besides, we propose a comprehensive assessment protocol for multi-stage CL of NMT systems. Experiments conducted following this new protocol showcase the superior performance of F-MALLOC, evidenced by higher BLEU scores and almost zero forgetting.

pdf bib
Towards Reducing Diagnostic Errors with Interpretable Risk Prediction
Denis McInerney | William Dickinson | Lucy Flynn | Andrea Young | Geoffrey Young | Jan-Willem van de Meent | Byron Wallace

Many diagnostic errors occur because clinicians cannot easily access relevant information in patient Electronic Health Records (EHRs). In this work we propose a method to use LLMs to identify pieces of evidence in patient EHR data that indicate increased or decreased risk of specific diagnoses; our ultimate aim is to increase access to evidence and reduce diagnostic errors. In particular, we propose a Neural Additive Model to make predictions backed by evidence with individualized risk estimates at time-points where clinicians are still uncertain, aiming to specifically mitigate delays in diagnosis and errors stemming from an incomplete differential. To train such a model, it is necessary to infer temporally fine-grained retrospective labels of eventual “true” diagnoses. We do so with LLMs, to ensure that the input text is from before a confident diagnosis can be made. We use an LLM to retrieve an initial pool of evidence, but then refine this set of evidence according to correlations learned by the model. We conduct an in-depth evaluation of the usefulness of our approach by simulating how it might be used by a clinician to decide between a pre-defined list of differential diagnoses.

pdf bib
Generalizable Multilingual Hate Speech Detection on Low Resource Indian Languages using Fair Selection in Federated Learning
Akshay Singh | Rahul Thakur

Social media, originally meant for peaceful communication, now faces issues with hate speech. Detecting hate speech from social media in Indian languages with linguistic diversity and cultural nuances presents a complex and challenging task. Furthermore, traditional methods involve sharing of users’ sensitive data with a server for model training making it undesirable and involving potential risk to their privacy remained under-studied. In this paper, we combined various low-resource language datasets and propose MultiFED, a federated approach that performs effectively to detect hate speech. MultiFED utilizes continuous adaptation and fine-tuning to aid generalization using subsets of multilingual data overcoming the limitations of data scarcity. Extensive experiments are conducted on 13 Indic datasets across five different pre-trained models. The results show that MultiFED outperforms the state-of-the-art baselines by 8% (approx.) in terms of Accuracy and by 12% (approx.) in terms of F-Score.

pdf bib
Key ingredients for effective zero-shot cross-lingual knowledge transfer in generative tasks
Nadezhda Chirkova | Vassilina Nikoulina

Zero-shot cross-lingual transfer, which implies finetuning of the multilingual pretrained language model on input-output pairs in one language and using it to make task predictions for inputs in other languages, was widely studied for natural language understanding but is understudied for generation. Previous works notice a frequent problem of generation in a wrong language and propose approaches to address it, usually using mT5 as a backbone model. In this work we compare various approaches proposed from the literature in unified settings, also including alternative backbone models, namely mBART and NLLB-200. We first underline the importance of tuning learning rate used for finetuning, which helps to substantially alleviate the problem of generation in the wrong language. Then, we show that with careful learning rate tuning, the simple full finetuning of the model acts as a very strong baseline and alternative approaches bring only marginal improvements. Finally, we find that mBART performs similarly to mT5 of the same size, and NLLB-200 can be competitive in some cases. Our final zero-shot models reach the performance of the approach based on data translation which is usually considered as an upper baseline for zero-shot cross-lingual transfer in generation.

pdf bib
The Impact of Depth on Compositional Generalization in Transformer Language Models
Jackson Petty | Sjoerd Steenkiste | Ishita Dasgupta | Fei Sha | Dan Garrette | Tal Linzen

To process novel sentences, language models (LMs) must generalize compositionally—combine familiar elements in new ways. What aspects of a model’s structure promote compositional generalization? Focusing on transformers, we test the hypothesis, motivated by theoretical and empirical work, that deeper transformers generalize more compositionally. Simply adding layers increases the total number of parameters; to address this confound between depth and size, we construct three classes of models which trade off depth for width such that the total number of parameters is kept constant (41M, 134M and 374M parameters). We pretrain all models as LMs and fine-tune them on tasks that test for compositional generalization. We report three main conclusions: (1) after fine-tuning, deeper models generalize more compositionally than shallower models do, but the benefit of additional layers diminishes rapidly; (2) within each family, deeper models show better language modeling performance, but returns are similarly diminishing; (3) the benefits of depth for compositional generalization cannot be attributed solely to better performance on language modeling. Because model latency is approximately linear in the number of layers, these results lead us to the recommendation that, with a given total parameter budget, transformers can be made shallower than is typical without sacrificing performance.

pdf bib
Pregnant Questions: The Importance of Pragmatic Awareness in Maternal Health Question Answering
Neha Srikanth | Rupak Sarkar | Heran Mane | Elizabeth Aparicio | Quynh Nguyen | Rachel Rudinger | Jordan Boyd-Graber

Questions posed by information-seeking users often contain implicit false or potentially harmful assumptions. In a high-risk domain such as maternal and infant health, a question-answering system must recognize these pragmatic constraints and go beyond simply answering user questions, examining them in context to respond helpfully. To achieve this, we study assumptions and implications, or pragmatic inferences, made when mothers ask questions about pregnancy and infant care by collecting a dataset of 2,727 inferences from 500 questions across three diverse sources. We study how health experts naturally address these inferences when writing answers, and illustrate that informing existing QA pipelines with pragmatic inferences produces responses that are more complete, mitigating the propagation of harmful beliefs.

pdf bib
Towards Explainability in Legal Outcome Prediction Models
Josef Valvoda | Ryan Cotterell

Current legal outcome prediction models - a staple of legal NLP - do not explain their reasoning. However, to employ these models in the real world, human legal actors need to be able to understand the model’s decisions. In the case of common law, legal practitioners reason towards the outcome of a case by referring to past case law, known as precedent. We contend that precedent is, therefore, a natural way of facilitating explainability for legal NLP models. In this paper, we contribute a novel method for identifying the precedent employed by legal outcome prediction models. Furthermore, by developing a taxonomy of legal precedent, we are able to compare human judges and neural models with respect to the different types of precedent they rely on. We find that while the models learn to predict outcomes reasonably well, their use of precedent is unlike that of human judges.

pdf bib
The steerability of large language models toward data-driven personas
Junyi Li | Charith Peris | Ninareh Mehrabi | Palash Goyal | Kai-Wei Chang | Aram Galstyan | Richard Zemel | Rahul Gupta

Large language models (LLMs) are known to generate biased responses where the opinions of certain groups and populations are underrepresented. Here, we present a novel approach to achieve controllable generation of specific viewpoints using LLMs, that can be leveraged to produce multiple perspectives and to reflect the diverse opinions. Moving beyond the traditional reliance on demographics like age, gender, or party affiliation, we introduce a data-driven notion of persona grounded in collaborative filtering, which is defined as either a single individual or a cohort of individuals manifesting similar views across specific inquiries. As individuals in the same demographic group may have different personas, our data-driven persona definition allows for a more nuanced understanding of different (latent) social groups present in the population. In addition to this, we also explore an efficient method to steer LLMs toward the personas that we define. We show that our data-driven personas significantly enhance model steerability, with improvements of between 57%-77% over our best performing baselines.

pdf bib
CCSum: A Large-Scale and High-Quality Dataset for Abstractive News Summarization
Xiang Jiang | Markus Dreyer

Training a supervised news summarization model requires large amounts of high-quality training data consisting of news articles paired with reference summaries. However, obtaining such data is costly, and existing datasets contain considerable amount of noise. We present a new large-scale and high-quality dataset for supervised abstractive news summarization containing 1.3 million training samples, which we call CCSum. In creating this dataset, we take advantage of the journalistic inverted-pyramid style in news writing: In some articles, the first sentence can be considered a summary of the reported story. Accordingly, among 35 million CommonCrawl News articles, we identify pairs of articles about the same news story and use one article’s first sentence as the summary for the other article. To ensure high quality, we apply strict filters whose parameters we optimize using Bayesian optimization. We show that the resulting dataset is more factual and informative than established summarization datasets; less than 1% of the summaries have major factual inconsistencies with the corresponding news articles, compared to 5.5% to 15.4% in existing datasets, according to our human evaluation. Summarization models trained on our dataset are more favored compared to those trained on CNN/Daily Mail. The proposed dataset can open new opportunities for future research in abstractive summarization.

pdf bib
Capturing Perspectives of Crowdsourced Annotators in Subjective Learning Tasks
Negar Mokhberian | Myrl Marmarelis | Frederic Hopp | Valerio Basile | Fred Morstatter | Kristina Lerman

Supervised classification heavily depends on datasets annotated by humans. However, in subjective tasks such as toxicity classification, these annotations often exhibit low agreement among raters. Annotations have commonly been aggregated by employing methods like majority voting to determine a single ground truth label. In subjective tasks, aggregating labels will result in biased labeling and, consequently, biased models that can overlook minority opinions. Previous studies have shed light on the pitfalls of label aggregation and have introduced a handful of practical approaches to tackle this issue. Recently proposed multi-annotator models, which predict labels individually per annotator, are vulnerable to under-determination for annotators with few samples. This problem is exacerbated in crowdsourced datasets. In this work, we propose Annotator Aware Representations for Texts (AART) for subjective classification tasks. Our approach involves learning representations of annotators, allowing for exploration of annotation behaviors. We show the improvement of our method on metrics that assess the performance on capturing individual annotators’ perspectives. Additionally, we demonstrate fairness metrics to evaluate our model’s equability of performance for marginalized annotators compared to others.

pdf bib
Improving Factual Accuracy of Neural Table-to-Text Output by Addressing Input Problems in ToTTo
Barkavi Sundararajan | Yaji Sripada | Ehud Reiter

Neural Table-to-Text models tend to hallucinate, producing texts that contain factual errors. We investigate whether such errors in the output can be traced back to problems with the input. We manually annotated 1,837 texts generated by multiple models in the politics domain of the ToTTo dataset. We identify the input problems that are responsible for many output errors and show that fixing these inputs reduces factual errors by between 52% and 76% (depending on the model). In addition, we observe that models struggle in processing tabular inputs that are structured in a non-standard way, particularly when the input lacks distinct row and column values or when the column headers are not correctly mapped to corresponding values.

pdf bib
CERET: Cost-Effective Extrinsic Refinement for Text Generation
Jason Cai | Hang Su | Monica Sunkara | Igor Shalyminov | Saab Mansour

Large Language Models (LLMs) are powerful models for generation tasks, but they may not generate good quality outputs in their first attempt. Apart from model fine-tuning, existing approaches to improve prediction accuracy and quality typically involve LLM self-improvement / self-reflection that incorporate feedback from models themselves. Despite their effectiveness, these methods are hindered by their high computational cost and lack of scalability. In this work, we propose CERET, a method for refining text generations by considering semantic stability, entailment and inter-sample uncertainty measures. Experimental results show that CERET outperforms Self-consistency and Self-rerank baselines consistently under various task setups, by 1.6% in Rouge-1 for abstractive summarization and 3.5% in hit rate for question answering. Compared to LLM Self-rerank method, our approach only requires 9.4% of its latency and is more cost-effective.

pdf bib
Parameter-Efficient Instruction Tuning of Large Language Models For Extreme Financial Numeral Labelling
Subhendu Khatuya | Rajdeep Mukherjee | Akash Ghosh | Manjunath Hegde | Koustuv Dasgupta | Niloy Ganguly | Saptarshi Ghosh | Pawan Goyal

We study the problem of automatically annotating relevant numerals (GAAP metrics) occurring in the financial documents with their corresponding XBRL tags. Different from prior works, we investigate the feasibility of solving this extreme classification problem using a generative paradigm through instruction tuning of Large Language Models (LLMs). To this end, we leverage metric metadata informationto frame our target outputs while proposing a parameter efficient solution for the task using LoRA. We perform experiments on two recently released financial numeric labeling datasets. Our proposed model, **FLAN-FinXC**, achieves new state-of-the-art performances on both the datasets, outperforming several strong baselines. We explain the better scores of our proposed model by demonstrating its capability for zero-shot as well as the least frequently occurring tags. Also, even when we fail to predict the XBRL tags correctly, our generated output has substantial overlap with the ground-truth in majority of the cases.

pdf bib
Analysis of State-Level Legislative Process in Enhanced Linguistic and Nationwide Network Contexts
Maryam Davoodi | Dan Goldwasser

State bills have a significant impact on various aspects of society, including health, education, and the economy. Consequently, it is crucial to conduct systematic research on state bills before and after they are enacted to evaluate their benefits and drawbacks, thereby guiding future decision-making. In this work, we developed the first state-level deep learning framework that (1) handles the complex and inconsistent language of policies across US states using generative large language models and (2) decodes legislators’ behavior and implications of state policies by establishing a shared nationwide network, enriched with diverse contexts, such as information on interest groups influencing public policy and legislators’ courage test results, which reflect their political positions.

pdf bib
DeMuX: Data-efficient Multilingual Learning
Simran Khanuja | Srinivas Gowriraj | Lucio Dery | Graham Neubig

Pre-trained multilingual models have enabled deployment of NLP technologies for multiple languages. However, optimally fine-tuning these models under an annotation budget, such that performance on desired target languages is jointly maximized, still remains an open question. In this paper, we introduce DeMuX, a framework that prescribes the exact data-points to label from vast amounts of unlabelled multilingual data, having unknown degrees of overlap with the target set. Unlike most prior works, our end-to-end framework is language-agnostic, accounts for model representations, and supports multilingual target configurations. Our active learning strategies rely upon distance and uncertainty measures to select task-specific neighbors that are most informative to label, given a model. DeMuX outperforms strong baselines in 84% of the test cases, in the zero-shot setting of disjoint source and target language sets (including multilingual target pools), across three models and four tasks. Notably, in low-budget settings (5-100 examples), we observe gains of up to 8-11 F1 points. Our code is released here: https://github.com/simran-khanuja/demux.

pdf bib
DUQGen: Effective Unsupervised Domain Adaptation of Neural Rankers by Diversifying Synthetic Query Generation
Ramraj Chandradevan | Kaustubh Dhole | Eugene Agichtein

State-of-the-art neural rankers pre-trained on large task-specific training data such as MS-MARCO, have been shown to exhibit strong performance on various ranking tasks without domain adaptation, also called zero-shot. However, zero-shot neural ranking may be sub-optimal, as it does not take advantage of the target domain information. Unfortunately, acquiring sufficiently large and high quality target training data to improve a modern neural ranker can be costly and time-consuming. To address this problem, we propose a new approach to unsupervised domain adaptation for ranking, DUQGen, which addresses a critical gap in prior literature, namely how to automatically generate both effective and diverse synthetic training data to fine tune a modern neural ranker for a new domain. Specifically, DUQGen produces a more effective representation of the target domain by identifying clusters of similar documents; and generates a more diverse training dataset by probabilistic sampling over the resulting document clusters. Our extensive experiments, over the standard BEIR collection, demonstrate that DUQGen consistently outperforms all zero-shot baselines and substantially outperforms the SOTA baselines on 16 out of 18 datasets, for an average of 4% relative improvement across all datasets. We complement our results with a thorough analysis for more in-depth understanding of the proposed method’s performance and to identify promising areas for further improvements.

pdf bib
How did we get here? Summarizing conversation dynamics
Yilun Hua | Nicholas Chernogor | Yuzhe Gu | Seoyeon Jeong | Miranda Luo | Cristian Danescu-Niculescu-Mizil

Throughout a conversation, the way participants interact with each other is in constant flux: their tones may change, they may resort to different strategies to convey their points, or they might alter their interaction patterns. An understanding of these dynamics can complement that of the actual facts and opinions discussed, offering a more holistic view of the trajectory of the conversation: how it arrived at its current state and where it is likely heading.In this work, we introduce the task of summarizing the dynamics of conversations, by constructing a dataset of human-written summaries, and exploring several automated baselines. We evaluate whether such summaries can capture the trajectory of conversations via an established downstream task: forecasting whether an ongoing conversation will eventually derail into toxic behavior. We show that they help both humans and automated systems with this forecasting task. Humans make predictions three times faster, and with greater confidence, when reading the summaries than when reading the transcripts. Furthermore, automated forecasting systems are more accurate when constructing, and then predicting based on, summaries of conversation dynamics, compared to directly predicting on the transcripts.

pdf bib
Can Language Model Moderators Improve the Health of Online Discourse?
Hyundong Cho | Shuai Liu | Taiwei Shi | Darpan Jain | Basem Rizk | Yuyang Huang | Zixun Lu | Nuan Wen | Jonathan Gratch | Emilio Ferrara | Jonathan May

Conversational moderation of online communities is crucial to maintaining civility for a constructive environment, but it is challenging to scale and harmful to moderators. The inclusion of sophisticated natural language generation modules as a force multiplier to aid human moderators is a tantalizing prospect, but adequate evaluation approaches have so far been elusive. In this paper, we establish a systematic definition of conversational moderation effectiveness grounded on moderation literature and establish design criteria for conducting realistic yet safe evaluation. We then propose a comprehensive evaluation framework to assess models’ moderation capabilities independently of human intervention. With our framework, we conduct the first known study of language models as conversational moderators, finding that appropriately prompted models that incorporate insights from social science can provide specific and fair feedback on toxic behavior but struggle to influence users to increase their levels of respect and cooperation.

pdf bib
LeanReasoner: Boosting Complex Logical Reasoning with Lean
Dongwei Jiang | Marcio Fonseca | Shay Cohen

Large language models (LLMs) often struggle with complex logical reasoning due to logical inconsistencies and the inherent difficulty ofsuch reasoning. We use Lean, a theorem proving framework, to address these challenges. By formalizing logical reasoning problems intotheorems within Lean, we can solve them by proving or disproving the corresponding theorems. This method reduces the risk of logical inconsistencies with the help of Lean’s symbolic solver. It also enhances our ability to treat complex reasoning tasks using Lean’s extensive library of theorem proofs. Our method achieves state-of-the-art performance on the FOLIO dataset and achieves performance near this level on ProofWriter. Notably, these results were accomplished by fine-tuning on fewer than 100 in-domain samples for each dataset

pdf bib
UICoder: Finetuning Large Language Models to Generate User Interface Code through Automated Feedback
Jason Wu

Many large language models (LLMs) struggle to consistently generate UI code that compiles and produces visually relevant designs. Existing approaches to improve generation rely either on expensive human feedback or distilling a proprietary model. In this paper, we explore the use of automated feedback (compilers and multi-modal models) to guide LLMs to generate high-quality UI code. Our method starts with an existing LLM and iteratively produces improved models by self-generating a large synthetic dataset using an original model, applying automated tools to aggressively filter, score, and de-duplicate the data into a refined higher quality dataset, and producing a new LLM by finetuning the original on the refined dataset.We applied our approach to several open-source LLMs and compared the resulting performance to baseline models with both automated metrics and human preferences.Our results show the resulting models outperform all other downloadable baselines and approach the performance of larger proprietary models.

pdf bib
Measuring Cross-lingual Transfer in Bytes
Leandro De Souza | Thales Almeida | Roberto Lotufo | Rodrigo Frassetto Nogueira

Multilingual pretraining has been a successful solution to the challenges posed by the lack of resources for languages. These models can transfer knowledge to target languages with minimal or no examples. Recent research suggests that monolingual models also have a similar capability, but the mechanisms behind this transfer remain unclear. Some studies have explored factors like language contamination and syntactic similarity. An emerging line of research suggests that the representations learned by language models contain two components: a language-specific and a language-agnostic component. The latter is responsible for transferring a more universal knowledge. However, there is a lack of comprehensive exploration of these properties across diverse target languages. To investigate this hypothesis, we conducted an experiment inspired by the work on the Scaling Laws for Transfer. We measured the amount of data transferred from a source language to a target language and found that models initialized from diverse languages perform similarly to a target language in a cross-lingual setting. This was surprising because the amount of data transferred to 10 diverse target languages, such as Spanish, Korean, and Finnish, was quite similar. We also found evidence that this transfer is not related to language contamination or language proximity, which strengthens the hypothesis that the model also relies on language-agnostic knowledge. Our experiments have opened up new possibilities for measuring how much data represents the language-agnostic representations learned during pretraining.

pdf bib
MisgenderMender: A Community-Informed Approach to Interventions for Misgendering
Tamanna Hossain | Sunipa Dev | Sameer Singh

Content Warning: This paper contains examples of misgendering and erasure that could be offensive and potentially triggering.Misgendering, the act of incorrectly addressing someone’s gender, inflicts serious harm and is pervasive in everyday technologies, yet there is a notable lack of research to combat it. We are the first to address this lack of research into interventions for misgendering by conducting a survey of gender-diverse individuals in the US to understand perspectives about automated interventions for text-based misgendering. Based on survey insights on the prevalence of misgendering, desired solutions, and associated concerns, we introduce a misgendering interventions task and evaluation dataset, MisgenderMender. We define the task with two sub-tasks: (i) detecting misgendering, followed by (ii) correcting misgendering where misgendering is present, in domains where editing is appropriate. MisgenderMender comprises 3790 instances of social media content and LLM-generations about non-cisgender public figures, annotated for the presence of misgendering, with additional annotations for correcting misgendering in LLM-generated text. Using this dataset, we set initial benchmarks by evaluating existing NLP systems and highlighting challenges for future models to address. We release the full dataset, code, and demo at https://tamannahossainkay.github.io/misgendermender/

pdf bib
Interplay of Machine Translation, Diacritics, and Diacritization
Wei-Rui Chen | Ife Adebara | Muhammad Abdul-Mageed

We investigate two research questions: (1) how do machine translation (MT) and diacritization influence the performance of each other in a multi-task learning setting (2) the effect of keeping (vs. removing) diacritics on MT performance. We examine these two questions in both high-resource (HR) and low-resource (LR) settings across 55 different languages (36 African languages and 19 European languages). For (1), results show that diacritization significantly benefits MT in the LR scenario, doubling or even tripling performance for some languages, but harms MT in the HR scenario. We find that MT harms diacritization in LR but benefits significantly in HR for some languages. For (2), MT performance is similar regardless of diacritics being kept or removed. In addition, we propose two classes of metrics to measure the complexity of a diacritical system, finding these metrics to correlate positively with the performance of our diacritization models. Overall, our work provides insights for developing MT and diacritization systems under different data size conditions and may have implications that generalize beyond the 55 languages we investigate.

pdf bib
From Quantity to Quality: Boosting LLM Performance with Self-Guided Data Selection for Instruction Tuning
Ming Li | Yong Zhang | Zhitao Li | Jiuhai Chen | Lichang Chen | Ning Cheng | Jianzong Wang | Tianyi Zhou | Jing Xiao

In the realm of Large Language Models (LLMs), the balance between instruction data quality and quantity is a focal point. Recognizing this, we introduce a self-guided methodology for LLMs to autonomously discern and select cherry samples from open-source datasets, effectively minimizing manual curation and potential cost for instruction tuning an LLM. Our key innovation, the Instruction-Following Difficulty (IFD) metric, emerges as a pivotal metric to identify discrepancies between a model’s expected responses and its intrinsic generation capability. Through the application of IFD, cherry samples can be pinpointed, leading to a marked uptick in model training efficiency. Empirical validations on datasets like Alpaca and WizardLM underpin our findings; with a mere 10% of original data input, our strategy showcases improved results. This synthesis of self-guided cherry-picking and the IFD metric signifies a transformative leap in the instruction tuning of LLMs, promising both efficiency and resource-conscious advancements. Codes, data, and models are available.

pdf bib
Safer-Instruct: Aligning Language Models with Automated Preference Data
Taiwei Shi | Kai Chen | Jieyu Zhao

Reinforcement learning from human feedback (RLHF) is a vital strategy for enhancing model capability in language models. However, annotating preference data for RLHF is a resource-intensive and creativity-demanding process, while existing automatic generation methods face limitations in data diversity and quality. In response, we present Safer-Instruct, a novel pipeline for automatically constructing large-scale preference data. Our approach leverages reversed instruction tuning, instruction induction, and expert model evaluation to efficiently generate high-quality preference data without human annotators. To verify the effectiveness of Safer-Instruct, we apply the pipeline to construct a safety preference dataset as a case study. Finetuning an Alpaca model on this synthetic dataset not only demonstrates improved harmlessness but also outperforms models fine-tuned on human-annotated safety preference data, all the while maintaining a competitive edge in downstream tasks. Importantly, our Safer-Instruct framework is versatile and can be applied to generate preference data across various domains, extending its utility beyond safety preferences. It addresses the challenges in preference data acquisition and advances the development of more capable and responsible AI systems. For dataset and code implementation, see https://github.com/uscnlp-lime/safer-instruct/.

pdf bib
PELMS: Pre-training for Effective Low-Shot Multi-Document Summarization
Joseph Peper | Wenzhao Qiu | Lu Wang

We investigate pre-training techniques for abstractive multi-document summarization (MDS), which is much less studied than summarizing single documents. Though recent work has demonstrated the effectiveness of highlighting information salience for pre-training strategy design, they struggle to generate abstractive and reflective summaries, which are critical properties for MDS. To this end, we present **PELMS**, a pre-trained model that uses pre-training objectives based on semantic coherence heuristics and faithfulness constraints together with unlabeled multi-document inputs, to promote the generation of concise, fluent, and faithful summaries. To support the training of PELMS, we compile **MultiPT**, a multi-document pre-training corpus containing over 93 million documents to form more than 3million unlabeled topic-centric document clusters, covering diverse genres such as product reviews, news, and general knowledge. We perform extensive evaluation of PELMS in low-shot settings on a wide range of MDS datasets. Our approach consistently outperforms competitive comparisons with respect to overall informativeness, abstractiveness, coherence, and faithfulness, and with minimal fine-tuning can match performance of language models at a much larger scale (e.g., GPT-4).

pdf bib
Deceptive Semantic Shortcuts on Reasoning Chains: How Far Can Models Go without Hallucination?
Bangzheng Li | Ben Zhou | Fei Wang | Xingyu Fu | Dan Roth | Muhao Chen

Despite the high performances of large language models (LLMs) across numerous benchmarks, recent research has unveiled their suffering from hallucinations and unfaithful reasoning. This work studies a type of hallucination induced by semantic associations. We investigate to what extent LLMs take shortcuts from certain keyword/entity biases in the prompt instead of following correct reasoning paths. To quantify this phenomenon, we propose a novel probing method and benchmark called EUREQA. EUREQA is an entity-searching task where a model finds a missing entity based on described multi-hop relations with other entities. These deliberately designed multi-hop relations create deceptive semantic associations, and models must stick to the correct reasoning path instead of incorrect shortcuts to find the correct answer.Experiments show that existing LLMs cannot follow correct reasoning paths and resist the attempt of greedy shortcuts, with GPT-4 only achieving 62% accuracy. Analyses provide further evidence that LLMs rely on semantic biases to solve the task instead of proper reasoning, questioning the validity and generalizability of current LLMs’ high performances.

pdf bib
IndiSentiment140: Sentiment Analysis Dataset for Indian Languages with Emphasis on Low-Resource Languages using Machine Translation
Saurabh Kumar | Ranbir Sanasam | Sukumar Nandi

Sentiment analysis, a fundamental aspect of Natural Language Processing (NLP), involves the classification of emotions, opinions, and attitudes in text data. In the context of India, with its vast linguistic diversity and low-resource languages, the challenge is to support sentiment analysis in numerous Indian languages. This study explores the use of machine translation to bridge this gap. The investigation examines the feasibility of machine translation for creating sentiment analysis datasets in 22 Indian languages. Google Translate, with its extensive language support, is employed for this purpose in translating the Sentiment140 dataset. The study aims to provide insights into the practicality of using machine translation in the context of India’s linguistic diversity for sentiment analysis datasets. Our findings indicate that a dataset generated using Google Translate has the potential to serve as a foundational framework for tackling the low-resource challenges commonly encountered in sentiment analysis for Indian languages.

pdf bib
Leveraging LLMs for Synthesizing Training Data Across Many Languages in Multilingual Dense Retrieval
Nandan Thakur | Jianmo Ni | Gustavo Hernandez Abrego | John Wieting | Jimmy Lin | Daniel Cer

There has been limited success for dense retrieval models in multilingual retrieval, due to uneven and scarce training data available across multiple languages. Synthetic training data generation is promising (e.g., InPars or Promptagator), but has been investigated only for English. Therefore, to study model capabilities across both cross-lingual and monolingual retrieval tasks, we develop **SWIM-IR**, a synthetic retrieval training dataset containing 33 (high to very-low resource) languages for fine-tuning multilingual dense retrievers without requiring any human supervision. To construct SWIM-IR, we propose SAP (summarize-then-ask prompting), where the large language model (LLM) generates a textual summary prior to the query generation step. SAP assists the LLM in generating informative queries in the target language. Using SWIM-IR, we explore synthetic fine-tuning of multilingual dense retrieval models and evaluate them robustly on three retrieval benchmarks: XOR-Retrieve (cross-lingual), MIRACL (monolingual) and XTREME-UP (cross-lingual). Our models, called SWIM-X, are competitive with human-supervised dense retrieval models, e.g., mContriever-X, finding that SWIM-IR can cheaply substitute for expensive human-labeled retrieval training data. SWIM-IR dataset and SWIM-X models are available at: https://github.com/google-research-datasets/SWIM-IR.

pdf bib
SCANNER: Knowledge-Enhanced Approach for Robust Multi-modal Named Entity Recognition of Unseen Entities
Hyunjong Ok | Taeho Kil | Sukmin Seo | Jaeho Lee

Recent advances in named entity recognition (NER) have pushed the boundary of the task to incorporate visual signals, leading to many variants, including multi-modal NER (MNER) or grounded MNER (GMNER). A key challenge to these tasks is that the model should be able to generalize to the entities unseen during the training, and should be able to handle the training samples with noisy annotations.To address this obstacle, we propose SCANNER (Span CANdidate detection and recognition for NER), a model capable of effectively handling all three NER variants.SCANNER is a two-stage structure; we extract entity candidates in the first stage and use it as a query to get knowledge, effectively pulling knowledge from various sources.We can boost our performance by utilizing this entity-centric extracted knowledge to address unseen entities.Furthermore, to tackle the challenges arising from noisy annotations in NER datasets, we introduce a novel self-distillation method, enhancing the robustness and accuracy of our model in processing training data with inherent uncertainties.Our approach demonstrates competitive performance on the NER benchmark and surpasses existing methods on both MNER and GMNER benchmarks.Further analysis shows that the proposed distillation and knowledge utilization methods improve the performance of our model on various benchmarks.

pdf bib
A Theory Guided Scaffolding Instruction Framework for LLM-Enabled Metaphor Reasoning
Yuan Tian | Nan Xu | Wenji Mao

Metaphor detection is a challenging task in figurative language processing, which aims to distinguish between metaphorical and literal expressions in text. Existing methods tackle metaphor detection via training or fine-tuning discriminative models on labeled data. However, these approaches struggle to explain the underlying reasoning process behind the metaphorical/literal judgment. Recently, large language models (LLMs) have shown promise in language reasoning tasks. Although promising, LLM-based methods for metaphor detection and reasoning are still faced with the challenging issue of bringing the explainable concepts for metaphor reasoning and their linguistic manifestation. To fill this gap, we propose a novel Theory guided Scaffolding Instruction (TSI) framework that instructs an LLM to infer the underlying reasoning process of metaphor detection guided by metaphor theories for the first time. Our work is inspired by a pedagogical strategy called scaffolding instruction, which encourages educators to provide questioning and support as scaffolding so as to assist learners in constructing the understanding of pedagogical goals step by step. We first construct a metaphor knowledge graph grounded in metaphor theory which serves as the instructional structure to obtain a series of scaffolding questions, directing the LLM to incrementally generate the reasoning process for metaphor understanding through dialogue interactions. During this theory guided instruction process, we explore the LLM’s mastery boundary and provide the relevant knowledge as scaffolding support when the question is beyond the LLM’s capability. Experimental results verify that our method significantly outperforms both the LLM-based reasoning methods and the SOTA methods in metaphor detection, indicating the facilitation of metaphor and instruction theories in guiding LLM-based reasoning process.

pdf bib
Learning to Compress Prompt in Natural Language Formats
Yu-Neng Chuang | Tianwei Xing | Chia-Yuan Chang | Zirui Liu | Xun Chen | Xia Hu

Large language models (LLMs) are great at processing multiple natural language processing tasks, but their abilities are constrained by inferior performance with long context, slow inference speed, and the high cost of computing the results. Deploying LLMs with precise and informative context helps users process large-scale datasets more effectively and cost-efficiently. Existing works rely on compressing long prompt contexts into soft prompts. However, soft prompt compression encounters limitations in transferability across different LLMs, especially API-based LLMs. To this end, this work aims to compress lengthy prompts in the form of natural language with LLM transferability. This poses two challenges: (i) Natural Language (NL) prompts are incompatible with back-propagation, and (ii) NL prompts lack flexibility in imposing length constraints. In this work, we propose a Natural Language Prompt Encapsulation (Nano-Capsulator) framework compressing original prompts into NL formatted Capsule Prompt while maintaining prompt utility and transferability. Specifically, to tackle the first challenge, the Nano-Capsulator is optimized by a reward function that interacts with the proposed semantics preserving loss. To address the second question, the Nano-Capsulator is optimized by a reward function featuring length constraints. Experimental results demonstrate that the Capsule Prompt can reduce 81.4% of the original length, decrease inference latency up to 4.5x, and save 80.1% of budget overheads while providing transferability across diverse LLMs and different datasets.

pdf bib
Automatic, Meta and Human Evaluation for Multimodal Summarization with Multimodal Output
Haojie Zhuang | Wei Emma Zhang | Leon Xie | Weitong Chen | Jian Yang | Quan Sheng

Multimodal summarization with multimodal output (MSMO) has attracted increasing research interests recently as multimodal summary could provide more comprehensive information compared to text-only summary, effectively improving the user experience and satisfaction. As one of the most fundamental components for the development of MSMO, evaluation is an emerging yet underexplored research topic. In this paper, we fill this gap and propose a research framework that studies three research questions of MSMO evaluation: (1) Automatic Evaluation: We propose a novel metric mLLM-EVAL, which utilizes multimodal Large Language Model for MSMO EVALuation. (2) Meta-Evaluation: We create a meta-evaluation benchmark dataset by collecting human-annotated scores for multimodal summaries. With our benchmark, we conduct meta-evaluation analysis to assess the quality of different evaluation metrics and show the effectiveness of our proposed mLLM-EVAL. (3) Human Evaluation: To provide more objective and unbiased human annotations for meta-evaluation, we hypothesize and verify three types of cognitive biases in human evaluation. We also incorporate our findings into the human annotation process in the meta-evaluation benchmark. Overall, our research framework provides an evaluation metric, a meta-evaluation benchmark dataset annotated by humans and an analysis of cognitive biases in human evaluation, which we believe would serve as a valuable and comprehensive resource for the MSMO research community.

pdf bib
Naive Bayes-based Context Extension for Large Language Models
Jianlin Su | Murtadha Ahmed | Bo Wen | Luo Ao | Mingren Zhu | Yunfeng Liu

Large Language Models (LLMs) have shown promising in-context learning abilities. However, conventional In-Context Learning (ICL) approaches are often impeded by length limitations of transformer architecture, which pose challenges when attempting to effectively integrate supervision from a substantial number of demonstration examples. In this paper, we introduce a novel framework, called Naive Bayes-based Context Extension (NBCE), to enable existing LLMs to perform ICL with an increased number of demonstrations by significantly expanding their context size. Importantly, this expansion does not require fine-tuning or dependence on particular model architectures, all the while preserving linear efficiency. NBCE initially splits the context into equal-sized windows fitting the target LLM’s maximum length. Then, it introduces a voting mechanism to select the most relevant window, regarded as the posterior context. Finally, it employs Bayes’ theorem to generate the test task. Our experimental results demonstrate that NBCE substantially enhances performance, particularly as the number of demonstration examples increases, consistently outperforming alternative methods. The NBCE code will be made publicly accessible. The code NBCE is available at: https://github.com/amurtadha/NBCE-master

pdf bib
Leitner-Guided Memory Replay for Cross-lingual Continual Learning
Meryem M’hamdi | Jonathan May

Cross-lingual continual learning aims to continuously fine-tune a downstream model on emerging data from new languages. One major challenge in cross-lingual continual learning is catastrophic forgetting: a stability-plasticity dilemma, where performance on previously seen languages decreases as the model learns to transfer to new languages. Experience replay, which revisits data from a fixed-size memory of old languages while training on new ones, is among the most successful approaches for solving this dilemma. Faced with the challenge of dynamically storing the memory with high-quality examples while complying with its fixed size limitations, we consider Leitner queuing, a human-inspired spaced-repetition technique, to determine what should be replayed at each phase of learning. Via a controlled set of quantitative and qualitative analyses across different memory strategies, we show that, just like humans, carefully picking informative examples to be prioritized in cross-lingual memory replay helps tame the stability-plasticity dilemma. Compared to vanilla and strong memory replay baselines, our Leitner-guided approach significantly and consistently decreases forgetting while maintaining accuracy across natural language understanding tasks, language orders, and languages.

pdf bib
Multilingual Nonce Dependency Treebanks: Understanding how Language Models Represent and Process Syntactic Structure
David Arps | Laura Kallmeyer | Younes Samih | Hassan Sajjad

We introduce SPUD (Semantically Perturbed Universal Dependencies), a framework for creating nonce treebanks for the multilingual Universal Dependencies (UD) corpora. SPUD data satisfies syntactic argument structure, provides syntactic annotations, and ensures grammaticality via language-specific rules. We create nonce data in Arabic, English, French, German, and Russian, and demonstrate two use cases of SPUD treebanks. First, we investigate the effect of nonce data on word co-occurrence statistics, as measured by perplexity scores of autoregressive (ALM) and masked language models (MLM). We find that ALM scores are significantly more affected by nonce data than MLM scores. Second, we show how nonce data affects the performance of syntactic dependency probes. We replicate the findings of Müller-Eberstein et al. (2022) on nonce test data and show that the performance declines on both MLMs and ALMs wrt. original test data. However, a majority of the performance is kept, suggesting that the probe indeed learns syntax independently from semantics.

pdf bib
Actively Learn from LLMs with Uncertainty Propagation for Generalized Category Discovery
Jinggui Liang | Lizi Liao | Hao Fei | Bobo Li | Jing Jiang

Generalized category discovery faces a key issue: the lack of supervision for new and unseen data categories. Traditional methods typically combine supervised pretraining with self-supervised learning to create models, and then employ clustering for category identification. However, these approaches tend to become overly tailored to known categories, failing to fully resolve the core issue. Hence, we propose to integrate the feedback from LLMs into an active learning paradigm. Specifically, our method innovatively employs uncertainty propagation to select data samples from high-uncertainty regions, which are then labeled using LLMs through a comparison-based prompting scheme. This not only eases the labeling task but also enhances accuracy in identifying new categories. Additionally, a soft feedback propagation mechanism is introduced to minimize the spread of inaccurate feedback. Experiments on various datasets demonstrate our framework’s efficacy and generalizability, significantly improving baseline models at a nominal average cost.

pdf bib
Explaining Text Similarity in Transformer Models
Alexandros Vasileiou | Oliver Eberle

As Transformers have become state-of-the-art models for natural language processing (NLP) tasks, the need to understand and explain their predictions is increasingly apparent. Especially in unsupervised applications, such as information retrieval tasks, similarity models built on top of foundation model representations have been widely applied. However, their inner prediction mechanisms have mostly remained opaque. Recent advances in explainable AI have made it possible to mitigate these limitations by leveraging improved explanations for Transformers through layer-wise relevance propagation (LRP). Using BiLRP, an extension developed for computing second-order explanations in bilinear similarity models, we investigate which feature interactions drive similarity in NLP models. We validate the resulting explanations and demonstrate their utility in three corpus-level use cases, analyzing grammatical interactions, multilingual semantics, and biomedical text retrieval. Our findings contribute to a deeper understanding of different semantic similarity tasks and models, highlighting how novel explainable AI methods enable in-depth analyses and corpus-level insights.

pdf bib
Large Language Models can Contrastively Refine their Generation for Better Sentence Representation Learning
Huiming Wang | Zhaodonghui Li | Liying Cheng | De Wen Soh | Lidong Bing

Recently, large language models (LLMs) have emerged as a groundbreaking technology and their unparalleled text generation capabilities have sparked interest in their application to the fundamental sentence representation learning task. Existing methods have explored utilizing LLMs as data annotators to generate synthesized data for training contrastive learning based sentence embedding models such as SimCSE. However, since contrastive learning models are sensitive to the quality of sentence pairs, the effectiveness of these methods is largely influenced by the content generated from LLMs, highlighting the need for more refined generation in the context of sentence representation learning. Building upon this premise, we propose MultiCSR, a multi-level contrastive sentence representation learning framework that decomposes the process of prompting LLMs to generate a corpus for training base sentence embedding models into three stages (i.e., sentence generation, sentence pair construction, in-batch training) and refines the generated content at these three distinct stages, ensuring only high-quality sentence pairs are utilized to train a base contrastive learning model. Our extensive experiments reveal that MultiCSR enables a less advanced LLM to surpass the performance of ChatGPT, while applying it to ChatGPT achieves better state-of-the-art results. Comprehensive analyses further underscore the potential of our framework in various application scenarios and achieving better sentence representation learning with LLMs.

pdf bib
HIL: Hybrid Isotropy Learning for Zero-shot Performance in Dense retrieval
Jaeyoung Kim | Dohyeon Lee | Seung-won Hwang

Advancements in dense retrieval models have brought ColBERT to prominence in Information Retrieval (IR) with its advanced interaction techniques.However, ColBERT is reported to frequently underperform in zero-shot scenarios, where traditional techniques such as BM25 still exceed it.Addressing this, we propose to balance representation isotropy and anisotropy for zero-shot model performance, based on our observations that isotropy can enhance cosine similarity computations and anisotropy may aid in generalizing to unseen data.Striking a balance between these isotropic and anisotropic qualities stands as a critical objective to refine model efficacy.Based on this, we present ours, a Hybrid Isotropy Learning (HIL) architecture that integrates isotropic and anisotropic representations.Our experiments with the BEIR benchmark show that our model significantly outperforms the baseline ColBERT model, highlighting the importance of harmonized isotropy in improving zero-shot retrieval performance.

pdf bib
SuperGLEBer: German Language Understanding Evaluation Benchmark
Jan Pfister | Andreas Hotho

We assemble a broad Natural Language Understanding benchmark suite for the German language and consequently evaluate a wide array of existing German-capable models in order to create a better understanding of the current state of German LLMs. Our benchmark consists of 29 different tasks ranging over different types such as document classification, sequence tagging, sentence similarity, and question answering, on which we evaluate 10 different German-pretrained models, thereby charting the landscape of German LLMs. In our comprehensive evaluation we find that encoder models are a good choice for most tasks, but also that the largest encoder model does not necessarily perform best for all tasks. We make our benchmark suite and a leaderboard publically available at https://supergleber.professor-x.de and encourage the community to contribute new tasks and evaluate more models on it (https://github.com/LSX-UniWue/SuperGLEBer).

pdf bib
“You are an expert annotator”: Automatic Best–Worst-Scaling Annotations for Emotion Intensity Modeling
Christopher Bagdon | Prathamesh Karmalkar | Harsha Gurulingappa | Roman Klinger

Labeling corpora constitutes a bottleneck to create models for new tasks or domains. Large language models mitigate the issue with automatic corpus labeling methods, particularly for categorical annotations. Some NLP tasks such as emotion intensity prediction, however, require text regression, but there is no work on automating annotations for continuous label assignments. Regression is considered more challenging than classification: The fact that humans perform worse when tasked to choose values from a rating scale lead to comparative annotation methods, including best–worst scaling. This raises the question if large language model-based annotation methods show similar patterns, namely that they perform worse on rating scale annotation tasks than on comparative annotation tasks. To study this, we automate emotion intensity predictions and compare direct rating scale predictions, pairwise comparisons and best–worst scaling. We find that the latter shows the highest reliability. A transformer regressor fine-tuned on these data performs nearly on par with a model trained on the original manual annotations.

pdf bib
What Matters in Training a GPT4-Style Language Model with Multimodal Inputs?
Yan Zeng | Hanbo Zhang | Jiani Zheng | Jiangnan Xia | Guoqiang Wei | Yang Wei | Yuchen Zhang | Tao Kong | Ruihua Song

Recent advancements in GPT-4V have displayed remarkable multi-modal capabilities in processing image inputs and following open-ended instructions. Despite these advancements, there is considerable scope for enhancing open-source multi-modal LLMs, especially in terms of multi-modal understanding accuracy and instruction-following proficiency. In this paper, we conduct a comprehensive study on training GPT4-style models. We introduce Lynx a multi-modal LLM developed through a series of controlled experiments comparing various model variants. This process allowed us to identify and implement an optimal training strategy tailored for multi-modal LLMs. In addition to our model development, we propose a plug-and-play technique designed to augment the instruction-following capabilities of multi-modal LLMs. We have validated the performance of Lynx on multiple benchmarks. Results demonstrate that Lynx not only achieves strong image understanding accuracy but also excels in instruction-following tasks, paving the path for ongoing enhancements in multi-modal LLMs.

pdf bib
Defining and Detecting Vulnerability in Human Evaluation Guidelines: A Preliminary Study Towards Reliable NLG Evaluation
Jie Ruan | WangWenqing WangWenqing | Xiaojun Wan

Human evaluation serves as the gold standard for assessing the quality of Natural Language Generation (NLG) systems. Nevertheless, the evaluation guideline, as a pivotal element ensuring reliable and reproducible human assessment, has received limited attention. Our investigation revealed that only 29.84% of recent papers involving human evaluation at top conferences release their evaluation guidelines, with vulnerabilities identified in 77.09% of these guidelines. Unreliable evaluation guidelines can yield inaccurate assessment outcomes, potentially impeding the advancement of NLG in the right direction. To address these challenges, we take an initial step towards reliable evaluation guidelines and propose the first human evaluation guideline dataset by collecting annotations of guidelines extracted from existing papers as well as generated via Large Language Models (LLMs). We then introduce a taxonomy of eight vulnerabilities and formulate a principle for composing evaluation guidelines. Furthermore, a method for detecting guideline vulnerabilities has been explored using LLMs, and we offer a set of recommendations to enhance reliability in human evaluation. The annotated human evaluation guideline dataset and code for the vulnerability detection method are publicly available online.

pdf bib
MOSAICo: a Multilingual Open-text Semantically Annotated Interlinked Corpus
Simone Conia | Edoardo Barba | Abelardo Carlos Martinez Lorenzo | Pere-Lluís Huguet Cabot | Riccardo Orlando | Luigi Procopio | Roberto Navigli

Several Natural Language Understanding (NLU) tasks focus on linking text to explicit knowledge, including Word Sense Disambiguation, Semantic Role Labeling, Semantic Parsing, and Relation Extraction. In addition to the importance of connecting raw text with explicit knowledge bases, the integration of such carefully curated knowledge into deep learning models has been shown to be beneficial across a diverse range of applications, including Language Modeling and Machine Translation. Nevertheless, the scarcity of semantically-annotated corpora across various tasks and languages limits the potential advantages significantly. To address this issue, we put forward MOSAICo, the first endeavor aimed at equipping the research community with the key ingredients to model explicit semantic knowledge at a large scale, providing hundreds of millions of silver yet high-quality annotations for four NLU tasks across five languages. We describe the creation process of MOSAICo, demonstrate its quality and variety, and analyze the interplay between different types of semantic information. MOSAICo, available at https://github.com/SapienzaNLP/mosaico, aims to drop the requirement of closed, licensed datasets and represents a step towards a level playing field across languages and tasks in NLU.

pdf bib
SemRoDe: Macro Adversarial Training to Learn Representations that are Robust to Word-Level Attacks
Brian Formento | Wenjie Feng | Chuan-Sheng Foo | Anh Tuan Luu | See-Kiong Ng

Language models (LMs) are indispensable tools for natural language processing tasks, but their vulnerability to adversarial attacks remains a concern. While current research has explored adversarial training techniques, their improvements to defend against word-level attacks have been limited. In this work, we propose a novel approach called Semantic Robust Defence (SemRoDe), a Macro Adversarial Training strategy to enhance the robustness of LMs. Drawing inspiration from recent studies in the image domain, we investigate and later confirm that in a discrete data setting such as language, adversarial samples generated via word substitutions do indeed belong to an adversarial domain exhibiting a high Wasserstein distance from the base domain. Our method learns a robust representation that bridges these two domains. We hypothesize that if samples were not projected into an adversarial domain, but instead to a domain with minimal shift, it would improve attack robustness. We align the domains by incorporating a new distance-based objective. With this, our model is able to learn more generalized representations by aligning the model’s high-level output features and therefore better handling unseen adversarial samples. This method can be generalized across word embeddings, even when they share minimal overlap at both vocabulary and word-substitution levels. To evaluate the effectiveness of our approach, we conduct experiments on BERT and RoBERTa models on three datasets. The results demonstrate promising state-of-the-art robustness.

pdf bib
BUST: Benchmark for the evaluation of detectors of LLM-Generated Text
Joseph Cornelius | Oscar Lithgow-Serrano | Sandra Mitrovic | Ljiljana Dolamic | Fabio Rinaldi

We introduce BUST, a comprehensive benchmark designed to evaluate detectors of texts generated by instruction-tuned large language models (LLMs). Unlike previous benchmarks, our focus lies on evaluating the performance of detector systems, acknowledging the inevitable influence of the underlying tasks and different LLM generators. Our benchmark dataset consists of 25K texts from humans and 7 LLMs responding to instructions across 10 tasks from 3 diverse sources. Using the benchmark, we evaluated 5 detectors and found substantial performance variance across tasks. A meta-analysis of the dataset characteristics was conducted to guide the examination of detector performance. The dataset was analyzed using diverse metrics assessing linguistic features like fluency and coherence, readability scores, and writer attitudes, such as emotions, convincingness, and persuasiveness. Features impacting detector performance were investigated with surrogate models, revealing emotional content in texts enhanced some detectors, yet the most effective detector demonstrated consistent performance, irrespective of writer’s attitudes and text styles. Our approach focused on investigating relationships between the detectors’ performance and two key factors: text characteristics and LLM generators. We believe BUST will provide valuable insights into selecting detectors tailored to specific text styles and tasks and facilitate a more practical and in-depth investigation of detection systems for LLM-generated text.

pdf bib
Improving In-context Learning of Multilingual Generative Language Models with Cross-lingual Alignment
Chong Li | Shaonan Wang | Jiajun Zhang | Chengqing Zong

Multilingual generative models obtain remarkable cross-lingual in-context learning capabilities through pre-training on large-scale corpora. However, they still exhibit a performance bias toward high-resource languages and learn isolated distributions of multilingual sentence representations, which may hinder knowledge transfer across languages. To bridge this gap, we propose a simple yet effective cross-lingual alignment framework exploiting pairs of translation sentences. It aligns the internal sentence representations across different languages via multilingual contrastive learning and aligns outputs by following cross-lingual instructions in the target language. Experimental results show that even with less than 0.1\textperthousand of pre-training tokens, our alignment framework significantly boosts the cross-lingual abilities of generative language models and mitigates the performance gap. Further analyses reveal that it results in a better internal multilingual representation distribution of multilingual models.

pdf bib
MaCSC: Towards Multimodal-augmented Pre-trained Language Models via Conceptual Prototypes and Self-balancing Calibration
Xianwei Zhuang | Zhichang Wang | Xuxin Cheng | Yuxin Xie | Liming Liang | Yuexian Zou

Pre-trained language models (PLMs) that rely solely on textual data may exhibit limitations in multimodal semantics comprehension. Existing solutions attempt to alleviate this issue by incorporating explicit image retrieval or generation techniques.However, these methods: (1) focus exclusively on the static image modality; (2) inevitably encounter modality gaps and noise; (3) indiscriminately treat all modalities.In this paper, we propose a novel multimodal-augmented framework termed MaCSC, which can infuse multimodal semantics into PLMs and facilitate a self-balancing calibration of information allocation.Specifically, MaCSC obtains modal-specific conceptual prototypes from contrastive pre-training models (e.g., CLIP),and aggregates the intra- and inter-modal semantics of the conceptual prototype to enhance PLMs.In addition, we utilize a novel self-balancing contrastive loss to achieve multi-scale self-balancing calibration of multimodal information during fine-tuning PLMs.Experimental results show that MaCSC consistently improves the performance of PLMs across various architectures and scales, and outperforms competitive baselines on multiple NLP tasks.

pdf bib
Does Pre-trained Language Model Actually Infer Unseen Links in Knowledge Graph Completion?
Yusuke Sakai | Hidetaka Kamigaito | Katsuhiko Hayashi | Taro Watanabe

Knowledge graphs (KGs) consist of links that describe relationships between entities. Due to the difficulty of manually enumerating all relationships between entities, automatically completing them is essential for KGs. Knowledge Graph Completion (KGC) is a task that infers unseen relationships between entities in a KG. Traditional embedding-based KGC methods (e.g. RESCAL, TransE, DistMult, ComplEx, RotatE, HAKE, HousE, etc.) infer missing links using only the knowledge from training data. In contrast, the recent Pre-trained Language Model (PLM)-based KGC utilizes knowledge obtained during pre-training, which means it can estimate missing links between entities by reusing memorized knowledge from pre-training without inference. This part is problematic because building KGC models aims to infer unseen links between entities. However, conventional evaluations in KGC do not consider inference and memorization abilities separately. Thus, a PLM-based KGC method, which achieves high performance in current KGC evaluations, may be ineffective in practical applications. To address this issue, we analyze whether PLM-based KGC methods make inferences or merely access memorized knowledge. For this purpose, we propose a method for constructing synthetic datasets specified in this analysis and conclude that PLMs acquire the inference abilities required for KGC through pre-training, even though the performance improvements mostly come from textual information of entities and relations.

pdf bib
Discovering Lobby-Parliamentarian Alignments through NLP
Aswin Suresh | Lazar Radojević | Francesco Salvi | Antoine Magron | Victor Kristof | Matthias Grossglauser

We discover alignments of views between interest groups (lobbies) and members of the European Parliament (MEPs) by automatically analyzing their texts. Specifically, we do so by collecting novel datasets of lobbies’ position papers and MEPs’ speeches, and comparing these texts on the basis of semantic similarity and entailment. In the absence of ground-truth, we perform an indirect validation by comparing the discovered alignments with a dataset, which we curate, of retweet links between MEPs and lobbies, and with the publicly disclosed meetings of MEPs. Our best method performs significantly better than several baselines. Moreover, an aggregate analysis of the discovered alignments, between groups of related lobbies and political groups of MEPs, correspond to the expectations from the ideology of the groups (e.g., groups on the political left are more aligned with humanitarian and environmental organisations). We believe that this work is a step towards enhancing the transparency of the intricate decision-making processes within democratic institutions.

pdf bib
IterCQR: Iterative Conversational Query Reformulation with Retrieval Guidance
Yunah Jang | Kang-il Lee | Hyunkyung Bae | Hwanhee Lee | Kyomin Jung

Conversational search aims to retrieve passages containing essential information to answer queries in a multi-turn conversation. In conversational search, reformulating context-dependent conversational queries into stand-alone forms is imperative to effectively utilize off-the-shelf retrievers. Previous methodologies for conversational query reformulation frequently depend on human-annotated rewrites.However, these manually crafted queries often result in sub-optimal retrieval performance and require high collection costs.To address these challenges, we propose **Iter**ative **C**onversational **Q**uery **R**eformulation (**IterCQR**), a methodology that conducts query reformulation without relying on human rewrites. IterCQR iteratively trains the conversational query reformulation (CQR) model by directly leveraging information retrieval (IR) signals as a reward.Our IterCQR training guides the CQR model such that generated queries contain necessary information from the previous dialogue context.Our proposed method shows state-of-the-art performance on two widely-used datasets, demonstrating its effectiveness on both sparse and dense retrievers. Moreover, IterCQR exhibits superior performance in challenging settings such as generalization on unseen datasets and low-resource scenarios.

pdf bib
AceGPT, Localizing Large Language Models in Arabic
Huang Huang | Fei Yu | Jianqing Zhu | Xuening Sun | Hao Cheng | Song Dingjie | Zhihong Chen | Mosen Alharthi | Bang An | Juncai He | Ziche Liu | Junying Chen | Jianquan Li | Benyou Wang | Lian Zhang | Ruoyu Sun | Xiang Wan | Haizhou Li | Jinchao Xu

This paper is devoted to the development of a localized Large Language Model (LLM) specifically for Arabic, a language imbued with unique cultural characteristics inadequately addressed by current mainstream models. Significant concerns emerge when addressing cultural sensitivity and local values. To address this, the paper proposes a comprehensive solution that includes further pre-training with Arabic texts, Supervised Fine-Tuning (SFT) utilizing native Arabic instructions, and GPT-4 responses in Arabic, alongside Reinforcement Learning with AI Feedback (RLAIF) employing a reward model attuned to local culture and values. The goal is to cultivate culturally cognizant and value-aligned Arabic LLMs capable of accommodating the diverse, application-specific needs of Arabic-speaking communities. Comprehensive evaluations reveal that the resulting model, dubbed ‘AceGPT’, sets the state-of-the-art standard for open Arabic LLMs across various benchmarks. Codes, data, and models are in https://github.com/FreedomIntelligence/AceGPT.

pdf bib
Improving Machine Translation with Human Feedback: An Exploration of Quality Estimation as a Reward Model
Zhiwei He | Xing Wang | Wenxiang Jiao | Zhuosheng Zhang | Rui Wang | Shuming Shi | Zhaopeng Tu

Insufficient modeling of human preferences within the reward model is a major obstacle for leveraging human feedback to improve translation quality. Fortunately, quality estimation (QE), which predicts the quality of a given translation without reference, has achieved impressive alignment with human evaluations in the last two years. In this work, we investigate the potential of employing the QE model as the reward model to predict human preferences for feedback training. We first identify the overoptimization problem during QE-based feedback training, manifested as an increase in reward while translation quality declines. We examine the problem and argue that the vulnerability of the QE model might lead to high rewards for incorrect translations, resulting in overoptimization and error propagation. To address the problem, we adopt a simple yet effective method that uses heuristic rules to detect the incorrect translations and assigns a penalty term to the reward scores of them. Experimental results show that the proposed QE-based feedback training achieves consistent and significant improvements across various settings, further verified through human preference studies. Our subsequent analysis demonstrates the high data efficiency of the proposed QE-based feedback training: it outperforms systems using larger parallel corpora by a small amount of monolingual data. Our code is available at: https://github.com/zwhe99/FeedbackMT

pdf bib
Depression Detection in Clinical Interviews with LLM-Empowered Structural Element Graph
Zhuang Chen | Jiawen Deng | Jinfeng Zhou | Jincenzi Wu | Tieyun Qian | Minlie Huang

Depression is a widespread mental health disorder affecting millions globally. Clinical interviews are the gold standard for assessing depression, but they heavily rely on scarce professional clinicians, highlighting the need for automated detection systems. However, existing methods only capture part of the relevant elements in clinical interviews, unable to incorporate all depressive cues. Moreover, the scarcity of participant data, due to privacy concerns and collection challenges, intrinsically constrains interview modeling. To address these limitations, in this paper, we propose a structural element graph (SEGA), which transforms the clinical interview into an expertise-inspired directed acyclic graph for comprehensive modeling. Additionally, we further empower SEGA by devising novel principle-guided data augmentation with large language models (LLMs) to supplement high-quality synthetic data and enable graph contrastive learning. Extensive evaluations on two real-world clinical datasets, in both English and Chinese, show that SEGA significantly outperforms baseline methods and powerful LLMs like GPT-3.5 and GPT-4.

pdf bib
SQATIN: Supervised Instruction Tuning Meets Question Answering for Improved Dialogue NLU
Evgeniia Razumovskaia | Goran Glavaš | Anna Korhonen | Ivan Vulić

Task-oriented dialogue (TOD) systems help users execute well-defined tasks across a variety of domains (e.g., flight booking or food ordering), with their Natural Language Understanding (NLU) components being dedicated to the analysis of user utterances, predicting users’ intents (Intent Detection, ID) and extracting values for informational slots (Value Extraction, VE). In most domains, labelled NLU data is scarce, making sample-efficient learning – enabled with effective transfer paradigms – paramount. In this work, we introduce SQATIN, a new framework for dialog NLU based on (i) instruction tuning and (ii) question-answering-based formulation of ID and VE tasks. According to the evaluation on established NLU benchmarks, SQATIN sets the new state of the art in dialogue NLU, substantially surpassing the performance of current models based on standard fine-tuning objectives in both in-domain training and cross-domain transfer, and it also surpasses off-the-shelf large language models for the same task, both in terms of performance and inference efficiency. Furthermore, SQATIN yields particularly large performance gains in cross-domain transfer, owing to the fact that our QA-based instruction tuning leverages similarities between natural language descriptions of classes (i.e., slots and intents) across domains.

pdf bib
Enhancing Argument Summarization: Prioritizing Exhaustiveness in Key Point Generation and Introducing an Automatic Coverage Evaluation Metric
Mohammad Khosravani | Chenyang Huang | Amine Trabelsi

The proliferation of social media platforms has given rise to the amount of online debates and arguments. Consequently, the need for automatic summarization methods for such debates is imperative, however this area of summarization is rather understudied. The Key Point Analysis (KPA) task formulates argument summarization as representing the summary of a large collection of arguments in the form of concise sentences in bullet-style format, called key points. A sub-task of KPA, called Key Point Generation (KPG), focuses on generating these key points given the arguments. This paper introduces a novel extractive approach for key point generation, that outperforms previous state-of-the-art methods for the task. Our method utilizes an extractive clustering based approach that offers concise, high quality generated key points with higher coverage of reference summaries, and less redundant outputs. In addition, we show that the existing evaluation metrics for summarization such as ROUGE are incapable of differentiating between generated key points of different qualities. To this end, we propose a new evaluation metric for assessing the generated key points by their coverage. Our code can be accessed online.

pdf bib
ARM: Alignment with Residual Energy-Based Model
Bo Pang | Caiming Xiong | Yingbo Zhou

While large language models (LLMs) trained with large-scale unsupervised learning acquire a wide variety of world knowledge and skills, its behavior does not necessarily align with human preferences. RLHF methods achieve successes in aligning LLM responses with human preferences and improving the controllability of LLM behavior with human instruction. However, RLHF methods are considerably complicated to implement, computationally expensive to train, and notoriously tricky to tune. In this work, we propose Alignment with Residual Energy-Based Model (ARM), as a simple and flexible alternative to RLHF methods. Our method is driven by an observation that we can learn an aligned policy by minimizing a forward Kullback–Leibler (KL) divergence from a target policy (in the form of a residual energy-based model) to a parameteric policy (LLM), instead of a reverse KL as in RLHF methods. With samples from the energy-based target policy, we can leverage the power of DPO (or other offline methods) to learn an aligned policy efficiently. ARM is simple to implement and applicable in various data settings. Our extensive experiments demonstrate its strong performance across multiple datasets, compared to strong baselines like PPO, DPO.

pdf bib
HumanRankEval: Automatic Evaluation of LMs as Conversational Assistants
Milan Gritta | Gerasimos Lampouras | Ignacio Iacobacci

Language models (LMs) as conversational assistants recently became popular tools that help people accomplish a variety of tasks. These typically result from adapting LMs pretrained on general domain text sequences through further instruction-tuning and possibly preference optimisation methods. The evaluation of such LMs would ideally be performed using human judgement, however, this is not scalable. On the other hand, automatic evaluation featuring auxiliary LMs as judges and/or knowledge-based tasks is scalable but struggles with assessing conversational ability and adherence to instructions. To help accelerate the development of LMs as conversational assistants, we propose a novel automatic evaluation task: HumanRankEval (HRE). It consists of a large-scale, diverse and high-quality set of questions, each with several answers authored and scored by humans. To perform evaluation, HRE ranks these answers based on their log-likelihood under the LM’s distribution, and subsequently calculates their correlation with the corresponding human rankings. We support HRE’s efficacy by investigating how efficiently it separates pretrained and instruction-tuned LMs of various sizes. We show that HRE correlates well with human judgements and is particularly responsive to model changes following instruction-tuning.

pdf bib
FAMuS: Frames Across Multiple Sources
Siddharth Vashishtha | Alexander Martin | William Gantt | Benjamin Van Durme | Aaron White

Understanding event descriptions is a central aspect of language processing, but current approaches focus overwhelmingly on single sentences or documents. Aggregating information about an event across documents can offer a much richer understanding. To this end, we present FAMuS, a new corpus of Wikipedia passages that report on some event, paired with underlying, genre-diverse (non-Wikipedia) source articles for the same event. Events and (cross-sentence) arguments in both report and source are annotated against FrameNet, providing broad coverage of different event types. We present results on two key event understanding tasks enabled by FAMuS: source validation—determining whether a document is a valid source for a target report event—and cross-document argument extraction—full-document argument extraction for a target event from both its report and the correct source article.

pdf bib
Rationale-based Opinion Summarization
Haoyuan Li | Snigdha Chaturvedi

Opinion summarization aims to generate concise summaries that present popular opinions of a large group of reviews. However, these summaries can be too generic and lack supporting details. To address these issues, we propose a new paradigm for summarizing reviews, rationale-based opinion summarization. Rationale-based opinion summaries output the representative opinions as well as one or more corresponding rationales. To extract good rationales, we define four desirable properties: relatedness, specificity, popularity, and diversity and present a Gibbs-sampling-based method to extract rationales. Overall, we propose RATION, an unsupervised extractive system that has two components: an Opinion Extractor (to extract representative opinions) and Rationales Extractor (to extract corresponding rationales). We conduct automatic and human evaluations to show that rationales extracted by RATION have the proposed properties and its summaries are more useful than conventional summaries. The implementation of our work is available at https://github.com/leehaoyuan/RATION.

pdf bib
Mustango: Toward Controllable Text-to-Music Generation
Jan Melechovsky | Zixun Guo | Deepanway Ghosal | Navonil Majumder | Dorien Herremans | Soujanya Poria

The quality of the text-to-music models has reached new heights due to recent advancements in diffusion models. The controllability of various musical aspects, however, has barely been explored. In this paper, we propose Mustango: a music-domain-knowledge-inspired text-to-music system based on diffusion. Mustango aims to control the generated music, not only with general text captions, but with more rich captions that can include specific instructions related to chords, beats, tempo, and key. At the core of Mustango is MuNet, a Music-Domain-Knowledge-Informed UNet guidance module that steers the generated music to include the music-specific conditions, which we predict from the text prompt, as well as the general text embedding, during the reverse diffusion process. To overcome the limited availability of open datasets of music with text captions, we propose a novel data augmentation method that includes altering the harmonic, rhythmic, and dynamic aspects of music audio and using state-of-the-art Music Information Retrieval methods to extract the music features which will then be appended to the existing descriptions in text format. We release the resulting MusicBench dataset which contains over 52K instances and includes music-theory-based descriptions in the caption text. Through extensive experiments, we show that the quality of the music generated by Mustango is state-of-the-art, and the controllability through music-specific text prompts greatly outperforms other models such as MusicGen and AudioLDM2.

pdf bib
Adaptive Cross-lingual Text Classification through In-Context One-Shot Demonstrations
Emilio Cueva | Adrian Lopez Monroy | Fernando Sánchez-Vega | Thamar Solorio

Zero-Shot Cross-lingual Transfer (ZS-XLT) utilizes a model trained in a source language to make predictions in another language, often with a performance loss. To alleviate this, additional improvements can be achieved through subsequent adaptation using examples in the target language. In this paper, we exploit In-Context Tuning (ICT) for One-Shot Cross-lingual transfer in the classification task by introducing In-Context Cross-lingual Transfer (IC-XLT). The novel concept involves training a model to learn from context examples and subsequently adapting it during inference to a target language by prepending a One-Shot context demonstration in that language. Our results show that IC-XLT successfully leverages target-language examples to improve the cross-lingual capabilities of the evaluated mT5 model, outperforming prompt-based models in the Zero and Few-shot scenarios adapted through fine-tuning. Moreover, we show that when source-language data is limited, the fine-tuning framework employed for IC-XLT performs comparably to prompt-based fine-tuning with significantly more training data in the source language.

pdf bib
CNER: Concept and Named Entity Recognition
Giuliano Martinelli | Francesco Molfese | Simone Tedeschi | Alberte Fernández-Castro | Roberto Navigli

Named entities – typically expressed via proper nouns – play a key role in Natural Language Processing, as their identification and comprehension are crucial in tasks such as Relation Extraction, Coreference Resolution and Question Answering, among others. Tasks like these also often entail dealing with concepts – typically represented by common nouns – which, however, have not received as much attention. Indeed, the potential of their identification and understanding remains underexplored, as does the benefit of a synergistic formulation with named entities. To fill this gap, we introduce Concept and Named Entity Recognition (CNER), a new unified task that handles concepts and entities mentioned in unstructured texts seamlessly. We put forward a comprehensive set of categories that can be used to model concepts and named entities jointly, and propose new approaches for the creation of CNER datasets. We evaluate the benefits of performing CNER as a unified task extensively, showing that a CNER model gains up to +5.4 and +8 macro F1 points when compared to specialized named entity and concept recognition systems, respectively. Finally, to encourage the development of CNER systems, we release our datasets and models at https://github.com/Babelscape/cner.

pdf bib
Branch-Solve-Merge Improves Large Language Model Evaluation and Generation
Swarnadeep Saha | Omer Levy | Asli Celikyilmaz | Mohit Bansal | Jason Weston | Xian Li

Large Language Models (LLMs) are frequently used for multi-faceted language generation and evaluation tasks that involve satisfying intricate user constraints or taking into account multiple aspects and criteria. However, their performance can fall short, due to the model’s lack of coherence and inability to plan and decompose the problem. We propose Branch-Solve-Merge (BSM), a Large Language Model program (Schlag et al., 2023) for tackling such challenging natural language tasks. It consists of branch, solve, and merge modules that are parameterized with specific prompts to the base LLM. These three modules plan a decomposition of the task into multiple parallel sub-tasks, independently solve them, and fuse the solutions to the sub-tasks. We apply our method to the tasks of LLM response evaluation and constrained text generation and evaluate its effectiveness with multiple LLMs, including Vicuna, LLaMA-2-chat, and GPT-4. BSM improves the evaluation correctness and consistency for each LLM by enhancing human-LLM agreement by up to 26%, reducing length and pairwise position biases by up to 50%, and allowing LLaMA-2-chat to match or outperform GPT-4 on most domains. On a constraint story generation task, BSM improves the coherence of stories while also improving constraint satisfaction by 12%.

pdf bib
REPLUG: Retrieval-Augmented Black-Box Language Models
Weijia Shi | Sewon Min | Michihiro Yasunaga | Minjoon Seo | Richard James | Mike Lewis | Luke Zettlemoyer | Wen-tau Yih

We introduce REPLUG, a retrieval-augmented language modeling framework that treats the language model (LM) as a black box and augments it with a tuneable retrieval model. Unlike prior retrieval-augmented LMs that train language models with special cross-attention mechanisms to encode the retrieved text, REPLUG simply prepends retrieved documents to the input for the frozen black-box LM. This simple design can be easily applied to any existing language models. Furthermore, we show that the LM can be used to supervise the retrieval model, which can then find documents that help the LM make better predictions. Our experiments demonstrate that REPLUG with the tuned retriever significantly improves the performance of GPT-3 (175B) on language modeling by 6.3%, as well as the performance of Codex on five-shot MMLU by 5.1%. Code is publicly released at github.com/swj0419/REPLUG.

pdf bib
David helps Goliath: Inference-Time Collaboration Between Small Specialized and Large General Diffusion LMs
Xiaochuang Han | Sachin Kumar | Yulia Tsvetkov | Marjan Ghazvininejad

Diffusion-based language models are emerging as a promising alternative to autoregressive LMs: they approach the competence of autoregressive LMs while offering nuanced controllability at inference time. While autoregressive LMs have benefited immensely from scaling and instruction-based learning, existing studies of diffusion LMs have been conducted on a smaller scale. Starting with a recently proposed diffusion model SSD-LM, in this work we first explore methods to scale it from 0.4B to 13B parameters, proposing techniques to improve its training and inference efficiency, and to finetune the model to follow instructions. Armed with a more powerful, general purpose diffusion LM, we introduce the primary contribution of this work – SSD-2 – an approach to easily ensemble at inference time a large general-purpose diffusion LM with smaller, but specialized and contextualized diffusion LMs. We show that SSD-2 facilitates novel ensembles with 100x smaller models that can be customized and deployed by individual users. We find that compared to autoregressive models, the collaboration between diffusion LMs is more effective, leading to higher-quality model responses due to their ability to dynamically incorporate bi-directional contexts.

pdf bib
Efficient End-to-End Visual Document Understanding with Rationale Distillation
Wang Zhu | Alekh Agarwal | Mandar Joshi | Robin Jia | Jesse Thomason | Kristina Toutanova

Understanding visually situated language requires interpreting complex layouts of textual and visual elements. Pre-processing tools, such as optical character recognition (OCR), can map document image inputs to textual tokens, then large language models (LLMs) can reason over text.However, such methods have high computational and engineering complexity. Can small pretrained image-to-text models accurately understand visual documents through similar recognition and reasoning steps instead?We propose Rationale Distillation (RD), which incorporates the outputs of OCR tools, LLMs, and larger multimodal models as intermediate “rationales”, and trains a small student model to predict both rationales and answers. On three visual document understanding benchmarks representing infographics, scanned documents, and figures, our Pix2Struct (282M parameters) student model finetuned with RD outperforms the base model by 4-5% absolute accuracy with only 1% higher computational cost.

pdf bib
A Systematic Comparison of Syllogistic Reasoning in Humans and Language Models
Tiwalayo Eisape | Michael Tessler | Ishita Dasgupta | Fei Sha | Sjoerd Steenkiste | Tal Linzen

A central component of rational behavior is logical inference: the process of determining which conclusions follow from a set of premises. Psychologists have documented several ways in which humans’ inferences deviate from the rules of logic. Do language models, which are trained on text generated by humans, replicate such human biases, or are they able to overcome them? Focusing on the case of syllogisms—inferences from two simple premises—we show that, within the PaLM 2 family of transformer language models, larger models are more logical than smaller ones, and also more logical than humans. At the same time, even the largest models make systematic errors, some of which mirror human reasoning biases: they show sensitivity to the (irrelevant) ordering of the variables in the syllogism, and draw confident but incorrect inferences from particular syllogisms (syllogistic fallacies). Overall, we find that language models often mimic the human biases included in their training data, but are able to overcome them in some cases.

pdf bib
AnchorAL: Computationally Efficient Active Learning for Large and Imbalanced Datasets
Pietro Lesci | Andreas Vlachos

Active learning for imbalanced classification tasks is challenging as the minority classes naturally occur rarely. Gathering a large pool of unlabelled data is thus essential to capture minority instances. Standard pool-based active learning is computationally expensive on large pools and often reaches low accuracy by overfitting the initial decision boundary, thus failing to explore the input space and find minority instances. To address these issues we propose AnchorAL. At each iteration, AnchorAL chooses class-specific instances from the labelled set, or *anchors*, and retrieves the most similar unlabelled instances from the pool. This resulting *subpool* is then used for active learning. Using a small, fixed-sized subpool AnchorAL allows scaling any active learning strategy to large pools. By dynamically selecting different anchors at each iteration it promotes class balance and prevents overfitting the initial decision boundary, thus promoting the discovery of new clusters of minority instances. Experiments across different classification tasks, active learning strategies, and model architectures AnchorAL is *(i)* faster, often reducing runtime from hours to minutes, *(ii)* trains more performant models, *(iii)* and returns more balanced datasets than competing methods.

pdf bib
ICLE++: Modeling Fine-Grained Traits for Holistic Essay Scoring
Shengjie Li | Vincent Ng

The majority of the recently developed models for automated essay scoring (AES) are evaluated solely on the ASAP corpus. However, ASAP is not without its limitations. For instance, it is not clear whether models trained on ASAP can generalize well when evaluated on other corpora. In light of these limitations, we introduce ICLE++, a corpus of persuasive student essays annotated with both holistic scores and trait-specific scores. Not only can ICLE++ be used to test the generalizability of AES models trained on ASAP, but it can also facilitate the evaluation of models developed for newer AES problems such as multi-trait scoring and cross-prompt scoring. We believe that ICLE++, which represents a culmination of our long-term effort in annotating the essays in the ICLE corpus, contributes to the set of much-needed annotated corpora for AES research.

pdf bib
UNcommonsense Reasoning: Abductive Reasoning about Uncommon Situations
Wenting Zhao | Justin Chiu | Jena Hwang | Faeze Brahman | Jack Hessel | Sanjiban Choudhury | Yejin Choi | Xiang Li | Alane Suhr

Language technologies that accurately model the dynamics of events must perform commonsense reasoning. Existing work evaluating commonsense reasoning focuses on making inferences about common, everyday situations. To instead investigate the ability to model unusual, unexpected, and unlikely situations, we explore the task of uncommonsense abductive reasoning. Given a piece of context with an unexpected outcome, this task requires reasoning abductively to generate an explanation that makes the unexpected outcome more likely in the context. To this end, we curate and release a new English language corpus called UNcommonsense. We characterize the performance differences between human explainers and the best-performing large language models, finding that model-enhanced human-written explanations achieve the highest quality by trading off between specificity and diversity. Finally, we experiment with several imitation learning algorithms to train open and accessible language models on this task. When compared with the vanilla supervised fine-tuning approach, these methods consistently reduce lose rates on both common and uncommonsense abductive reasoning judged by human evaluators.

pdf bib
To Tell The Truth: Language of Deception and Language Models
Sanchaita Hazra | Bodhisattwa Prasad Majumder

Text-based false information permeates online discourses, yet evidence of people’s ability to discern truth from such deceptive textual content is scarce. We analyze a novel TV game show data where conversations in a high-stake environment between individuals with conflicting objectives result in lies. We investigate the manifestation of potentially verifiable language cues of deception in the presence of objective truth, a distinguishing feature absent in previous text-based deception datasets. We show that there exists a class of detectors (algorithms) that have similar truth detection performance compared to human subjects, even when the former accesses only the language cues while the latter engages in conversations with complete access to all potential sources of cues (language and audio-visual). Our model, built on a large language model, employs a bottleneck framework to learn discernible cues to determine truth, an act of reasoning in which human subjects often perform poorly, even with incentives. Our model detects novel but accurate language cues in many cases where humans failed to detect deception, opening up the possibility of humans collaborating with algorithms and ameliorating their ability to detect the truth.

pdf bib
Multilingual Models for ASR in Chibchan Languages
Rolando Coto-Solano | Tai Wan Kim | Alexander Jones | Sharid Loáiciga

We present experiments on Automatic Speech Recognition (ASR) for Bribri and Cabécar, two languages from the Chibchan family. We fine-tune four ASR algorithms (Wav2Vec2, Whisper, MMS & WavLM) to create monolingual models, with the Wav2Vec2 model demonstrating the best performance. We then proceed to use Wav2Vec2 for (1) experiments on training joint and transfer learning models for both languages, and (2) an analysis of the errors, with a focus on the transcription of tone. Results show effective transfer learning for both Bribri and Cabécar, but especially for Bribri. A post-processing spell checking step further reduced character and word error rates. As for the errors, tone is where the Bribri models make the most errors, whereas the simpler tonal system of Cabécar is better transcribed by the model. Our work contributes to developing better ASR technology, an important tool that could facilitate transcription, one of the major bottlenecks in language documentation efforts. Our work also assesses how existing pre-trained models and algorithms perform for genuine extremely low resource-languages.

pdf bib
LegalDiscourse: Interpreting When Laws Apply and To Whom
Alexander Spangher | Zihan Xue | Te-Lin Wu | Mark Hansen | Jonathan May

While legal AI has made strides in recent years, it still struggles with basic legal concepts: _when_ does a law apply? _Who_ does it applies to? _What_ does it do? We take a _discourse_ approach to addressing these problems and introduce a novel taxonomy for span-and-relation parsing of legal texts. We create a dataset, _LegalDiscourse_ of 602 state-level law paragraphs consisting of 3,715 discourse spans and 1,671 relations. Our trained annotators have an agreement-rate 𝜅>.8, yet few-shot GPT3.5 performs poorly at span identification and relation classification. Although fine-tuning improves performance, GPT3.5 still lags far below human level. We demonstrate the usefulness of our schema by creating a web application with journalists. We collect over 100,000 laws for 52 U.S. states and territories using 20 scrapers we built, and apply our trained models to 6,000 laws using U.S. Census population numbers. We describe two journalistic outputs stemming from this application: (1) an investigation into the increase in liquor licenses following population growth and (2) a decrease in applicable laws under different under-count projections.

pdf bib
X-Eval: Generalizable Multi-aspect Text Evaluation via Augmented Instruction Tuning with Auxiliary Evaluation Aspects
Minqian Liu | Ying Shen | Zhiyang Xu | Yixin Cao | Eunah Cho | Vaibhav Kumar | Reza Ghanadan | Lifu Huang

Natural Language Generation (NLG) typically involves evaluating the generated text in various aspects (e.g., consistency and naturalness) to obtain a comprehensive assessment. However, multi-aspect evaluation remains challenging as it may require the evaluator to generalize to any given evaluation aspect even if it’s absent during training. In this paper, we introduce X-Eval, a two-stage instruction tuning framework to evaluate text in both seen and unseen aspects customized by end users. X-Eval consists of two learning stages: the vanilla instruction tuning stage that improves the model’s ability to follow evaluation instructions, and an enhanced instruction tuning stage that exploits the connections between fine-grained evaluation aspects to better assess text quality. To support the training of X-Eval, we collect AspectInstruct, the first instruction tuning dataset tailored for multi-aspect NLG evaluation spanning 27 diverse evaluation aspects with 65 tasks. To enhance task diversity, we devise an augmentation strategy that converts human rating annotations into diverse forms of NLG evaluation tasks, including scoring, comparison, ranking, and Boolean question answering. Extensive experiments across three essential categories of NLG tasks: dialogue generation, summarization, and data-to-text coupled with 21 aspects in meta-evaluation, demonstrate that X-Eval enables even a lightweight language model to achieve a comparable if not higher correlation with human judgments compared to the state-of-the-art NLG evaluators like GPT-4.

pdf bib
Is Reference Necessary in the Evaluation of NLG Systems? When and Where?
Shuqian Sheng | Yi Xu | Luoyi Fu | Jiaxin Ding | Lei Zhou | Xinbing Wang | Chenghu Zhou

The majority of automatic metrics for evaluating NLG systems are reference-based. However, the challenge of collecting human annotation results in a lack of reliable references in numerous application scenarios. Despite recent advancements in reference-free metrics, it has not been well understood when and where they can be used as an alternative to reference-based metrics. In this study, by employing diverse analytical approaches, we comprehensively assess the performance of both metrics across a wide range of NLG tasks, encompassing eight datasets and eight evaluation models. Based on solid experiments, the results show that reference-free metrics exhibit a higher correlation with human judgment and greater sensitivity to deficiencies in language quality. However, their effectiveness varies across tasks and is influenced by the quality of candidate texts. Therefore, it’s important to assess the performance of reference-free metrics before applying them to a new task, especially when inputs are in uncommon form or when the answer space is highly variable. Our study can provide insight into the appropriate application of automatic metrics and the impact of metric choice on evaluation performance.

pdf bib
Semi-Structured Chain-of-Thought: Integrating Multiple Sources of Knowledge for Improved Language Model Reasoning
Xin Su | Tiep Le | Steven Bethard | Phillip Howard

An important open question in the use of large language models for knowledge-intensive tasks is how to effectively integrate knowledge from three sources: the model’s parametric memory, external structured knowledge, and external unstructured knowledge. Most existing prompting methods either rely on one or two of these sources, or require repeatedly invoking large language models to generate similar or identical content. In this work, we overcome these limitations by introducing a novel semi-structured prompting approach that seamlessly integrates the model’s parametric memory with unstructured knowledge from text documents and structured knowledge from knowledge graphs. Experimental results on open-domain multi-hop question answering datasets demonstrate that our prompting method significantly surpasses existing techniques, even exceeding those that require fine-tuning.

pdf bib
Evaluating the Deductive Competence of Large Language Models
S Seals | Valerie Shalin

The development of highly fluent large language models (LLMs) has prompted increased interest in assessing their reasoning and problem-solving capabilities. We investigate whether several LLMs can solve a classic type of deductive reasoning problem from the cognitive science literature. The tested LLMs have limited abilities to solve these problems in their conventional form. We performed follow up experiments to investigate if changes to the presentation format and content improve model performance. We do find performance differences between conditions; however, they do not improve overall performance. Moreover, we find that performance interacts with presentation format and content in unexpected ways that differ from human performance. Overall, our results suggest that LLMs have unique reasoning biases that are only partially predicted from human reasoning performance and the human-generated language corpora that informs them.

pdf bib
Large Human Language Models: A Need and the Challenges
Nikita Soni | H. Schwartz | João Sedoc | Niranjan Balasubramanian

As research in human-centered NLP advances, there is a growing recognition of the importance of incorporating human and social factors into NLP models. At the same time, our NLP systems have become heavily reliant on LLMs, most of which do not model authors. To build NLP systems that can truly understand human language, we must better integrate human contexts into LLMs. This brings to the fore a range of design considerations and challenges in terms of what human aspects to capture, how to represent them, and what modeling strategies to pursue. To address these, we advocate for three positions toward creating large human language models (LHLMs) using concepts from psychological and behavioral sciences: First, LM training should include the human context. Second, LHLMs should recognize that people are more than their group(s). Third, LHLMs should be able to account for the dynamic and temporally-dependent nature of the human context. We refer to relevant advances and present open challenges that need to be addressed and their possible solutions in realizing these goals.

pdf bib
On Learning to Summarize with Large Language Models as References
Yixin Liu | Kejian Shi | Katherine He | Longtian Ye | Alexander Fabbri | Pengfei Liu | Dragomir Radev | Arman Cohan

Recent studies have found that summaries generated by large language models (LLMs) are favored by human annotators over the original reference summaries in commonly used summarization datasets. Therefore, we study an LLM-as-reference learning setting for smaller text summarization models to investigate whether their performance can be substantially improved. To this end, we use LLMs as both oracle summary generators for standard supervised fine-tuning and oracle summary evaluators for efficient contrastive learning that leverages the LLMs’ supervision signals. We conduct comprehensive experiments with source news articles and find that (1) summarization models trained under the LLM-as-reference setting achieve significant performance improvement in both LLM and human evaluations; (2) contrastive learning outperforms standard supervised fine-tuning under both low and high resource settings. Our experimental results also enable a meta-analysis of LLMs’ summary evaluation capacities under a challenging setting, showing that LLMs are not well-aligned with human evaluators. Particularly, our expert human evaluation reveals remaining nuanced performance gaps between LLMs and our fine-tuned models, which LLMs fail to capture. Thus, we call for further studies into both the potential and challenges of using LLMs in summarization model development.

pdf bib
Hallucination Diversity-Aware Active Learning for Text Summarization
Yu Xia | Xu Liu | Tong Yu | Sungchul Kim | Ryan Rossi | Anup Rao | Tung Mai | Shuai Li

Large Language Models (LLMs) have shown propensity to generate hallucinated outputs, i.e., texts that are factually incorrect or unsupported. Existing methods for alleviating hallucinations typically require costly human annotations to identify and correct hallucinations in LLM outputs. Moreover, most of these methods focus on a specific type of hallucination, e.g., entity or token errors, which limits their effectiveness in addressing various types of hallucinations exhibited in LLM outputs. To our best knowledge, in this paper we propose the first active learning framework to alleviate LLM hallucinations, reducing costly human annotations of hallucination needed. By measuring fine-grained hallucinations from errors in semantic frame, discourse and content verifiability in text summarization, we propose HAllucination Diversity-Aware Sampling (HADAS) to select diverse hallucinations for annotations in active learning for LLM finetuning. Extensive experiments on three datasets and different backbone models demonstrate advantages of our method in effectively and efficiently mitigating LLM hallucinations.

pdf bib
Keep it Private: Unsupervised Privatization of Online Text
Calvin Bao | Marine Carpuat

Authorship obfuscation techniques hold the promise of helping people protect their privacy in online communications by automatically rewriting text to hide the identity of the original author. However, obfuscation has been evaluated in narrow settings in the NLP literature and has primarily been addressed with superficial edit operations that can lead to unnatural outputs. In this work, we introduce an automatic text privatization framework that fine-tunes a large language model via reinforcement learning to produce rewrites that balance soundness, sense, and privacy. We evaluate it extensively on a large-scale test set of English Reddit posts by 68k authors composed of short-medium length texts. We study how the performance changes among evaluative conditions including authorial profile length and authorship detection strategy. Our method maintains high text quality according to both automated metrics and human evaluation, and successfully evades several automated authorship attacks.

pdf bib
Tied-LoRA: Enhancing parameter efficiency of LoRA with Weight Tying
Adithya Renduchintala | Tugrul Konuk | Oleksii Kuchaiev

We introduce Tied-LoRA, a novel paradigm leveraging weight tying and selective training to enhance the parameter efficiency of Low-rank Adaptation (LoRA). Our exploration encompasses different plausible combinations of parameter training and freezing, coupled with weight tying, aimed at identifying the optimal trade-off between performance and the count of trainable parameters. Across 5 diverse tasks and two foundational language models with different parameter counts, our experiments provide comprehensive insights into the inherent trade-offs between efficiency and performance.Our findings reveal a specific Tied-LoRA configuration that distinguishes itself by showcasing comparable performance to LoRA across multiple tasks while utilizing only a fraction of the parameters employed by the standard LoRA method, particularly at elevated ranks. This underscores the efficacy of Tied-LoRA in achieving impressive results with significantly reduced model complexity.

pdf bib
Investigating Data Contamination in Modern Benchmarks for Large Language Models
Chunyuan Deng | Yilun Zhao | Xiangru Tang | Mark Gerstein | Arman Cohan

Recent observations have underscored a disparity between the inflated benchmark scores and the actual performance of LLMs, raising concerns about potential contamination of evaluation benchmarks. This issue is especially critical for closed-source models and certain open-source models where training data transparency is lacking. In this paper we study data contamination by proposing two methods tailored for both open-source and proprietary LLMs. We first introduce a retrieval-based system to explore potential overlaps between evaluation benchmarks and pretraining corpora. We further present a novel investigation protocol named Testset Slot Guessing (TS-Guessing), applicable to both open and proprietary models. This approach entails masking a wrong answer in a multiple-choice question and prompting the model to fill in the gap. Additionally, it involves obscuring an unlikely word in an evaluation example and asking the model to produce it. We find that certain commercial LLMs could surprisingly guess the missing option in various test sets. Specifically, in the MMLU benchmark, ChatGPT and GPT-4 demonstrated an exact match rate of 52% and 57%, respectively, in guessing the missing options in benchmark test data. We hope these results underscore the need for more robust evaluation methodologies and benchmarks in the field.

pdf bib
Pre-trained Language Models for Entity Blocking: A Reproducibility Study
Runhui Wang | Yongfeng Zhang

Entity Resolution (ER) is an essential task in data integration and its goal is to find records that represent the same entity in a dataset. Deep learning models, especially large pre-trained language models, have achieved state-of-the-art results on this task. A typical ER pipeline consists of Entity Blocking and Entity Matching: Entity Blocking finds candidate record pairs that potentially match and Entity Matching determines if the pairs match. The goal of the entity blocking step is to include as many matching pairs as possible while including as few non-matching pairs as possible. On the other hand, the blocking task can also be considered as an Information Retrieval (IR) task. However, state-of-the-art neural IR models that are based on large language models have not been evaluated on the ER task. What’s more, the generalization ability of state-of-the-art methods for entity blocking is not well-studied but an import aspect in real-world applications. In this work, we evaluate state-of-the-art models for Entity Blocking along with neural IR models on a wide range of real-world datasets, and also study their in-distribution and out-of-distribution generalization abilities.

pdf bib
RE2: Region-Aware Relation Extraction from Visually Rich Documents
Pritika Ramu | Sijia Wang | Lalla Mouatadid | Joy Rimchala | Lifu Huang

Current research in form understanding predominantly relies on large pre-trained language models, necessitating extensive data for pre-training. However, the importance of layout structure (i.e., the spatial relationship between the entity blocks in the visually rich document) to relation extraction has been overlooked. In this paper, we propose REgion-Aware Relation Extraction (\bf{RE^2}) that leverages region-level spatial structure among the entity blocks to improve their relation prediction. We design an edge-aware graph attention network to learn the interaction between entities while considering their spatial relationship defined by their region-level representations. We also introduce a constraint objective to regularize the model towards consistency with the inherent constraints of the relation extraction task. To support the research on relation extraction from visually rich documents and demonstrate the generalizability of \bf{RE^2}, we build a new benchmark dataset, DiverseForm, that covers a wide range of domains. Extensive experiments on DiverseForm and several public benchmark datasets demonstrate significant superiority and transferability of \bf{RE^2} across various domains and languages, with up to 18.88% absolute F-score gain over all high-performing baselines

pdf bib
Mix-Initiative Response Generation with Dynamic Prefix Tuning
Yuxiang Nie | Heyan Huang | Xian-Ling Mao | Lizi Liao

Mixed initiative serves as one of the key factors in controlling conversation directions. For a speaker, responding passively or leading proactively would result in rather different responses. However, most dialogue systems focus on training a holistic response generation model without any distinction among different initiatives. It leads to the cross-contamination problem, where the model confuses different initiatives and generates inappropriate responses. Moreover, obtaining plenty of human annotations for initiative labels can be expensive. To address this issue, we propose a general mix-Initiative Dynamic Prefix Tuning framework (IDPT) to decouple different initiatives from the generation model, which learns initiative-aware prefixes in both supervised and unsupervised settings. Specifically, IDPT decouples initiative factors into different prefix parameters and uses the attention mechanism to adjust the selection of initiatives in guiding generation dynamically. The prefix parameters can be tuned towards accurate initiative prediction as well as mix-initiative response generation. Extensive experiments on two public dialogue datasets show that the proposed IDPT outperforms previous baselines on both automatic metrics and human evaluations. It also manages to generate appropriate responses with manipulated initiatives.

pdf bib
Value FULCRA: Mapping Large Language Models to the Multidimensional Spectrum of Basic Human Value
Jing Yao | Xiaoyuan Yi | Yifan Gong | Xiting Wang | Xing Xie

Value alignment is crucial for the responsible development of Large Language Models (LLMs). However, how to define values in this context remains largely unexplored. Existing work mainly specifies values as risk criteria formulated in the AI community, e.g., fairness and privacy protection, suffering from poor clarity, adaptability and transparency. Leveraging basic values established in humanity and social science that are compatible with values across cultures, this paper introduces a novel value space spanned by multiple basic value dimensions and proposes BaseAlign, a corresponding value alignment paradigm. Applying the representative Schwartz’s Theory of Basic Values as an instantiation, we construct FULCRA, a dataset consisting of 20k (LLM output, value vector) pairs. LLMs’ outputs are mapped into the K-dim value space beyond simple binary labels, by identifying their underlying priorities for these value dimensions. Extensive analysis and experiments on FULCRA: (1) reveal the essential relation between basic values and LLMs’ behaviors, (2) demonstrate that our paradigm with basic values not only covers existing risks but also anticipates the unidentified ones, and (3) manifest BaseAlign’s superiority in alignment performance with less data, paving the way for addressing the above three challenges.

pdf bib
IndiBias: A Benchmark Dataset to Measure Social Biases in Language Models for Indian Context
Nihar Sahoo | Pranamya Kulkarni | Arif Ahmad | Tanu Goyal | Narjis Asad | Aparna Garimella | Pushpak Bhattacharyya

The pervasive influence of social biases in language data has sparked the need for benchmark datasets that capture and evaluate these biases in Large Language Models (LLMs). Existing efforts predominantly focus on English language and the Western context, leaving a void for a reliable dataset that encapsulates India’s unique socio-cultural nuances. To bridge this gap, we introduce IndiBias, a comprehensive benchmarking dataset designed specifically for evaluating social biases in the Indian context. We filter and translate the existing CrowS-Pairs dataset to create a benchmark dataset suited to the Indian context in Hindi language. Additionally, we leverage LLMs including ChatGPT and InstructGPT to augment our dataset with diverse societal biases and stereotypes prevalent in India. The included bias dimensions encompass gender, religion, caste, age, region, physical appearance, and occupation. We also build a resource to address intersectional biases along three intersectional dimensions. Our dataset contains 800 sentence pairs and 300 tuples for bias measurement across different demographics. The dataset is available in English and Hindi, providing a size comparable to existing benchmark datasets. Furthermore, using IndiBias we compare ten different language models on multiple bias measurement metrics. We observed that the language models exhibit more bias across a majority of the intersectional groups. All the scripts utilized and datasets created in this study are publicly available.
up

pdf (full)
bib (full) Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers)

pdf bib
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 2: Short Papers)
Kevin Duh | Helena Gomez | Steven Bethard

pdf bib
Revisiting Zero-Shot Abstractive Summarization in the Era of Large Language Models from the Perspective of Position Bias
Anshuman Chhabra | Hadi Askari | Prasant Mohapatra

We characterize and study zero-shot abstractive summarization in Large Language Models (LLMs) by measuring position bias, which we propose as a general formulation of the more restrictive lead bias phenomenon studied previously in the literature. Position bias captures the tendency of a model unfairly prioritizing information from certain parts of the input text over others, leading to undesirable behavior. Through numerous experiments on four diverse real-world datasets, we study position bias in multiple LLM models such as GPT 3.5-Turbo, Llama-2, and Dolly-v2, as well as state-of-the-art pretrained encoder-decoder abstractive summarization models such as Pegasus and BART. Our findings lead to novel insights and discussion on performance and position bias of models for zero-shot summarization tasks.

pdf bib
Struc-Bench: Are Large Language Models Good at Generating Complex Structured Tabular Data?
Xiangru Tang | Yiming Zong | Jason Phang | Yilun Zhao | Wangchunshu Zhou | Arman Cohan | Mark Gerstein

Despite the remarkable capabilities of Large Language Models (LLMs) like GPT-4, producing complex, structured tabular data remains challenging. Our study assesses LLMs’ proficiency in structuring tables and introduces a novel fine-tuning method, cognizant of data structures, to bolster their performance. We unveil Struc-Bench, a comprehensive benchmark featuring prominent LLMs (GPT-NeoX-20B, GPT-3.5, GPT-4, and Vicuna), which spans text tables, HTML, and LaTeX formats. Our proposed FormatCoT aids in crafting format-specific instructions from the intended outputs to populate this benchmark. Addressing the gap in task-centered evaluation, we propose two innovative metrics, P-Score (Prompting Score) and H-Score (Heuristical Score), to more accurately gauge LLM performance. Our experiments show that applying our structure-aware fine-tuning to LLaMA-7B leads to substantial performance gains, outshining its LLM counterparts across most measures. In-depth error analysis and creating an ability map across six dimensions, coverage, formatting, reasoning, comprehension, pragmatics, and hallucination, highlight areas for future enhancements and suggest forthcoming research trajectories. Our code and models can be found at https://github.com/gersteinlab/Struc-Bench.

pdf bib
Improving Toponym Resolution by Predicting Attributes to Constrain Geographical Ontology Entries
Zeyu Zhang | Egoitz Laparra | Steven Bethard

Geocoding is the task of converting location mentions in text into structured geospatial data.We propose a new prompt-based paradigm for geocoding, where the machine learning algorithm encodes only the location mention and its context.We design a transformer network for predicting the country, state, and feature class of a location mention, and a deterministic algorithm that leverages the country, state, and feature class predictions as constraints in a search for compatible entries in the ontology.Our architecture, GeoPLACE, achieves new state-of-the-art performance on multiple datasets.Code and models are available at https://github.com/clulab/geonorm.

pdf bib
Advancing Regular Language Reasoning in Linear Recurrent Neural Networks
Ting-Han Fan | Ta-Chung Chi | Alexander Rudnicky

In recent studies, linear recurrent neural networks (LRNNs) have achieved Transformer-level performance in natural language and long-range modeling, while offering rapid parallel training and constant inference cost. With the resurgence of interest in LRNNs, we study whether they can learn the hidden rules in training sequences, such as the grammatical structures of regular language. We theoretically analyze some existing LRNNs and discover their limitations in modeling regular language. Motivated by this analysis, we propose a new LRNN equipped with a block-diagonal and input-dependent transition matrix. Experiments suggest that the proposed model is the only LRNN capable of performing length extrapolation on regular language tasks such as Sum, Even Pair, and Modular Arithmetic. The code is released at https://github.com/tinghanf/RegluarLRNN.

pdf bib
Extracting Lexical Features from Dialects via Interpretable Dialect Classifiers
Roy Xie | Orevaoghene Ahia | Yulia Tsvetkov | Antonios Anastasopoulos

Identifying linguistic differences between dialects of a language often requires expert knowledge and meticulous human analysis. This is largely due to the complexity and nuance involved in studying various dialects. We present a novel approach to extract distinguishing lexical features of dialects by utilizing interpretable dialect classifiers, even in the absence of human experts. We explore both post-hoc and intrinsic approaches to interpretability, conduct experiments on Mandarin, Italian, and Low Saxon, and experimentally demonstrate that our method successfully identifies key language-specific lexical features that contribute to dialectal variations.

pdf bib
Clear Up Confusion: Advancing Cross-Domain Few-Shot Relation Extraction through Relation-Aware Prompt Learning
Ge Bai | Chenji Lu | Daichi Guo | Shilong Li | Ying Liu | Zhang Zhang | Guanting Dong | Ruifang Liu | Sun Yong

Cross-domain few-shot Relation Extraction (RE) aims to transfer knowledge from a source domain to a different target domain to address low-resource problems.Previous work utilized label descriptions and entity information to leverage the knowledge of the source domain.However, these models are prone to confusion when directly applying this knowledge to a target domain with entirely new types of relations, which becomes particularly pronounced when facing similar relations.In this work, we propose a relation-aware prompt learning method with pre-training.Specifically, we empower the model to clear confusion by decomposing various relation types through an innovative label prompt, while a context prompt is employed to capture differences in different scenarios, enabling the model to further discern confusion. Two pre-training tasks are designed to leverage the prompt knowledge and paradigm.Experiments show that our method outperforms previous sota methods, yielding significantly better results on cross-domain few-shot RE tasks.

pdf bib
Fusion Makes Perfection: An Efficient Multi-Grained Matching Approach for Zero-Shot Relation Extraction
Shilong Li | Ge Bai | Zhang Zhang | Ying Liu | Chenji Lu | Daichi Guo | Ruifang Liu | Sun Yong

Predicting unseen relations that cannot be observed during the training phase is a challenging task in relation extraction. Previous works have made progress by matching the semantics between input instances and label descriptions. However, fine-grained matching often requires laborious manual annotation, and rich interactions between instances and label descriptions come with significant computational overhead. In this work, we propose an efficient multi-grained matching approach that uses virtual entity matching to reduce manual annotation cost, and fuses coarse-grained recall and fine-grained classification for rich interactions with guaranteed inference speed.Experimental results show that our approach outperforms the previous State Of The Art (SOTA) methods, and achieves a balance between inference efficiency and prediction accuracy in zero-shot relation extraction tasks.Our code is available at https://github.com/longls777/EMMA.

pdf bib
Personalized Review Recommendation based on Implicit dimension mining
Bei Xu | Yifan Xu

Users usually browse product reviews before buying products from e-commerce websites. Lots of e-commerce websites can recommend reviews. However, existing research on review recommendation mainly focuses on the general usefulness of reviews and ignores personalized and implicit requirements. To address the issue, we propose a Large language model driven Personalized Review Recommendation model based on Implicit dimension mining (PRR-LI). The model mines implicit dimensions from reviews and requirements, and encodes them in the form of “text + dimension”. The experiments show that our model significantly outperforms other state-of-the-art textual models on the Amazon-MRHP dataset, with some of the metrics outperforming the state-of-the-art multimodal models. And we prove that encoding “text + dimension” is better than encoding “text” and “dimension” separately in review recommendation.

pdf bib
Unlocking Structure Measuring: Introducing PDD, an Automatic Metric for Positional Discourse Coherence
Yinhong Liu | Yixuan Su | Ehsan Shareghi | Nigel Collier

Recent large language models (LLMs) have shown remarkable performance in aligning generated text with user intentions across various tasks. When it comes to long-form text generation, there has been a growing interest in generation from a discourse coherence perspective.However, existing lexical or semantic metrics such as BLEU, ROUGE, BertScore cannot effectively capture the discourse coherence.The development of discourse-specific automatic evaluation methods for assessing the output of LLMs warrants greater focus and exploration. In this paper, we present a novel automatic metric designed to quantify the discourse divergence between two long-form articles.Extensive experiments on three datasets from representative domains demonstrate that our metric aligns more closely with human preferences and GPT-4 coherence evaluation, outperforming existing evaluation methods.

pdf bib
Returning to the Start: Generating Narratives with Related Endpoints
Anneliese Brei | Chao Zhao | Snigdha Chaturvedi

Human writers often *bookend* their writing with ending sentences that relate back to the beginning sentences in order to compose a satisfying narrative that “closes the loop.” Motivated by this observation, we propose RENarGen, a controllable story-generation paradigm that generates narratives by ensuring the first and last sentences are related and then infilling the middle sentences. Our contributions include an initial exploration of how various methods of bookending from Narratology affect language modeling for stories. Automatic and human evaluations indicate RENarGen produces better stories with more narrative closure than current autoregressive models.

pdf bib
Unified Examination of Entity Linking in Absence of Candidate Sets
Nicolas Ong | Hassan Shavarani | Anoop Sarkar

Despite remarkable strides made in the development of entity linking systems in recent years, a comprehensive comparative analysis of these systems using a unified framework is notably absent. This paper addresses this oversight by introducing a new black-box benchmark and conducting a comprehensive evaluation of all state-of-the-art entity linking methods. We use an ablation study to investigate the impact of candidate sets on the performance of entity linking. Our findings uncover exactly how much such entity linking systems depend on candidate sets, and how much this limits the general applicability of each system. We present an alternative approach to candidate sets, demonstrating that leveraging the entire in-domain candidate set can serve as a viable substitute for certain models. We show the trade-off between less restrictive candidate sets, increased inference time and memory footprint for some models.

pdf bib
MultiParaDetox: Extending Text Detoxification with Parallel Data to New Languages
Daryna Dementieva | Nikolay Babakov | Alexander Panchenko

Text detoxification is a textual style transfer (TST) task where a text is paraphrased from a toxic surface form, e.g. featuring rude words, to the neutral register. Recently, text detoxification methods found their applications in various task such as detoxification of Large Language Models (LLMs) (Leong et al., 2023; He et al., 2024; Tang et al., 2023) and toxic speech combating in social networks (Deng et al., 2023; Mun et al., 2023; Agarwal et al., 2023). All these applications are extremely important to ensure safe communication in modern digital worlds. However, the previous approaches for parallel text detoxification corpora collection—ParaDetox (Logacheva et al., 2022) and APPADIA (Atwell et al., 2022)—were explored only in monolingual setup. In this work, we aim to extend ParaDetox pipeline to multiple languages presenting MultiParaDetox to automate parallel detoxification corpus collection for potentially any language. Then, we experiment with different text detoxification models—from unsupervised baselines to LLMs and fine-tuned models on the presented parallel corpora—showing the great benefit of parallel corpus presence to obtain state-of-the-art text detoxification models for any language.

pdf bib
SKICSE: Sentence Knowable Information Prompted by LLMs Improves Contrastive Sentence Embeddings
Fangwei Ou | Jinan Xu

Contrastive learning, which utilizes positive pairs and in-batch negatives to optimize the loss objective, has been proven to be an effective method for learning sentence embeddings. However, we argue that the previous methods of constructing positive pairs only through dropout perturbation or entailment relation are limited. Since there is more sentence knowable information (SKI) to be mined, such as sentence external knowledge, semantic analysis, and grammatical description. In this work, we first hand-craft a simple and effective prompt template that is able to obtain the knowable information of input sentences from LLMs (e.g., LLaMA). Then we combine the original sentence and its knowable information to form a positive pair for contrastive learning. We evaluate our method on standard semantic textual similarity (STS) tasks. Experimental results show that our unsupervised and supervised models using BERTbase achieve an average of 78.65% and 82.45% Spearman’s correlation respectively, a 2.40% and 0.88% improvement compared to SimCSE. Our model outperforms the previous state-of-the-art model PromptBERT in both unsupervised and supervised settings and specifically yields a new state-of-the-art performance in supervised setting.

pdf bib
A Multi-Aspect Framework for Counter Narrative Evaluation using Large Language Models
Jaylen Jones | Lingbo Mo | Eric Fosler-Lussier | Huan Sun

Counter narratives - informed responses to hate speech contexts designed to refute hateful claims and de-escalate encounters - have emerged as an effective hate speech intervention strategy. While previous work has proposed automatic counter narrative generation methods to aid manual interventions, the evaluation of these approaches remains underdeveloped. Previous automatic metrics for counter narrative evaluation lack alignment with human judgment as they rely on superficial reference comparisons instead of incorporating key aspects of counter narrative quality as evaluation criteria. To address prior evaluation limitations, we propose a novel evaluation framework prompting LLMs to provide scores and feedback for generated counter narrative candidates using 5 defined aspects derived from guidelines from counter narrative specialized NGOs. We found that LLM evaluators achieve strong alignment to human-annotated scores and feedback and outperform alternative metrics, indicating their potential as multi-aspect, reference-free and interpretable evaluators for counter narrative evaluation.

pdf bib
How does Multi-Task Training Affect Transformer In-Context Capabilities? Investigations with Function Classes
Harmon Bhasin | Timothy Ossowski | Yiqiao Zhong | Junjie Hu

Large language models (LLM) have recently shown the extraordinary ability to perform unseen tasks based on few-shot examples provided as text, also known as in-context learning (ICL). While recent works have attempted to understand the mechanisms driving ICL, few have explored training strategies that incentivize these models to generalize to multiple tasks. Multi-task learning (MTL) for generalist models is a promising direction that offers transfer learning potential, enabling large parameterized models to be trained from simpler, related tasks. In this work, we investigate the combination of MTL with ICL to build models that efficiently learn tasks while being robust to out-of-distribution examples. We propose several effective curriculum learning strategies that allow ICL models to achieve higher data efficiency and more stable convergence. Our experiments reveal that ICL models can effectively learn difficult tasks by training on progressively harder tasks while mixing in prior tasks, denoted as mixed curriculum in this work.

pdf bib
CELI: Simple yet Effective Approach to Enhance Out-of-Domain Generalization of Cross-Encoders.
Crystina Zhang | Minghan Li | Jimmy Lin

In text ranking, it is generally believed that the cross-encoders already gather sufficient token interaction information via the attention mechanism in the hidden layers. However, our results show that the cross-encoders can consistently benefit from additional token interaction in the similarity computation at the last layer. We introduce CELI (Cross-Encoder with Late Interaction), which incorporates a late interaction layer into the current cross-encoder models. This simple method brings 5% improvement on BEIR without compromising in-domain effectiveness or search latency. Extensive experiments show that this finding is consistent across different sizes of the cross-encoder models and the first-stage retrievers. Our findings suggest that boiling all information into the [CLS] token is a suboptimal use for cross-encoders, and advocate further studies to investigate its relevance score mechanism.

pdf bib
ContrastiveMix: Overcoming Code-Mixing Dilemma in Cross-Lingual Transfer for Information Retrieval
Junggeun Do | Jaeseong Lee | Seung-won Hwang

Multilingual pretrained language models (mPLMs) have been widely adopted in cross-lingual transfer, and code-mixing has demonstrated effectiveness across various tasks in the absence of target language data. Our contribution involves an in-depth investigation into the counterproductive nature of training mPLMs on code-mixed data for information retrieval (IR). Our finding is that while code-mixing demonstrates a positive effect in aligning representations across languages, it hampers the IR-specific objective of matching representations between queries and relevant passages. To balance between positive and negative effects, we introduce ContrastiveMix, which disentangles contrastive loss between these conflicting objectives, thereby enhancing zero-shot IR performance. Specifically, we leverage both English and code-mixed data and employ two contrastive loss functions, by adding an additional contrastive loss that aligns embeddings of English data with their code-mixed counterparts in the query encoder. Our proposed ContrastiveMix exhibits statistically significant outperformance compared to mDPR, particularly in scenarios involving lower linguistic similarity, where the conflict between goals is more pronounced.

pdf bib
SLIDE: Reference-free Evaluation for Machine Translation using a Sliding Document Window
Vikas Raunak | Tom Kocmi | Matt Post

Reference-based metrics that operate at the sentence-level typically outperform quality estimation metrics, which have access only to the source and system output.This is unsurprising, since references resolve ambiguities that may be present in the source.In this paper, we investigate whether additional source context can effectively substitute for a reference.We present a metric named SLIDE (SLIding Document Evaluator), which operates on blocks of sentences. SLIDE leverages a moving window that slides over each document in the test set, feeding each chunk of sentences into an unmodified, off-the-shelf quality estimation model.We find that SLIDE obtains significantly higher pairwise system accuracy than its sentence-level baseline, in some cases even eliminating the gap with reference-base metrics.This suggests that source context may provide the same information as a human reference in disambiguating source ambiguities. This finding is especially pertinent for reference-free document-level evaluation, wherein SLIDE could provide higher-quality pairwise system assessments while only requiring document boundary annotations.

pdf bib
Separately Parameterizing Singleton Detection Improves End-to-end Neural Coreference Resolution
Xiyuan Zou | Yiran Li | Ian Porada | Jackie Cheung

Current end-to-end coreference resolution models combine detection of singleton mentions and antecedent linking into a single step. In contrast, singleton detection was often treated as a separate step in the pre-neural era. In this work, we show that separately parameterizing these two sub-tasks also benefits end-to-end neural coreference systems. Specifically, we add a singleton detector to the coarse-to-fine (C2F) coreference model, and design an anaphoricity-aware span embedding and singleton detection loss. Our method significantly improves model performance on OntoNotes and four additional datasets.

pdf bib
Unveiling Divergent Inductive Biases of LLMs on Temporal Data
Sindhu Kishore | Hangfeng He

Unraveling the intricate details of events in natural language necessitates a subtle understanding of temporal dynamics. Despite the adeptness of Large Language Models (LLMs) in discerning patterns and relationships from data, their inherent comprehension of temporal dynamics remains a formidable challenge. This research meticulously explores these intrinsic challenges within LLMs, with a specific emphasis on evaluating the performance of GPT-3.5 and GPT-4 models in the analysis of temporal data. Employing two distinct prompt types, namely Question Answering (QA) format and Textual Entailment (TE) format, our analysis probes into both implicit and explicit events. The findings underscore noteworthy trends, revealing disparities in the performance of GPT-3.5 and GPT-4. Notably, biases toward specific temporal relationships come to light, with GPT-3.5 demonstrating a preference for “AFTER” in the QA format for both implicit and explicit events, while GPT-4 leans towards “BEFORE”. Furthermore, a consistent pattern surfaces wherein GPT-3.5 tends towards “TRUE”, and GPT-4 exhibits a preference for “FALSE” in the TE format for both implicit and explicit events. This persistent discrepancy between GPT-3.5 and GPT-4 in handling temporal data highlights the intricate nature of inductive bias in LLMs, suggesting that the evolution of these models may not merely mitigate bias but may introduce new layers of complexity.

pdf bib
On Retrieval Augmentation and the Limitations of Language Model Training
Ting-Rui Chiang | Xinyan Yu | Joshua Robinson | Ollie Liu | Isabelle Lee | Dani Yogatama

Augmenting a language model (LM) with k-nearest neighbors (kNN) retrieval on its training data alone can decrease its perplexity, though the underlying reasons for this remain elusive. In this work, we rule out one previously posited possibility — the “softmax bottleneck.” We then create a new dataset to evaluate LM generalization ability in the setting where training data contains additional information that is not causally relevant. This task is challenging even for GPT-3.5 Turbo. We show that, for both GPT-2 and Mistral 7B, kNN retrieval augmentation consistently improves per formance in this setting. Finally, to make kNN retrieval more accessible, we propose using amulti-layer perceptron model that maps datastore keys to values as a drop-in replacement for traditional retrieval. This reduces storage costsby over 25x.

pdf bib
GenDecider: Integrating “None of the Candidates” Judgments in Zero-Shot Entity Linking Re-ranking
Kang Zhou | Yuepei Li | Qing Wang | Qiao Qiao | Qi Li

We introduce GenDecider, a novel re-ranking approach for Zero-Shot Entity Linking (ZSEL), built on the Llama model. It innovatively detects scenarios where the correct entity is not among the retrieved candidates, a common oversight in existing re-ranking methods. By autoregressively generating outputs based on the context of the entity mention and the candidate entities, GenDecider significantly enhances disambiguation, improving the accuracy and reliability of ZSEL systems, as demonstrated on the benchmark ZESHEL dataset. Our code is available at https://github.com/kangISU/GenDecider.

pdf bib
Advancing the Robustness of Large Language Models through Self-Denoised Smoothing
Jiabao Ji | Bairu Hou | Zhen Zhang | Guanhua Zhang | Wenqi Fan | Qing Li | Yang Zhang | Gaowen Liu | Sijia Liu | Shiyu Chang

Although large language models (LLMs) have achieved significant success, their vulnerability to adversarial perturbations, including recent jailbreak attacks, has raised considerable concerns. However, the increasing size of these models and their limited access make improving their robustness a challenging task. Among various defense strategies, randomized smoothing has shown great potential for LLMs, as it does not require full access to the model’s parameters or fine-tuning via adversarial training. However, randomized smoothing involves adding noise to the input before model prediction, and the final model’s robustness largely depends on the model’s performance on these noise-corrupted data. Its effectiveness is often limited by the model’s sub-optimal performance on noisy data. To address this issue, we propose to leverage the multitasking nature of LLMs to first denoise the noisy inputs and then to make predictions based on these denoised versions. We call this procedure self-denoised smoothing. Unlike previous denoised smoothing techniques in computer vision, which require training a separate model to enhance the robustness of LLMs, our method offers significantly better efficiency and flexibility. Our experimental results indicate that our method surpasses existing methods in both empirical and certified robustness in defending against adversarial attacks for both downstream tasks and human alignments (i.e., jailbreak attacks). Our code is publicly available at https://github.com/UCSB-NLP-Chang/SelfDenoise.

pdf bib
Can LLM’s Generate Human-Like Wayfinding Instructions? Towards Platform-Agnostic Embodied Instruction Synthesis
Vishnu Sashank Dorbala | Sanjoy Chowdhury | Dinesh Manocha

We present a novel approach to automatically synthesize “wayfinding instructions” for an embodied robot agent. In contrast to prior approaches that are heavily reliant on human-annotated datasets designed exclusively for specific simulation platforms, our algorithm uses in-context learning to condition an LLM to generate instructions using just a few references. Using an LLM-based Visual Question Answering strategy, we gather detailed information about the environment which is used by the LLM for instruction synthesis. We implement our approach on multiple simulation platforms including Matterport3D, AI Habitat and ThreeDWorld, thereby demonstrating its platform-agnostic nature. We subjectively evaluate our approach via a user study and observe that 83.3% of users find the synthesized instructions accurately capture the details of the environment and show characteristics similar to those of human-generated instructions. Further, we conduct zero-shot navigation with multiple approaches on the REVERIE dataset using the generated instructions, and observe very close correlation with the baseline on standard success metrics (< 1% change in SR), quantifying the viability of generated instructions in replacing human-annotated data. We finally discuss the applicability of our approach in enabling a generalizable evaluation of embodied navigation policies. To the best of our knowledge, ours is the first LLM-driven approach capable of generating “human-like” instructions in a platform-agnostic manner, without training.

pdf bib
On the Role of Summary Content Units in Text Summarization Evaluation
Marcel Nawrath | Agnieszka Nowak | Tristan Ratz | Danilo Walenta | Juri Opitz | Leonardo Ribeiro | João Sedoc | Daniel Deutsch | Simon Mille | Yixin Liu | Sebastian Gehrmann | Lining Zhang | Saad Mahamood | Miruna Clinciu | Khyathi Chandu | Yufang Hou

At the heart of the Pyramid evaluation method for text summarization lie human written summary content units (SCUs). These SCUs areconcise sentences that decompose a summary into small facts. Such SCUs can be used to judge the quality of a candidate summary, possibly partially automated via natural language inference (NLI) systems. Interestingly, with the aim to fully automate the Pyramid evaluation, Zhang and Bansal (2021) show that SCUs can be approximated by automatically generated semantic role triplets (STUs). However, several questions currently lack answers, in particular: i) Are there other ways of approximating SCUs that can offer advantages?ii) Under which conditions are SCUs (or their approximations) offering the most value? In this work, we examine two novel strategiesto approximate SCUs: generating SCU approximations from AMR meaning representations (SMUs) and from large language models (SGUs), respectively. We find that while STUs and SMUs are competitive, the best approximation quality is achieved by SGUs. We also show through a simple sentence-decomposition baseline (SSUs) that SCUs (and their approximations) offer the most value when rankingshort summaries, but may not help as much when ranking systems or longer summaries.

pdf bib
More room for language: Investigating the effect of retrieval on language models
David Samuel | Lucas Charpentier | Sondre Wold

Retrieval-augmented language models pose a promising alternative to standard language modeling. During pretraining, these models search in a corpus of documents for contextually relevant information that could aid the language modeling objective. We introduce an ‘ideal retrieval’ methodology to study these models in a fully controllable setting. We conduct an extensive evaluation to examine how retrieval augmentation affects the behavior of the underlying language model. Among other things, we observe that these models: (i) save substantially less world knowledge in their weights, (ii) are better at understanding local context and inter-word dependencies, but (iii) are worse at comprehending global context.

pdf bib
Discourse-Aware In-Context Learning for Temporal Expression Normalization
Akash Gautam | Lukas Lange | Jannik Strötgen

Temporal expression (TE) normalization is a well-studied problem. However, the predominately used rule-based systems are highly restricted to specific settings, and upcoming machine learning approaches suffer from a lack of labeled data. In this work, we explore the feasibility of proprietary and open-source large language models (LLMs) for TE normalization using in-context learning to inject task, document, and example information into the model. We explore various sample selection strategies to retrieve the most relevant set of examples. By using a window-based prompt design approach, we can perform TE normalization across sentences, while leveraging the LLM knowledge without training the model.Our experiments show competitive results to models designed for this task. In particular, our method achieves large performance improvements for non-standard settings by dynamically including relevant examples during inference.

pdf bib
Contextualizing Argument Quality Assessment with Relevant Knowledge
Darshan Deshpande | Zhivar Sourati | Filip Ilievski | Fred Morstatter

Automatic assessment of the quality of arguments has been recognized as a challenging task with significant implications for misinformation and targeted speech. While real-world arguments are tightly anchored in context, existing computational methods analyze their quality in isolation, which affects their accuracy and generalizability. We propose SPARK: a novel method for scoring argument quality based on contextualization via relevant knowledge. We devise four augmentations that leverage large language models to provide feedback, infer hidden assumptions, supply a similar-quality argument, or give a counter-argument. SPARK uses a dual-encoder Transformer architecture to enable the original argument and its augmentation to be considered jointly. Our experiments in both in-domain and zero-shot setups show that SPARK consistently outperforms existing techniques across multiple metrics

pdf bib
Selective Perception: Learning Concise State Descriptions for Language Model Actors
Kolby Nottingham | Yasaman Razeghi | Kyungmin Kim | Jb Lanier | Pierre Baldi | Roy Fox | Sameer Singh

The latest large language models (LMs) support increasingly longer contexts. While this trend permits using substantial amounts of text with SOTA LMs, requiring these large LMs to process potentially redundant or irrelevant data needlessly increases inference time and cost. To remedy this problem, we propose BLINDER, a method that leverages a small finetuned LM to sample the minimal set of input features that maximizes the performance of a downstream LM. BLINDER trains an LM with a value head to estimate the likelihood of optimal outputs from a downstream LM given an input. We evaluate BLINDER on embodied decision making tasks with notoriously verbose state descriptions: NetHack and robot planning. BLINDER reduces the length of LM actor input by 87% and 99% while improving task success rates by 158% and 54% on NetHack and robot planning respectively which represents substantial inference cost savings while actually increasing performance.

pdf bib
ALOHa: A New Measure for Hallucination in Captioning Models
Suzanne Petryk | David Chan | Anish Kachinthaya | Haodi Zou | John Canny | Joseph Gonzalez | Trevor Darrell

Despite recent advances in multimodal pre-training for visual description, state-of-the-art models still produce captions containing errors, such as hallucinating objects not present in a scene. The existing prominent metric for object hallucination, CHAIR, is limited to a fixed set of MS COCO objects and synonyms. In this work, we propose a modernized open-vocabulary metric, ALOHa, which leverages large language models (LLMs) to measure object hallucinations. Specifically, we use an LLM to extract groundable objects from a candidate caption, measure their semantic similarity to reference objects from captions and object detections, and use Hungarian matching to produce a final hallucination score. We show that ALOHa correctly identifies 13.6% more hallucinated objects than CHAIR on HAT, a new gold-standard subset of MS COCO Captions annotated for hallucinations, and 30.8% more on nocaps, where objects extend beyond MS COCO categories.

pdf bib
Beyond Yes and No: Improving Zero-Shot LLM Rankers via Scoring Fine-Grained Relevance Labels
Honglei Zhuang | Zhen Qin | Kai Hui | Junru Wu | Le Yan | Xuanhui Wang | Michael Bendersky

Zero-shot text rankers powered by recent LLMs achieve remarkable ranking performance by simply prompting. Existing prompts for pointwise LLM rankers mostly ask the model to choose from binary relevance labels like “Yes” and “No”. However, the lack of intermediate relevance label options may cause the LLM to provide noisy or biased answers for documents that are partially relevant to the query. We propose to incorporate fine-grained relevance labels into the prompt for LLM rankers, enabling them to better differentiate among documents with different levels of relevance to the query and thus derive a more accurate ranking. We study two variants of the prompt template, coupled with different numbers of relevance levels. Our experiments on 8 BEIR data sets show that adding fine-grained relevance labels significantly improves the performance of LLM rankers.

pdf bib
LLM-Driven Knowledge Injection Advances Zero-Shot and Cross-Target Stance Detection
Zhao Zhang | Yiming Li | Jin Zhang | Hui Xu

Stance detection aims at inferring an author’s attitude towards a specific target in a text. Prior methods mainly consider target-related background information for a better understanding of targets while neglecting the accompanying input texts. In this study, we propose to prompt Large Language Models (LLMs) to explicitly extract the relationship between paired text and target as contextual knowledge. We then inject such LLM-driven knowledge into a generation model BART to exploit the rich contexts and semantics. Moreover, to further enhance the decoding capability of BART, a novel prototypical contrastive scheme is designed to align input contents with stance labels. Our experimental results demonstrate the state-of-the-art performance across several publicly available datasets, showcasing effectiveness in both zero-shot and cross-target stance detection scenarios. We publicly release our code to facilitate future research.

pdf bib
Leveraging Prototypical Representations for Mitigating Social Bias without Demographic Information
Shadi Iskander | Kira Radinsky | Yonatan Belinkov

Mitigating social biases typically requires identifying the social groups associated with each data sample. In this paper, we present DAFair, a novel approach to address social bias in language models. Unlike traditional methods that rely on explicit demographic labels, our approach does not require any such information. Instead, we leverage predefined prototypical demographic texts and incorporate a regularization term during the fine-tuning process to mitigate bias in the model’s representations. Our empirical results across two tasks and two models demonstrate the effectiveness of our method compared to previous approaches that do not rely on labeled data. Moreover, with limited demographic-annotated data, our approach outperforms common debiasing approaches.

pdf bib
Direct Preference Optimization for Neural Machine Translation with Minimum Bayes Risk Decoding
Guangyu Yang | Jinghong Chen | Weizhe Lin | Bill Byrne

Minimum Bayes Risk (MBR) decoding can significantly improve translation performance of Multilingual Large Language Models (MLLMs). However, MBR decoding is computationally expensive. We show how the recently developed Reinforcement Learning technique, Direct Preference Optimization (DPO), can fine-tune MLLMs to get the gains of MBR without any additional computation in inference. Our method uses only a small monolingual fine-tuning set and yields significantly improved performance on multiple NMT test sets compared to MLLMs without DPO.

pdf bib
EchoPrompt: Instructing the Model to Rephrase Queries for Improved In-context Learning
Raja Sekhar Reddy Mekala | Yasaman Razeghi | Sameer Singh

Language models are achieving impressive performance on various tasks by aggressively adopting inference-time prompting techniques,such as zero-shot and few-shot prompting. In this work, we introduce EchoPrompt, a simple yet effective approach that prompts the model to rephrase its queries before answering them. EchoPrompt is tailored for four scenarios, including standard and chain-of-thought prompting, in both zero-shot and few-shot settings. Experimental results show that EchoPrompt yields substantial improvementsacross all these settings for four families of causal language models. These improvements are observed across various numerical reasoning (e.g., GSM8K, SVAMP), reading comprehension (e.g., DROP), and logical reasoning (e.g., Coin flipping) tasks. On average, EchoPrompt improves the Zero-shot-CoT performance of code-davinci-002 by 5% in numerical tasks and 13% in reading comprehension tasks. Our empirical results indicate that EchoPrompt is an effective technique that enhances in-context learning performance.

pdf bib
LEAF: Language Learners’ English Essays and Feedback Corpus
Shabnam Behzad | Omid Kashefi | Swapna Somasundaran

This paper addresses the issue of automated feedback generation for English language learners by presenting a corpus of English essays and their corresponding feedback, called LEAF, collected from the “essayforum” website. The corpus comprises approximately 6K essay-feedback pairs, offering a diverse and valuable resource for developing personalized feedback generation systems that address the critical deficiencies within essays, spanning from rectifying grammatical errors to offering insights on argumentative aspects and organizational coherence. Using this corpus, we present and compare multiple feedback generation baselines. Our findings shed light on the challenges of providing personalized feedback and highlight the potential of the LEAF corpus in advancing automated essay evaluation.

pdf bib
Zero-Shot vs. Translation-Based Cross-Lingual Transfer: The Case of Lexical Gaps
Abteen Ebrahimi | Katharina Wense

Cross-lingual transfer can be achieved through two main approaches: zero-shot transfer or machine translation (MT). While the former has been the dominant approach, both have been shown to be competitive. In this work, we compare the current performance and long-term viability of these methods. We leverage lexical gaps to create a multilingual question answering dataset, which provides a difficult domain for evaluation. Both approaches struggle in this setting, though zero-shot transfer performs better, as current MT outputs are not specific enough for the task. Using oracle translation offers the best performance, showing that this approach can perform well long-term, however current MT quality is a bottleneck. We also conduct an exploratory study to see if humans produce translations sufficient for the task with only general instructions. We find this to be true for the majority of translators, but not all. This indicates that while translation has the potential to outperform zero-shot approaches, creating MT models that generate accurate task-specific translations may not be straightforward.

pdf bib
On the True Distribution Approximation of Minimum Bayes-Risk Decoding
Atsumoto Ohashi | Ukyo Honda | Tetsuro Morimura | Yuu Jinnai

Minimum Bayes-risk (MBR) decoding has recently gained renewed attention in text generation.MBR decoding considers texts sampled from a model as pseudo-references and selects the text with the highest similarity to the others.Therefore, sampling is one of the key elements of MBR decoding, and previous studies reported that the performance varies by sampling methods.From a theoretical standpoint, this performance variation is likely tied to how closely the samples approximate the true distribution of references.However, this approximation has not been the subject of in-depth study.In this study, we propose using anomaly detection to measure the degree of approximation.We first closely examine the performance variation and then show that previous hypotheses about samples do not correlate well with the variation, but our introduced anomaly scores do.The results are the first to empirically support the link between the performance and the core assumption of MBR decoding.

pdf bib
Rehearsal-Free Modular and Compositional Continual Learning for Language Models
Mingyang Wang | Heike Adel | Lukas Lange | Jannik Strötgen | Hinrich Schuetze

Continual learning aims at incrementally acquiring new knowledge while not forgetting existing knowledge. To overcome catastrophic forgetting, methods are either rehearsal-based, i.e., store data examples from previous tasks for data replay, or isolate parameters dedicated to each task. However, rehearsal-based methods raise privacy and memory issues, and parameter-isolation continual learning does not consider interaction between tasks, thus hindering knowledge transfer. In this work, we propose MoCL, a rehearsal-free **Mo**dular and **C**ompositional Continual **L**earning framework which continually adds new modules to language models and composes them with existing modules. Experiments on various benchmarks show that MoCL outperforms state of the art and effectively facilitates knowledge transfer.

pdf bib
Llama meets EU: Investigating the European political spectrum through the lens of LLMs
Ilias Chalkidis | Stephanie Brandl

Instruction-finetuned Large Language Models inherit clear political leanings that have been shown to influence downstream task performance. We expand this line of research beyond the two-party system in the US and audit Llama Chat in the context of EU politics in various settings to analyze the model’s political knowledge and its ability to reason in context. We adapt, i.e., further fine-tune, Llama Chat on speeches of individual euro-parties from debates in the European Parliament to reevaluate its political leaning based on the EUandI questionnaire. Llama Chat shows considerable knowledge of national parties’ positions and is capable of reasoning in context. The adapted, party-specific, models are substantially re-aligned towards respective positions which we see as a starting point for using chat-based LLMs as data-driven conversational engines to assist research in political science.

pdf bib
M3T: A New Benchmark Dataset for Multi-Modal Document-Level Machine Translation
Benjamin Hsu | Xiaoyu Liu | Huayang Li | Yoshinari Fujinuma | Maria Nadejde | Xing Niu | Ron Litman | Yair Kittenplon | Raghavendra Pappagari

Document translation poses a challenge for Neural Machine Translation (NMT) systems. Most document-level NMT systems rely on meticulously curated sentence-level parallel data, assuming flawless extraction of text from documents along with their precise reading order. These systems also tend to disregard additional visual cues such as the document layout, deeming it irrelevant. However, real-world documents often possess intricate text layouts that defy these assumptions. Extracting information from Optical Character Recognition (OCR) or heuristic rules can result in errors, and the layout (e.g., paragraphs, headers) may convey relationships between distant sections of text. This complexity is particularly evident in widely used PDF documents, which represent information visually. This paper addresses this gap by introducing M3T a novel benchmark dataset tailored to evaluate NMT systems on the comprehensive task of translating semi-structured documents. This dataset aims to bridge the evaluation gap in document-level NMT systems, acknowledging the challenges posed by rich text layouts in real-world applications.

pdf bib
Control-DAG: Constrained Decoding for Non-Autoregressive Directed Acyclic T5 using Weighted Finite State Automata
Jinghong Chen | Weizhe Lin | Jingbiao Mei | Bill Byrne

The Directed Acyclic Transformer is a fast non-autoregressive (NAR) model that performs well in Neural Machine Translation. Two issues prevent its application to general Natural Language Generation (NLG) tasks: frequent Out-Of-Vocabulary (OOV) errors and the inability to faithfully generate entity names. We introduce Control-DAG, a constrained decoding algorithm for our Directed Acyclic T5 (DA-T5) model which offers lexical, vocabulary and length control. We show that Control-DAG significantly enhances DA-T5 on the Schema Guided Dialogue and the DART datasets, establishing strong NAR results for Task-Oriented Dialogue and Data-to-Text NLG.

pdf bib
Do Vision-Language Models Understand Compound Nouns?
Sonal Kumar | Sreyan Ghosh | S Sakshi | Utkarsh Tyagi | Dinesh Manocha

Open-vocabulary vision-language models (VLMs) like CLIP, trained using contrastive loss, have emerged as a promising new paradigm for text-to-image retrieval. However, do VLMs understand compound nouns (CNs) (e.g., *lab coat*) as well as they understand nouns (e.g., *lab*)? We curate Compun, a novel benchmark with 400 unique and commonly used CNs, to evaluate the effectiveness of VLMs in interpreting CNs. The Compun benchmark challenges a VLM for text-to-image retrieval where, given a text prompt with a CN, the task is to select the correct image that shows the CN among a pair of distractor images that show the constituent nouns that make up the CN. Next, we perform an in-depth analysis to highlight CLIPs’ limited understanding of certain types of CNs. Finally, we present an alternative framework that moves beyond hand-written templates for text prompts widely used by CLIP-like models. We employ a Large Language Model to generate multiple diverse captions that include the CN as an object in the scene described by the caption. Our proposed method improves CN understanding of CLIP by 8.25% on Compun. Code and benchmark are available.

pdf bib
Is Prompt Transfer Always Effective? An Empirical Study of Prompt Transfer for Question Answering
Minji Jung | Soyeon Park | Jeewoo Sul | Yong Suk Choi

Prompt tuning, which freezes all parameters of a pre-trained model and only trains a soft prompt, has emerged as a parameter-efficient approach. For the reason that the prompt initialization becomes sensitive when the model size is small, the prompt transfer that uses the trained prompt as an initialization for the target task has recently been introduced. Since previous works have compared tasks in large categories (e.g., summarization, sentiment analysis), the factors that influence prompt transfer have not been sufficiently explored. In this paper, we characterize the question answering task based on features such as answer format and empirically investigate the transferability of soft prompts for the first time. We analyze the impact of initialization during prompt transfer and find that the train dataset size of source and target tasks have the influence significantly. Furthermore, we propose a novel approach for measuring catastrophic forgetting and investigate how it occurs in terms of the amount of evidence. Our findings can help deeply understand transfer learning in prompt tuning.

pdf bib
Lost in Space: Probing Fine-grained Spatial Understanding in Vision and Language Resamplers
Georgios Pantazopoulos | Alessandro Suglia | Oliver Lemon | Arash Eshghi

An effective method for combining frozen large language models (LLM) and visual encoders involves a resampler module that creates a ‘visual prompt’ which is provided to the LLM, along with the textual prompt. While this approach has enabled impressive performance across many coarse-grained tasks like image captioning and visual question answering, more fine-grained tasks that require spatial understanding have not been thoroughly examined. In this paper, we use diagnostic classifiers to measure the extent to which the visual prompt produced by the resampler encodes spatial information. Our results show that this information is largely absent from the resampler output when kept frozen during training of the classifiers. However, when the resampler and classifier are trained jointly, we observe a significant performance boost. This shows that the compression achieved by the resamplers can in principle encode the requisite spatial information, but that more object-aware objectives are needed at the pretraining stage to facilitate this capability.

pdf bib
Do Multilingual Language Models Think Better in English?
Julen Etxaniz | Gorka Azkune | Aitor Soroa | Oier Lacalle | Mikel Artetxe

Translate-test is a popular technique to improve the performance of multilingual language models. This approach works by translating the input into English using an external machine translation system before running inference. However, these improvements can be attributed to the use of a separate translation system, which is typically trained on large amounts of parallel data not seen by the language model. In this work, we introduce a new approach called self-translate that leverages the few-shot translation capabilities of multilingual language models. This allows us to analyze the effect of translation in isolation. Experiments over 5 tasks show that self-translate consistently outperforms direct inference, demonstrating that language models are unable to leverage their full multilingual potential when prompted in non-English languages. Our code is available at https://github.com/juletx/self-translate.

pdf bib
A Continued Pretrained LLM Approach for Automatic Medical Note Generation
Dong Yuan | Eti Rastogi | Gautam Naik | Sree Prasanna Rajagopal | Sagar Goyal | Fen Zhao | Bharath Chintagunta | Jeffrey Ward

LLMs are revolutionizing NLP tasks. However, the use of the most advanced LLMs, such as GPT-4, is often prohibitively expensive for most specialized fields. We introduce HEAL, the first continuously trained 13B LLaMA2-based LLM that is purpose-built for medical conversations and measured on automated scribing. Our results demonstrate that HEAL outperforms GPT-4 and PMC-LLaMA in PubMedQA, with an accuracy of 78.4%. It also achieves parity with GPT-4 in generating medical notes. Remarkably, HEAL surpasses GPT-4 and Med-PaLM 2 in identifying more correct medical concepts and exceeds the performance of human scribes and other comparable models in correctness and completeness.

pdf bib
Lost in Translation? Translation Errors and Challenges for Fair Assessment of Text-to-Image Models on Multilingual Concepts
Michael Saxon | Yiran Luo | Sharon Levy | Chitta Baral | Yezhou Yang | William Yang Wang

Benchmarks of the multilingual capabilities of text-to-image (T2I) models compare generated images prompted in a test language to an expected image distribution over a concept set. One such benchmark, “Conceptual Coverage Across Languages” (CoCo-CroLa), assesses the tangible noun inventory of T2I models by prompting them to generate pictures from a concept list translated to seven languages and comparing the output image populations. Unfortunately, we find that this benchmark contains translation errors of varying severity in Spanish, Japanese, and Chinese. We provide corrections for these errors and analyze how impactful they are on the utility and validity of CoCo-CroLa as a benchmark. We reassess multiple baseline T2I models with the revisions, compare the outputs elicited under the new translations to those conditioned on the old, and show that a correction’s impactfulness on the image-domain benchmark results can be predicted in the text domain with similarity scores. Our findings will guide the future development of T2I multilinguality metrics by providing analytical tools for practical translation decisions.

pdf bib
Self-Improving for Zero-Shot Named Entity Recognition with Large Language Models
Tingyu Xie | Qi Li | Yan Zhang | Zuozhu Liu | Hongwei Wang

Exploring the application of powerful large language models (LLMs) on the named entity recognition (NER) task has drawn much attention recently. This work pushes the performance boundary of zero-shot NER with LLMs by proposing a training-free self-improving framework, which utilizes an unlabeled corpus to stimulate the self-learning ability of LLMs. First, we use the LLM to make predictions on the unlabeled corpus using self-consistency and obtain a self-annotated dataset. Second, we explore various strategies to select reliable annotations to form a reliable self-annotated dataset. Finally, for each test input, we retrieve demonstrations from the reliable self-annotated dataset and perform inference via in-context learning. Experiments on four benchmarks show substantial performance improvements achieved by our framework. Through comprehensive experimental analysis, we find that increasing the size of unlabeled corpus or iterations of self-improving does not guarantee further improvement, but the performance might be boosted via more advanced strategies for reliable annotation selection.

pdf bib
Lifelong Event Detection with Embedding Space Separation and Compaction
Chengwei Qin | Ruirui Chen | Ruochen Zhao | Wenhan Xia | Shafiq Joty

To mitigate forgetting, existing lifelong event detection methods typically maintain a memory module and replay the stored memory data during the learning of a new task. However, the simple combination of memory data and new-task samples can still result in substantial forgetting of previously acquired knowledge, which may occur due to the potential overlap between the feature distribution of new data and the previously learned embedding space. Moreover, the model suffers from overfitting on the few memory samples rather than effectively remembering learned patterns. To address the challenges of forgetting and overfitting, we propose a novel method based on embedding space separation and compaction. Our method alleviates forgetting of previously learned tasks by forcing the feature distribution of new data away from the previous embedding space. It also mitigates overfitting by a memory calibration mechanism that encourages memory data to be close to its prototype to enhance intra-class compactness. In addition, the learnable parameters of the new task are initialized by drawing upon acquired knowledge from the previously learned task to facilitate forward knowledge transfer. With extensive experiments, we demonstrate that our method can significantly outperform previous state-of-the-art approaches.

pdf bib
Language Models (Mostly) Do Not Consider Emotion Triggers When Predicting Emotion
Smriti Singh | Cornelia Caragea | Junyi Jessy Li

Situations and events evoke emotions in humans, but to what extent do they inform the prediction of emotion detection models? This work investigates how well human-annotated emotion triggers correlate with features that models deemed salient in their prediction of emotions. First, we introduce a novel dataset EmoTrigger, consisting of 900 social media posts sourced from three different datasets; these were annotated by experts for emotion triggers with high agreement. Using EmoTrigger, we evaluate the ability of large language models (LLMs) to identify emotion triggers, and conduct a comparative analysis of the features considered important for these tasks between LLMs and fine-tuned models. Our analysis reveals that emotion triggers are largely not considered salient features for emotion prediction models, instead there is intricate interplay between various features and the task of emotion detection.

pdf bib
CPopQA: Ranking Cultural Concept Popularity by LLMs
Ming Jiang | Mansi Joshi

Many recent studies examining the knowledge capacity of large language models (LLM) have focused on knowledge explicitly learned from the pretraining data or implicitly inferable from similar contexts. However, the extent to which an LLM effectively captures corpus-level statistical trends of concepts for reasoning, especially long-tail ones, is largely underexplored. In this study, we introduce a novel few-shot question-answering task (CPopQA) that examines LLMs’ statistical ranking abilities for long-tail cultural concepts (e.g., holidays), particularly focusing on these concepts’ popularity in the United States and the United Kingdom, respectively. We curate a dataset of 457 holidays across 58 countries, generating a total of 9,000 QA testing pairs. Experiments on four strong LLMs show that open-sourced LLMs still lag way behind close LLM API (e.g., GPT-3.5) in statistical ranking of cultural concepts. Notably, GPT-3.5 exhibited its potential to identify geo-cultural proximity across continents.

pdf bib
The Impact of Language on Arithmetic Proficiency: A Multilingual Investigation with Cross-Agent Checking Computation
Chung-Chi Chen | Hiroya Takamura | Ichiro Kobayashi | Yusuke Miyao

This paper critically examines the arithmetic capabilities of Large Language Models (LLMs), uncovering significant limitations in their performance. Our research reveals a notable decline in accuracy for complex calculations involving large numbers, with addition and subtraction tasks showing varying degrees of proficiency. Additionally, we challenge the notion that arithmetic is language-independent, finding up to a 10% difference in performance across twenty languages. The study also compares self-verification methods with cross-agent collaborations, showing that a single model often outperforms collaborative approaches in basic arithmetic tasks. These findings suggest a need to reassess the effectiveness of LLMs in tasks requiring numerical accuracy and precision.

pdf bib
Efficient Information Extraction in Few-Shot Relation Classification through Contrastive Representation Learning
Philipp Borchert | Jochen De Weerdt | Marie-Francine Moens

Differentiating relationships between entity pairs with limited labeled instances poses a significant challenge in few-shot relation classification. Representations of textual data extract rich information spanning the domain, entities, and relations. In this paper, we introduce a novel approach to enhance information extraction combining multiple sentence representations and contrastive learning. While representations in relation classification are commonly extracted using entity marker tokens, we argue that substantial information within the internal model representations remains untapped. To address this, we propose aligning multiple sentence representations, such as the CLS] token, the [MASK] token used in prompting, and entity marker tokens. Our method employs contrastive learning to extract complementary discriminative information from these individual representations. This is particularly relevant in low-resource settings where information is scarce. Leveraging multiple sentence representations is especially effective in distilling discriminative information for relation classification when additional information, like relation descriptions, are not available. We validate the adaptability of our approach, maintaining robust performance in scenarios that include relation descriptions, and showcasing its flexibility to adapt to different resource constraints.

pdf bib
A diverse Multilingual News Headlines Dataset from around the World
Felix Leeb | Bernhard Schölkopf

Babel Briefings is a novel dataset featuring 4.7 million news headlines from August 2020 to November 2021, across 30 languages and 54 locations worldwide with English translations of all articles included. Designed for natural language processing and media studies, it serves as a high-quality dataset for training or evaluating language models as well as offering a simple, accessible collection of articles, for example, to analyze global news coverage and cultural narratives. As a simple demonstration of the analyses facilitated by this dataset, we use a basic procedure using a TF-IDF weighted similarity metric to group articles into clusters about the same event. We then visualize the event signatures of the event showing articles of which languages appear over time, revealing intuitive features based on the proximity of the event and unexpectedness of the event. The dataset is available on [Kaggle](https://www.kaggle.com/datasets/felixludos/babel-briefings) and [HuggingFace](https://huggingface.co/datasets/felixludos/babel-briefings) with accompanying [GitHub](https://github.com/felixludos/babel-briefings) code.

pdf bib
The Unreasonable Effectiveness of Random Target Embeddings for Continuous-Output Neural Machine Translation
Evgeniia Tokarchuk | Vlad Niculae

Continuous-output neural machine translation (CoNMT) replaces the discrete next-word prediction problem with an embedding prediction.The semantic structure of the target embedding space (*i.e.*, closeness of related words) is intuitively believed to be crucial. We challenge this assumption and show that completely random output embeddings can outperform laboriously pre-trained ones, especially on larger datasets. Further investigation shows this surprising effect is strongest for rare words, due to the geometry of their embeddings. We shed further light on this finding by designing a mixed strategy that combines random and pre-trained embeddings, and that performs best overall.

pdf bib
Efficient Sample-Specific Encoder Perturbations
Yassir Fathullah | Mark Gales

Encoder-decoder foundation models have displayed state-of-the-art performance on a range of autoregressive sequence tasks. This paper proposes a simple and lightweight modification to such systems to control the behaviour according to a specific attribute of interest. This paper proposes a novel inference-efficient approach to modifying the behaviour of an encoder-decoder system according to a specific attribute of interest. Specifically, we show that a small proxy network can be used to find a sample-by-sample perturbation of the encoder output of a frozen foundation model to trigger the decoder to generate improved decodings. This work explores a specific realization of this framework focused on improving the COMET performance of Flan-T5 on Machine Translation and the WER of Whisper foundation models on Speech Recognition. Results display consistent improvements in performance evaluated through COMET and WER respectively. Furthermore, experiments also show that the proxies are robust to the exact nature of the data used to train them and can extend to other domains.

pdf bib
Diverse Perspectives, Divergent Models: Cross-Cultural Evaluation of Depression Detection on Twitter
Nuredin Ali Abdelkadir | Charles Zhang | Ned Mayo | Stevie Chancellor

Social media data has been used for detecting users with mental disorders, such as depression. Despite the global significance of cross-cultural representation and its potential impact on model performance, publicly available datasets often lack crucial metadata relatedto this aspect. In this work, we evaluate the generalization of benchmark datasets to build AI models on cross-cultural Twitter data. We gather a custom geo-located Twitter dataset of depressed users from seven countries as a test dataset. Our results show that depressiondetection models do not generalize globally. The models perform worse on Global South users compared to Global North. Pre-trainedlanguage models achieve the best generalization compared to Logistic Regression, though still show significant gaps in performance on depressed and non-Western users. We quantify our findings and provide several actionable suggestions to mitigate this issue

pdf bib
Removing RLHF Protections in GPT-4 via Fine-Tuning
Qiusi Zhan | Richard Fang | Rohan Bindu | Akul Gupta | Tatsunori Hashimoto | Daniel Kang

As large language models (LLMs) have increased in their capabilities, so doestheir potential for dual use. To reduce harmful outputs, produces and vendors ofLLMs have used reinforcement learning with human feedback (RLHF). In tandem,LLM vendors have been increasingly enabling fine-tuning of their most powerfulmodels. However, concurrent work has shown that fine-tuning can remove RLHFprotections. We may expect that the most powerful models currently available(GPT-4) are less susceptible to fine-tuning attacks. In this work, we show the contrary: fine-tuning allows attackers to remove RLHFprotections with as few as 340 examples and a 95% success rate. These trainingexamples can be automatically generated with weaker models. We further show thatremoving RLHF protections does not decrease usefulness on non-censored outputs,providing evidence that our fine-tuning strategy does not decrease usefulnessdespite using weaker models to generate training data. Our results show the needfor further research on protections on LLMs.

pdf bib
LifeTox: Unveiling Implicit Toxicity in Life Advice
Minbeom Kim | Jahyun Koo | Hwanhee Lee | Joonsuk Park | Hwaran Lee | Kyomin Jung

As large language models become increasingly integrated into daily life, detecting implicit toxicity across diverse contexts is crucial. To this end, we introduce LifeTox, a dataset designed for identifying implicit toxicity within a broad range of advice-seeking scenarios. Unlike existing safety datasets, LifeTox comprises diverse contexts derived from personal experiences through open-ended questions. Our experiments demonstrate that RoBERTa fine-tuned on LifeTox matches or surpasses the zero-shot performance of large language models in toxicity classification tasks. These results underscore the efficacy of LifeTox in addressing the complex challenges inherent in implicit toxicity. We open-sourced the dataset and the LifeTox moderator family; 350M, 7B, and 13B.

pdf bib
Arithmetic Reasoning with LLM: Prolog Generation & Permutation
Xiaocheng Yang | Bingsen Chen | Yik-Cheung Tam

Instructing large language models (LLMs) to solve elementary school math problems has shown great success using Chain of Thought (CoT). However, the CoT approach relies on an LLM to generate a sequence of arithmetic calculations which can be prone to cascaded calculation errors. We hypothesize that an LLM should focus on extracting predicates and generating symbolic formulas from the math problem description so that the underlying calculation can be done via an external code interpreter. We investigate using LLM to generate Prolog programs to solve mathematical questions. Experimental results show that our Prolog-based arithmetic problem-solving outperforms CoT generation in the GSM8K benchmark across three distinct LLMs. In addition, given the insensitive ordering of predicates and symbolic formulas in Prolog, we propose to permute the ground truth predicates for more robust LLM training via data augmentation.

pdf bib
Verifying Claims About Metaphors with Large-Scale Automatic Metaphor Identification
Kotaro Aono | Ryohei Sasano | Koichi Takeda

There are several linguistic claims about situations where words are more likely to be used as metaphors.However, few studies have sought to verify such claims with large corpora.This study entails a large-scale, corpus-based analysis of certain existing claims about verb metaphors, by applying metaphor detection to sentences extracted from Common Crawl and using the statistics obtained from the results.The verification results indicate that the direct objects of verbs used as metaphors tend to have lower degrees of concreteness, imageability, and familiarity, and that metaphors are more likely to be used in emotional and subjective sentences.

pdf bib
InstructABSA: Instruction Learning for Aspect Based Sentiment Analysis
Kevin Scaria | Himanshu Gupta | Siddharth Goyal | Saurabh Sawant | Swaroop Mishra | Chitta Baral

We introduce InstructABSA, an instruction learning paradigm for Aspect-Based Sentiment Analysis (ABSA) subtasks.Our method introduces positive, negative, and neutral examples to each training sample, and instruction tune the model (Tk-Instruct) for ABSA subtasks, yielding significant performance improvements. Experimental results on the Sem Eval 2014, 15, and 16 datasets demonstrate that InstructABSA outperforms the previous state-of-the-art (SOTA) approaches on Term Extraction (ATE), Sentiment Classification(ATSC) and Sentiment Pair Extraction (ASPE) subtasks.In particular, InstructABSA outperforms the previous state-of-the-art (SOTA) on the Rest14 ATE subtask by 5.69% points, the Rest15 ATSC subtask by 9.59% points, and the Lapt14 AOPE subtask by 3.37% points, surpassing 7x larger models.We get competitive results on AOOE, AOPE, AOSTE, and ACOSQE subtasks indicating strong generalization ability to all subtasks. Exploring sample efficiency reveals that just 50% train data is required to get competitive results with other instruction tuning approaches. Lastly, we assess the quality of instructions and observe that InstructABSA’s performance experiences a decline of ~10% when adding misleading examples

pdf bib
MEMORY-VQ: Compression for Tractable Internet-Scale Memory
Yury Zemlyanskiy | Michiel de Jong | Luke Vilnis | Santiago Ontanon | William Cohen | Sumit Sanghai | Joshua Ainslie

Retrieval augmentation is a powerful but expensive method to make language models more knowledgeable about the world. Memory-based methods like LUMEN (de Jong et al., 2023a) pre-compute token representations for retrieved passages to drastically speed up inference. However, memory also leads to much greater storage requirements from storing pre-computed representations. We propose MEMORY-VQ, a new method to reduce storage requirements of memory-augmented models without sacrificing performance. Our method uses a vector quantization variational autoencoder (VQ-VAE) to compress token representations. We apply MEMORY-VQ to the LUMEN model to obtain LUMEN-VQ, a memory model that achieves a 16x compression rate with comparable performance on the KILT benchmark. LUMEN-VQ enables practical retrieval augmentation even for extremely large retrieval corpora.

pdf bib
Unveiling the Magic: Investigating Attention Distillation in Retrieval-Augmented Generation
Zizhong Li | Haopeng Zhang | Jiawei Zhang

Retrieval-augmented generation framework addresses the limitations of large language models by enabling real-time knowledge updates for more accurate answers. An efficient way in the training phase of retrieval-augmented models is attention distillation, which uses attention scores as supervision signals instead of manually annotated query-document pairs. Despite its growing popularity, the detailed mechanisms behind the success of attention distillation remain unexplored, particularly the specific patterns it leverages to benefit training. In this paper, we address this gap by conducting a comprehensive investigation of attention distillation workflow and identifying key factors influencing the learning performance of retrieval-augmented language models. We further propose several insightful indicators for optimizing models’ training methods and avoiding ineffective training.

pdf bib
Improving Factuality in Clinical Abstractive Multi-Document Summarization by Guided Continued Pre-training
Ahmed Elhady | Khaled Elsayed | Eneko Agirre | Mikel Artetxe

Factual accuracy is an important property of neural abstractive summarization models, especially in fact-critical domains such as the clinical literature. In this work, we introduce a guided continued pre-training stage for encoder-decoder models that improves their understanding of the factual attributes of documents, which is followed by supervised fine-tuning on summarization. Our approach extends the pre-training recipe of BART to incorporate 3 additional objectives based on PICO spans, which capture the population, intervention, comparison, and outcomes related to a clinical study. Experiments on multi-document summarization in the clinical domain demonstrate that our approach is competitive with prior work, improving the quality and factuality of the summaries and achieving the best-published results in factual accuracy on the MSLR task.

pdf bib
MuLan: A Study of Fact Mutability in Language Models
Constanza Fierro | Nicolas Garneau | Emanuele Bugliarello | Yova Kementchedjhieva | Anders Søgaard

Facts are subject to contingencies and can be true or false in different circumstances. One such contingency is time, wherein some facts mutate over a given period, e.g., the president of a country or the winner of a championship. Trustworthy language models ideally identify mutable facts as such and process them accordingly. We create MuLan, a benchmark for evaluating the ability of English language models to anticipate time-contingency, covering both 1:1 and 1:N relations. We hypothesize that mutable facts are encoded differently than immutable ones, hence being easier to update. In a detailed evaluation of six popular large language models, we consistently find differences in the LLMs’ confidence, representations, and update behavior, depending on the mutability of a fact. Our findings should inform future work on the injection of and induction of time-contingent knowledge to/from LLMs.

pdf bib
Language-Independent Representations Improve Zero-Shot Summarization
Vladimir Solovyev | Danni Liu | Jan Niehues

Finetuning pretrained models on downstream generation tasks often leads to catastrophic forgetting in zero-shot conditions. In this work, we focus on summarization and tackle the problem through the lens of language-independent representations. After training on monolingual summarization, we perform zero-shot transfer to new languages or language pairs. We first show naively finetuned models are highly language-specific in both output behavior and internal representations, resulting in poor zero-shot performance. Next, we propose query-key (QK) finetuning to decouple task-specific knowledge from the pretrained language generation abilities. Then, after showing downsides of the standard adversarial language classifier, we propose a balanced variant that more directly enforces language-agnostic representations. Moreover, our qualitative analyses show removing source language identity correlates to zero-shot summarization performance. Our code is openly available.

pdf bib
Trusting Your Evidence: Hallucinate Less with Context-aware Decoding
Weijia Shi | Xiaochuang Han | Mike Lewis | Yulia Tsvetkov | Luke Zettlemoyer | Wen-tau Yih

Language models (LMs) often struggle to pay enough attention to the input context, and generate texts that are unfaithful or contain hallucinations. To mitigate this issue, we present context-aware decoding (CAD), which follows a contrastive output distribution that amplifies the difference between the output probabilities when a model is used with and without context. Our experiments show that CAD, without additional training, significantly improves the faithfulness of different LM families, including OPT, GPT, LLaMA, and FLAN-T5 for summarization tasks (e.g., 14.3% gain for LLaMA in factuality metrics). Furthermore, CAD is particularly effective in overriding a model’s prior knowledge when it contradicts the provided context, leading to substantial improvements in tasks where resolving the knowledge conflict is essential. Our code is publicly released at https://github.com/xhan77/context-aware-decoding.

pdf bib
GuyLingo: The Republic of Guyana Creole Corpora
Christopher Clarke | Roland Daynauth | Jason Mars | Charlene Wilkinson | Hubert Devonish

While major languages often enjoy substantial attention and resources, the linguistic diversity across the globe encompasses a multitude of smaller, indigenous, and regional languages that lack the same level of computational support. One such region is the Caribbean. While commonly labeled as “English speaking”, the ex-British Caribbean region consists of a myriad of Creole languages thriving alongside English. In this paper, we present Guylingo: a comprehensive corpus designed for advancing NLP research in the domain of Creolese (Guyanese English-lexicon Creole), the most widely spoken language in the culturally rich nation of Guyana. We first outline our framework for gathering and digitizing this diverse corpus, inclusive of colloquial expressions, idioms, and regional variations in a low-resource language. We then demonstrate the challenges of training and evaluating NLP models for machine translation for Creolese. Lastly, we discuss the unique opportunities presented by recent NLP advancements for accelerating the formal adoption of Creole languages as official languages in the Caribbean.

pdf bib
DoubleLingo: Causal Estimation with Large Language Models
Marko Veljanovski | Zach Wood-Doughty

Estimating causal effects from non-randomized data requires assumptions about the underlying data-generating process. To achieve unbiased estimates of the causal effect of a treatment on an outcome, we typically adjust for any confounding variables that influence both treatment and outcome. When such confounders include text data, existing causal inference methods struggle due to the high dimensionality of the text. The simple statistical models which have sufficient convergence criteria for causal estimation are not well-equipped to handle noisy unstructured text, but flexible large language models that excel at predictive tasks with text data do not meet the statistical assumptions necessary for causal estimation. Our method enables theoretically consistent estimation of causal effects using LLM-based nuisance models by incorporating them within the framework of Double Machine Learning. On the best available dataset for evaluating such methods, we obtain a 10.4% reduction in the relative absolute error for the estimated causal effect over existing methods.

pdf bib
Improved Text Emotion Prediction Using Combined Valence and Arousal Ordinal Classification
Michail Mitsios | Georgios Vamvoukakis | Georgia Maniati | Nikolaos Ellinas | Georgios Dimitriou | Konstantinos Markopoulos | Panos Kakoulidis | Alexandra Vioni | Myrsini Christidou | Junkwang Oh | Gunu Jho | Inchul Hwang | Georgios Vardaxoglou | Aimilios Chalamandaris | Pirros Tsiakoulis | Spyros Raptis

Emotion detection in textual data has received growing interest in recent years, as it is pivotal for developing empathetic human-computer interaction systems.This paper introduces a method for categorizing emotions from text, which acknowledges and differentiates between the diversified similarities and distinctions of various emotions.Initially, we establish a baseline by training a transformer-based model for standard emotion classification, achieving state-of-the-art performance. We argue that not all misclassifications are of the same importance, as there are perceptual similarities among emotional classes.We thus redefine the emotion labeling problem by shifting it from a traditional classification model to an ordinal classification one, where discrete emotions are arranged in a sequential order according to their valence levels.Finally, we propose a method that performs ordinal classification in the two-dimensional emotion space, considering both valence and arousal scales.The results show that our approach not only preserves high accuracy in emotion prediction but also significantly reduces the magnitude of errors in cases of misclassification.

pdf bib
On Narrative Question Answering Skills
Emil Kalbaliyev | Kairit Sirts

Narrative Question Answering is an important task for evaluating and improving reading comprehension abilities in both humans and machines. However, there is a lack of consensus on the skill taxonomy that would enable systematic and comprehensive assessment and learning of the various aspects of Narrative Question Answering. Existing task-level skill views oversimplify the multidimensional nature of tasks, while question-level taxonomies face issues in evaluation and methodology. To address these challenges, we introduce a more inclusive skill taxonomy that synthesizes and redefines narrative understanding skills from previous taxonomies and includes a generation skill dimension from the answering perspective.

pdf bib
Order-Based Pre-training Strategies for Procedural Text Understanding
Abhilash Nandy | Yash Kulkarni | Pawan Goyal | Niloy Ganguly

In this paper, we propose sequence-based pre-training methods to enhance procedural understanding in natural language processing. Procedural text, containing sequential instructions to accomplish a task, is difficult to understand due to the changing attributes of entities in the context. We focus on recipes as they are commonly represented as ordered instructions, and use this order as a supervision signal. Our work is one of the first to compare several ‘order-as-supervision’ transformer pre-training methods, including Permutation Classification, Embedding Regression, and Skip-Clip, and show that these methods give improved results compared to baselines and SoTA LLMs on two downstream Entity-Tracking datasets: NPN-Cooking dataset in recipe domain and ProPara dataset in open domain. Our proposed methods address the non-trivial Entity Tracking Task that requires prediction of entity states across procedure steps, which requires understanding the order of steps. These methods show an improvement over the best baseline by 1.6% and 7-9% on NPN-Cooking and ProPara Datasets respectively across metrics.

pdf bib
Breaking the Language Barrier: Can Direct Inference Outperform Pre-Translation in Multilingual LLM Applications?
Yotam Intrator | Matan Halfon | Roman Goldenberg | Reut Tsarfaty | Matan Eyal | Ehud Rivlin | Yossi Matias | Natalia Aizenberg

Large language models hold significant promise in multilingual applications. However, inherent biases stemming from predominantly English-centric pre-training have led to the widespread practice of pre-translation, i.e., translating non-English inputs to English before inference, leading to complexity and information loss. This study re-evaluates the need for pre-translation in the context of PaLM2 models, which have been established as highly performant in multilingual tasks. We offer a comprehensive investigation across 108 languages and 6 diverse benchmarks, including open-end generative tasks, which were excluded from previous similar studies. Our findings challenge the pre-translation paradigm established in prior research, highlighting the advantages of direct inference in PaLM2. Specifically, PaLM2-L consistently outperforms pre-translation in 94 out of 108 languages. These findings pave the way for more efficient and effective multilingual applications, alleviating the limitations associated with pre-translation and unlocking linguistic authenticity.
up

pdf (full)
bib (full) Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 3: System Demonstrations)

pdf bib
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 3: System Demonstrations)
Kai-Wei Chang | Annie Lee | Nazneen Rajani

pdf bib
TOPICAL: TOPIC Pages AutomagicaLly
John Giorgi | Amanpreet Singh | Doug Downey | Sergey Feldman | Lucy Wang

Topic pages aggregate useful information about an entity or concept into a single succinct and accessible article. Automated creation of topic pages would enable their rapid curation as information resources, providing an alternative to traditional web search. While most prior work has focused on generating topic pages about biographical entities, in this work, we develop a completely automated process to generate high-quality topic pages for scientific entities, with a focus on biomedical concepts. We release TOPICAL, a web app and associated open-source code, comprising a model pipeline combining retrieval, clustering, and prompting, that makes it easy for anyone to generate topic pages for a wide variety of biomedical entities on demand. In a human evaluation of 150 diverse topic pages generated using TOPICAL, we find that the vast majority were considered relevant, accurate, and coherent, with correct supporting citations. We make all code publicly available and host a free-to-use web app at: https://s2-topical.apps.allenai.org.

pdf bib
Low-code LLM: Graphical User Interface over Large Language Models
Yuzhe Cai | Shaoguang Mao | Wenshan Wu | Zehua Wang | Yaobo Liang | Tao Ge | Chenfei Wu | WangYou WangYou | Ting Song | Yan Xia | Nan Duan | Furu Wei

Utilizing Large Language Models (LLMs) for complex tasks is challenging, often involving a time-consuming and uncontrollable prompt engineering process. This paper introduces a novel human-LLM interaction framework, Low-code LLM. It incorporates six types of simple low-code visual programming interactions to achieve more controllable and stable responses. Through visual interaction with a graphical user interface, users can incorporate their ideas into the process without writing trivial prompts. The proposed Low-code LLM framework consists of a Planning LLM that designs a structured planning workflow for complex tasks, which can be correspondingly edited and confirmed by users through low-code visual programming operations, and an Executing LLM that generates responses following the user-confirmed workflow. We highlight three advantages of the low-code LLM: user-friendly interaction, controllable generation, and wide applicability. We demonstrate its benefits using four typical applications. By introducing this framework, we aim to bridge the gap between humans and LLMs, enabling more effective and efficient utilization of LLMs for complex tasks. The code, prompts, and experimental details are available at https://github.com/moymix/TaskMatrix/tree/main/LowCodeLLM. A system demonstration video can be found at https://www.youtube.com/watch?v=jb2C1vaeO3E.

pdf bib
EdTec-QBuilder: A Semantic Retrieval Tool for Assembling Vocational Training Exams in German Language
Alonso Palomino | Andreas Fischer | Jakub Kuzilek | Jarek Nitsch | Niels Pinkwart | Benjamin Paassen

Selecting and assembling test items from a validated item database into comprehensive exam forms is an under-researched but significant challenge in education. Search and retrieval methods provide a robust framework to assist educators when filtering and assembling relevant test items. In this work, we present EdTec-QBuilder, a semantic search tool developed to assist vocational educators in assembling exam forms. To implement EdTec-QBuilder’s core search functionality, we evaluated eight retrieval strategies and twenty-five popular pre-trained sentence similarity models. Our evaluation revealed that employing cross-encoders to re-rank an initial list of relevant items is best for assisting vocational trainers in assembling examination forms. Beyond topic-based exam assembly, EdTec-QBuilder aims to provide a crowdsourcing infrastructure enabling manual exam assembly data collection, which is critical for future research and development in assisted and automatic exam assembly models.

pdf bib
DIALIGHT: Lightweight Multilingual Development and Evaluation of Task-Oriented Dialogue Systems with Large Language Models
Songbo Hu | Xiaobin Wang | Moy Yuan | Anna Korhonen | Ivan Vulić

We present DIALIGHT, a toolkit for developing and evaluating multilingual Task-Oriented Dialogue (ToD) systems which facilitates systematic evaluations and comparisons between ToD systems using fine-tuning of Pretrained Language Models (PLMs) and those utilising the zero-shot and in-context learning capabilities of Large Language Models (LLMs). In addition to automatic evaluation, this toolkit features (i) a secure, user-friendly web interface for fine-grained human evaluation at both local utterance level and global dialogue level, and (ii) a microservice-based backend, improving efficiency and scalability. Our evaluations reveal that while PLM fine-tuning leads to higher accuracy and coherence, LLM-based systems excel in producing diverse and likeable responses. However, we also identify significant challenges of LLMs in adherence to task-specific instructions and generating outputs in multiple languages, highlighting areas for future research. We hope this open-sourced toolkit will serve as a valuable resource for researchers aiming to develop and properly evaluate multilingual ToD systems and will lower, currently still high, entry barriers in the field.

pdf bib
RTSUM: Relation Triple-based Interpretable Summarization with Multi-level Salience Visualization
Seonglae Cho | Myungha Jang | Jinyoung Yeo | Dongha Lee

In this paper, we present RTSum, an unsupervised summarization framework that utilizes relation triples as the basic unit for summarization. Given an input document, RTSum first selects salient relation triples via multi-level salience scoring and then generates a concise summary from the selected relation triples by using a text-to-text language model. On the basis of RTSum, we also develop a web demo for an interpretable summarizing tool, providing fine-grained interpretations with the output summary. With support for customization options, our tool visualizes the salience for textual units at three distinct levels: sentences, relation triples, and phrases. The code, demo, and video are publicly available.

pdf bib
Edu-ConvoKit: An Open-Source Library for Education Conversation Data
Rose Wang | Dorottya Demszky

We introduce Edu-ConvoKit, an open-source library designed to handle pre-processing, annotation and analysis of conversation data in education. Resources for analyzing education conversation data are scarce, making the research challenging to perform and therefore hard to access. We address these challenges with Edu-ConvoKit. Edu-ConvoKit is open-source [1], pip-installable [2], with comprehensive documentation [3]. Our demo video is available at: https://youtu.be/zdcI839vAko?si=h9qlnl76ucSuXb8-. We include additional resources, such as Colab applications of Edu-ConvoKit to three diverse education datasets [4] and a repository of Edu-ConvoKit-related papers [5].[1] https://github.com/stanfordnlp/edu-convokit[2] https://pypi.org/project/edu-convokit/[3] https://edu-convokit.readthedocs.io/en/latest/[4] https://github.com/stanfordnlp/edu-convokit?tab=readme-ov-file#datasets-with-edu-convokit[5] https://github.com/stanfordnlp/edu-convokit/blob/main/papers.md

pdf bib
jp-evalb: Robust Alignment-based PARSEVAL Measures
Jungyeul Park | Junrui Wang | Eunkyul Jo | Angela Park

We introduce an evaluation system designed to compute PARSEVAL measures, offering a viable alternative to evalb commonly used for constituency parsing evaluation. The widely used evalb script has traditionally been employed for evaluating the accuracy of constituency parsing results, albeit with the requirement for consistent tokenization and sentence boundaries. In contrast, our approach, named jp-evalb, is founded on an alignment method. This method aligns sentences and words when discrepancies arise. It aims to overcome several known issues associated with evalb by utilizing the ‘jointly preprocessed (JP)’ alignment-based method. We introduce a more flexible and adaptive framework, ultimately contributing to a more accurate assessment of constituency parsing performance.

pdf bib
OpinionGPT: Modelling Explicit Biases in Instruction-Tuned LLMs
Patrick Haller | Ansar Aynetdinov | Alan Akbik

Instruction-tuned Large Language Models (LLMs) have recently showcased remarkable ability to generate fitting responses to natural language instructions. However, an open research question concerns the inherent biases of trained models and their responses. For instance, if the data used to tune an LLM is dominantly written by persons with a specific political bias, we might expect generated answers to share this bias. Current research work seeks to de-bias such models, or suppress potentially biased answers.With this demonstration, we take a different view on biases in instruction-tuning: Rather than aiming to suppress them, we aim to make them explicit and transparent. To this end, we present OpinionGPT, a web demo in which users can ask questions and select all biases they wish to investigate. The demo will answer this question using a model fine-tuned on text representing each of the selected biases, allowing side-by-side comparison. To train the underlying model, we identified 11 different biases (political, geographic, gender, age) and derived an instruction-tuning corpus in which each answer was written by members of one of these demographics. This paper presents OpinionGPT, illustrates how we trained the bias-aware model and showcases the web application (available at https://opiniongpt.informatik.hu-berlin.de).

pdf bib
ATLAS: A System for PDF-centric Human Interaction Data Collection
Alexa Siu | Zichao Wang | Joshua Hoeflich | Naman Kapasi | Ani Nenkova | Tong Sun

The Portable Document Format (PDF) is a popular format for distributing digital documents. Datasets on PDF reading behaviors and interactions remain limited due to the challenges of instrumenting PDF readers for these data collection tasks. We present ATLAS, a data collection tool designed to better support researchers in collecting rich PDF-centric datasets from users. ATLAS supports researchers in programmatically creating a user interface for data collection that is ready to share with annotators. It includes a toolkit and an extensible schema to easily customize the data collection tasks for a variety of purposes, allowing collection of PDF annotations (e.g., highlights, drawings) as well as reading behavior analytics (e.g., page scroll, text selections). We open-source ATLAS1 to support future research efforts and review use cases of ATLAS that showcase our system’s broad applicability.

pdf bib
BeLeaf: Belief Prediction as Tree Generation
John Murzaku | Owen Rambow

We present a novel approach to predicting source-and-target factuality by transforming it into a linearized tree generation task. Unlike previous work, our model and representation format fully account for the factuality tree structure, generating the full chain of nested sources instead of the last source only. Furthermore, our linearized tree representation significantly compresses the amount of tokens needed compared to other representations, allowing for fully end-to-end systems. We achieve state-of-the-art results on FactBank and the Modal Dependency Corpus, which are both corpora annotating source-and-target event factuality. Our results on fine-tuning validate the strong generality of the proposed linearized tree generation task, which can be easily adapted to other corpora with a similar structure. We then present BeLeaf, a system which directly leverages the linearized tree representation to create both sentence level and document level visualizations. Our system adds several missing pieces to the source-and-target factuality task such as coreference resolution and event head word to syntactic span conversion. Our demo code is available on https://github.com/yurpl/beleaf and our video is available on https://youtu.be/SpbMNnin-Po.

pdf bib
QueryExplorer: An Interactive Query Generation Assistant for Search and Exploration
Kaustubh Dhole | Shivam Bajaj | Ramraj Chandradevan | Eugene Agichtein

Formulating effective search queries remains a challenging task, particularly when users lack expertise in a specific domain or are not proficient in the language of the content. Providing example documents of interest might be easier for a user. However, such query-by-example scenarios are prone to concept drift, and the retrieval effectiveness is highly sensitive to the query generation method, without a clear way to incorporate user feedback. To enable exploration and to support Human-In-The-Loop experiments we propose QueryExplorer– an interactive query generation, reformulation, and retrieval interface with support for Hug-gingFace generation models and PyTerrier’sretrieval pipelines and datasets, and extensivelogging of human feedback. To allow users to create and modify effective queries, our demo supports complementary approaches of using LLMs interactively, assisting the user with edits and feedback at multiple stages of the query formulation process. With support for recording fine-grained interactions and user annotations, QueryExplorer can serve as a valuable experimental and research platform for annotation, qualitative evaluation, and conducting Human-in-the-Loop (HITL) experiments for complex search tasks where users struggle to formulate queries.

pdf bib
LMFlow: An Extensible Toolkit for Finetuning and Inference of Large Foundation Models
Shizhe Diao | Rui Pan | Hanze Dong | KaShun Shum | Jipeng Zhang | Wei Xiong | Tong Zhang

Foundation models have demonstrated a great ability to achieve general human-level intelligence far beyond traditional approaches. As the technique keeps attracting attention from the AI community, more and more foundation models have become publicly available.However, most of those models exhibit a major deficiency in specialized-domain and specialized-task applications, where the step of domain- and task-aware finetuning is still required to obtain scientific language models. As the number of available foundation models and specialized tasks keeps growing, the job of training scientific language models becomes highly nontrivial. In this paper, we take the first step to address this issue. We introduce an extensible and lightweight toolkit, LMFlow, which aims to simplify the domain- and task-aware finetuning of general foundation models.LMFlow offers a complete finetuning workflow for a foundation model to support specialized training with limited computing resources.Furthermore, it supports continuous pretraining, instruction tuning, parameter-efficient finetuning, alignment tuning, inference acceleration, long context generalization, model customization, and even multimodal finetuning, along with carefully designed and extensible APIs. This toolkit has been thoroughly tested and is available at https://github.com/OptimalScale/LMFlow.

pdf bib
DOCMASTER: A Unified Platform for Annotation, Training, & Inference in Document Question-Answering
Alex Nguyen | Zilong Wang | Jingbo Shang | Dheeraj Mekala

The application of natural language processing models to PDF documents is pivotal for various business applications yet the challenge of training models for this purpose persists in businesses due to specific hurdles. These include the complexity of working with PDF formats that necessitate parsing text and layout information for curating training data and the lack of privacy-preserving annotation tools. This paper introduces DOCMASTER, a unified platform designed for annotating PDF documents, model training, and inference, tailored to document question-answering. The annotation interface enables users to input questions and highlight text spans within the PDF file as answers, saving layout information and text spans accordingly. Furthermore, DOCMASTER supports both state-of-the-art layout-aware and text models for comprehensive training purposes. Importantly, as annotations, training, and inference occur on-device, it also safeguards privacy. The platform has been instrumental in driving several research prototypes concerning document analysis such as the AI assistant utilized by University of California San Diego’s (UCSD) International Services and Engagement Office (ISEO) for processing a substantial volume of PDF documents.

pdf bib
RedCoast: A Lightweight Tool to Automate Distributed Training of LLMs on Any GPU/TPUs
Bowen Tan | Yun Zhu | Lijuan Liu | Hongyi Wang | Yonghao Zhuang | Jindong Chen | Eric Xing | Zhiting Hu

The recent progress of AI can be largely attributed to large language models (LLMs). However, their escalating memory requirements introduce challenges for machine learning (ML) researchers and engineers. Addressing this requires developers to partition a large model to distribute it across multiple GPUs or TPUs. This necessitates considerable coding and intricate configuration efforts with existing model parallel tools, such as Megatron-LM, DeepSpeed, and Alpa. These tools require users’ expertise in machine learning systems (MLSys), creating a bottleneck in LLM development, particularly for developers without MLSys background. In this work, we present RedCoast (Redco), a lightweight and user-friendly tool crafted to automate distributed training and inference for LLMs, as well as to simplify ML pipeline development. The design of Redco emphasizes two key aspects. Firstly, to automate model parallelism, our study identifies two straightforward rules to generate tensor parallel strategies for any given LLM. Integrating these rules into Redco facilitates effortless distributed LLM training and inference, eliminating the need of additional coding or complex configurations. We demonstrate the effectiveness by applying Redco on a set of LLM architectures, such as GPT-J, LLaMA, T5, and OPT, up to the size of 66B. Secondly, we propose a mechanism that allows for the customization of diverse ML pipelines through the definition of merely three functions, avoiding redundant and formulaic code like multi-host related processing. This mechanism proves adaptable across a spectrum of ML algorithms, from foundational language modeling to complex algorithms like meta-learning and reinforcement learning. As a result, Redco implementations exhibit significantly fewer lines of code compared to their official counterparts. RedCoast (Redco) has been released under Apache 2.0 license at https://github.com/tanyuqian/redco.

pdf bib
Concept Over Time Analysis: Unveiling Temporal Patterns for Qualitative Data Analysis
Tim Fischer | Florian Schneider | Robert Geislinger | Florian Helfer | Gertraud Koch | Chris Biemann

In this system demonstration paper, we present the Concept Over Time Analysis extension for the Discourse Analysis Tool Suite.The proposed tool empowers users to define, refine, and visualize their concepts of interest within an interactive interface. Adhering to the Human-in-the-loop paradigm, users can give feedback through sentence annotations. Utilizing few-shot sentence classification, the system employs Sentence Transformers to compute representations of sentences and concepts. Through an iterative process involving semantic similarity searches, sentence annotation, and fine-tuning with contrastive data, the model continuously refines, providing users with enhanced analysis outcomes. The final output is a timeline visualization of sentences classified to concepts. Especially suited for the Digital Humanities, Concept Over Time Analysis serves as a valuable tool for qualitative data analysis within extensive datasets. The chronological overview of concepts enables researchers to uncover patterns, trends, and shifts in discourse over time.

pdf bib
pyvene: A Library for Understanding and Improving PyTorch Models via Interventions
Zhengxuan Wu | Atticus Geiger | Aryaman Arora | Jing Huang | Zheng Wang | Noah Goodman | Christopher Manning | Christopher Potts

Interventions on model-internal states are fundamental operations in many areas of AI, including model editing, steering, robustness, and interpretability. To facilitate such research, we introduce pyvene, an open-source Python library that supports customizable interventions on a range of different PyTorch modules. pyvene supports complex intervention schemes with an intuitive configuration format, and its interventions can be static or include trainable parameters. We show how pyvene provides a unified and extensible framework for performing interventions on neural models and sharing the intervened upon models with others. We illustrate the power of the library via interpretability analyses using causal abstraction and knowledge localization. We publish our library through Python Package Index (PyPI) and provide code, documentation, and tutorials at ‘https://github.com/stanfordnlp/pyvene‘.

pdf bib
Newspaper Signaling for Crisis Prediction
Prajvi Saxena | Sabine Janzen | Wolfgang Maass

To establish sophisticated monitoring of newspaper articles for detecting crisis-related signals, natural language processing has to cope with unstructured data, media, and cultural bias as well as multiple languages. So far, research on detecting signals in newspaper articles is focusing on structured data, restricted language settings, and isolated application domains. When considering complex crisis-related signals, a high number of diverse newspaper articles in terms of language and culture reduces potential biases. We demonstrate MENDEL – a model for multi-lingual and open-domain newspaper signaling for detecting crisis-related indicators in newspaper articles. The model works with unstructured news data and combines multiple transformer-based models for pre-processing (STANZA) and content filtering (RoBERTa, GPT-3.5). Embedded in a Question-Answering (QA) setting, MENDEL supports multiple languages (>66) and can detect early newspaper signals for open crisis domains in real-time.

pdf bib
FastFit: Fast and Effective Few-Shot Text Classification with a Multitude of Classes
Asaf Yehudai | Elron Bandel

We present FastFit, a Python package designed to provide fast and accurate few-shot classification, especially for scenarios with many semantically similar classes. FastFit utilizes a novel approach integrating batch contrastive learning and token-level similarity score. Compared to existing few-shot learning packages, such as SetFit, Transformers, or few-shot prompting of large language models via API calls, FastFit significantly improves multi-class classification performance in speed and accuracy across various English and Multilingual datasets. FastFit demonstrates a 3-20x improvement in training speed, completing training in just a few seconds. The FastFit package is now available on GitHub, presenting a user-friendly solution for NLP practitioners.

pdf bib
AgentQuest: A Modular Benchmark Framework to Measure Progress and Improve LLM Agents
Luca Gioacchini | Giuseppe Siracusano | Davide Sanvito | Kiril Gashteovski | David Friede | Roberto Bifulco | Carolin Lawrence

The advances made by Large Language Models (LLMs) have led to the pursuit of LLM agents that can solve intricate, multi-step reasoning tasks. As with any research pursuit, benchmarking and evaluation are key corner stones to efficient and reliable progress. However, existing benchmarks are often narrow and simply compute overall task success. To face these issues, we propose AgentQuest – a framework where (i) both benchmarks and metrics are modular and easily extensible through well documented and easy-to-use APIs; (ii) we offer two new evaluation metrics that can reliably track LLM agent progress while solving a task. We exemplify the utility of the metrics on two use cases wherein we identify common failure points and refine the agent architecture to obtain a significant performance increase. Together with the research community, we hope to extend AgentQuest further and therefore we make it available under https://github.com/nec-research/agentquest.

pdf bib
ZhuJiu-Knowledge: A Fairer Platform for Evaluating Multiple Knowledge Types in Large Language Models
Pengfan Du | Sirui Liang | Baoli Zhang | Pengfei Cao | Yubo Chen | Kang Liu | Jun Zhao

The swift advancement in large language models (LLMs) has heightened the importance of model evaluations. LLMs have acquired a substantial amount of knowledge, and evaluating the knowledge of these LLMs is crucial. To address this, we introduce the ZhuJiu-Knowledge benchmark which carefully considers the following factors: (1) For knowledge scope, we concentrate on three domains: commonsense knowledge, world knowledge, language knowledge, which comes from ATOMIC, Conceptnet, Wikidata, and Wordnet. (2) For data construction, to prevent data contamination, we utilize knowledge derived from corpora and knowledge graphs to formulate novel questions which are ensured not to appear in the training corpus. A multitude of prompts is purposefully devised to mitigate the impact of prompt design on evaluation and to further analyze the LLMs’ sensitivity to various prompts. (3) For evaluation criteria, we propose a novel voting methodology for assessing generative text, aligning the model’s evaluation with human preferences to reduce biases inherent in individual model assessments. We evaluate 14 current mainstream LLMs and conduct a comprehensive discussion and analysis of their results. The ZhuJiu-Knowledge benchmark and open-participation leaderboard are publicly released at http://zhujiu-knowledge.top and we also provide a demo video at https://youtu.be/QJp4qlEHVH8.

pdf bib
Unitxt: Flexible, Shareable and Reusable Data Preparation and Evaluation for Generative AI
Elron Bandel | Yotam Perlitz | Elad Venezian | Roni Friedman | Ofir Arviv | Matan Orbach | Shachar Don-Yehiya | Dafna Sheinwald | Ariel Gera | Leshem Choshen | Michal Shmueli-Scheuer | Yoav Katz

In the dynamic landscape of generative NLP, traditional text processing pipelines limit research flexibility and reproducibility, as they are tailored to specific dataset, task, and model combinations. The escalating complexity, involving system prompts, model-specific formats, instructions, and more, calls for a shift to a structured, modular, and customizable solution.Addressing this need, we present Unitxt, an innovative library for customizable textual data preparation and evaluation tailored to generative language models. Unitxt natively integrates with common libraries like HuggingFace and LM-eval-harness and deconstructs processing flows into modular components, enabling easy customization and sharing between practitioners. These components encompass model-specific formats, task prompts, and many other comprehensive dataset processing definitions. The Unitxt Catalog centralizes these components, fostering collaboration and exploration in modern textual data workflows. Beyond being a tool, Unitxt is a community-driven platform, empowering users to build, share, and advance their pipelines collaboratively. Join the Unitxt community at https://github.com/IBM/unitxt
up

pdf (full)
bib (full) Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 4: Student Research Workshop)

pdf bib
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 4: Student Research Workshop)
Yang (Trista) Cao | Isabel Papadimitriou | Anaelia Ovalle

pdf bib
Systematic Analysis for Pretrained Language Model Priming for Parameter-Efficient Fine-tuning
Shih-Cheng Huang | Shih-Heng Wang | Min-Han Shih | Saurav Sahay | Hung-yi Lee

Parameter-efficient (PE) methods (like Prompts or Adapters) for adapting pre-trained language models (PLM) to downstream tasks have been popular recently. However, hindrances still prevent these methods from reaching their full potential. For example, two significant challenges are few-shot adaptation and cross-task generalization. To tackle these issues, we propose a general PE priming framework to enhance and explore the few-shot adaptation and generalization ability of PE methods. In this framework, PLMs are primed with PE methods for rapidly adapting to various target tasks. To evaluate the generalization ability of these PE methods, we conduct experiments on a few-shot cross-domain benchmark containing 160 diverse NLP tasks. Our experiment not only reveals the best priming strategy but also verifies that priming facilitates the adaptation to target tasks.

pdf bib
Rephrasing Invokes Better Generations for Large Language Models
Haoran Yang | Hongyuan Lu | Wai Lam

In the realm of emerging multitasking abilities of Large language models (LLMs), methodologies like prompt tuning enable low-cost adaptation to downstream tasks without retraining the model. However, automatic input pre-processing when LLMs are unavailable is currently under-studied. This paper proposes ReLLM (Rephrasing for LLMs), a method that automatically paraphrases input content for better output generations. ReLLM replaces low-frequency lexical items with their high-frequency counterparts. This substitution is particularly beneficial for low-resource language tasks that lack sufficient training data and resources. ReLLM is user-friendly and requires no additional LLM training. Experimental results in cross-lingual summarization, and natural language inference demonstrate the effectiveness of ReLLM.

pdf bib
Exploring Compositional Generalization of Large Language Models
Haoran Yang | Hongyuan Lu | Wai Lam | Deng Cai

In this paper, we study the generalization ability of large language models (LLMs) with respect to compositional instructions, which are instructions that can be decomposed into several sub-instructions. We argue that the ability to generalize from simple instructions to more intricate compositional instructions represents a key aspect of the out-of-distribution generalization for LLMs. Since there are no specialized datasets for studying this phenomenon, we first construct a dataset with the help of ChatGPT, guided by the self-instruct technique. Then, we fine-tune and evaluate LLMs on these datasets. Interestingly, our experimental results indicate that training LLMs on higher-order compositional instructions enhances their performance on lower-order ones, but the reverse does not hold true.

pdf bib
Explainable CED: A Dataset for Explainable Critical Error Detection in Machine Translation
Dahyun Jung | Sugyeong Eo | Chanjun Park | Heuiseok Lim

Critical error detection (CED) in machine translation is a task that aims to detect errors that significantly distort the intended meaning. However, the existing study of CED lacks explainability due to the absence of content addressing the reasons for catastrophic errors. To address this limitation, we propose Explainable CED, a dataset that introduces the attributes of error explanation and correction regarding critical errors. Considering the advantage of reducing time costs and mitigating human annotation bias, we leverage a large language model in the data construction process. To improve the quality of the dataset and mitigate hallucination, we compare responses from the model and introduce an additional data filtering method through feedback scoring. The experiment demonstrates that the dataset appropriately reflects a consistent explanation and revision for errors, validating the reliability of the dataset.

pdf bib
SMARTR: A Framework for Early Detection using Survival Analysis of Longitudinal Texts
Jean-Thomas Baillargeon | Luc Lamontagne

This paper presents an innovative approach to the early detection of expensive insurance claims by leveraging survival analysis concepts within a deep learning framework exploiting textual information from claims notes. Our proposed SMARTR model addresses limitations of state-of-the-art models, such as handling data-label mismatches and non-uniform data frequency, to enhance a posteriori classification and early detection. Our results suggest that incorporating temporal dynamics and empty period representation improves model performance, highlighting the importance of considering time in insurance claim analysis. The approach appears promising for application to other insurance datasets.

pdf bib
Fast Exact Retrieval for Nearest-neighbor Lookup (FERN)
Richard Zhu

Exact nearest neighbor search is a computationally intensive process, and even its simpler sibling — vector retrieval — can be computationally complex. This is exacerbated when retrieving vectors which have high-dimension d relative to the number of vectors, N, in the database. Exact nearest neighbor retrieval has been generally acknowledged to be a O(Nd) problem with no sub-linear solutions. Attention has instead shifted towards Approximate Nearest-Neighbor (ANN) retrieval techniques, many of which have sub-linear or even logarithmic time complexities. However, if our intuition from binary search problems (e.g. d=1 vector retrieval) carries, there ought to be a way to retrieve an organized representation of vectors without brute-forcing our way to a solution. For low dimension (e.g. d=2 or d=3 cases), kd-trees provide a O(dlog N) algorithm for retrieval. Unfortunately the algorithm deteriorates rapidly to a O(dN) solution at high dimensions (e.g. k=128), in practice. We propose a novel algorithm for logarithmic Fast Exact Retrieval for Nearest-neighbor lookup (FERN), inspired by kd-trees. The algorithm achieves O(dlog N) look-up with 100% recall on 10 million d=128 uniformly randomly generated vectors.

pdf bib
Start Simple: Progressive Difficulty Multitask Learning
Yunfei Luo | Yuyang Liu | Rukai Cai | Tauhidur Rahman

The opaque nature of neural networks, often described as black boxes, poses significant challenges in understanding their learning mechanisms, which limit our ability to fully optimize and trust these models.Inspired by how humans learn, this paper proposes a novel neural network training strategy that employs multitask learning with progressive difficulty subtasks, which we believe can potentially shed light on the internal learning mechanisms of neural networks.We implemented this strategy across a range of NLP tasks, data sets, and neural network architectures and observed notable improvements in model performance.This suggests that neural networks may be able to extract common features and internalize shared representations across similar subtasks that differ in their difficulty.Analyzing this strategy could lead us to more interpretable and robust neural networks, enhancing both their performance and our understanding of their nature.

pdf bib
LUCID: LLM-Generated Utterances for Complex and Interesting Dialogues
Joe Stacey | Jianpeng Cheng | John Torr | Tristan Guigue | Joris Driesen | Alexandru Coca | Mark Gaynor | Anders Johannsen

Spurred by recent advances in Large Language Models (LLMs), virtual assistants are poised to take a leap forward in terms of their dialogue capabilities. Yet a major bottleneck to achieving genuinely transformative task-oriented dialogue capabilities remains the scarcity of high quality data. Existing datasets, while impressive in scale, have limited domain coverage and contain few genuinely challenging conversational phenomena; those which are present are typically unlabelled, making it difficult to assess the strengths and weaknesses of models without time-consuming and costly human evaluation. Moreover, creating high quality dialogue data has until now required considerable human input, limiting both the scale of these datasets and the ability to rapidly bootstrap data for a new target domain. We aim to overcome these issues with LUCID, a modularised and highly automated LLM-driven data generation system that produces realistic, diverse and challenging dialogues. We use LUCID to generate a seed dataset of 4,277 conversations across 100 intents to demonstrate its capabilities, with a human review finding consistently high quality labels in the generated data.

pdf bib
Fine-tuning Pre-trained Named Entity Recognition Models For Indian Languages
Sankalp Bahad | Pruthwik Mishra | Parameswari Krishnamurthy | Dipti Sharma

Named Entity Recognition (NER) is a use-ful component in Natural Language Process-ing (NLP) applications. It is used in varioustasks such as Machine Translation, Summa-rization, Information Retrieval, and Question-Answering systems. The research on NER iscentered around English and some other ma-jor languages, whereas limited attention hasbeen given to Indian languages. We analyze thechallenges and propose techniques that can betailored for Multilingual Named Entity Recog-nition for Indian Languages. We present a hu-man annotated named entity corpora of ∼40Ksentences for 4 Indian languages from two ofthe major Indian language families. Addition-ally, we show the transfer learning capabilitiesof pre-trained transformer models from a highresource language to multiple low resource lan-guages through a series of experiments. Wealso present a multilingual model fine-tunedon our dataset, which achieves an F1 score of∼0.80 on our dataset on average. We achievecomparable performance on completely unseenbenchmark datasets for Indian languages whichaffirms the usability of our model.

pdf bib
Knowledge-centered conversational agents with a drive to learn
Selene Baez Santamaria

We create an adaptive conversational agent that assesses the quality of its knowledge and is driven to become more knowledgeable. Unlike agents with predefined tasks, ours can leverage people as diverse sources to meet its knowledge needs. We test the agent in social contexts, where personal and subjective information can be obtained through dialogue. We provide the agent both with generic methods for assessing its knowledge quality (e.g. correctness, completeness, redundancy, interconnectedness, and diversity), as well as with generic capabilities to improve its knowledge by leveraging external sources. We demonstrate that the agent can learn effective policies to acquire the knowledge needed by assessing the efficiency of these capabilities during interaction. Our framework enables on-the-fly learning, offering a dynamic and adaptive approach to shaping conversational interactions.

pdf bib
Exploring Inherent Biases in LLMs within Korean Social Context: A Comparative Analysis of ChatGPT and GPT-4
Seungyoon Lee | Dong Kim | Dahyun Jung | Chanjun Park | Heuiseok Lim

Large Language Models (LLMs) have significantly impacted various fields requiring advanced linguistic understanding, yet concerns regarding their inherent biases and ethical considerations have also increased. Notably, LLMs have been critiqued for perpetuating stereotypes against diverse groups based on race, sexual orientation, and other attributes. However, most research analyzing these biases has predominantly focused on communities where English is the primary language, neglecting to consider the cultural and linguistic nuances of other societies. In this paper, we aim to explore the inherent biases and toxicity of LLMs, specifically within the social context of Korea. We devise a set of prompts that reflect major societal issues in Korea and assign varied personas to both ChatGPT and GPT-4 to assess the toxicity of the generated sentences. Our findings indicate that certain personas or prompt combinations consistently yield harmful content, highlighting the potential risks associated with specific persona-issue alignments within the Korean cultural framework. Furthermore, we discover that GPT-4 can produce more than twice the level of toxic content than ChatGPT under certain conditions.

pdf bib
To Clarify or not to Clarify: A Comparative Analysis of Clarification Classification with Fine-Tuning, Prompt Tuning, and Prompt Engineering
Alina Leippert | Tatiana Anikina | Bernd Kiefer | Josef Genabith

Misunderstandings occur all the time in human conversation but deciding on when to ask for clarification is a challenging task for conversational systems that requires a balance between asking too many unnecessary questions and running the risk of providing incorrect information. This work investigates clarification identification based on the task and data from (Xu et al., 2019), reproducing their Transformer baseline and extending it by comparing pre-trained language model fine-tuning, prompt tuning and manual prompt engineering on the task of clarification identification. Our experiments show strong performance with LM and a prompt tuning approach with BERT and RoBERTa, outperforming standard LM fine-tuning, while manual prompt engineering with GPT-3.5 proved to be less effective, although informative prompt instructions have the potential of steering the model towards generating more accurate explanations for why clarification is needed.

pdf bib
Detecting Response Generation Not Requiring Factual Judgment
Ryohei Kamei | Daiki Shiono | Reina Akama | Jun Suzuki

With the remarkable development of large language models (LLMs), ensuring the factuality of output has become a challenge.However, having all the contents of the response with given knowledge or facts is not necessarily a good thing in dialogues.This study aimed to achieve both attractiveness and factuality in a dialogue response for which a task was set to predict sentences that do not require factual correctness judgment such as agreeing, or personal opinions/feelings.We created a dataset, dialogue dataset annotated with fact-check-needed label (DDFC), for this task via crowdsourcing, and classification tasks were performed on several models using this dataset.The model with the highest classification accuracy could yield about 88% accurate classification results.

pdf bib
Unknown Script: Impact of Script on Cross-Lingual Transfer
Wondimagegnhue Tufa | Ilia Markov | Piek Vossen

Cross-lingual transfer has become an effective way of transferring knowledge between languages. In this paper, we explore an often overlooked aspect in this domain: the influence of the source language of a language model on language transfer performance. We consider a case where the target language and its script are not part of the pre-trained model. We conduct a series of experiments on monolingual and multilingual models that are pre-trained on different tokenization methods to determine factors that affect cross-lingual transfer to a new language with a unique script. Our findings reveal the importance of the tokenizer as a stronger factor than the shared script, language similarity, and model size.

pdf bib
Improving Repository-level Code Search with Text Conversion
Mizuki Kondo | Daisuke Kawahara | Toshiyuki Kurabayashi

The ability to generate code using large language models (LLMs) has been increasing year by year. However, studies on code generation at the repository level are not very active. In repository-level code generation, it is necessary to refer to related code snippets among multiple files. By taking the similarity between code snippets, related files are searched and input into an LLM, and generation is performed. This paper proposes a method to search for related files (code search) by taking similarities not between code snippets but between the texts converted from the code snippets by the LLM. We confirmed that converting to text improves the accuracy of code search.

pdf bib
Improving Multi-lingual Alignment Through Soft Contrastive Learning
Minsu Park | Seyeon Choi | Chanyeol Choi | Jun-Seong Kim | Jy-yong Sohn

Making decent multi-lingual sentence representations is critical to achieve high performances in cross-lingual downstream tasks. In this work, we propose a novel method to align multi-lingual embeddings based on the similarity of sentences measured by a pre-trained mono-lingual embedding model. Given translation sentence pairs, we train a multi-lingual model in a way that the similarity between cross-lingual embeddings follows the similarity of sentences measured at the mono-lingual teacher model. Our method can be considered as contrastive learning with soft labels defined as the similarity between sentences. Our experimental results on five languages show that our contrastive loss with soft labels far outperforms conventional constrastive loss with hard labels in various benchmarks for bitext mining tasks and STS tasks. In addition, our method outperforms existing multi-lingual embeddings including LaBSE, for Tatoeba dataset.

pdf bib
Few-Shot Event Argument Extraction Based on a Meta-Learning Approach
Aboubacar Tuo | Romaric Besançon | Olivier Ferret | Julien Tourille

Few-shot learning techniques for Event Extraction are developed to alleviate the cost of data annotation. However, most studies on few-shot event extraction only focus on event trigger detection and no study has been proposed on argument extraction in a meta-learning context. In this paper, we investigate few-shot event argument extraction using prototypical networks, casting the task as a relation classification problem. Furthermore, we propose to enhance the relation embeddings by injecting syntactic knowledge into the model using graph convolutional networks. Our experimental results show that our proposed approach achieves strong performance on ACE 2005 in several few-shot configurations, and highlight the importance of syntactic knowledge for this task. More generally, our paper provides a unified evaluation framework for meta-learning approaches for argument extraction.

pdf bib
Investigating Web Corpus Filtering Methods for Language Model Development in Japanese
Rintaro Enomoto | Arseny Tolmachev | Takuro Niitsuma | Shuhei Kurita | Daisuke Kawahara

The development of large language models (LLMs) is becoming increasingly significant, and there is a demand for high-quality, large-scale corpora for their pretraining.The quality of a web corpus is especially essential to improve the performance of LLMs because it accounts for a large proportion of the whole corpus. However, filtering methods for Web corpora have yet to be established.In this paper, we present empirical studies to reveal which filtering methods are indeed effective and analyze why they are.We build classifiers and language models in Japanese that can process large amounts of corpora rapidly enough for pretraining LLMs in limited computational resources. By evaluating these filtering methods based on a Web corpus quality evaluation benchmark, we reveal that the most accurate method is the N-gram language model. Indeed, we empirically present that strong filtering methods can rather lead to lesser performance in downstream tasks.We also report that the proportion of some specific topics in the processed documents decreases significantly during the filtering process.

pdf bib
Referring Expressions in Human-Robot Common Ground: A Thesis Proposal
Jaap Kruijt

In this PhD, we investigate the processes through which common ground shapes the pragmatic use of referring expressions in Human-Robot Interaction. A central point in our investigation is the interplay between a growing common ground and changes in the surrounding context, which can create ambiguity, variation and the need for pragmatic interpretations. We outline three objectives that define the scope of our work: 1) obtaining data with common ground interactions, 2) examining reference-making, and 3) evaluating the robot interlocutor. We use datasets as well as a novel interactive experimental framework to investigate the linguistic processes involved in shaping referring expressions. We also design an interactive robot model, which models these linguistic processes and can use pragmatic inference to resolve referring expressions. With this work, we contribute to existing work in HRI, reference resolution and the study of common ground.

pdf bib
Source Code is a Graph, Not a Sequence: A Cross-Lingual Perspective on Code Clone Detection
Mohammed Rahaman | Julia Ive

Code clone detection is challenging, as sourcecode can be written in different languages, do-mains, and styles. In this paper, we arguethat source code is inherently a graph, not asequence, and that graph-based methods aremore suitable for code clone detection thansequence-based methods. We compare the per-formance of two state-of-the-art models: Code-BERT (Feng et al., 2020), a sequence-basedmodel, and CodeGraph (Yu et al., 2023), agraph-based model, on two benchmark data-sets: BCB (Svajlenko et al., 2014) and PoolC(PoolC, no date). We show that CodeGraphoutperforms CodeBERT on both data-sets, es-pecially on cross-lingual code clones. To thebest of our knowledge, this is the first work todemonstrate the cross-lingual code clone detec-tion showing superiority on graph-based meth-ods over sequence-based methods

pdf bib
Distilling Text Style Transfer With Self-Explanation From LLMs
Chiyu Zhang | Honglong Cai | Yuezhang Li | Yuexin Wu | Le Hou | Muhammad Abdul-Mageed

Text Style Transfer (TST) seeks to alter the style of text while retaining its core content. Given the constraints of limited parallel datasets for TST, we propose CoTeX, a framework that leverages large language models (LLMs) alongside chain-of-thought (CoT) prompting to facilitate TST. CoTeX distills the complex rewriting and reasoning capabilities of LLMs into more streamlined models capable of working with both non-parallel and parallel data. Through experimentation across four TST datasets, CoTeX is shown to surpass traditional supervised fine-tuning and knowledge distillation methods, particularly in low-resource settings. We conduct a comprehensive evaluation, comparing CoTeX against current unsupervised, supervised, in-context learning (ICL) techniques, and instruction-tuned LLMs. Furthermore, CoTeX distinguishes itself by offering transparent explanations for its style transfer process.

pdf bib
Reinforcement Learning for Edit-Based Non-Autoregressive Neural Machine Translation
Hao Wang | Tetsuro Morimura | Ukyo Honda | Daisuke Kawahara

Non-autoregressive (NAR) language models are known for their low latency in neural machine translation (NMT). However, a performance gap exists between NAR and autoregressive models due to the large decoding space and difficulty in capturing dependency between target words accurately. Compounding this, preparing appropriate training data for NAR models is a non-trivial task, often exacerbating exposure bias. To address these challenges, we apply reinforcement learning (RL) to Levenshtein Transformer, a representative edit-based NAR model, demonstrating that RL with self-generated data can enhance the performance of edit-based NAR models. We explore two RL approaches: stepwise reward maximization and episodic reward maximization. We discuss the respective pros and cons of these two approaches and empirically verify them. Moreover, we experimentally investigate the impact of temperature setting on performance, confirming the importance of proper temperature setting for NAR models’ training.

pdf bib
Evaluation Dataset for Japanese Medical Text Simplification
Koki Horiguchi | Tomoyuki Kajiwara | Yuki Arase | Takashi Ninomiya

We create a parallel corpus for medical text simplification in Japanese, which simplifies medical terms into expressions that patients can understand without effort.While text simplification in the medial domain is strongly desired by society, it is less explored in Japanese because of the lack of language resources.In this study, we build a parallel corpus for Japanese text simplification evaluation in the medical domain using patients’ weblogs.This corpus consists of 1,425 pairs of complex and simple sentences with or without medical terms.To tackle medical text simplification without a training corpus of the corresponding domain, we repurpose a Japanese text simplification model of other domains.Furthermore, we propose a lexically constrained reranking method that allows to avoid technical terms to be output.Experimental results show that our method contributes to achieving higher simplification performance in the medical domain.

pdf bib
Multi-Source Text Classification for Multilingual Sentence Encoder with Machine Translation
Reon Kajikawa | Keiichiro Yamada | Tomoyuki Kajiwara | Takashi Ninomiya

To reduce the cost of training models for each language for developers of natural language processing applications, pre-trained multilingual sentence encoders are promising.However, since training corpora for such multilingual sentence encoders contain only a small amount of text in languages other than English, they suffer from performance degradation for non-English languages.To improve the performance of pre-trained multilingual sentence encoders for non-English languages, we propose a method of machine translating a source sentence into English and then inputting it together with the source sentence in a multi-source manner.Experimental results on sentiment analysis and topic classification tasks in Japanese revealed the effectiveness of the proposed method.

pdf bib
A Reproducibility Study on Quantifying Language Similarity: The Impact of Missing Values in the URIEL Knowledge Base
Hasti Toossi | Guo Huai | Jinyu Liu | Eric Khiu | A. Doğruöz | En-Shiun Lee

In the pursuit of supporting more languages around the world, tools that characterize properties of languages play a key role in expanding the existing multilingual NLP research. In this study, we focus on a widely used typological knowledge base, URIEL, which aggregates linguistic information into numeric vectors. Specifically, we delve into the soundness and reproducibility of the approach taken by URIEL in quantifying language similarity. Our analysis reveals URIEL’s ambiguity in calculating language distances and in handling missing values. Moreover, we find that URIEL does not provide any information about typological features for 31% of the languages it represents, undermining the reliabilility of the database, particularly on low-resource languages. Our literature review suggests URIEL and lang2vec are used in papers on diverse NLP tasks, which motivates us to rigorously verify the database as the effectiveness of these works depends on the reliability of the information the tool provides.

pdf bib
Coding Open-Ended Responses using Pseudo Response Generation by Large Language Models
Yuki Zenimoto | Ryo Hasegawa | Takehito Utsuro | Masaharu Yoshioka | Noriko Kando

Survey research using open-ended responses is an important method thatcontributes to the discovery of unknown issues and new needs. However,survey research generally requires time and cost-consuming manual dataprocessing, indicating that it is difficult to analyze large dataset.To address this issue, we propose an LLM-based method to automate partsof the grounded theory approach (GTA), a representative approach of thequalitative data analysis. We generated and annotated pseudo open-endedresponses, and used them as the training data for the coding proceduresof GTA. Through evaluations, we showed that the models trained withpseudo open-ended responses are quite effective compared with thosetrained with manually annotated open-ended responses. We alsodemonstrate that the LLM-based approach is highly efficient andcost-saving compared to human-based approach.

pdf bib
Cross-Task Generalization Abilities of Large Language Models
Qinyuan Ye

Humans can learn a new language task efficiently with only few examples, by leveraging their knowledge and experience obtained when learning prior tasks. Enabling similar cross-task generalization abilities in NLP systems is fundamental for approaching the goal of general intelligence and expanding the reach of language technology in the future.In this thesis proposal, I will present my work on (1) benchmarking cross-task generalization abilities with diverse NLP tasks; (2) developing model architectures for improving cross-task generalization abilities; (3) analyzing and predicting the generalization landscape of current state-of-the-art large language models. Additionally, I will outline future research directions, along with preliminary thoughts on addressing them.

pdf bib
Commentary Generation from Data Records of Multiplayer Strategy Esports Game
Zihan Wang | Naoki Yoshinaga

Esports, a sports competition on video games, has become one of the most important sporting events. Although esports play logs have been accumulated, only a small portion of them accompany text commentaries for the audience to retrieve and understand the plays. In this study, we therefore introduce the task of generating game commentaries from esports’ data records. We first build large-scale esports data-to-text datasets that pair structured data and commentaries from a popular esports game, League of Legends. We then evaluate Transformer-based models to generate game commentaries from structured data records, while examining the impact of the pre-trained language models. Evaluation results on our dataset revealed the challenges of this novel task. We will release our dataset to boost potential research in the data-to-text generation community.

pdf bib
Facilitating Opinion Diversity through Hybrid NLP Approaches
Michiel Van Der Meer

Modern democracies face a critical issue of declining citizen participation in decision-making. Online discussion forums are an important avenue for enhancing citizen participation. This thesis proposal 1) identifies the challenges involved in facilitating large-scale online discussions with Natural Language Processing (NLP), 2) suggests solutions to these challenges by incorporating hybrid human-AI technologies, and 3) investigates what these technologies can reveal about individual perspectives in online discussions. We propose a three-layered hierarchy for representing perspectives that can be obtained by a mixture of human intelligence and large language models. We illustrate how these representations can draw insights into the diversity of perspectives and allow us to investigate interactions in online discussions.

pdf bib
HybridBERT - Making BERT Pretraining More Efficient Through Hybrid Mixture of Attention Mechanisms
Gokul Srinivasagan | Simon Ostermann

Pretrained transformer-based language models have produced state-of-the-art performance in most natural language understanding tasks. These models undergo two stages of training: pretraining on a huge corpus of data and fine-tuning on a specific downstream task. The pretraining phase is extremely compute-intensive and requires several high-performance computing devices like GPUs and several days or even months of training, but it is crucial for the model to capture global knowledge and also has a significant impact on the fine-tuning task. This is a major roadblock for researchers without access to sophisticated computing resources. To overcome this challenge, we propose two novel hybrid architectures called HybridBERT (HBERT), which combine self-attention and additive attention mechanisms together with sub-layer normalization. We introduce a computing budget to the pretraining phase, limiting the training time and usage to a single GPU. We show that HBERT attains twice the pretraining accuracy of a vanilla-BERT baseline. We also evaluate our proposed models on two downstream tasks, where we outperform BERT-base while accelerating inference. Moreover, we study the effect of weight initialization with a limited pretraining budget. The code and models are publicly available at: www.github.com/gokulsg/HBERT/.
up

pdf (full)
bib (full) Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 5: Tutorial Abstracts)

pdf bib
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 5: Tutorial Abstracts)
Rui Zhang | Nathan Schneider | Snigdha Chaturvedi

pdf bib
Catch Me If You GPT: Tutorial on Deepfake Texts
Adaku Uchendu | Saranya Venkatraman | Thai Le | Dongwon Lee

In recent years, Natural Language Generation (NLG) techniques have greatly advanced, especially in the realm of Large Language Models (LLMs). With respect to the quality of generated texts, it is no longer trivial to tell the difference between human-written and LLMgenerated texts (i.e., deepfake texts). While this is a celebratory feat for NLG, it poses new security risks (e.g., the generation of misinformation). To combat this novel challenge, researchers have developed diverse techniques to detect deepfake texts. While this niche field of deepfake text detection is growing, the field of NLG is growing at a much faster rate, thus making it difficult to understand the complex interplay between state-of-the-art NLG methods and the detectability of their generated texts. To understand such inter-play, two new computational problems emerge: (1) Deepfake Text Attribution (DTA) and (2) Deepfake Text Obfuscation (DTO) problems, where the DTA problem is concerned with attributing the authorship of a given text to one of k NLG methods, while the DTO problem is to evade the authorship of a given text by modifying parts of the text. In this cutting-edge tutorial, therefore, we call attention to the serious security risk both emerging problems pose and give a comprehensive review of recent literature on the detection and obfuscation of deepfake text authorships. Our tutorial will be 3 hours long with a mix of lecture and hands-on examples for interactive audience participation. You can find our tutorial materials here: https://tinyurl.com/naacl24-tutorial.

pdf bib
Combating Security and Privacy Issues in the Era of Large Language Models
Muhao Chen | Chaowei Xiao | Huan Sun | Lei Li | Leon Derczynski | Anima Anandkumar | Fei Wang

This tutorial seeks to provide a systematic summary of risks and vulnerabilities in security, privacy and copyright aspects of large language models (LLMs), and most recent solutions to address those issues. We will discuss a broad thread of studies that try to answer the following questions: (i) How do we unravel the adversarial threats that attackers may leverage in the training time of LLMs, especially those that may exist in recent paradigms of instruction tuning and RLHF processes? (ii) How do we guard the LLMs against malicious attacks in inference time, such as attacks based on backdoors and jailbreaking? (iii) How do we ensure privacy protection of user information and LLM decisions for Language Model as-a-Service (LMaaS)? (iv) How do we protect the copyright of an LLM? (v) How do we detect and prevent cases where personal or confidential information is leaked during LLM training? (vi) How should we make policies to control against improper usage of LLM-generated content? In addition, will conclude the discussions by outlining emergent challenges in security, privacy and reliability of LLMs that deserve timely investigation by the community

pdf bib
Explanation in the Era of Large Language Models
Zining Zhu | Hanjie Chen | Xi Ye | Qing Lyu | Chenhao Tan | Ana Marasovic | Sarah Wiegreffe

Explanation has long been a part of communications, where humans use language to elucidate each other and transmit information about the mechanisms of events. There have been numerous works that study the structures of the explanations and their utility to humans. At the same time, explanation relates to a collection of research directions in natural language processing (and more broadly, computer vision and machine learning) where researchers develop computational approaches to explain the (usually deep neural network) models. Explanation has received rising attention. In recent months, the advance of large language models (LLMs) provides unprecedented opportunities to leverage their reasoning abilities, both as tools to produce explanations and as the subjects of explanation analysis. On the other hand, the sheer sizes and the opaque nature of LLMs introduce challenges to the explanation methods. In this tutorial, we intend to review these opportunities and challenges of explanations in the era of LLMs, connect lines of research previously studied by different research groups, and hopefully spark thoughts of new research directions

pdf bib
From Text to Context: Contextualizing Language with Humans, Groups, and Communities for Socially Aware NLP
Adithya V Ganesan | Siddharth Mangalik | Vasudha Varadarajan | Nikita Soni | Swanie Juhng | João Sedoc | H. Andrew Schwartz | Salvatore Giorgi | Ryan L Boyd

Aimed at the NLP researchers or practitioners who would like to integrate human - individual, group, or societal level factors into their analyses, this tutorial will cover recent techniques and libraries for doing so at each level of analysis. Starting with human-centered techniques that provide benefit to traditional document- or word-level NLP tasks (Garten et al., 2019; Lynn et al., 2017), we undertake a thorough exploration of critical human-level aspects as they pertain to NLP, gradually moving up to higher levels of analysis: individual persons, individual with agent (chat/dialogue), groups of people, and finally communities or societies.

pdf bib
Human-AI Interaction in the Age of LLMs
Diyi Yang | Sherry Tongshuang Wu | Marti A. Hearst

Recently, the development of Large Language Models (LLMs) has revolutionized the capabilities of AI systems. These models possess the ability to comprehend and generate human-like text, enabling them to engage in sophisticated conversations, generate content, and even perform tasks that once seemed beyond the reach of machines. As a result, the way we interact with technology and each other — an established field called “Human-AI Interaction” and have been studied for over a decade — is undergoing a profound transformation. This tutorial will provide an overview of the interaction between humans and LLMs, exploring the challenges, opportunities, and ethical considerations that arise in this dynamic landscape. It will start with a review of the types of AI models we interact with, and a walkthrough of the core concepts in Human-AI Interaction. We will then emphasize the emerging topics shared between HCI and NLP communities in light of LLMs.

pdf bib
Spatial and Temporal Language Understanding: Representation, Reasoning, and Grounding
Parisa Kordjamshidi | Qiang Ning | James Pustejovsky | Marie-Francine Moens

This tutorial provides an overview of the cutting edge research on spatial and temporal language understanding. We also cover some essential background material from various subdisciplines to this topic, which we believe will enrich the CL community’s appreciation of the complexity of spatiotemporal reasoning.
up

pdf (full)
bib (full) Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 6: Industry Track)

pdf bib
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 6: Industry Track)
Yi Yang | Aida Davani | Avi Sil | Anoop Kumar

pdf bib
HPipe: Large Language Model Pipeline Parallelism for Long Context on Heterogeneous Cost-effective Devices
Ruilong Ma | Xiang Yang | Jingyu Wang | Qi Qi | Haifeng Sun | Jing Wang | Zirui Zhuang | Jianxin Liao

Micro-enterprises and individual developers emerge analysis demands for long sequence with powerful Large Language Models (LLMs). They try to deploy the LLMs at local, but only possess various commodity devices and the unreliable interconnection between devices. Existing parallel techniques do not lead to the same effectiveness in limited environment. The heterogeneity of devices, coupled with their limited capacity and expensive communication, brings challenges to private deployment for maximized utilization of available devices while masking latency. Hence, we introduce HPipe, a pipeline inference framework that successfully mitigates LLMs from high-performance clusters to heterogeneous commodity devices. By ensuring a balanced distribution of workloads, HPipe facilitates the parallel execution of LLMs through pipelining the sequences on the token dimension. The evaluation conducted on LLaMA-7B and GPT3-2B demonstrates that HPipe holds the potential for context analysis on LLM with heterogeneity devices, achieving an impressive speedup in latency and throughput up to 2.28 times.

pdf bib
Lossless Acceleration of Large Language Model via Adaptive N-gram Parallel Decoding
Jie Ou | Yueming Chen | Prof. Tian

While Large Language Models (LLMs) have shown remarkable abilities, they are hindered by significant resource consumption and considerable latency due to autoregressive processing. In this study, we introduce Adaptive N-gram Parallel Decoding (ANPD), an innovative and lossless approach that accelerates inference by allowing the simultaneous generation of multiple tokens. ANPD incorporates a two-stage approach: it begins with a rapid drafting phase that employs an N-gram module, which adapts based on the current interactive context, followed by a verification phase, during which the original LLM assesses and confirms the proposed tokens. Consequently, ANPD preserves the integrity of the LLM’s original output while enhancing processing speed. We further leverage a multi-level architecture for the N-gram module to enhance the precision of the initial draft, consequently reducing inference latency. ANPD eliminates the need for retraining or extra GPU memory, making it an efficient and plug-and-play enhancement. In our experiments, models such as LLaMA and its fine-tuned variants have shown speed improvements up to 3.67x, validating the effectiveness of our proposed ANPD.

pdf bib
SOLAR 10.7B: Scaling Large Language Models with Simple yet Effective Depth Up-Scaling
Sanghoon Kim | Dahyun Kim | Chanjun Park | Wonsung Lee | Wonho Song | Yunsu Kim | Hyeonwoo Kim | Yungi Kim | Hyeonju Lee | Jihoo Kim | Changbae Ahn | Seonghoon Yang | Sukyung Lee | Hyunbyung Park | Gyoungjin Gim | Mikyoung Cha | Hwalsuk Lee | Sunghun Kim

We introduce SOLAR 10.7B, a large language model (LLM) with 10.7 billion parameters, demonstrating superior performance in various natural language processing (NLP) tasks. Inspired by recent efforts to efficiently up-scale LLMs, we present a method for scaling LLMs called depth up-scaling (DUS), which encompasses depthwise scaling and continued pretraining. In contrast to other LLM up-scaling methods that use mixture-of-experts, DUS does not require complex changes to train and inference efficiently. We show experimentally that DUS is simple yet effective in scaling up high-performance LLMs from small ones. Building on the DUS model, we additionally present SOLAR 10.7B-Instruct, a variant fine-tuned for instruction-following capabilities, surpassing Mixtral-8x7B-Instruct. SOLAR 10.7B is publicly available under the Apache 2.0 license, promoting broad access and application in the LLM field.

pdf bib
UINav: A Practical Approach to Train On-Device Automation Agents
Wei Li | Fu-Lin Hsu | William Bishop | Folawiyo Campbell-Ajala | Max Lin | Oriana Riva

Automation systems that can autonomously drive application user interfaces to complete user tasks are of great benefit, especially when users are situationally or permanently impaired. Prior automation systems do not produce generalizable models while AI-based automation agents work reliably only in simple, hand-crafted applications or incur high computation costs. We propose UINav, a demonstration-based approach to train automation agents that fit mobile devices, yet achieving high success rates with modest numbers of demonstrations. To reduce the demonstration overhead, UINav uses a referee model that provides users with immediate feedback on tasks where the agent fails, and automatically augments human demonstrations to increase diversity in training data. Our evaluation shows that with only 10 demonstrations can achieve 70% accuracy, and that with enough demonstrations it can surpass 90% accuracy.

pdf bib
Efficiently Distilling LLMs for Edge Applications
Achintya Kundu | Yu Chin Fabian Lim | Aaron Chew | Laura Wynter | Penny Chong | Rhui Lee

Supernet training of LLMs is of great interest in industrial applications as it confers the ability to produce a palette of smaller models at constant cost, regardless of the number of models (of different size / latency) produced. We propose a new method called Multistage Low-rank Fine-tuning of Super-transformers (MLFS) for parameter-efficient supernet training. We show that it is possible to obtain high-quality encoder models that are suitable for commercial edge applications, and that while decoder-only models are resistant to a comparable degree of compression, decoders can be effectively sliced for a significant reduction in training time.

pdf bib
Modeling and Detecting Company Risks from News
Jiaxin Pei | Soumya Vadlamannati | Liang-Kang Huang | Daniel Preotiuc-Pietro | Xinyu Hua

Identifying risks associated with a company is important to investors and the wellbeing of the overall financial markets. In this study, we build a computational framework to automatically extract company risk factors from news articles. Our newly proposed schema comprises seven distinct aspects, such as supply chain, regulations, and competition. We annotate 666 news articles and benchmark various machine learning models. While large language mod- els have achieved remarkable progress in various types of NLP tasks, our experiment shows that zero-shot and few-shot prompting state-of- the-art LLMs (e.g., Llama-2) can only achieve moderate to low performances in identifying risk factors. In contrast, fine-tuning pre-trained language models yields better results on most risk factors. Using this model, we analyze over 277K Bloomberg News articles and demonstrate that identifying risk factors from news could provide extensive insights into the operations of companies and industries.

pdf bib
Multiple-Question Multiple-Answer Text-VQA
Peng Tang | Srikar Appalaraju | R. Manmatha | Yusheng Xie | Vijay Mahadevan

We present Multiple-Question Multiple-Answer (MQMA), a novel approach to do text-VQA in encoder-decoder transformer models. To the best of our knowledge, almost all previous approaches for text-VQA process a single question and its associated content to predict a single answer. However, in industry applications, users may come up with multiple questions about a single image. In order to answer multiple questions from the same image, each question and content are fed into the model multiple times. In contrast, our proposed MQMA approach takes multiple questions and content as input at the encoder and predicts multiple answers at the decoder in an auto-regressive manner at the same time. We make several novel architectural modifications to standard encoder-decoder transformers to support MQMA. We also propose a novel MQMA denoising pre-training task which is designed to teach the model to align and delineate multiple questions and content with associated answers. MQMA pre-trained model achieves state-of-the-art results on multiple text-VQA datasets, each with strong baselines. Specifically, on OCR-VQA (+2.5%), TextVQA (+1.4%), ST-VQA (+0.6%), DocVQA (+1.1%) absolute improvements over the previous state-of-the-art approaches.

pdf bib
An NLP-Focused Pilot Training Agent for Safe and Efficient Aviation Communication
Xiaochen Liu | Bowei Zou | AiTi Aw

Aviation communication significantly influences the success of flight operations, ensuring safety of lives and efficient air transportation. In day-to-day flight operations, air traffic controllers (ATCos) would timely communicate instructions to pilots using specific phraseology for aircraft manipulation . However, pilots, originating from diverse backgrounds and understanding of English language, have struggled with conforming to strict phraseology for readback and communication in the live operation, this problem had not been effectively addressed over the past decades. Traditionally, aviation communication training involved expensive setups and resources, often relying on human-in-the-loop (HIL) air traffic simulations that demand allocating a specific environment, domain experts for participation, and substantial amount of annotated data for simulation. Therefore, we would like to propose an NLP-oriented training agent and address these challenges. Our approach involves leveraging only natural language capabilities and fine-tuning on communication data to generate instructions based on input scenarios (keywords). Given the absence of prior references for this business problem, we investigated the feasibility of our proposed solution by 1) generating all instructions at once and 2) generating one instruction while incorporating conversational history in each input. Our findings affirm the feasibility of this approach, highlighting the effectiveness of fine-tuning pre-trained models and large language models in advancing aviation communication training.

pdf bib
Visual Grounding for User Interfaces
Yijun Qian | Yujie Lu | Alexander Hauptmann | Oriana Riva

Enabling autonomous language agents to drive application user interfaces (UIs) as humans do can significantly expand the capability of today’s API-based agents. Essential to this vision is the ability of agents to ground natural language commands to on-screen UI elements. Prior UI grounding approaches work by relaying on developer-provided UI metadata (UI trees, such as web DOM, and accessibility labels) to detect on-screen elements. However, such metadata is often unavailable or incomplete. Object detection techniques applied to UI screens remove this dependency, by inferring location and types of UI elements directly from the UI’s visual appearance. The extracted semantics, however, are too limited to directly enable grounding. We overcome the limitations of both approaches by introducing the task of visual UI grounding, which unifies detection and grounding. A model takes as input a UI screenshot and a free-form language expression, and must identify the referenced UI element. We propose a solution to this problem, LVG, which learns UI element detection and grounding using a new technique called layout-guided contrastive learning, where the semantics of individual UI objects are learned also from their visual organization. Due to the scarcity of UI datasets, LVG integrates synthetic data in its training using multi-context learning. LVG outperforms baselines pre-trained on much larger datasets by over 4.9 points in top-1 accuracy, thus demonstrating its effectiveness.

pdf bib
Prompt Tuned Embedding Classification for Industry Sector Allocation
Valentin Buchner | Lele Cao | Jan-Christoph Kalo | Vilhelm Von Ehrenheim

We introduce Prompt Tuned Embedding Classification (PTEC) for classifying companies within an investment firm’s proprietary industry taxonomy, supporting their thematic investment strategy. PTEC assigns companies to the sectors they primarily operate in, conceptualizing this process as a multi-label text classification task. Prompt Tuning, usually deployed as a text-to-text (T2T) classification approach, ensures low computational cost while maintaining high task performance. However, T2T classification has limitations on multi-label tasks due to the generation of non-existing labels, permutation invariance of the label sequence, and a lack of confidence scores. PTEC addresses these limitations by utilizing a classification head in place of the Large Language Models (LLMs) language head. PTEC surpasses both baselines and human performance while lowering computational demands. This indicates the continuing need to adapt state-of-the-art methods to domain-specific tasks, even in the era of LLMs with strong generalization abilities.

pdf bib
REXEL: An End-to-end Model for Document-Level Relation Extraction and Entity Linking
Nacime Bouziani | Shubhi Tyagi | Joseph Fisher | Jens Lehmann | Andrea Pierleoni

Extracting structured information from unstructured text is critical for many downstream NLP applications and is traditionally achieved by closed information extraction (cIE). However, existing approaches for cIE suffer from two limitations: (i) they are often pipelines which makes them prone to error propagation, and/or (ii) they are restricted to sentence level which prevents them from capturing long-range dependencies and results in expensive inference time. We address these limitations by proposing REXEL, a highly efficient and accurate model for the joint task of document level cIE (DocIE). REXEL performs mention detection, entity typing, entity disambiguation, coreference resolution and document-level relation classification in a single forward pass to yield facts fully linked to a reference knowledge graph. It is on average 11 times faster than competitive existing approaches in a similar setting and performs competitively both when optimised for any of the individual sub-task and a variety of combinations of different joint tasks, surpassing the baselines by an average of more than 6 F1 points. The combination of speed and accuracy makes REXEL an accurate cost-efficient system for extracting structured information at web-scale. We also release an extension of the DocRED dataset to enable benchmarking of future work on DocIE, which will be available at https://github.com/amazon-science/e2e-docie.

pdf bib
Conformer-Based Speech Recognition On Extreme Edge-Computing Devices
Mingbin Xu | Alex Jin | Sicheng Wang | Mu Su | Tim Ng | Henry Mason | Shiyi Han | Zhihong Lei | Yaqiao Deng | Zhen Huang | Mahesh Krishnamoorthy

With increasingly more powerful compute capabilities and resources in today’s devices, traditionally compute-intensive automatic speech recognition (ASR) has been moving from the cloud to devices to better protect user privacy. However, it is still challenging to implement on-device ASR on resource-constrained devices, such as smartphones, smart wearables, and other small home automation devices. In this paper, we propose a series of model architecture adaptions, neural network graph transformations, and numerical optimizations to fit an advanced Conformer based end-to-end streaming ASR system on resource-constrained devices without accuracy degradation. We achieve over 5.26 times faster than realtime (0.19 RTF) speech recognition on small wearables while minimizing energy consumption and achieving state-of-the-art accuracy. The proposed methods are widely applicable to other transformer-based server-free AI applications. In addition, we provide a complete theory on optimal pre-normalizers that numerically stabilize layer normalization in any Lp-norm using any floating point precision.

pdf bib
Generating Signed Language Instructions in Large-Scale Dialogue Systems
Mert Inan | Katherine Atwell | Anthony Sicilia | Lorna Quandt | Malihe Alikhani

We introduce a goal-oriented conversational AI system enhanced with American Sign Language (ASL) instructions, presenting the first implementation of such a system on a worldwide multimodal conversational AI platform. Accessible through a touch-based interface, our system receives input from users and seamlessly generates ASL instructions by leveraging retrieval methods and cognitively based gloss translations. Central to our design is a sign translation module powered by Large Language Models, alongside a token-based video retrieval system for delivering instructional content from recipes and wikiHow guides. Our development process is deeply rooted in a commitment to community engagement, incorporating insights from the Deaf and Hard-of-Hearing community, as well as experts in cognitive and ASL learning sciences. The effectiveness of our signing instructions is validated by user feedback, achieving ratings on par with those of the system in its non-signing variant. Additionally, our system demonstrates exceptional performance in retrieval accuracy and text-generation quality, measured by metrics such as BERTScore. We have made our codebase and datasets publicly accessible at https://github.com/Merterm/signed-dialogue, and a demo of our signed instruction video retrieval system is available at https://huggingface.co/spaces/merterm/signed-instructions.

pdf bib
Leveraging Natural Language Processing and Large Language Models for Assisting Due Diligence in the Legal Domain
Myeongjun Jang | Gábor Stikkel

Due diligence is a crucial legal process that mitigates potential risks of mergers and acquisitions (M&A). However, despite its prominent importance, there has been a lack of research regarding leveraging NLP techniques for due diligence. In this study, our aim is to explore the most efficient deep-learning model architecture for due diligence in terms of performance and latency, and evaluate the potential of large language models (LLMs) as an efficient due diligence assistant. To our knowledge, this is the first study that employs pre-trained language models (PLMs) and LLMs for the due diligence problem. Our experimental results suggest that methodologies that have demonstrated promising performance in the general domain encounter challenges when applied in due diligence due to the inherent lengthy nature of legal documents. We also ascertain that LLMs can be a useful tool for helping lawyers who perform due diligence.

pdf bib
AnnoLLM: Making Large Language Models to Be Better Crowdsourced Annotators
Xingwei He | Zhenghao Lin | Yeyun Gong | A-Long Jin | Hang Zhang | Chen Lin | Jian Jiao | Siu Ming Yiu | Nan Duan | Weizhu Chen

Many natural language processing (NLP) tasks rely on labeled data to train machine learning models with high performance. However, data annotation is time-consuming and expensive, especially when the task involves a large amount of data or requires specialized domains. Recently, GPT-3.5 series models have demonstrated remarkable few-shot and zero-shot ability across various NLP tasks. In this paper, we first claim that large language models (LLMs), such as GPT-3.5, can serve as an excellent crowdsourced annotator when provided with sufficient guidance and demonstrated examples. Accordingly, we propose AnnoLLM, an annotation system powered by LLMs, which adopts a two-step approach, explain-then-annotate. Concretely, we first prompt LLMs to provide explanations for why the specific ground truth answer/label was assigned for a given example. Then, we construct the few-shot chain-of-thought prompt with the self-generated explanation and employ it to annotate the unlabeled data with LLMs. Our experiment results on three tasks, including user input and keyword relevance assessment, BoolQ, and WiC, demonstrate that AnnoLLM surpasses or performs on par with crowdsourced annotators. Furthermore, we build the first conversation-based information retrieval dataset employing AnnoLLM. This dataset is designed to facilitate the development of retrieval models capable of retrieving pertinent documents for conversational text. Human evaluation has validated the dataset’s high quality.

pdf bib
An Automatic Prompt Generation System for Tabular Data Tasks
Ashlesha Akella | Abhijit Manatkar | Brijkumar Chavda | Hima Patel

Efficient processing of tabular data is important in various industries, especially when working with datasets containing a large number of columns. Large language models (LLMs) have demonstrated their ability on several tasks through carefully crafted prompts. However, creating effective prompts for tabular datasets is challenging due to the structured nature of the data and the need to manage numerous columns. This paper presents an innovative auto-prompt generation system suitable for multiple LLMs, with minimal training. It proposes two novel methods; 1) A Reinforcement Learning-based algorithm for identifying and sequencing task-relevant columns 2) cell-level similarity-based approach for enhancing few-shot example selection. Our approach has been extensively tested across 66 datasets, demonstrating improved performance in three downstream tasks: data imputation, error detection, and entity matching using two distinct LLMs; Google/flant-t5xxl and Mixtral 8x7B.

pdf bib
Fighting crime with Transformers: Empirical analysis of address parsing methods in payment data
Haitham Hammami | Louis Baligand | Bojan Petrovski

In the financial industry, identifying the location of parties involved in payments is a major challenge in the context of Anti-Money Laundering transaction monitoring. For this purpose address parsing entails extracting fields such as street, postal code, or country from free text message attributes. While payment processing platforms are updating their standards with more structured formats such as SWIFT with ISO 20022, address parsing remains essential for a considerable volume of messages. With the emergence of Transformers and Generative Large Language Models (LLM), we explore the performance of state-of-the-art solutions given the constraint of processing a vast amount of daily data. This paper also aims to show the need for training robust models capable of dealing with real-world noisy transactional data. Our results suggest that a well fine-tuned Transformer model using early-stopping significantly outperforms other approaches. Nevertheless, generative LLMs demonstrate strong zero_shot performance and warrant further investigations.

pdf bib
Language Models are Alignable Decision-Makers: Dataset and Application to the Medical Triage Domain
Brian Hu | Bill Ray | Alice Leung | Amy Summerville | David Joy | Christopher Funk | Arslan Basharat

In difficult decision-making scenarios, it is common to have conflicting opinions among expert human decision-makers as there may not be a single right answer. Such decisions may be guided by different attributes that can be used to characterize an individual’s decision. We introduce a novel dataset for medical triage decision-making, labeled with a set of decision-maker attributes (DMAs). This dataset consists of 62 scenarios, covering six different DMAs, including ethical principles such as fairness and moral desert. We present a novel software framework for human-aligned decision-making by utilizing these DMAs, paving the way for trustworthy AI with better guardrails. Specifically, we demonstrate how large language models (LLMs) can serve as ethical decision-makers, and how their decisions can be aligned to different DMAs using zero-shot prompting. Our experiments focus on different open-source models with varying sizes and training techniques, such as Falcon, Mistral, and Llama 2. Finally, we also introduce a new form of weighted self-consistency that improves the overall quantified performance. Our results provide new research directions in the use of LLMs as alignable decision-makers. The dataset and open-source software are publicly available at: https://github.com/ITM-Kitware/llm-alignable-dm.

pdf bib
Reducing hallucination in structured outputs via Retrieval-Augmented Generation
Orlando Ayala | Patrice Bechard

A current limitation of Generative AI (GenAI) is its propensity to hallucinate. While Large Language Models (LLM) have taken the world by storm, without eliminating or at least reducing hallucination, real-world GenAI systems will likely continue to face challenges in user adoption. In the process of deploying an enterprise application that produces workflows from natural language requirements, we devised a system leveraging Retrieval-Augmented Generation (RAG) to improve the quality of the structured output that represents such workflows. Thanks to our implementation of RAG, our proposed system significantly reduces hallucination and allows the generalization of our LLM to out-of-domain settings. In addition, we show that using a small, well-trained retriever can reduce the size of the accompanying LLM at no loss in performance, thereby making deployments of LLM-based systems less resource-intensive.

pdf bib
Towards Translating Objective Product Attributes Into Customer Language
Ram Yazdi | Oren Kalinsky | Alexander Libov | Dafna Shahaf

When customers search online for a product they are not familiar with, their needs are often expressed through subjective product attributes, such as ”picture quality” for a TV or ”easy to clean” for a sofa. In contrast, the product catalog in online stores includes objective attributes such as ”screen resolution” or ”material”. In this work, we aim to find a link between the objective product catalog and the subjective needs of the customers, to help customers better understand the product space using their own words. We apply correlation-based methods to the store’s product catalog and product reviews in order to find the best potential links between objective and subjective attributes; next, Large Language Models (LLMs) reduce spurious correlations by incorporating common sense and world knowledge (e.g., picture quality is indeed affected by screen resolution, and 8k is the best one). We curate a dataset for this task and show that our combined approach outperforms correlation-only and causation-only approaches.

pdf bib
Automating the Generation of a Functional Semantic Types Ontology with Foundational Models
Sachin Konan | Larry Rudolph | Scott Affens

The rise of data science, the inherent dirtiness of data, and the proliferation of vast data providers have increased the value proposition of Semantic Types. Semantic Types are a way of encoding contextual information onto a data schema that informs the user about the definitional meaning of data, its broader context, and relationships to other types. We increasingly see a world where providing structure to this information, attached directly to data, will enable both people and systems to better understand the content of a dataset and the ability to efficiently automate data tasks such as validation, mapping/joins, and eventually machine learning. While ontological systems exist, they have not had widespread adoption due to challenges in mapping to operational datasets and lack of specificity of entity-types. Additionally, the validation checks associated with data are stored in code bases separate from the datasets that are distributed. In this paper, we address both challenges holistically by proposing a system that efficiently maps and encodes functional meaning on Semantic Types.

pdf bib
Leveraging Customer Feedback for Multi-modal Insight Extraction
Sandeep Mukku | Abinesh Kanagarajan | Pushpendu Ghosh | Chetan Aggarwal

Businesses can benefit from customer feedback in different modalities, such as text and images, to enhance their products and services. However, it is difficult to extract actionable and relevant pairs of text segments and images from customer feedback in a single pass. In this paper, we propose a novel multi-modal method that fuses image and text information in a latent space and decodes it to extract the relevant feedback segments using an image-text grounded text decoder. We also introduce a weakly-supervised data generation technique that produces training data for this task. We evaluate our model on unseen data and demonstrate that it can effectively mine actionable insights from multi-modal customer feedback, outperforming the existing baselines by 14 points in F1 score.

pdf bib
Optimizing LLM Based Retrieval Augmented Generation Pipelines in the Financial Domain
Yiyun Zhao | Prateek Singh | Hanoz Bhathena | Bernardo Ramos | Aviral Joshi | Swaroop Gadiyaram | Saket Sharma

Retrieval Augmented Generation (RAG) is a prominent approach in real-word applications for grounding large language model (LLM) generations in up to date and domain-specific knowledge. However, there is a lack of systematic investigations of the impact of each component (retrieval quality, prompts, generation models) on the generation quality of a RAG pipeline in real world scenarios. In this study, we benchmark 6 LLMs in 15 retrieval scenarios exploring 9 prompts over 2 real world financial domain datasets. We thoroughly discuss the impact of each component in RAG pipeline on answer generation quality and formulate specific recommendations for the design of RAG systems.

pdf bib
Scaling Up Authorship Attribution
Jacob Striebel | Abishek Edikala | Ethan Irby | Alex Rosenfeld | J. Gage | Daniel Dakota | Sandra Kübler

We describe our system for authorship attribution in the IARPA HIATUS program. We describe the model and compute infrastructure developed to satisfy the set of technical constraints imposed by IARPA, including runtime limits as well as other constraints related to the ultimate use case. One use-case constraint concerns the explainability of the features used in the system. For this reason, we integrate features from frame semantic parsing, as they are both interpretable and difficult for adversaries to evade. One trade-off with using such features, however, is that more sophisticated feature representations require more complicated architectures, which limit usefulness in time-sensitive and constrained compute environments. We propose an approach to increase the efficiency of frame semantic parsing through an analysis of parallelization and beam search sizes. Our approach results in a system that is approximately 8.37x faster than the base system with a minimal effect on accuracy.

pdf bib
Multimodal Contextual Dialogue Breakdown Detection for Conversational AI Models
Md Messal Monem Miah | Ulie Schnaithmann | Arushi Raghuvanshi | Youngseo Son

Detecting dialogue breakdown in real time is critical for conversational AI systems, because it enables taking corrective action to successfully complete a task. In spoken dialog systems, this breakdown can be caused by a variety of unexpected situations including high levels of background noise, causing STT mistranscriptions, or unexpected user flows.In particular, industry settings like healthcare, require high precision and high flexibility to navigate differently based on the conversation history and dialogue states. This makes it both more challenging and more critical to accurately detect dialog breakdown. To accurately detect breakdown, we found it requires processing audio inputs along with downstream NLP model inferences on transcribed text in real time. In this paper, we introduce a Multimodal Contextual Dialogue Breakdown (MultConDB) model. This model significantly outperforms other known best models by achieving an F1 of 69.27.

pdf bib
Deferred NAM: Low-latency Top-K Context Injection via Deferred Context Encoding for Non-Streaming ASR
Zelin Wu | Gan Song | Christopher Li | Pat Rondon | Zhong Meng | Xavier Velez | Weiran Wang | Diamantino Caseiro | Golan Pundak | Tsendsuren Munkhdalai | Angad Chandorkar | Rohit Prabhavalkar

Contextual biasing enables speech recognizers to transcribe important phrases in the speaker’s context, such as contact names, even if they are rare in, or absent from, the training data. Attention-based biasing is a leading approach which allows for full end-to-end cotraining of the recognizer and biasing system and requires no separate inference-time components. Such biasers typically consist of a context encoder; followed by a context filter which narrows down the context to apply, improving per-step inference time; and, finally, context application via cross attention. Though much work has gone into optimizing per-frame performance, the context encoder is at least as important: recognition cannot begin before context encoding ends. Here, we show the lightweight phrase selection pass can be moved before context encoding, resulting in a speedup of up to 16.1 times and enabling biasing to scale to 20K phrases with a maximum pre-decoding delay under 33ms. With the addition of phrase- and wordpiece-level cross-entropy losses, our technique also achieves up to a 37.5% relative WER reduction over the baseline without the losses and lightweight phrase selection pass.

pdf bib
Less is More for Improving Automatic Evaluation of Factual Consistency
Tong Wang | Ninad Kulkarni | Yanjun Qi

Assessing the factual consistency of automatically generated texts in relation to source context is crucial for developing reliable natural language generation applications. Recent literature proposes AlignScore which uses a unified alignment model to evaluate factual consistency and substantially outperforms previous methods across many benchmark tasks. In this paper, we take a closer look of datasets used in AlignScore and uncover an unexpected finding: utilizing a smaller number of data points can actually improve performance. We process the original AlignScore training dataset to remove noise, augment with robustness-enhanced samples, and utilize a subset comprising 10% of the data to train an improved factual consistency evaluation model, we call LIM-RA (Less Is More for Robust AlignScore). LIM-RA demonstrates superior performance, consistently outperforming AlignScore and other strong baselines like ChatGPT across four benchmarks (two utilizing traditional natural language generation datasets and two focused on large language model outputs). Our experiments show that LIM-RA achieves the highest score on 24 of the 33 test datasets, while staying competitive on the rest, establishing the new state-of-the-art benchmarks.

pdf bib
DriftWatch: A Tool that Automatically Detects Data Drift and Extracts Representative Examples Affected by Drift
Myeongjun Jang | Antonios Georgiadis | Yiyun Zhao | Fran Silavong

Data drift, which denotes a misalignment between the distribution of reference (i.e., training) and production data, constitutes a significant challenge for AI applications, as it undermines the generalisation capacity of machine learning (ML) models. Therefore, it is imperative to proactively identify data drift before users meet with performance degradation. Moreover, to ensure the successful execution of AI services, endeavours should be directed not only toward detecting the occurrence of drift but also toward effectively addressing this challenge. % considering the limited resources prevalent in practical industrial domains. In this work, we introduce a tool designed to detect data drift in text data. In addition, we propose an unsupervised sampling technique for extracting representative examples from drifted instances. This approach bestows a practical advantage by significantly reducing expenses associated with annotating the labels for drifted instances, an essential prerequisite for retraining the model to sustain its performance on production data.

pdf bib
Graph Integrated Language Transformers for Next Action Prediction in Complex Phone Calls
Amin Marani | Ulie Schnaithmann | Youngseo Son | Akil Iyer | Manas Paldhe | Arushi Raghuvanshi

Current Conversational AI systems employ different machine learning pipelines, as well as external knowledge sources and business logic to predict the next action. Maintaining various components in dialogue managers’ pipeline adds complexity in expansion and updates, increases processing time, and causes additive noise through the pipeline that can lead to incorrect next action prediction. This paper investigates graph integration into language transformers to improve understanding the relationships between humans’ utterances, previous, and next actions without the dependency on external sources or components. Experimental analyses on real calls indicate that the proposed Graph Integrated Language Transformer models can achieve higher performance compared to other production level conversational AI systems in driving interactive calls with human users in real-world settings.

pdf bib
Leveraging LLMs for Dialogue Quality Measurement
Jinghan Jia | Abi Komma | Timothy Leffel | Xujun Peng | Ajay Nagesh | Tamer Soliman | Aram Galstyan | Anoop Kumar

In task-oriented conversational AI evaluation, unsupervised methods poorly correlate with human judgments, and supervised approaches lack generalization. Recent advances in large language models (LLMs) show robust zero- and few-shot capabilities across NLP tasks. Our paper explores using LLMs for automated dialogue quality evaluation, experimenting with various configurations on public and proprietary datasets. Manipulating factors such as model size, in-context examples, and selection techniques, we examine “chain-of-thought” (CoT) reasoning and label extraction procedures. Our results show that (1) larger models yield more accurate dialogue labels; (2) algorithmic selection of in-context examples outperforms random selection,; (3) CoT reasoning where an LLM is asked to provide justifications before outputting final labels improves performance; and (4) fine-tuned LLMs outperform out-of-the-box ones. In addition, we find that suitably tuned LLMs exhibit high accuracy in dialogue evaluation compared to human judgments.

pdf bib
Uncertainty Estimation in Large Language Models to Support Biodiversity Conservation
Maria Mora-Cross | Saul Calderon-Ramirez

Large Language Models (LLM) provide significant value in question answering (QA) scenarios and have practical application in complex decision-making contexts, such as biodiversity conservation. However, despite substantial performance improvements, they may still produce inaccurate outcomes. Consequently, incorporating uncertainty quantification alongside predictions is essential for mitigating the potential risks associated with their use. This study introduces an exploratory analysis of the application of Monte Carlo Dropout (MCD) and Expected Calibration Error (ECE) to assess the uncertainty of generative language models. To that end, we analyzed two publicly available language models (Falcon-7B and DistilGPT-2). Our findings suggest the viability of employing ECE as a metric to estimate uncertainty in generative LLM. The findings from this research contribute to a broader project aiming at facilitating free and open access to standardized and integrated data and services about Costa Rica’s biodiversity to support the development of science, education, and biodiversity conservation.

pdf bib
AMA-LSTM: Pioneering Robust and Fair Financial Audio Analysis for Stock Volatility Prediction
Shengkun Wang | Taoran Ji | Jianfeng He | Mariam ALMutairi | Dan Wang | Linhan Wang | Min Zhang | Chang-Tien Lu

Stock volatility prediction is an important task in the financial industry. Recent multimodal methods have shown advanced results by combining text and audio information, such as earnings calls. However, these multimodal methods have faced two drawbacks. First, they often fail to yield reliable models and overfit the data due to their absorption of stochastic information from the stock market. Moreover, using multimodal models to predict stock volatility suffers from gender bias and lacks an efficient way to eliminate such bias. To address these aforementioned problems, we use adversarial training to generate perturbations that simulate the inherent stochasticity and bias, by creating areas resistant to random information around the input space to improve model robustness and fairness. Our comprehensive experiments on two real-world financial audio datasets reveal that this method exceeds the performance of current state-of-the-art solution. This confirms the value of adversarial training in reducing stochasticity and bias for stock volatility prediction tasks.

pdf bib
Tiny Titans: Can Smaller Large Language Models Punch Above Their Weight in the Real World for Meeting Summarization?
Xue-Yong Fu | Md Tahmid Rahman Laskar | Elena Khasanova | Cheng Chen | Shashi Tn

Large Language Models (LLMs) have demonstrated impressive capabilities to solve a wide range of tasks without being explicitly fine-tuned on task-specific datasets. However, deploying LLMs in the real world is not trivial, as it requires substantial computing resources. In this paper, we investigate whether smaller, Compact LLMs are a good alternative to the comparatively Larger LLMs to address significant costs associated with utilizing LLMs in the real world. In this regard, we study the meeting summarization task in a real-world industrial environment and conduct extensive experiments by comparing the performance of fine-tuned compact LLMs (FLAN-T5, TinyLLaMA, LiteLLaMA, etc.) with zero-shot larger LLMs (LLaMA-2, GPT-3.5, PaLM-2). We observe that most smaller LLMs, even after fine-tuning, fail to outperform larger zero-shot LLMs in meeting summarization datasets. However, a notable exception is FLAN-T5 (780M parameters), which achieves performance on par with zero-shot Larger LLMs (from 7B to above 70B parameters), while being significantly smaller. This makes compact LLMs like FLAN-T5 a suitable cost-efficient LLM for real-world industrial deployment.

pdf bib
Shears: Unstructured Sparsity with Neural Low-rank Adapter Search
Juan Munoz | Jinjie Yuan | Nilesh Jain

Recently, several approaches successfully demonstrated that weight-sharing Neural Architecture Search (NAS) can effectively explore a search space of elastic low-rank adapters (LoRA), allowing the parameter-efficient fine-tuning (PEFT) and compression of large language models. In this paper, we introduce a novel approach called Shears, demonstrating how the integration of cost-effective sparsity and a proposed Neural Low-rank adapter Search (NLS) algorithm can further improve the efficiency of PEFT approaches. Results demonstrate the benefits of Shears compared to other methods, reaching high sparsity levels while improving or with little drop in accuracy, utilizing a single GPU for a pair of hours.

pdf bib
Tree-of-Question: Structured Retrieval Framework for Korean Question Answering Systems
Dongyub Lee | Younghun Jeong | Hwa-Yeon Kim | Hongyeon Yu | Seunghyun Han | Taesun Whang | Seungwoo Cho | Chanhee Lee | Gunsu Lee | Youngbum Kim

We introduce Korean language-specific RAG-based QA systems, primarily through the innovative Tree-of-Question (ToQ) methodology and enhanced query generation techniques. We address the complex, multi-hop nature of real-world questions by effectively integrating advanced LLMs with nuanced query planning. Our comprehensive evaluations, including a newly created Korean multi-hop QA dataset, demonstrate our method’s ability to elevate response validity and accuracy, especially in deeper levels of reasoning. This paper not only showcases significant progress in handling the intricacies of Korean linguistic structures but also sets a new standard in the development of context-aware and linguistically sophisticated QA systems.

pdf bib
LLM-based Frameworks for API Argument Filling in Task-Oriented Conversational Systems
Jisoo Mok | Mohammad Kachuee | Shuyang Dai | Shayan Ray | Tara Taghavi | Sungroh Yoon

Task-orientated conversational agents interact with users and assist them via leveraging external APIs. A typical task-oriented conversational system can be broken down into three phases: external API selection, argument filling, and response generation. The focus of our work is the task of argument filling, which is in charge of accurately providing arguments required by the selected API. Upon comprehending the dialogue history and the pre-defined API schema, the argument filling task is expected to provide the external API with the necessary information to generate a desirable agent action. In this paper, we study the application of Large Language Models (LLMs) for the problem of API argument filling task. Our initial investigation reveals that LLMs require an additional grounding process to successfully perform argument filling, inspiring us to design training and prompting frameworks to ground their responses. Our experimental results demonstrate that when paired with proposed techniques, the argument filling performance of LLMs noticeably improves, paving a new way toward building an automated argument filling framework.

pdf bib
Large Language Models Encode the Practice of Medicine
Teja Kanchinadam | Gauher Shaheen

Healthcare tasks such as predicting clinical outcomes across medical and surgical populations, disease prediction, predicting patient health journeys, are typically approached with supervised learning on task-specific datasets. We demonstrate that language models begin to learn these tasks without any explicit supervision when trained on a new dataset of billions of administrative claims, which essentially encapsulates the practice of medicine, offering a unique perspective on patient care and treatment patterns. Our model, MediClaimGPT, a 125M parameter Transformer demonstrates strong zero-shot predictive capabilities, accurately forecasting patient health events across four evaluation datasets, with its capabilities further demonstrated in various downstream tasks. A significant application of MediClaimGPT is in generating high-quality, clinically plausible synthetic claims data, enhancing healthcare data utility while preserving patient privacy. This research underscores the potential of language models in handling complex datasets and their strategic application in healthcare and related fields.

pdf bib
Leveraging Interesting Facts to Enhance User Engagement with Conversational Interfaces
Nikhita Vedula | Giuseppe Castellucci | Eugene Agichtein | Oleg Rokhlenko | Shervin Malmasi

Conversational Task Assistants (CTAs) guide users in performing a multitude of activities, such as making recipes. However, ensuring that interactions remain engaging, interesting, and enjoyable for CTA users is not trivial, especially for time-consuming or challenging tasks. Grounded in psychological theories of human interest, we propose to engage users with contextual and interesting statements or facts during interactions with a multi-modal CTA, to reduce fatigue and task abandonment before a task is complete. To operationalize this idea, we train a high-performing classifier (82% F1-score) to automatically identify relevant and interesting facts for users. We use it to create an annotated dataset of task-specific interesting facts for the domain of cooking. Finally, we design and validate a dialogue policy to incorporate the identified relevant and interesting facts into a conversation, to improve user engagement and task completion. Live testing on a leading multi-modal voice assistant shows that 66% of the presented facts were received positively, leading to a 40% gain in the user satisfaction rating, and a 37% increase in conversation length. These findings emphasize that strategically incorporating interesting facts into the CTA experience can promote real-world user participation for guided task interactions.

pdf bib
Search Query Refinement for Japanese Named Entity Recognition in E-commerce Domain
Yuki Nakayama | Ryutaro Tatsushima | Erick Mendieta | Koji Murakami | Keiji Shinzato

In the E-Commerce domain, search query refinement reformulates malformed queries into canonicalized forms by preprocessing operations such as “term splitting” and “term merging”. Unfortunately, most relevant research is rather limited to English. In particular, there is a severe lack of study on search query refinement for the Japanese language. Furthermore, no attempt has ever been made to apply refinement methods to data improvement for downstream NLP tasks in real-world scenarios.This paper presents a novel query refinement approach for the Japanese language. Experimental results show that our method achieves significant improvement by 3.5 points through comparison with BERT-CRF as a baseline. Further experiments are also conducted to measure beneficial impact of query refinement on named entity recognition (NER) as the downstream task. Evaluations indicate that the proposed query refinement method contributes to better data quality, leading to performance boost on E-Commerce specific NER tasks by 11.7 points, compared to search query data preprocessed by MeCab, a very popularly adopted Japanese tokenizer.

pdf bib
EIVEN: Efficient Implicit Attribute Value Extraction using Multimodal LLM
Henry Zou | Gavin Yu | Ziwei Fan | Dan Bu | Han Liu | Peng Dai | Dongmei Jia | Cornelia Caragea

In e-commerce, accurately extracting product attribute values from multimodal data is crucial for improving user experience and operational efficiency of retailers. However, previous approaches to multimodal attribute value extraction often struggle with implicit attribute values embedded in images or text, rely heavily on extensive labeled data, and can easily confuse similar attribute values. To address these issues, we introduce EIVEN, a data- and parameter-efficient generative framework that pioneers the use of multimodal LLM for implicit attribute value extraction. EIVEN leverages the rich inherent knowledge of a pre-trained LLM and vision encoder to reduce reliance on labeled data. We also introduce a novel Learning-by-Comparison technique to reduce model confusion by enforcing attribute value comparison and difference identification. Additionally, we construct initial open-source datasets for multimodal implicit attribute value extraction. Our extensive experiments reveal that EIVEN significantly outperforms existing methods in extracting implicit attribute values while requiring less labeled data.

pdf bib
Exploring the Impact of Table-to-Text Methods on Augmenting LLM-based Question Answering with Domain Hybrid Data
Dehai Min | Nan Hu | Rihui Jin | Nuo Lin | Jiaoyan Chen | Yongrui Chen | Yu Li | Guilin Qi | Yun Li | Nijun Li | Qianren Wang

Augmenting Large Language Models (LLMs) for Question Answering (QA) with domain specific data has attracted wide attention. However, domain data often exists in a hybrid format, including text and semi-structured tables, posing challenges for the seamless integration of information. Table-to-Text Generation is a promising solution by facilitating the transformation of hybrid data into a uniformly text-formatted corpus. Although this technique has been widely studied by the NLP community, there is currently no comparative analysis on how corpora generated by different table-to-text methods affect the performance of QA systems.In this paper, we address this research gap in two steps. First, we innovatively integrate table-to-text generation into the framework of enhancing LLM-based QA systems with domain hybrid data. Then, we utilize this framework in real-world industrial data to conduct extensive experiments on two types of QA systems (DSFT and RAG frameworks) with four representative methods: Markdown format, Template serialization, TPLM-based method, and LLM-based method. Based on the experimental results, we draw some empirical findings and explore the underlying reasons behind the success of some methods. We hope the findings of this work will provide a valuable reference for the academic and industrial communities in developing robust QA systems.

pdf bib
Solving General Natural-Language-Description Optimization Problems with Large Language Models
Jihai Zhang | Wei Wang | Siyan Guo | Li Wang | Fangquan Lin | Cheng Yang | Wotao Yin

Optimization problems seek to find the best solution to an objective under a set of constraints, and have been widely investigated in real-world applications. Modeling and solving optimization problems in a specific domain typically require a combination of domain knowledge, mathematical skills, and programming ability, making it difficult for general users and even domain professionals. In this paper, we propose a novel framework called OptLLM that augments LLMs with external solvers. Specifically, OptLLM accepts user queries in natural language, convert them into mathematical formulations and programming codes, and calls the solvers to calculate the results for decision-making. In addition, OptLLM supports multi-round dialogues to gradually refine the modeling and solving of optimization problems. To illustrate the effectiveness of OptLLM, we provide tutorials on three typical optimization applications and conduct experiments on both prompt-based GPT models and a fine-tuned Qwen model using a large-scale self-developed optimization dataset. Experimental results show that OptLLM works with various LLMs, and the fine-tuned model achieves an accuracy boost compared to the prompt-based models. Some features of OptLLM framework have been available for trial since June 2023 (https://opt.alibabacloud.com/chat or https://opt.aliyun.com/chat).

pdf bib
Self-Regulated Data-Free Knowledge Amalgamation for Text Classification
Prashanth Vijayaraghavan | Hongzhi Wang | Luyao Shi | Tyler Baldwin | David Beymer | Ehsan Degan

Recently, there has been a growing availability of pre-trained text models on various model repositories. These models greatly reduce the cost of training new models from scratch as they can be fine-tuned for specific tasks or trained on large datasets. However, these datasets may not be publicly accessible due to the privacy, security, or intellectual property issues. In this paper, we aim to develop a lightweight student network that can learn from multiple teacher models without accessing their original training data. Hence, we investigate Data-Free Knowledge Amalgamation (DFKA), a knowledge-transfer task that combines insights from multiple pre-trained teacher models and transfers them effectively to a compact student network. To accomplish this, we propose STRATANET, a modeling framework comprising: (a) a steerable data generator that produces text data tailored to each teacher and (b) an amalgamation module that implements a self-regulative strategy using confidence estimates from the teachers’ different layers to selectively integrate their knowledge and train a versatile student. We evaluate our method on three benchmark text classification datasets with varying labels or domains. Empirically, we demonstrate that the student model learned using our STRATANET outperforms several baselines significantly under data-driven and data-free constraints.
up

pdf (full)
bib (full) Findings of the Association for Computational Linguistics: NAACL 2024

pdf bib
Findings of the Association for Computational Linguistics: NAACL 2024
Kevin Duh | Helena Gomez | Steven Bethard

pdf bib
Structured Pruning for Large Language Models Using Coupled Components Elimination and Minor Fine-tuning
Honghe Zhang | XiaolongShi XiaolongShi | Jingwei Sun | Guangzhong Sun

Large language models (LLMs) have demonstrated powerful capabilities in natural language processing, yet their vast number of parameters poses challenges for deployment and inference efficiency. Structured model pruning emerges as a viable approach to reduce model size and accelerate inference, without requiring specialized operators and libraries for deployment. However, structured pruning often severely weakens the model’s capability.Despite repetitive fine-tuning can restore the capability to a certain extent, it impairs LLMs’ utility as versatile problem solvers.To address this issue, we propose a novel structured pruning algorithm tailored for LLMs. It derives the importance of different components, namely rows and columns in parameter matrices, based on intermediate data dependencies. Then it removes coupled components across different layers simultaneously and preserves dependency relationships within remaining parameters, avoiding significant performance degradation. The pruned model requires only few epochs of fine-tuning to restore its performance, ensuring the model’s ability to generalize.Empirical evaluations on LLaMA, Vicuna, and ChatGLM3 demonstrate our algorithm’s efficacy, yielding 20% parameter reduction while retaining at least 94.4% of original performance metrics.

pdf bib
Weight-Inherited Distillation for Task-Agnostic BERT Compression
Taiqiang Wu | Cheng Hou | Shanshan Lao | Jiayi Li | Ngai Wong | Zhe Zhao | Yujiu Yang

Knowledge Distillation (KD) is a predominant approach for BERT compression.Previous KD-based methods focus on designing extra alignment losses for the student model to mimic the behavior of the teacher model.These methods transfer the knowledge in an indirect way.In this paper, we propose a novel Weight-Inherited Distillation (WID), which directly transfers knowledge from the teacher.WID does not require any additional alignment loss and trains a compact student by inheriting the weights, showing a new perspective of knowledge distillation.Specifically, we design the row compactors and column compactors as mappings and then compress the weights via structural re-parameterization.Experimental results on the GLUE and SQuAD benchmarks show that WID outperforms previous state-of-the-art KD-based baselines.Further analysis indicates that WID can also learn the attention patterns from the teacher model without any alignment loss on attention distributions.The code is available at https://github.com/wutaiqiang/WID-NAACL2024.

pdf bib
Ignore Me But Don’t Replace Me: Utilizing Non-Linguistic Elements for Pretraining on the Cybersecurity Domain
Eugene Jang | Jian Cui | Dayeon Yim | Youngjin Jin | Jin-Woo Chung | Seungwon Shin | Yongjae Lee

Cybersecurity information is often technically complex and relayed through unstructured text, making automation of cyber threat intelligence highly challenging. For such text domains that involve high levels of expertise, pretraining on in-domain corpora has been a popular method for language models to obtain domain expertise. However, cybersecurity texts often contain non-linguistic elements (such as URLs and hash values) that could be unsuitable with the established pretraining methodologies. Previous work in other domains have removed or filtered such text as noise, but the effectiveness of these methods have not been investigated, especially in the cybersecurity domain. We experiment with different pretraining methodologies to account for non-linguistic elements (NLEs) and evaluate their effectiveness through downstream tasks and probing tasks. Our proposed strategy, a combination of selective MLM and jointly training NLE token classification, outperforms the commonly taken approach of replacing NLEs. We use our domain-customized methodology to train CyBERTuned, a cybersecurity domain language model that outperforms other cybersecurity PLMs on most tasks.

pdf bib
Extremely efficient online query encoding for dense retrieval
Nachshon Cohen | Yaron Fairstein | Guy Kushilevitz

Existing dense retrieval systems utilize the same model architecture for encoding both the passages and the queries, even though queries are much shorter and simpler than passages. This leads to high latency of the query encoding, which is performed online and therefore might impact user experience. We show that combining a standard large passage encoder with a small efficient query encoder can provide significant latency drops with only a small decrease in quality. We offer a pretraining and training solution for multiple small query encoder architectures. Using a small transformer architecture we are able to decrease latency by up to ∼12×, while MRR@10 on the MS MARCO dev set only decreases from 38.2 to 36.2. If this solution does not reach the desired latency requirements, we propose an efficient RNN as the query encoder, which processes the query prefix incrementally and only infers the last word after the query is issued. This shortens latency by ∼38× with only a minor drop in quality, reaching 35.5 MRR@10 score.

pdf bib
DIVKNOWQA: Assessing the Reasoning Ability of LLMs via Open-Domain Question Answering over Knowledge Base and Text
Wenting Zhao | Ye Liu | Tong Niu | Yao Wan | Philip Yu | Shafiq Joty | Yingbo Zhou | Semih Yavuz

Large Language Models (LLMs) have exhibited impressive generation capabilities, but they suffer from hallucinations when solely relying on their internal knowledge, especially when answering questions that require less commonly known information. Retrievalaugmented LLMs have emerged as a potential solution to ground LLMs in external knowledge. Nonetheless, recent approaches have primarily emphasized retrieval from unstructured text corpora, owing to its seamless integration into prompts. When using structured data such as knowledge graphs, most methods simplify it into natural text, neglecting the underlying structures. Moreover, a significant gap in the current landscape is the absence of a realistic benchmark for evaluating the effectiveness of grounding LLMs on heterogeneous knowledge sources (e.g., knowledge base and text). To fill this gap, we have curated a comprehensive dataset that poses two unique challenges: (1) Two-hop multi-source questions that require retrieving information from both open-domain structured and unstructured knowledge sources; retrieving information from structured knowledge sources is a critical component in correctly answering the questions. (2) Generation of symbolic queries (e.g., SPARQL for Wikidata) is a key requirement, which adds another layer of challenge. Our dataset is created using a combination of automatic generation through predefined reasoning chains and human annotation. We also introduce a novel approach that leverages multiple retrieval tools, including text passage retrieval and symbolic language-assisted retrieval. Our model outperforms previous approaches by a significant margin, demonstrating its effectiveness in addressing the above-mentioned reasoning challenges.

pdf bib
SpeedE: Euclidean Geometric Knowledge Graph Embedding Strikes Back
Aleksandar Pavlović | Emanuel Sallinger

Geometric knowledge graph embedding models (gKGEs) have shown great potential for knowledge graph completion (KGC), i.e., automatically predicting missing triples. However, contemporary gKGEs require high embedding dimensionalities or complex embedding spaces for good KGC performance, drastically limiting their space and time efficiency. Facing these challenges, we propose SpeedE, a lightweight Euclidean gKGE that (1) provides strong inference capabilities, (2) is competitive with state-of-the-art gKGEs, even significantly outperforming them on YAGO3-10 and WN18RR, and (3) dramatically increases their efficiency, in particular, needing solely a fifth of the training time and a fourth of the parameters of the state-of-the-art ExpressivE model on WN18RR to reach the same KGC performance.

pdf bib
Language Guided Exploration for RL Agents in Text Environments
Hitesh Golchha | Sahil Yerawar | Dhruvesh Patel | Soham Dan | Keerthiram Murugesan

Real-world sequential decision making is characterized by sparse rewards and large decision spaces, posing significant difficulty for experiential learning systems like tabula rasa reinforcement learning (RL) agents. Large Language Models (LLMs), with a wealth of world knowledge, can help RL agents learn quickly and adapt to distribution shifts. In this work, we introduce Language Guided Exploration (LGE) framework, which uses a pre-trained language model (called GUIDE ) to provide decision-level guidance to an RL agent (called EXPLORER ). We observe that on ScienceWorld (Wang et al., 2022), a challenging text environment, LGE outperforms vanilla RL agents significantly and also outperforms other sophisticated methods like Behaviour Cloning and Text Decision Transformer.

pdf bib
GPT-who: An Information Density-based Machine-Generated Text Detector
Saranya Venkatraman | Adaku Uchendu | Dongwon Lee

The Uniform Information Density (UID) principle posits that humans prefer to spread information evenly during language production. We examine if this UID principle can help capture differences between Large Language Models (LLMs)-generated and human-generated texts. We propose GPT-who, the first psycholinguistically-inspired domain-agnostic statistical detector. This detector employs UID-based featuresto model the unique statistical signature of each LLM and human author for accurate detection. We evaluate our method using 4 large-scale benchmark datasets and find that GPT-who outperforms state-of-the-art detectors (both statistical- & non-statistical) such as GLTR, GPTZero, DetectGPT, OpenAI detector, and ZeroGPT by over 20% across domains.In addition to better performance, it is computationally inexpensive and utilizes an interpretable representation of text articles. We find that GPT-who can distinguish texts generated by very sophisticated LLMs, even when the overlying text is indiscernible.UID-based measures for all datasets and code are available at https://github.com/saranya-venkatraman/gpt-who.

pdf bib
DEED: Dynamic Early Exit on Decoder for Accelerating Encoder-Decoder Transformer Models
Peng Tang | Pengkai Zhu | Tian Li | Srikar Appalaraju | Vijay Mahadevan | R. Manmatha

Encoder-decoder transformer models have achieved great success on various vision-language (VL) and language tasks, but they suffer from high inference latency. Typically, the decoder takes up most of the latency because of the auto-regressive decoding. To accelerate the inference, we propose an approach of performing Dynamic Early Exit on Decoder (DEED). We build a multi-exit encoder-decoder transformer model which is trained with deep supervision so that each of its decoder layers is capable of generating plausible predictions. In addition, we leverage simple yet practical techniques, including shared generation head and adaptation modules, to keep accuracy when exiting at shallow decoder layers. Based on the multi-exit model, we perform step-level dynamic early exit during inference, where the model may decide to use fewer decoder layers based on its confidence of the current layer at each individual decoding step. Considering different number of decoder layers may be used at different decoding steps, we compute deeper-layer decoder features of previous decoding steps just-in-time, which ensures the features from different decoding steps are semantically aligned. We evaluate our approach with three state-of-the-art encoder-decoder transformer models on various VL and language tasks. We show our approach can reduce overall inference latency by 20%-74% with comparable or even higher accuracy compared to baselines.

pdf bib
Attention Alignment and Flexible Positional Embeddings Improve Transformer Length Extrapolation
Ta-Chung Chi | Ting-Han Fan | Alexander Rudnicky

An ideal length-extrapolatable Transformer language model can handle sequences longer than the training length without any fine-tuning. Such long-context utilization capability relies heavily on a flexible positional embedding design. Upon investigating the flexibility of existing large pre-trained Transformer language models, we find that the T5 family deserves a closer look, as its positional embeddings capture rich and flexible attention patterns. However, T5 suffers from the dispersed attention issue: the longer the input sequence, the flatter the attention distribution. To alleviate the issue, we propose two attention alignment strategies via temperature scaling. Our findings show improvement on the long-context utilization capability of T5 on language modeling, retrieval, multi-document question answering, and code completion tasks without any fine-tuning. This suggests that a flexible positional embedding design and attention alignment can go a long way toward Transformer length extrapolation. The code is released at: https://github.com/chijames/T5-Attention-Alignment

pdf bib
Automatic Pair Construction for Contrastive Post-training
Canwen Xu | Corby Rosset | Ethan Chau | Luciano Corro | Shweti Mahajan | Julian McAuley | Jennifer Neville | Ahmed Awadallah | Nikhil Rao

Alignment serves as an important step to steer large language models (LLMs) towards human preferences. In this paper, we propose an automatic way to construct contrastive data for LLM, using preference pairs from multiple models of varying strengths (e.g., InstructGPT, ChatGPT and GPT-4). We compare the contrastive techniques of SLiC and DPO to SFT baselines and find that DPO provides a step-function improvement even after continuing SFT saturates. We also explore a data curriculum learning scheme for contrastive post-training, which starts by learning from “easier” pairs and transitioning to “harder” ones, which further improves alignment. Finally, we scale up our experiments to train with more data and larger models like Orca. Remarkably, our automatic contrastive post-training further improves the performance of Orca, already a state-of-the-art instruction learning model tuned with GPT-4 outputs, to outperform ChatGPT.

pdf bib
Self-Checker: Plug-and-Play Modules for Fact-Checking with Large Language Models
Miaoran Li | Baolin Peng | Michel Galley | Jianfeng Gao | Zhu Zhang

Fact-checking is an essential task in NLP that is commonly utilized to validate the factual accuracy of a piece of text. Previous approaches mainly involve the resource-intensive process of fine-tuning pre-trained language models on specific datasets. In addition, there is a notable gap in datasets that focus on fact-checking texts generated by large language models (LLMs). In this paper, we introduce Self-Checker, a plug-and-play framework that harnesses LLMs for efficient and rapid fact-checking in a few-shot manner. We also present the BingCheck dataset, specifically designed for fact-checking texts generated by LLMs. Empirical results demonstrate the potential of Self-Checker in the use of LLMs for fact-checking. Compared to state-of-the-art fine-tuned models, there is still significant room for improvement, indicating that adopting LLMs could be a promising direction for future fact-checking research.

pdf bib
Low-resource neural machine translation with morphological modeling
Antoine Nzeyimana

Morphological modeling in neural machine translation (NMT) is a promising approach to achieving open-vocabulary machine translation for morphologically-rich languages. However, existing methods such as sub-word tokenization and character-based models are limited to the surface forms of the words. In this work, we propose a framework-solution for modeling complex morphology in low-resource settings. A two-tier transformer architecture is chosen to encode morphological information at the inputs. At the target-side output, a multi-task multi-label training scheme coupled with a beam search-based decoder are found to improve machine translation performance. An attention augmentation scheme to the transformer model is proposed in a generic form to allow integration of pre-trained language models and also facilitate modeling of word order relationships between the source and target languages. Several data augmentation techniques are evaluated and shown to increase translation performance in low-resource settings. We evaluate our proposed solution on Kinyarwanda ↔ English translation using public-domain parallel text. Our final models achieve competitive performance in relation to large multi-lingual models. We hope that our results will motivate more use of explicit morphological information and the proposed model and data augmentations in low-resource NMT.

pdf bib
Self-Cleaning: Improving a Named Entity Recognizer Trained on Noisy Data with a Few Clean Instances
Zhendong Chu | Ruiyi Zhang | Tong Yu | Rajiv Jain | Vlad Morariu | Jiuxiang Gu | Ani Nenkova

To achieve state-of-the-art performance, one still needs to train NER models on large-scale, high-quality annotated data, an asset that is both costly and time-intensive to accumulate. In contrast, real-world applications often resort to massive low-quality labeled data through non-expert annotators via crowdsourcing and external knowledge bases via distant supervision as a cost-effective alternative. However, these annotation methods result in noisy labels, which in turn lead to a notable decline in performance. Hence, we propose to denoise the noisy NER data with guidance from a small set of clean instances. Along with the main NER model we train a discriminator model and use its outputs to recalibrate the sample weights. The discriminator is capable of detecting both span and category errors with different discriminative prompts. Results on public crowdsourcing and distant supervision datasets show that the proposed method can consistently improve performance with a small guidance set.

pdf bib
VLUE: A New Benchmark and Multi-task Knowledge Transfer Learning for Vietnamese Natural Language Understanding
Phong Do | Son Tran | Phu Hoang | Kiet Nguyen | Ngan Nguyen

The success of Natural Language Understanding (NLU) benchmarks in various languages, such as GLUE for English, CLUE for Chinese, KLUE for Korean, and IndoNLU for Indonesian, has facilitated the evaluation of new NLU models across a wide range of tasks. To establish a standardized set of benchmarks for Vietnamese NLU, we introduce the first Vietnamese Language Understanding Evaluation (VLUE) benchmark. The VLUE benchmark encompasses five datasets covering different NLU tasks, including text classification, span extraction, and natural language understanding. To provide an insightful overview of the current state of Vietnamese NLU, we then evaluate seven state-of-the-art pre-trained models, including both multilingual and Vietnamese monolingual models, on our proposed VLUE benchmark. Furthermore, we present CafeBERT, a new state-of-the-art pre-trained model that achieves superior results across all tasks in the VLUE benchmark. Our model combines the proficiency of a multilingual pre-trained model with Vietnamese linguistic knowledge. CafeBERT is developed based on the XLM-RoBERTa model, with an additional pretraining step utilizing a significant amount of Vietnamese textual data to enhance its adaptation to the Vietnamese language. For the purpose of future research, CafeBERT is made publicly available for research purposes.

pdf bib
LETI: Learning to Generate from Textual Interactions
Xingyao Wang | Hao Peng | Reyhaneh Jabbarvand | Heng Ji

Fine-tuning pre-trained language models (LMs) is essential for enhancing their capabilities.Existing techniques commonly fine-tune on input-output pairs (e.g., instruction tuning) or with numerical rewards that gauge the output quality (e.g., RLHF). We explore LMs’ potential to **le**arn from **t**extual **i**nteractions (**LETI**) that not only check their correctness with *binary labels* but also pinpoint and explain errors in their outputs through *textual feedback*.Our focus is the code generation task, where the model produces code based on natural language instructions. This setting invites a natural and scalable way to acquire textual feedback: the error messages and stack traces from code execution using a Python interpreter. LETI iteratively fine-tunes the model, using the LM objective, on a concatenation of natural language instructions, LM-generated programs, and textual feedback. Prepended to this fine-tuning text, a binary reward token is used to differentiate correct and buggy solutions.LETI requires *no* ground-truth outputs for training and even outperforms a fine-tuned baseline that does. LETI not only improves the performance of LMs on a code generation dataset MBPP, but also generalizes to other datasets. Trained on MBPP, it achieves comparable or better performance than the base LMs on unseen problems in HumanEval. Furthermore, compared to binary feedback, we observe that textual feedback leads to improved generation quality and sample efficiency, achieving the same performance with fewer than half of the gradient steps.LETI is equally applicable in natural language tasks when they can be formulated as code generation, which we empirically verified on event argument extraction.

pdf bib
Bilateral Masking with prompt for Knowledge Graph Completion
Yonghui Kong | Cunhang Fan | Yujie Chen | Shuai Zhang | Zhao Lv | Jianhua Tao

The pre-trained language model (PLM) has achieved significant success in the field of knowledge graph completion (KGC) by effectively modeling entity and relation descriptions. In recent studies, the research in this field has been categorized into methods based on word matching and sentence matching, with the former significantly lags behind. However, there is a critical issue in word matching methods, which is that these methods fail to obtain satisfactory single embedding representations for entities.To address this issue and enhance entity representation, we propose the Bilateral Masking with prompt for Knowledge Graph Completion (BMKGC) approach.Our methodology employs prompts to narrow the distance between the predicted entity and the known entity. Additionally, the BMKGC model incorporates a bi-encoder architecture, enabling simultaneous predictions at both the head and tail. Furthermore, we propose a straightforward technique to augment positive samples, mitigating the problem of degree bias present in knowledge graphs and thereby improving the model’s robustness. Experimental results conclusively demonstrate that BMKGC achieves state-of-the-art performance on the WN18RR dataset.

pdf bib
MiLe Loss: a New Loss for Mitigating the Bias of Learning Difficulties in Generative Language Models
Zhenpeng Su | Zijia Lin | Baixue Baixue | Hui Chen | Songlin Hu | Wei Zhou | Guiguang Ding | Xing W

Generative language models are usually pre-trained on large text corpus via predicting the next token (i.e., sub-word/word/phrase) given the previous ones. Recent works have demonstrated the impressive performance of large generative language models on downstream tasks. However, existing generative language models generally neglect an inherent challenge in text corpus during training, i.e., the imbalance between frequent tokens and infrequent ones. It can lead a language model to be dominated by common and easy-to-learn tokens, thereby overlooking the infrequent and difficult-to-learn ones. To alleviate that, we propose a **MiLe Loss** function for **mi**tigating the bias of **le**arning difficulties with tokens. During training, it can dynamically assess the learning difficulty of a to-be-learned token, according to the information entropy of the corresponding predicted probability distribution over the vocabulary. Then it scales the training loss adaptively, trying to lead the model to focus more on the difficult-to-learn tokens. On the Pile dataset, we train generative language models at different scales of 468M, 1.2B, and 6.7B parameters. Experiments reveal that models incorporating the proposed MiLe Loss can gain consistent performance improvement on downstream benchmarks.

pdf bib
GOLD: Geometry Problem Solver with Natural Language Description
Jiaxin Zhang | Yashar Moshfeghi

Addressing the challenge of automated geometry math problem-solving in artificial intelligence (AI) involves understanding multi-modal information and mathematics. blackCurrent methods struggle with accurately interpreting geometry diagrams, which hinders effective problem-solving. To tackle this issue, we present the Geometry problem sOlver with natural Language Description (GOLD) model. GOLD enhances the extraction of geometric relations by separately processing symbols and geometric primitives within the diagram. Subsequently, it converts the extracted relations into natural language descriptions, efficiently utilizing large language models to solve geometry math problems. Experiments show that the GOLD model outperforms the Geoformer model, the previous best method on the UniGeo dataset, by achieving accuracy improvements of 12.7% and 42.1% in calculation and proving subsets. Additionally, it surpasses the former best model on the PGPS9K and Geometry3K datasets, PGPSNet, by obtaining accuracy enhancements of 1.8% and 3.2%, respectively.

pdf bib
RoDia: A New Dataset for Romanian Dialect Identification from Speech
Rotaru Codruț | Nicolae Ristea | Radu Ionescu

We introduce RoDia, the first dataset for Romanian dialect identification from speech. The RoDia dataset includes a varied compilation of speech samples from five distinct regions of Romania, covering both urban and rural environments, totaling 2 hours of manually annotated speech data. Along with our dataset, we introduce a set of competitive models to be used as baselines for future research. The top scoring model achieves a macro F1 score of 59.83% and a micro F1 score of 62.08%, indicating that the task is challenging. We thus believe that RoDia is a valuable resource that will stimulate research aiming to address the challenges of Romanian dialect identification. We release our dataset at https://github.com/codrut2/RoDia.

pdf bib
Examining Modularity in Multilingual LMs via Language-Specialized Subnetworks
Rochelle Choenni | Ekaterina Shutova | Dan Garrette

Recent work has proposed explicitly inducing language-wise modularity in multilingual LMs via sparse fine-tuning (SFT) on per-language subnetworks as a means of better guiding cross-lingual sharing. In this paper, we investigate (1) the degree to which language-wise modularity *naturally* arises within models with no special modularity interventions, and (2) how cross-lingual sharing and interference differ between such models and those with explicit SFT-guided subnetwork modularity. In order to do so, we use XLM-R as our multilingual LM. Moreover, to quantify language specialization and cross-lingual interaction, we use a Training Data Attribution method that estimates the degree to which a model’s predictions are influenced by in-language or cross-language training examples. Our results show that language-specialized subnetworks do naturally arise, and that SFT, rather than always increasing modularity, can decrease language specialization of subnetworks in favor of more cross-lingual sharing.

pdf bib
Reverse Chain: A Generic-Rule for LLMs to Master Multi-API Planning
Yinger Zhang | Hui Cai | Xierui Song | Yicheng Chen | Rui Sun | Jing Zheng

While enabling large language models to implement function calling (known as APIs) can greatly enhance the performance of Large Language Models (LLMs), function calling is still a challenging task due to the complicated relations between different APIs, especially in a context-learning setting without fine-tuning. This paper introduces “Reverse Chain”, a controllable, target-driven approach designed to empower LLMs with the capability to operate external APIs only via prompts. Recognizing that most LLMs have limited tool-use capabilities, Reverse Chain limits LLMs to executing simple tasks, e.g., API Selection and Argument Completion. Furthermore, to manage a controllable multi-function calling, Reverse Chain adopts a generic rule-based on a backward reasoning process. This rule determines when to do API selection or Argument completion. To evaluate the multi-tool-use capability of LLMs, we have released a compositional multi-tool task dataset, available at https://github.com/zhangyingerjelly/reverse-chain. Extensive numerical experiments validate the remarkable proficiency of Reverse Chain in managing multiple API calls.

pdf bib
Incorporating Exponential Smoothing into MLP: a Simple but Effective Sequence Model
JiqunChu JiqunChu | Zuoquan Lin

Modeling long-range dependencies in sequential data is a crucial step in sequence learning. A recently developed model, the Structured State Space (S4), demonstrated significant effectiveness in modeling long-range sequences. However, It is unclear whether the success of S4 can be attributed to its intricate parameterization and HiPPO initialization or simply due to State Space Models (SSMs). To further investigate the potential of the deep SSMs, we start with exponential smoothing (ETS), a simple SSM, and propose a stacked architecture by directly incorporating it into an element-wise MLP. We augment simple ETS with additional parameters and complex field to reduce the inductive bias. Despite increasing less than 1% of parameters of element-wise MLP, our models achieve comparable results to S4 on the LRA benchmark.

pdf bib
OpenFMNav: Towards Open-Set Zero-Shot Object Navigation via Vision-Language Foundation Models
Yuxuan Kuang | Hai Lin | Meng Jiang

Object navigation (ObjectNav) requires an agent to navigate through unseen environments to find queried objects. Many previous methods attempted to solve this task by relying on supervised or reinforcement learning, where they are trained on limited household datasets with close-set objects. However, two key challenges are unsolved: understanding free-form natural language instructions that demand open-set objects, and generalizing to new environments in a zero-shot manner. Aiming to solve the two challenges, in this paper, we propose **OpenFMNav**, an **Open**-set **F**oundation **M**odel based framework for zero-shot object **Nav**igation. We first unleash the reasoning abilities of large language models (LLMs) to extract proposed objects from natural language instructions that meet the user’s demand. We then leverage the generalizability of large vision language models (VLMs) to actively discover and detect candidate objects from the scene, building a *Versatile Semantic Score Map (VSSM)*. Then, by conducting common sense reasoning on *VSSM*, our method can perform effective language-guided exploration and exploitation of the scene and finally reach the goal. By leveraging the reasoning and generalizing abilities of foundation models, our method can understand free-form human instructions and perform effective open-set zero-shot navigation in diverse environments. Extensive experiments on the HM3D ObjectNav benchmark show that our method surpasses all the strong baselines on all metrics, proving our method’s effectiveness. Furthermore, we perform real robot demonstrations to validate our method’s open-set-ness and generalizability to real-world environments.

pdf bib
Comparing Two Model Designs for Clinical Note Generation; Is an LLM a Useful Evaluator of Consistency?
Nathan Brake | Thomas Schaaf

Following an interaction with a patient, physicians are responsible for the submission of clinical documentation, often organized as a SOAP note. A clinical note is not simply a summary of the conversation but requires the use of appropriate medical terminology. The relevant information can then be extracted and organized according to the structure of the SOAP note. In this paper we analyze two different approaches to generate the different sections of a SOAP note based on the audio recording of the conversation, and specifically examine them in terms of note consistency. The first approach generates the sections independently, while the second method generates them all together. In this work we make use of PEGASUS-X Transformer models and observe that both methods lead to similar ROUGE values (less than 1% difference) and have no difference in terms of the Factuality metric. We perform a human evaluation to measure aspects of consistency and demonstrate that LLMs like Llama2 can be used to perform the same tasks with roughly the same agreement as the human annotators. Between the Llama2 analysis and the human reviewers we observe a Cohen Kappa inter-rater reliability of 0.79, 1.00, and 0.32 for consistency of age, gender, and body part injury, respectively. With this we demonstrate the usefulness of leveraging an LLM to measure quality indicators that can be identified by humans but are not currently captured by automatic metrics. This allows scaling evaluation to larger data sets, and we find that clinical note consistency improves by generating each new section conditioned on the output of all previously generated sections.

pdf bib
VOLTA: Improving Generative Diversity by Variational Mutual Information Maximizing Autoencoder
Yueen Ma | DaFeng Chi | Jingjing Li | Kai Song | Yuzheng Zhuang | Irwin King

The natural language generation domain has witnessed great success thanks to Transformer models. Although they have achieved state-of-the-art generative quality, they often neglect generative diversity. Prior attempts to tackle this issue suffer from either low model capacity or over-complicated architectures. Some recent methods employ the VAE framework to enhance diversity, but their latent variables fully depend on the input context, restricting exploration of the latent space. In this paper, we introduce VOLTA, a framework that elevates generative diversity by bridging Transformer with VAE via a more effective cross-attention-based connection, departing from conventional embedding concatenation or summation. Additionally, we propose integrating InfoGAN-style latent codes to enable input-independent variability, further diversifying the generation. Moreover, our framework accommodates discrete inputs alongside its existing support for continuous inputs. We perform comprehensive experiments with two types of Transformers on six datasets from three different NLG tasks to show that our approach can significantly improve generative diversity while maintaining generative quality.

pdf bib
EcoSpeak: Cost-Efficient Bias Mitigation for Partially Cross-Lingual Speaker Verification
Divya Sharma

Linguistic bias is a critical problem concerning the diversity, equity, and inclusiveness of Natural Language Processing tools. The severity of this problem intensifies in security systems, such as speaker verification, where fairness is paramount. Speaker verification systems are biometric systems that determine whether two speech recordings are of the same speaker. Such user-centric systems should be inclusive to bilingual speakers. However, Deep neural network models are linguistically biased. Linguistic bias can be full or partial. Partially cross-lingual bias occurs when one test trial pair recording is in the training set’s language, and the other is in an unseen target language. Such linguistic mismatch influences the speaker verification model’s decision, dissuading bilingual speakers from using the system. Domain adaptation can mitigate this problem. However, adapting to each existing language is expensive. This paper explores cost-efficient bias mitigation techniques for partially cross-lingual speaker verification. We study the behavior of five baselines in five partially cross-lingual scenarios. Using our baseline behavioral insights, we propose EcoSpeak, a low-cost solution to partially cross-lingual speaker verification. EcoSpeak incorporates contrastive linguistic (CL) attention. CL attention utilizes linguistic differences in trial pairs to emphasize relevant speaker verification embedding parts. Experimental results demonstrate EcoSpeak’s robustness to partially cross-lingual testing.

pdf bib
Leveraging Contextual Information for Effective Entity Salience Detection
Rajarshi Bhowmik | Marco Ponza | Atharva Tendle | Anant Gupta | Rebecca Jiang | Xingyu Lu | Qian Zhao | Daniel Preotiuc-Pietro

In text documents such as news articles, the content and key events usually revolve around a subset of all the entities mentioned in a document. These entities, often deemed as salient entities, provide useful cues of the aboutness of a document to a reader. Identifying the salience of entities was found helpful in several downstream applications such as search, ranking, and entity-centric summarization, among others. Prior work on salient entity detection mainly focused on machine learning models that require heavy feature engineering. We show that fine-tuning medium-sized language models with a cross-encoder style architecture yields substantial performance gains over feature engineering approaches. To this end, we conduct a comprehensive benchmarking of four publicly available datasets using models representative of the medium-sized pre-trained language model family. Additionally, we show that zero-shot prompting of instruction-tuned language models yields inferior results, indicating the task’s uniqueness and complexity.

pdf bib
LLM-as-a-Coauthor: Can Mixed Human-Written and Machine-Generated Text Be Detected?
Qihui Zhang | Chujie Gao | Dongping Chen | Yue Huang | Yixin Huang | Zhenyang Sun | Shilin Zhang | Weiye Li | Zhengyan Fu | Yao Wan | Lichao Sun

With the rapid development and widespread application of Large Language Models (LLMs), the use of Machine-Generated Text (MGT) has become increasingly common, bringing with it potential risks, especially in terms of quality and integrity in fields like news, education, and science. Current research mainly focuses on purely MGT detection, without adequately addressing mixed scenarios including AI-revised Human-Written Text (HWT) or human-revised MGT. To tackle this challenge, we define mixtext, a form of mixed text involving both AI and human-generated content. Then we introduce MixSet, the first dataset dedicated to studying these mixtext scenarios. Leveraging MixSet, we executed comprehensive experiments to assess the efficacy of prevalent MGT detectors in handling mixtext situations, evaluating their performance in terms of effectiveness, robustness, and generalization. Our findings reveal that existing detectors struggle to identify mixtext, particularly in dealing with subtle modifications and style adaptability. This research underscores the urgent need for more fine-grain detectors tailored for mixtext, offering valuable insights for future research. Code and Models are available at https://github.com/Dongping-Chen/MixSet.

pdf bib
A (More) Realistic Evaluation Setup for Generalisation of Community Models on Malicious Content Detection
Ivo Verhoeven | Pushkar Mishra | Rahel Beloch | Helen Yannakoudakis | Ekaterina Shutova

Community models for malicious content detection, which take into account the context from a social graph alongside the content itself, have shown remarkable performance on benchmark datasets. Yet, misinformation and hate speech continue to propagate on social media networks. This mismatch can be partially attributed to the limitations of current evaluation setups that neglect the rapid evolution of online content and the underlying social graph. In this paper, we propose a novel evaluation setup for model generalisation based on our few-shot subgraph sampling approach. This setup tests for generalisation through few labelled examples in local explorations of a larger graph, emulating more realistic application settings. We show this to be a challenging inductive setup, wherein strong performance on the training graph is not indicative of performance on unseen tasks, domains, or graph structures. Lastly, we show that graph meta-learners trained with our proposed few-shot subgraph sampling outperform standard community models in the inductive setup.

pdf bib
Citation: A Key to Building Responsible and Accountable Large Language Models
Jie Huang | Kevin Chang

Large Language Models (LLMs) bring transformative benefits alongside unique challenges, including intellectual property (IP) and ethical concerns. This position paper explores a novel angle to mitigate these risks, drawing parallels between LLMs and established web systems. We identify “citation”—the acknowledgement or reference to a source or evidence—as a crucial yet missing component in LLMs. Incorporating citation could enhance content transparency and verifiability, thereby confronting the IP and ethical issues in the deployment of LLMs. We further propose that a comprehensive citation mechanism for LLMs should account for both non-parametric and parametric content. Despite the complexity of implementing such a citation mechanism, along with the potential pitfalls, we advocate for its development. Building on this foundation, we outline several research problems in this area, aiming to guide future explorations towards building more responsible and accountable LLMs.

pdf bib
Graph-Induced Syntactic-Semantic Spaces in Transformer-Based Variational AutoEncoders
Yingji Zhang | Marco Valentino | Danilo Carvalho | Ian Pratt-Hartmann | Andre Freitas

The injection of syntactic information in Variational AutoEncoders (VAEs) can result in an overall improvement of performances and generalisation. An effective strategy to achieve such a goal is to separate the encoding of distributional semantic features and syntactic structures into heterogeneous latent spaces via multi-task learning or dual encoder architectures. However, existing works employing such techniques are limited to LSTM-based VAEs. This work investigates latent space separation methods for structural syntactic injection in Transformer-based VAE architectures (i.e., Optimus) through the integration of graph-based models. Our empirical evaluation reveals that the proposed end-to-end VAE architecture can improve theoverall organisation of the latent space, alleviating the information loss occurring in standard VAE setups, and resulting in enhanced performances on language modelling and downstream generation tasks.

pdf bib
Narrowing the Gap between Zero- and Few-shot Machine Translation by Matching Styles
Weiting Tan | Haoran Xu | Lingfeng Shen | Shuyue Stella Li | Kenton Murray | Philipp Koehn | Benjamin Van Durme | Yunmo Chen

Large language models trained primarily in a monolingual setting have demonstrated their ability to generalize to machine translation using zero- and few-shot examples with in-context learning. However, even though zero-shot translations are relatively good, there remains a discernible gap comparing their performance with the few-shot setting. In this paper, we investigate the factors contributing to this gap and find that this gap can largely be closed (for about 70%) by matching the writing styles of the target corpus. Additionally, we explore potential approaches to enhance zero-shot baselines without the need for parallel demonstration examples, providing valuable insights into how these methods contribute to improving translation metrics.

pdf bib
Which Modality should I use - Text, Motif, or Image? : Understanding Graphs with Large Language Models
Debarati Das | Ishaan Gupta | Jaideep Srivastava | Dongyeop Kang

Our research integrates graph data with Large Language Models (LLMs), which, despite their advancements in various fields using large text corpora, face limitations in encoding entire graphs due to context size constraints. This paper introduces a new approach to encoding a graph with diverse modalities, such as text, image, and motif, coupled with prompts to approximate a graph’s global connectivity, thereby enhancing LLMs’ efficiency in processing complex graph structures. The study also presents GraphTMI, a novel benchmark for evaluating LLMs in graph structure analysis, focusing on homophily, motif presence, and graph difficulty. Key findings indicate that the image modality, especially with vision-language models like GPT-4V, is superior to text in balancing token limits and preserving essential information and comes close to prior graph neural net (GNN) encoders. Furthermore, the research assesses how various factors affect the performance of each encoding modality and outlines the existing challenges and potential future developments for LLMs in graph understanding and reasoning tasks. Our code and data are publicly available on our project page - https://minnesotanlp.github.io/GraphLLM/

pdf bib
On-the-Fly Fusion of Large Language Models and Machine Translation
Hieu Hoang | Huda Khayrallah | Marcin Junczys-Dowmunt

We propose on-the-fly ensembling of a neural machine translation (NMT) model with a large language model (LLM), prompted on the same task and input. Through experiments on 4 language directions with varying data amounts, we find that a slightly weaker-at-translation LLM can improve translations of a NMT model, and such an ensemble can produce better translations than ensembling two stronger NMT models.We demonstrate that our ensemble method can be combined with various techniques from LLM prompting, such as in context learning and translation context.

pdf bib
READ: Improving Relation Extraction from an ADversarial Perspective
Dawei Li | William Hogan | Jingbo Shang

Recent works in relation extraction (RE) have achieved promising benchmark accuracy; however, our adversarial attack experiments show that these works excessively rely on entities, making their generalization capability questionable. To address this issue, we propose an adversarial training method specifically designed for RE. Our approach introduces both sequence- and token-level perturbations to the sample and uses a separate perturbation vocabulary to improve the search for entity and context perturbations.Furthermore, we introduce a probabilistic strategy for leaving clean tokens in the context during adversarial training. This strategy enables a larger attack budget for entities and coaxes the model to leverage relational patterns embedded in the context. Extensive experiments show that compared to various adversarial training methods, our method significantly improves both the accuracy and robustness of the model. Additionally, experiments on different data availability settings highlight the effectiveness of our method in low-resource scenarios.We also perform in-depth analyses of our proposed method and provide further hints.We will release our code at https://github.com/David-Li0406/READ.

pdf bib
REQUAL-LM: Reliability and Equity through Aggregation in Large Language Models
Sana Ebrahimi | Nima Shahbazi | Abolfazl Asudeh

The extensive scope of large language models (LLMs) across various domains underscores the critical importance of responsibility in their application, beyond natural language processing. In particular, the randomized nature of LLMs, coupled with inherent biases and historical stereotypes in data, raises critical concerns regarding reliability and equity. Addressing these challenges are necessary before using LLMs for applications with societal impact. Towards addressing this gap, we introduce REQUAL-LM, a novel method for finding reliable and equitable LLM outputs through aggregation. Specifically, we develop a Montecarlo method based on repeated sampling to find a reliable output close to the mean of the underlying distribution of possible outputs. We formally define the terms such as reliability and bias, and design an equity-aware aggregation to minimize harmful bias while finding a highly reliable output. REQUAL-LM does not require specialized hardware, does not impose a significant computing load, and uses LLMs as a blackbox. This design choice enables seamless scalability alongside the rapid advancement of LLM technologies. Our system does not require retraining the LLMs, which makes it deployment ready and easy to adapt. Our comprehensive experiments using various tasks and datasets demonstrate that REQUAL-LM effectively mitigates bias and selects a more equitable response, specifically the outputs that properly represents minority groups.

pdf bib
Addressing Both Statistical and Causal Gender Fairness in NLP Models
Hannah Chen | Yangfeng Ji | David Evans

Statistical fairness stipulates equivalent outcomes for every protected group, whereas causal fairness prescribes that a model makes the same prediction for an individual regardless of their protected characteristics. Counterfactual data augmentation (CDA) is effective for reducing bias in NLP models, yet models trained with CDA are often evaluated only on metrics that are closely tied to the causal fairness notion; similarly, sampling-based methods designed to promote statistical fairness are rarely evaluated for causal fairness. In this work, we evaluate both statistical and causal debiasing methods for gender bias in NLP models, and find that while such methods are effective at reducing bias as measured by the targeted metric, they do not necessarily improve results on other bias metrics. We demonstrate that combinations of statistical and causal debiasing techniques are able to reduce bias measured through both types of metrics.

pdf bib
LLM-Rec: Personalized Recommendation via Prompting Large Language Models
Hanjia Lyu | Song Jiang | Hanqing Zeng | Yinglong Xia | Qifan Wang | Si Zhang | Ren Chen | Chris Leung | Jiajie Tang | Jiebo Luo

Text-based recommendation holds a wide range of practical applications due to its versatility, as textual descriptions can represent nearly any type of item. However, directly employing the original item descriptions may not yield optimal recommendation performance due to the lack of comprehensive information to align with user preferences. Recent advances in large language models (LLMs) have showcased their remarkable ability to harness commonsense knowledge and reasoning. In this study, we introduce a novel approach, coined LLM-Rec, which incorporates four distinct prompting strategies of text enrichment for improving personalized text-based recommendations. Our empirical experiments reveal that using LLM-augmented text significantly enhances recommendation quality. Even basic MLP (Multi-Layer Perceptron) models achieve comparable or even better results than complex content-based methods. Notably, the success of LLM-Rec lies in its prompting strategies, which effectively tap into the language model’s comprehension of both general and specific item characteristics. This highlights the importance of employing diverse prompts and input augmentation techniques to boost the recommendation effectiveness of LLMs.

pdf bib
A Robust Semantics-based Watermark for Large Language Model against Paraphrasing
Jie Ren | Han Xu | Yiding Liu | Yingqian Cui | Shuaiqiang Wang | Dawei Yin | Jiliang Tang

Large language models (LLMs) have show their remarkable ability in various natural language tasks. However, there are concerns that LLMs are possible to be used improperly or even illegally. To prevent the malicious usage of LLMs, detecting LLM-generated text becomes crucial in the deployment of LLM applications. Watermarking is an effective strategy to detect the LLM-generated content by encoding a pre-defined secret watermark to facilitate the detection process. However, the majority of existing watermark methods leverage the simple hashes of precedent tokens to partition vocabulary. Such watermarks can be easily eliminated by paraphrase and, correspondingly, the detection effectiveness will be greatly compromised. Thus, to enhance the robustness against paraphrase, we propose a semantics-based watermark framework, SemaMark. It leverages the semantics as an alternative to simple hashes of tokens since the semantic meaning of the sentences will be likely preserved under paraphrase and the watermark can remain robust. Comprehensive experiments are conducted to demonstrate the effectiveness and robustness of SemaMark under different paraphrases.

pdf bib
Solving Data-centric Tasks using Large Language Models
Shraddha Barke | Christian Poelitz | Carina Negreanu | Benjamin Zorn | José Cambronero | Andrew Gordon | Vu Le | Elnaz Nouri | Nadia Polikarpova | Advait Sarkar | Brian Slininger | Neil Toronto | Jack Williams

Large language models are rapidly replacing help forums like StackOverflow, and are especially helpful to non-professional programmers and end users. These users are often interested in data-centric tasks, like spreadsheet manipulation and data wrangling, which are hard to solve if the intent is only communicated using a natural-language description, without including data. But how do we decide how much data and which data to include in the prompt?This paper makes two contributions towards answering this question. First, we create a dataset of real-world NL-to-code tasks manipulating tabular data, mined from StackOverflow posts. Second, we introduce a novel cluster-then-select prompting technique, which adds the most representative rows from the input data to the LLM prompt. Our experiments show that LLM performance is indeed sensitive to the amount of data passed in the prompt, and that for tasks with a lot of syntactic variation in the input table,our cluster-then-select technique outperforms a random selection baseline.

pdf bib
A Novel Paradigm Boosting Translation Capabilities of Large Language Models
Jiaxin Guo | Hao Yang | Zongyao Li | Daimeng Wei | Hengchao Shang | Xiaoyu Chen

This paper presents a study on strategies to enhance the translation capabilities of large language models (LLMs) in the context of machine translation (MT) tasks. The paper proposes a novel paradigm consisting of three stages: Secondary Pre-training using Extensive Monolingual Data, Continual Pre-training with Interlinear Text Format Documents, and Leveraging Source-Language Consistent Instruction for Supervised Fine-Tuning. Previous research on LLMs focused on various strategies for supervised fine-tuning (SFT), but their effectiveness has been limited. While traditional machine translation approaches rely on vast amounts of parallel bilingual data, our paradigm highlights the importance of using smaller sets of high-quality bilingual data. We argue that the focus should be on augmenting LLMs’ cross-lingual alignment abilities during pre-training rather than solely relying on extensive bilingual data during SFT. Experimental results conducted using the Llama2(CITATION)model, particularly on Chinese-Llama2(CITATION) after monolingual augmentation, demonstrate the improved translation capabilities of LLMs. A significant contribution of our approach lies in Stage2: Continual Pre-training with Interlinear Text Format Documents, which requires less than 1B training data, making our method highly efficient. Additionally, in Stage3, we observed that setting instructions consistent with the source language benefits the supervised fine-tuning process. Experimental results demonstrate that our approach surpasses previous work and achieves superior performance compared to models such as NLLB-54B(CITATION) and GPT3.5-text-davinci-003, despite having a significantly smaller parameter count of only 7B or 13B. This achievement establishes our method as a pioneering strategy in the field of machine translation.

pdf bib
Measuring Social Norms of Large Language Models
Ye Yuan | Kexin Tang | Jianhao Shen | Ming Zhang | Chenguang Wang

We present a new challenge to examine whether large language models understand social norms. In contrast to existing datasets, our dataset requires a fundamental understanding of social norms to solve. Our dataset features the largest set of social norm skills, consisting of 402 skills and 12,383 questions covering a wide set of social norms ranging from opinions and arguments to culture and laws. We design our dataset according to the K-12 curriculum. This enables the direct comparison of the social understanding of large language models to humans, more specifically, elementary students. While prior work generates nearly random accuracy on our benchmark, recent large language models such as GPT3.5-Turbo and LLaMA2-Chat are able to improve the performance significantly, only slightly below human performance. We then propose a multi-agent framework based on large language models to improve the models’ ability to understand social norms. This method further improves large language models to be on par with humans. Given the increasing adoption of large language models in real-world applications, our finding is particularly important and presents a unique direction for future improvements.

pdf bib
Source-Free Unsupervised Domain Adaptation for Question Answering via Prompt-Assisted Self-learning
Maxwell Yin | Boyu Wang | Charles Ling

This work addresses source-free domain adaptation (SFDA) for Question Answering (QA), wherein a model trained on a source domain is adapted to unlabeled target domains without additional source data. Existing SFDA methods only focus on the adaptation phase, overlooking the impact of source domain training on model generalizability. In this paper, we argue that source model training itself is also critical for improving the adaptation performance and stability. To this end, we investigate the role of prompt learning as an effective method to internalize domain-agnostic QA knowledge, which can be integrated into source training. After source training, an interactive self-learning strategy is proposed to further fine tune both model and prompt in the model adaptation phase. This leads to the Prompt-Assisted Self-Adaptive Learning (PASAL), an innovative SFDA approach for QA. Empirical evaluation on four benchmark datasets shows that PASAL surpasses existing methods in managing domain gaps and demonstrates greater stability across various target domains, validating the significance of source domain training for effective domain adaptation.

pdf bib
Hierarchical Attention Graph for Scientific Document Summarization in Global and Local Level
Chenlong Zhao | Xiwen Zhou | Xiaopeng Xie | Yong Zhang

Scientific document summarization has been a challenging task due to the long structure of the input text. The long input hinders the simultaneous effective modeling of both global high-order relations between sentences and local intra-sentence relations which is the most critical step in extractive summarization. However, existing methods mostly focus on one type of relation, neglecting the simultaneous effective modeling of both relations, which can lead to insufficient learning of semantic representations. In this paper, we propose HAESum, a novel approach utilizing graph neural networks to locally and globally model documents based on their hierarchical discourse structure. First, intra-sentence relations are learned using a local heterogeneous graph. Subsequently, a novel hypergraph self-attention layer is introduced to further enhance the characterization of high-order inter-sentence relations. We validate our approach on two benchmark datasets, and the experimental results demonstrate the effectiveness of HAESum and the importance of considering hierarchical structures in modeling long scientific documents.

pdf bib
LEEETs-Dial: Linguistic Entrainment in End-to-End Task-oriented Dialogue systems
Nalin Kumar | Ondrej Dusek

Linguistic entrainment, or alignment, represents a phenomenon where linguistic patterns employed by conversational participants converge to one another. While entrainment has been shown to produce a more natural user experience, most dialogue systems do not have any provisions for it. In this work, we introduce methods for achieving dialogue entrainment in a GPT-2-based end-to-end task-oriented dialogue system through the utilization of shared vocabulary. We experiment with training instance weighting, entrainment-specific loss, and additional conditioning to generate responses that align with the user. We demonstrate that all three approaches produce significantly better entrainment than the base, non-entrainment-optimized model, as confirmed by both automated and manual evaluation metrics.

pdf bib
Efficient Dependency Tree Sampling Without Replacement
Bogdan Dobre

In the context of computational models of dependency syntax, most dependency treebanks have the restriction that any valid dependency tree must have exactly one edge coming out of the root node in addition to respecting the spanning tree constraints. Many algorithms for dependency tree sampling were recently proposed, both for sampling with and without replacement.In this paper we propose a new algorithm called Wilson Reject SWOR for the case of sampling without replacement by adapting the Wilson Reject algorithm originally created for sampling with replacement and combining it with a Trie data structure. Experimental results indicate the efficiency of our approach in the scenario of sampling without replacement from dependency graphs with random weights.

pdf bib
Towards Better Generalization in Open-Domain Question Answering by Mitigating Context Memorization
Zixuan Zhang | Revanth Gangi Reddy | Kevin Small | Tong Zhang | Heng Ji

Open-domain Question Answering (OpenQA) aims at answering factual questions with an external large-scale knowledge corpus. However, real-world knowledge is not static; it updates and evolves continually. Such a dynamic characteristic of knowledge poses a vital challenge for these models, as the trained models need to constantly adapt to the latest information to make sure that the answers remain accurate. In addition, it is still unclear how well an OpenQA model can transfer to completely new knowledge domains. In this paper, we investigate the generalization performance of a retrieval-augmented QA model in two specific scenarios: 1) adapting to updated versions of the same knowledge corpus; 2) switching to completely different knowledge domains. We observe that the generalization challenges of OpenQA models stem from the reader’s over-reliance on memorizing the knowledge from the external corpus, which hinders the model from generalizing to a new knowledge corpus. We introduce Corpus-Invariant Tuning (CIT), a simple but effective training strategy, to mitigate the knowledge over-memorization by controlling the likelihood of retrieved contexts during training. Extensive experimental results on multiple OpenQA benchmarks show that CIT achieves significantly better generalizability without compromising the model’s performance in its original corpus and domain.

pdf bib
GEE! Grammar Error Explanation with Large Language Models
Yixiao Song | Kalpesh Krishna | Rajesh Bhatt | Kevin Gimpel | Mohit Iyyer

Existing grammatical error correction tools do not provide natural language explanations of the errors that they correct in user-written text. However, such explanations are essential for helping users learn the language by gaining a deeper understanding of its grammatical rules (DeKeyser, 2003; Ellis et al., 2006).To address this gap, we propose the task of grammar error explanation, where a system needs to provide one-sentence explanations for each grammatical error in a pair of erroneous and corrected sentences. The task is not easily solved by prompting LLMs: we find that, using one-shot prompting, GPT-4 only explains 40.6% of the errors and does not even attempt to explain 39.8% of the errors.Since LLMs struggle to identify grammar errors, we develop a two-step pipeline that leverages fine-tuned and prompted large language models to perform structured atomic token edit extraction, followed by prompting GPT-4 to explain each edit. We evaluate our pipeline on German, Chinese, and English grammar error correction data. Our atomic edit extraction achieves an F1 of 0.93 on German, 0.91 on Chinese, and 0.891 on English. Human evaluation of generated explanations reveals that 93.9% of German errors, 96.4% of Chinese errors, and 92.20% of English errors are correctly detected and explained. To encourage further research, we open-source our data and code.

pdf bib
AdaRefiner: Refining Decisions of Language Models with Adaptive Feedback
Wanpeng Zhang | Zongqing Lu

Large Language Models (LLMs) have demonstrated significant success across various domains. However, their application in complex decision-making tasks frequently necessitates intricate prompt engineering or fine-tuning, leading to challenges in unseen downstream tasks and heavy demands on computational resources. Meanwhile, Reinforcement Learning (RL) has been recognized as effective in decision-making problems but struggles in environments with sparse rewards, such as open-world games. To overcome these challenges, we introduce AdaRefiner, a novel framework designed to enhance the synergy between LLMs and RL feedback. The key component of AdaRefiner is a lightweight Adapter Language Model (LM), which automatically refines task comprehension based on feedback from RL agents. This method mitigates the need for intricate prompt engineering and intensive LLM fine-tuning while maintaining the LLMs’ generalization abilities and enhancing their decision-making capabilities in downstream tasks. Empirical evaluations of AdaRefiner on 22 diverse tasks within the open-world game Crafter have demonstrated its superior effectiveness, especially in guiding agents towards higher-level and common-sense skills. Our work makes contributions to the automatic self-refinement of LLMs with RL feedback, offering a more adaptable and efficient solution for complex decision-making problems. The code is available at https://github.com/PKU-RL/AdaRefiner.

pdf bib
DivTOD: Unleashing the Power of LLMs for Diversifying Task-Oriented Dialogue Representations
Weihao Zeng | Dayuan Fu | Keqing He | Yejie Wang | Yukai Xu | Weiran Xu

Language models pre-trained on general text have achieved impressive results in diverse fields. Yet, the distinct linguistic characteristics of task-oriented dialogues (TOD) compared to general text limit the practical utility of existing language models. Current task-oriented dialogue pre-training methods overlook the one-to-many property of conversations, where multiple responses can be appropriate given the same conversation context.In this paper, we propose a novel dialogue pre-training model called DivTOD, which collaborates with LLMs to learn diverse task-oriented dialogue representations. DivTOD guides LLMs in transferring diverse knowledge to smaller models while removing domain knowledge that contradicts task-oriented dialogues. Experiments show that our model outperforms strong TOD baselines on various downstream dialogue tasks and learns the intrinsic diversity of task-oriented dialogues.

pdf bib
Teaching a Multilingual Large Language Model to Understand Multilingual Speech via Multi-Instructional Training
Pavel Denisov | Thang Vu

Recent advancements in language modeling have led to the emergenceof Large Language Models (LLMs) capable ofvarious natural language processing tasks.Despite their success in text-based tasks, applying LLMs to the speech domainremains limited and challenging. This paper presents BLOOMZMMS, a novel modelthat integrates a multilingual LLM with a multilingual speech encoder,aiming to harness the capabilities of LLMs for speech recognition and beyond.Utilizing a multi-instructional training approach, we demonstrate the transferabilityof linguistic knowledge from the text to the speech modality.Our experiments, conducted on 1900 hours of transcribed data from 139 languages,establish that a multilingual speech representation can be effectivelylearned and aligned with a multilingual LLM. While this learned representationinitially shows limitations in task generalization, we address this issue bygenerating synthetic targets in a multi-instructional style.Our zero-shot evaluation results confirm the robustness of our approach acrossmultiple tasks, including speech translation and multilingual spoken languageunderstanding, thereby opening new avenues for applying LLMs in the speech domain.

pdf bib
CLEAN–EVAL: Clean Evaluation on Contaminated Large Language Models
Wenhong Zhu | Hongkun Hao | Zhiwei He | Yun-Ze Song | Jiao Yueyang | Yumeng Zhang | Hanxu Hu | Yiran Wei | Rui Wang | Hongyuan Lu

We are currently in an era of fierce competition among various large language models (LLMs), continuously pushing the boundaries of benchmark performance. However, genuinely assessing the capabilities of these LLMs has become a challenging and critical issue due to potential data contamination. In this paper, we propose a novel and valuable method, Clean-Eval, which mitigates the issue of data contamination and evaluates the LLMs more cleanly. Clean-Eval employs a neural-based model to paraphrase and back-translate the contaminated data into a candidate set, generating expressions with the same meaning but in different surface forms. A semantic detector is then used to filter those generated low-quality samples to narrow down this candidate set. Candidates with moderate BLEURT scores against the original samples are selected as the final evaluation set. According to human assessment, this set is almost semantically equivalent to the original contamination set but expressed differently. We conduct experiments on 20 existing benchmarks across diverse tasks, and results demonstrate that Clean-Eval substantially restores the actual evaluation results on contaminated LLMs under both few-shot learning and fine-tuning scenarios.

pdf bib
R-BASS : Relevance-aided Block-wise Adaptation for Speech Summarization
Roshan Sharma | Ruchira Sharma | Hira Dhamyal | Rita Singh | Bhiksha Raj

End-to-end speech summarization on long recordings is challenging because of the high computational cost. Block-wise Adaptation for Speech Summarization (BASS) summarizes arbitrarily long sequences by sequentially processing abutting chunks of audio. Despite the benefits of BASS, it has higher compute time due to sequential processing of all blocks, regardless of whether they are relevant to the final summary. In this paper, we propose R-BASS, a new relevance-aware block-wise adaptation method. First, we introduce two approaches to automatically estimate block relevance based on lexical and semantic similarity between the block-level transcript and the summary. Experiments on the How2 dataset show that using ground truth relevance during inference improves efficiency by 63.9 % by dropping irrelevant blocks. Finally, we incorporate relevance scores into training using a novel relevance loss and relevance predictor, and the proposed R-BASS model makes it possible to drop 86.3 % of the blocks while retaining comparable performance, resulting in a 2.2x speedup over BASS.

pdf bib
OVM, Outcome-supervised Value Models for Planning in Mathematical Reasoning
Fei Yu | Anningzhe Gao | Benyou Wang

Large language models (LLMs) often struggle with maintaining accuracy throughout multiple multiple reasoning steps, especially in mathematical reasoning where an error in earlier steps can propagate to subsequent ones and it ultimately leading to an incorrect answer.To reduce error propagation, guided decoding is employed to direct the LM decoding on a step-by-step basis. We argue that in guided decoding, assessing the potential of an incomplete reasoning path can be more advantageous than simply ensuring per-step correctness, as the former approach leads towards a correct final answer. This transforms the task into a value estimation problem in planning.Inspired by the findings that outcome supervision for guided decoding essentially acts as a value model, we propose Outcome-supervised Value Model (OVM) that employs outcome supervision for training a value model, which prioritizes steps that lead to accurate conclusions. Furthermore, the OVM eliminates the need for labor-intensive annotations of step-level correctness, thereby significantly enhancing its scalability. Our experiments on two multi-step mathematical reasoning datasets, GSM8K and Game of 24, demonstrate the superior performance of the OVM model. Notably, in GSM8K, our OVM-7B model achieves state-of-the-art results among LLMs up to 13B parameters; especially it does not utilize GPT-4 or code execution. These findings offer a novel perspective on the role of outcome supervision in training value models for multi-step reasoning tasks and provide theoretical justification for its advantage in value estimation for guided decoding.

pdf bib
The Whole is Better than the Sum: Using Aggregated Demonstrations in In-Context Learning for Sequential Recommendation
Lei Wang | Ee-Peng Lim

Large language models (LLMs) have shown excellent performance on various NLP tasks. To use LLMs as strong sequential recommenders, we explore the in-context learning approach to sequential recommendation. We investigate the effects of instruction format, task consistency, demonstration selection, and number of demonstrations. As increasing the number of demonstrations in ICL does not improve accuracy despite using a long prompt, we propose a novel method called LLMSRec-Syn that incorporates multiple demonstration users into one aggregated demonstration. Our experiments on three recommendation datasets show that LLMSRec-Syn outperforms state-of-the-art LLM-based sequential recommendation methods. In some cases, LLMSRec-Syn can perform on par with or even better than supervised learning methods. Our code is publicly available at https://github.com/demoleiwang/LLMSRec_Syn.

pdf bib
Bring Your Own KG: Self-Supervised Program Synthesis for Zero-Shot KGQA
Dhruv Agarwal | Rajarshi Das | Sopan Khosla | Rashmi Gangadharaiah

We present BYOKG, a universal question-answering (QA) system that can operate on any knowledge graph (KG), requires no human-annotated training data, and can be ready to use within a day—attributes that are out-of-scope for current KGQA systems. BYOKG draws inspiration from the remarkable ability of humans to comprehend information present in an unseen KG through exploration—starting at random nodes, inspecting the labels of adjacent nodes and edges, and combining them with their prior world knowledge. Exploration in BYOKG leverages an LLM-backed symbolic agent that generates a diverse set of query-program exemplars, which are then used to ground a retrieval-augmented reasoning procedure to synthesize programs for arbitrary questions. BYOKG is effective over both small- and large-scale graphs, showing dramatic gains in zero-shot QA accuracy of 27.89 and 59.88 F1 on GrailQA and MetaQA, respectively. We further find that performance of BYOKG reliably improves with continued exploration as well as improvements in the base LLM, notably outperforming a state-of-the-art fine-tuned model by 7.08 F1 on a sub-sampled zero-shot split of GrailQA. Lastly, we verify our universality claim by evaluating BYOKG on a domain-specific materials science KG and show that it improves zero-shot performance by 46.33 F1.

pdf bib
GraSAME: Injecting Token-Level Structural Information to Pretrained Language Models via Graph-guided Self-Attention Mechanism
Shuzhou Yuan | Michael Färber

Pretrained Language Models (PLMs) benefit from external knowledge stored in graph structures for various downstream tasks. However, bridging the modality gap between graph structures and text remains a significant challenge. Traditional methods like linearizing graphs for PLMs lose vital graph connectivity, whereas Graph Neural Networks (GNNs) require cumbersome processes for integration into PLMs. In this work, we propose a novel graph-guided self-attention mechanism, GraSAME. GraSAME seamlessly incorporates token-level structural information into PLMs without necessitating additional alignment or concatenation efforts. As an end-to-end, lightweight multimodal module, GraSAME follows a multi-task learning strategy and effectively bridges the gap between graph and textual modalities, facilitating dynamic interactions between GNNs and PLMs. Our experiments on the graph-to-text generation task demonstrate that GraSAME outperforms baseline models and achieves results comparable to state-of-the-art (SOTA) models on WebNLG datasets. Furthermore, compared to SOTA models, GraSAME eliminates the need for extra pre-training tasks to adjust graph inputs and reduces the number of trainable parameters by over 100 million.

pdf bib
Can Public Large Language Models Help Private Cross-device Federated Learning?
Boxin Wang | Yibo Zhang | Yuan Cao | Bo Li | Hugh McMahan | Sewoong Oh | Zheng Xu | Manzil Zaheer

We study (differentially) private federated learning (FL) of language models. The language models in cross-device FL are relatively small, which can be trained with meaningful formal user-level differential privacy (DP) guarantees when massive parallelism in training is enabled by the participation of a moderate size of users. Recently, public data has been used to improve privacy-utility trade-offs for both large and small language models. In this work, we provide a systematic study of using large-scale public data and LLMs to help differentially private training of on-device FL models, and further improve the privacy-utility tradeoff by techniques of distillation. Moreover, we propose a novel distribution matching algorithm with theoretical grounding to sample public data close to private data distribution, which significantly improves the sample efficiency of (pre-)training on public data. The proposed method is efficient and effective for training private models by taking advantage of public data, especially for customized on-device architectures that do not have ready-touse pre-trained models.

pdf bib
LangNav: Language as a Perceptual Representation for Navigation
Bowen Pan | Rameswar Panda | SouYoung Jin | Rogerio Feris | Aude Oliva | Phillip Isola | Yoon Kim

We explore the use of language as a perceptual representation for vision-and-language navigation (VLN), with a focus on low-data settings. Our approach uses off-the-shelf vision systems for image captioning and object detection to convert an agent’s egocentric panoramic view at each time step into natural language descriptions. We then finetune a pretrained language model to select an action, based on the current view and the trajectory history, that would best fulfill the navigation instructions. In contrast to the standard setup which adapts a pretrained language model to work directly with continuous visual features from pretrained vision models, our approach instead uses (discrete) language as the perceptual representation. We explore several use cases of our language-based navigation (LangNav) approach on the R2R VLN benchmark: generating synthetic trajectories from a prompted language model (GPT-4) with which to finetune a smaller language model; domain transfer where we transfer a policy learned on one simulated environment (ALFRED) to another (more realistic) environment (R2R); and combining both vision- and language-based representations for VLN. Our approach is found to improve upon baselines that rely on visual features in settings where only a few expert trajectories (10-100) are available, demonstrating the potential of language as a perceptual representation for navigation.

pdf bib
Planning and Editing What You Retrieve for Enhanced Tool Learning
Tenghao Huang | Dongwon Jung | Vaibhav Kumar | Mohammad Kachuee | Xiang Li | Puyang Xu | Muhao Chen

Recent advancements in integrating external tools with Large Language Models (LLMs) have opened new frontiers, with applications in mathematical reasoning, code generators, and smart assistants. However, existing methods, relying on simple one-time retrieval strategies, fall short on effectively and accurately shortlisting relevant tools. This paper introduces a novel PLUTO (Planning, Learning, and Understanding for TOols) approach, encompassing “Plan-and-Retrieve (P&R)” and “Edit-and-Ground (E&G)” paradigms. The P&R paradigm consists of a neural retrieval module for shortlisting relevant tools and an LLM-based query planner that decomposes complex queries into actionable tasks, enhancing the effectiveness of tool utilization. The E&G paradigm utilizes LLMs to enrich tool descriptions based on user scenarios, bridging the gap between user queries and tool functionalities. Experiment results demonstrate that these paradigms significantly improve the recall and NDCG in tool retrieval tasks, significantly surpassing current state-of-the-art models.

pdf bib
Chart-based Reasoning: Transferring Capabilities from LLMs to VLMs
Victor Carbune | Hassan Mansoor | Fangyu Liu | Rahul Aralikatte | Gilles Baechler | Jindong Chen | Abhanshu Sharma

Vision-language models (VLMs) are achieving increasingly strong performance on multimodal tasks. However, reasoning capabilities remain limited particularly for smaller VLMs, while those of large-language models (LLMs) have seen numerous improvements. We pro-pose a technique to transfer capabilities from LLMs to VLMs. On the recently introduced ChartQA, our method obtains state-of-the-artperformance when applied on the PaLI3-5B VLM by Chen et al. (2023c), while also enabling much better performance on PlotQA and FigureQA.We first improve the chart representation by continuing the pre-training stage using an improved version of the chart-to-table translation task by Liu et al. (2023a). We then propose constructing a 20x larger dataset than the original training set. To improve general reasoning capabilities and improve numerical operations, we synthesize reasoning traces using the table representation of charts. Lastly, our model is fine-tuned using the multitask loss introduced by Hsieh et al. (2023).Our variant ChartPaLI-5B outperforms even 10x larger models such as PaLIX-55B without using an upstream OCR system, while keeping inference time constant compared to the PaLI3-5B baseline. When rationales are further refined with a simple program-of-thought prompt (Chen et al., 2023a), our model outperforms the recently introduced Gemini Ultra and GPT-4V.

pdf bib
SLiM: Speculative Decoding with Hypothesis Reduction
Chi-Heng Lin | Shikhar Tuli | James Smith | Yen-Chang Hsu | Yilin Shen | Hongxia Jin

Speculative decoding has emerged as a prominent alternative to autoregressive decoding for expediting inference in large language models (LLMs). However, prevailing assumptions often focus solely on latency reduction, neglecting the computational expenses. In this paper, we present Speculate Less, validate More (SLiM), a speculative decoding enhancement to reduce the speculation set while validating more effective tokens. SLiM is designed to mitigate LLMs’ computation costs associated with the token verification by introducing hypothesis reduction based on a fast posterior estimation. It consistently surpasses counterparts lacking cost reduction across a spectrum from CPU to GPU. Our evaluation with diverse conversational datasets shows that SLiM can achieve a substantial 70% reduction in FLOPs while generating more effective predictions on top of prior arts.

pdf bib
REMATCH: Robust and Efficient Matching of Local Knowledge Graphs to Improve Structural and Semantic Similarity
Zoher Kachwala | Jisun An | Haewoon Kwak | Filippo Menczer

Knowledge graphs play a pivotal role in various applications, such as question-answering and fact-checking. Abstract Meaning Representation (AMR) represents text as knowledge graphs. Evaluating the quality of these graphs involves matching them structurally to each other and semantically to the source text. Existing AMR metrics are inefficient and struggle to capture semantic similarity. We also lack a systematic evaluation benchmark for assessing structural similarity between AMR graphs. To overcome these limitations, we introduce a novel AMR similarity metric, rematch, alongside a new evaluation for structural similarity called RARE. Among state-of-the-art metrics, rematch ranks second in structural similarity; and first in semantic similarity by 1–5 percentage points on the STS-B and SICK-R benchmarks. Rematch is also five times faster than the next most efficient metric.

pdf bib
Modeling the Sacred: Considerations when Using Religious Texts in Natural Language Processing
Ben Hutchinson

This position paper concerns the use of religious texts in Natural Language Processing (NLP), which is of special interest to the Ethics of NLP. Religious texts are expressions of culturally important values, and machine learned models have a propensity to reproduce cultural values encoded in their training data. Furthermore, translations of religious texts are frequently used by NLP researchers when language data is scarce. This repurposes the translations from their original uses and motivations, which often involve attracting new followers. This paper argues that NLP’s use of such texts raises considerations that go beyond model biases, including data provenance, cultural contexts, and their use in proselytism. We argue for more consideration of researcher positionality, and of the perspectives of marginalized linguistic and religious communities.

pdf bib
Testing the Effect of Code Documentation on Large Language Model Code Understanding
William Macke | Michael Doyle

Large Language Models (LLMs) have demonstrated impressive abilities in recent years with regards to code generation and understanding. However, little work has investigated how documentation and other code properties affect an LLM’s ability to understand and generate code or documentation. We present an empirical analysis of how underlying properties of code or documentation can affect an LLM’s capabilities. We show that providing an LLM with “incorrect” documentation can greatly hinder code understanding, while incomplete or missing documentation does not seem to significantly affect an LLM’s ability to understand code.

pdf bib
Aligning Large Language Models with Recommendation Knowledge
Yuwei Cao | Nikhil Mehta | Xinyang Yi | Raghunandan Hulikal Keshavan | Lukasz Heldt | Lichan Hong | Ed Chi | Maheswaran Sathiamoorthy

Large language models (LLMs) have recently been used as backbones for recommender systems. However, their performance often lags behind conventional methods in standard tasks like retrieval. We attribute this to a mismatch between LLMs’ knowledge and the knowledge crucial for effective recommendations. While LLMs excel at natural language reasoning, they cannot model complex user-item interactions inherent in recommendation tasks. We propose bridging the knowledge gap and equipping LLMs with recommendation-specific knowledge to address this. Operations such as Masked Item Modeling (MIM) and Bayesian Personalized Ranking (BPR) have found success in conventional recommender systems. Inspired by this, we simulate these operations through natural language to generate auxiliary-task data samples that encode item correlations and user preferences. Fine-tuning LLMs on such auxiliary-task data samples and incorporating more informative recommendation-task data samples facilitates the injection of recommendation-specific knowledge into LLMs. Extensive experiments across retrieval, ranking, and rating prediction tasks on LLMs such as FLAN-T5-Base and FLAN-T5-XL show the effectiveness of our technique in domains such as Amazon Toys & Games, Beauty, and Sports & Outdoors. Notably, our method outperforms conventional and LLM-based baselines, including the current SOTA, by significant margins in retrieval, showcasing its potential for enhancing recommendation quality.

pdf bib
OFA: A Framework of Initializing Unseen Subword Embeddings for Efficient Large-scale Multilingual Continued Pretraining
Yihong Liu | Peiqin Lin | Mingyang Wang | Hinrich Schuetze

Instead of pretraining multilingual language models from scratch, a more efficient method is to adapt existing pretrained language models (PLMs) to new languages via vocabulary extension and continued pretraining. However, this method usually randomly initializes the embeddings of new subwords and introduces substantially more embedding parameters to the model, thus weakening the efficiency. To address these issues, we propose a novel framework: One For All (OFA), which wisely initializes the embeddings of unseen subwords and thus can adapt a PLM to multiple languages efficiently and effectively. OFA takes advantage of external well-aligned multilingual static word vectors and injects the alignment knowledge into the subword embeddings. In addition, OFA applies matrix factorization and replaces the cumbersome embeddings with two lower-dimensional matrices, which largely reduces the number of parameters. We show OFA accelerates the convergence of continued pretraining, which is environmentally friendly as much fewer carbon footprints are generated. Through extensive experiments, we demonstrate OFA can achieve competitive or better performance than default continued pretraining baselines on a wide range of crosslingual downstream tasks. We make our code and models publicly available.

pdf bib
SELF-EXPERTISE: Knowledge-based Instruction Dataset Augmentation for a Legal Expert Language Model
Minju Kim | Haein Jung | Myoung-Wan Koo

The advent of instruction-tuned large language models (LLMs) has significantly advanced the field of automatic instruction dataset augmentation. However, the method of generating instructions and outputs from inherent knowledge of LLM can unintentionally produce hallucinations — instances of generating factually incorrect or misleading information. To overcome this, we propose SELF-EXPERTISE, automatically generating instruction dataset in the legal domain from a seed dataset. SELF-EXPERTISE extracts knowledge from the outputs of the seed dataset, and generates new instructions, inputs, and outputs. In this way, the proposed method reduces hallucination in automatic instruction augmentation. We trained an SELF-EXPERTISE augmented instruction dataset on the LLaMA-2 7B model to construct Korean legal specialized model, called LxPERT. LxPERT has demonstrated performance surpassing GPT-3.5-turbo in both in-domain and out-of-domain datasets. The SELF-EXPERTISE augmentation pipeline is not only applicable to the legal field but is also expected to be extendable to various domains, potentially advancing domain-specialized LLMs.

pdf bib
Re-evaluating the Need for Visual Signals in Unsupervised Grammar Induction
Boyi Li | Rodolfo Corona | Karttikeya Mangalam | Catherine Chen | Daniel Flaherty | Serge Belongie | Kilian Weinberger | Jitendra Malik | Trevor Darrell | Dan Klein

Are multimodal inputs necessary for grammar induction? Recent work has shown that multimodal training inputs can improve grammar induction. However, these improvements are based on comparisons to weak text-only baselines that were trained on relatively little textual data. To determine whether multimodal inputs are needed in regimes with large amounts of textual training data, we design a stronger text-only baseline, which we refer to as LC-PCFG. LC-PCFG is a C-PFCG that incorporates embeddings from text-only large language models (LLMs). We use a fixed grammar family to directly compare LC-PCFG to various multimodal grammar induction methods. We compare performance on four benchmark datasets. LC-PCFG provides an up to 17% relative improvement in Corpus-F1 compared to state-of-the-art multimodal grammar induction methods. LC-PCFG is also more computationally efficient, providing an up to 85% reduction in parameter count and 8.8× reduction in training time compared to multimodal approaches. These results suggest that multimodal inputs may not be necessary for grammar induction, and emphasize the importance of strong vision-free baselines for evaluating the benefit of multimodal approaches.

pdf bib
EDEntail: An Entailment-based Few-shot Text Classification with Extensional Definition
Zixiao Zhu | Junlang Qian | Zijian Feng | Hanzhang Zhou | Kezhi Mao

Few-shot text classification has seen significant advancements, particularly with entailment-based methods, which typically use either class labels or intensional definitions of class labels in hypotheses for label semantics expression. In this paper, we propose EDEntail, a method that employs extensional definition (EDef) of class labels in hypotheses, aiming to express the semantics of class labels more explicitly. To achieve the above goal, we develop an algorithm to gather and select extensional descriptive words of class labels and then order and format them into a sequence to form hypotheses. Our method has been evaluated and compared with state-of-the-art models on five classification datasets. The results demonstrate that our approach surpasses the supervised-learning methods and prompt-based methods under the few-shot setting, which underlines the potential of using an extensional definition of class labels for entailment-based few-shot text classification. Our code is available at https://github.com/MidiyaZhu/EDEntail.

pdf bib
What Makes Math Word Problems Challenging for LLMs?
Kv Aditya Srivatsa | Ekaterina Kochmar

This paper investigates the question of what makes math word problems (MWPs) in English challenging for large language models (LLMs). We conduct an in-depth analysis of the key linguistic and mathematical characteristics of MWPs. In addition, we train feature-based classifiers to better understand the impact of each feature on the overall difficulty of MWPs for prominent LLMs and investigate whether this helps predict how well LLMs fare against specific categories of MWPs.

pdf bib
SMILE: Multimodal Dataset for Understanding Laughter in Video with Language Models
Lee Hyun | Kim Sung-Bin | Seungju Han | Youngjae Yu | Tae-Hyun Oh

Despite the recent advances in artificial intelligence, building social intelligence remains a challenge.Among social signals, laughter is one of the distinctive expressions that occurs during social interactions between humans.In this work, we tackle a new challenge for machines to understand the rationale behind laughter in video, Video Laugh Reasoning.We introduce this new task to explain why people laugh in a particular video and a dataset for this task.Our proposed dataset, SMILE, comprises video clips and language descriptions of why people laugh. We propose a baseline by leveraging the reasoning capacity of large language models (LLMs) with textual video representation. Experiments show that our baseline can generate plausible explanations for laughter. We further investigate the scalability of our baseline by probing other video understanding tasks and in-the-wild videos. We release our dataset, code, and model checkpoints on https://github.com/postech-ami/SMILE-Dataset.

pdf bib
T3M: Text Guided 3D Human Motion Synthesis from Speech
Wenshuo Peng | Kaipeng Zhang | Sai Qian Zhang

Speech-driven 3D motion synthesis seeks to create lifelike animations based on human speech, with potential uses in virtual reality, gaming, and the film production. Existing approaches reply solely on speech audio for motion generation, leading to inaccurate and inflexible synthesis results. To mitigate this problem, we introduce a novel text-guided 3D human motion synthesis method, termed T3M. Unlike traditional approaches, T3M allows precise control over motion synthesis via textual input, enhancing the degree of diversity and user customization. The experiment results demonstrate that T3M can greatly outperform the state-of-the-art methods in both quantitative metrics and qualitative evaluations. We have publicly released our code at https://github.com/Gloria2tt/naacl2024.git

pdf bib
Deja vu: Contrastive Historical Modeling with Prefix-tuning for Temporal Knowledge Graph Reasoning
Miao Peng | Ben Liu | Wenjie Xu | Zihao Jiang | Jiahui Zhu | Min Peng

Temporal Knowledge Graph Reasoning (TKGR) is the task of inferring missing facts for incomplete TKGs in complex scenarios (e.g., transductive and inductive settings), which has been gaining increasing attention. Recently, to mitigate dependence on structured connections in TKGs, text-based methods have been developed to utilize rich linguistic information from entity descriptions. However, suffering from the enormous parameters and inflexibility of pre-trained language models, existing text-based methods struggle to balance the textual knowledge and temporal information with computationally expensive purpose-built training strategies. To tap the potential of text-based models for TKGR in various complex scenarios, we propose ChapTER, a Contrastive historical modeling framework with prefix-tuning for TEmporal Reasoning. ChapTER feeds history-contextualized text into the pseudo-Siamese encoders to strike a textual-temporal balance via contrastive estimation between queries and candidates. By introducing virtual time prefix tokens, it applies a prefix-based tuning method to facilitate the frozen PLM capable for TKGR tasks under different settings. We evaluate ChapTER on four transductive and three few-shot inductive TKGR benchmarks, and experimental results demonstrate that ChapTER achieves superior performance compared to competitive baselines with only 0.17% tuned parameters. We conduct thorough analysis to verify the effectiveness, flexibility and efficiency of ChapTER.

pdf bib
Explanation Extraction from Hierarchical Classification Frameworks for Long Legal Documents
Nishchal Prasad | Taoufiq Dkaki | Mohand Boughanem

Hierarchical classification frameworks have been widely used to process long sequences, especially in the legal domain for predictions from long legal documents. But being black-box models they are unable to explain their predictions making them less reliable for practical applications, more so in the legal domain. In this work, we develop an extractive explanation algorithm for hierarchical frameworks for long sequences based on the sensitivity of the trained model to its input perturbations. We perturb using occlusion and develop Ob-HEx; an Occlusion-based Hierarchical Explanation-extractor. We adapt Ob-HEx to Hierarchical Transformer models trained on long Indian legal texts. And use Ob-HEx to analyze them and extract their explanations for the ILDC-Expert dataset, achieving a minimum gain of 1 point over the previous benchmark on most of our performance evaluation metrics.

pdf bib
Low-Rank Adaptation for Multilingual Summarization: An Empirical Study
Chenxi Whitehouse | Fantine Huot | Jasmijn Bastings | Mostafa Dehghani | Chu-Cheng Lin | Mirella Lapata

Although the advancements of pre-trained Large Language Models have significantly accelerated recent progress in NLP, their ever-increasing size poses significant challenges for conventional fine-tuning, especially in memory-intensive tasks. We investigate the potential of Parameter-Efficient Fine-Tuning, focusing on Low-Rank Adaptation (LoRA), in the domain of multilingual summarization, a task that is both challenging (due to typically long inputs), and relatively unexplored. We conduct an extensive study across different data availability scenarios, including high- and low-data settings, and cross-lingual transfer, leveraging models of different sizes. Our findings reveal that LoRA is competitive with full fine-tuning when trained with high quantities of data, and excels in low-data scenarios and cross-lingual transfer. We also study different strategies for few-shot cross-lingual transfer, finding that continued LoRA tuning outperforms full fine-tuning and the dynamic composition of language-specific LoRA modules.

pdf bib
A Tree-of-Thoughts to Broaden Multi-step Reasoning across Languages
Leonardo Ranaldi | Giulia Pucci | Federico Ranaldi | Elena Sofia Ruzzetti | Fabio Massimo Zanzotto

Reasoning methods, best exemplified by the well-known Chain-of-Thought (CoT), empower the reasoning abilities of Large Language Models (LLMs) by eliciting them to solve complex tasks in a step-by-step manner. Although they are achieving significant success, the ability to deliver multi-step reasoning remains limited to English because of the imbalance in the distribution of pre-training data, which makes other languages a barrier. In this paper, we propose Cross-lingual Tree-of-Thoughts (Cross-ToT), a method for aligning Cross-lingual CoT reasoning across languages. The proposed method, through a self-consistent cross-lingual prompting mechanism inspired by the Tree-of-Thoughts approach, provides multi-step reasoning paths in different languages that, during the steps, lead to the final solution. Experimental evaluations show that our method significantly outperforms existing prompting methods by reducing the number of interactions and achieving state-of-the-art performance.

pdf bib
Emergent Abilities in Reduced-Scale Generative Language Models
Sherin Muckatira | Vijeta Deshpande | Vladislav Lialin | Anna Rumshisky

Large language models can solve new tasks without task-specific fine-tuning. This ability, also known as in-context learning (ICL), is considered an emergent ability and is primarily seen in large language models with billions of parameters. This study investigates if such emergent properties are strictly tied to model size or can be demonstrated by smaller models trained on reduced-scale data. To explore this, we simplify pre-training data and pre-train 36 causal language models with parameters varying from 1 million to 165 million parameters. We show that models trained on this simplified pre-training data demonstrate enhanced zero-shot capabilities across various tasks in simplified language, achieving performance comparable to that of pre-trained models six times larger on unrestricted language. This suggests that downscaling the language allows zero-shot learning capabilities to emerge in models with limited size.Additionally, we find that these smaller models pre-trained on simplified data demonstrate a power law relationship between the evaluation loss and the three scaling factors: compute, dataset size, and model size.

pdf bib
Context Does Matter: Implications for Crowdsourced Evaluation Labels in Task-Oriented Dialogue Systems
Clemencia Siro | Mohammad Aliannejadi | Maarten Rijke

Crowdsourced labels play a crucial role in evaluating task-oriented dialogue systems (TDSs). Obtaining high-quality and consistent ground-truth labels from annotators presents challenges. When evaluating a TDS, annotators must fully comprehend the dialogue before providing judgments. Previous studies suggest using only a portion of the dialogue context in the annotation process. However, the impact of this limitation on label quality remains unexplored. This study investigates the influence of dialogue context on annotation quality, considering the truncated context for relevance and usefulness labeling. We further propose to use large language models ( LLMs) to summarize the dialogue context to provide a rich and short description of the dialogue context and study the impact of doing so on the annotator’s performance. Reducing context leads to more positive ratings. Conversely, providing the entire dialogue context yields higher-quality relevance ratings but introduces ambiguity in usefulness ratings. Using the first user utterance as context leads to consistent ratings, akin to those obtained using the entire dialogue, with significantly reduced annotation effort. Our findings show how task design, particularly the availability of dialogue context, affects the quality and consistency of crowdsourced evaluation labels.

pdf bib
Matching Varying-Length Texts via Topic-Informed and Decoupled Sentence Embeddings
Xixi Zhou | Chunbin Gu | Xin Jie | Jiajun Bu | Haishuai Wang

Measuring semantic similarity between texts is a crucial task in natural language processing. While existing semantic text matching focuses on pairs of similar-length sequences, matching texts with non-comparable lengths has broader applications in specific domains, such as comparing professional document summaries and content. Current approaches struggle with text pairs of non-comparable lengths due to truncation issues. To address this, we split texts into natural sentences and decouple sentence representations using supervised contrastive learning (SCL). Meanwhile, we adopt the embedded topic model (ETM) for specific domain data. Our experiments demonstrate the effectiveness of our model, based on decoupled and topic-informed sentence embeddings, in matching texts of significantly different lengths across three well-studied datasets.

pdf bib
Instruction Tuning with Human Curriculum
Bruce W Lee | Hyunsoo Cho | Kang Min Yoo

In this work, we (1) introduce Curriculum Instruction Tuning, (2) explore the potential advantages of employing diverse curriculum strategies, and (3) delineate a synthetic instruction-response generation framework that complements our theoretical approach. Distinct from the existing instruction tuning dataset, our generation pipeline is systematically structured to emulate the sequential and orderly characteristic of human learning. Additionally, we describe a methodology for generating instruction-response datasets that extensively span the various stages of human education, from middle school through the graduate level, utilizing educational subject catalogs.Before training, we meticulously organize the instruction data to ensure that questions escalate in difficulty regarding (A) the subject matter and (B) the intricacy of the instructions. The findings of our study reveal that substantial improvements in performance can be achieved through the mere application of curriculum ordering to instruction data—achieving gains of +4.76 on TruthfulQA, +2.98 on MMLU, +2.8 on OpenbookQA, and +1.28 on ARC-hard—compared to random shuffling. This enhancement is achieved without incurring additional computational expenses. Through comprehensive experimentation, we observe that the advantages of our proposed method are consistently evident across nine benchmarks.

pdf bib
Natural Language-based State Representation in Deep Reinforcement Learning
Md Masudur Rahman | Yexiang Xue

This paper investigates the potential of using natural language descriptions as an alternative to direct image-based observations for learning policies in reinforcement learning. Due to the inherent challenges in managing image-based observations, which include abundant information and irrelevant features, we propose a method that compresses images into a natural language form for state representation. This approach allows better interpretability and leverages the processing capabilities of large-language models. We conducted several experiments involving tasks that required image-based observation. The results demonstrated that policies trained using natural language descriptions of images yield better generalization than those trained directly from images, emphasizing the potential of this approach in practical settings.

pdf bib
Learning Cross-Architecture Instruction Embeddings for Binary Code Analysis in Low-Resource Architectures
Junzhe Wang | Qiang Zeng | Lannan Luo

Binary code analysis is indispensable for a variety of software security tasks. Applying deep learning to binary code analysis has drawn great attention because of its notable performance. Today, source code is frequently compiled for various Instruction Set Architectures (ISAs). It is thus critical to expand binary analysis capabilities to multiple ISAs. Given a binary analysis task, the scale of available data on different ISAs varies. As a result, the rich datasets (e.g., malware) for certain ISAs, such as x86, lead to a disproportionate focus on these ISAs and a negligence of other ISAs, such as PowerPC, which suffer from the “data scarcity” problem. To address the problem, we propose to learn cross-architecture instruction embeddings (CAIE), where semantically-similar instructions, regardless of their ISAs, have close embeddings in a shared space. Consequently, we can transfer a model trained on a data-rich ISA to another ISA with less available data. We consider four ISAs (x86, ARM, MIPS, and PowerPC) and conduct both intrinsic and extrinsic evaluations (including malware detection and function similarity comparison). The results demonstrate the effectiveness of our approach to generate high-quality CAIE with good transferability.

pdf bib
ReEval: Automatic Hallucination Evaluation for Retrieval-Augmented Large Language Models via Transferable Adversarial Attacks
Xiaodong Yu | Hao Cheng | Xiaodong Liu | Dan Roth | Jianfeng Gao

Despite remarkable advancements in mitigating hallucinations in large language models (LLMs) by retrieval augmentation, it remains challenging to measure the reliability of LLMs using static question-answering (QA) data. Specifically, given the potential of data contamination (e.g., leading to memorization), good static benchmark performance does not ensure that model can reliably use the provided evidence for responding, which is essential to avoid hallucination when the required knowledge is new or private. Inspired by adversarial machine learning, we investigate the feasibility of automatically perturbing existing static one for dynamic evaluation. Specifically, this paper presents ReEval, an LLM-based framework using prompt chaining to perturb the original evidence for generating new test cases for evaluating the LLMs’ reliability in using new evidence for answering.We implement ReEval using ChatGPT and evaluate the resulting variants of two popular open-domain QA datasets on a collection ofLLMs under various prompting settings. Our generated data is human-readable and useful to trigger hallucination in LLM. Accurate models on static data are observed to produce unsupported answers from the perturbed evidence, with pronounced accuracy drops across LLMs including GPT-4. We find that our adversarial examples are transferable across all considered LLMs. The examples generated by a small model can be used to evaluate a much larger model, making our approach cost-effective.

pdf bib
An Effective Automated Speaking Assessment Approach to Mitigating Data Scarcity and Imbalanced Distribution
Tien-Hong Lo | Fu-An Chao | Tzu-i Wu | Yao-Ting Sung | Berlin Chen

Automated speaking assessment (ASA) typically involves automatic speech recognition (ASR) and hand-crafted feature extraction from the ASR transcript of a learner’s speech. Recently, self-supervised learning (SSL) has shown stellar performance compared to traditional methods. However, SSL-based ASA systems are faced with at least three data-related challenges: limited annotated data, uneven distribution of learner proficiency levels and non-uniform score intervals between different CEFR proficiency levels. To address these challenges, we explore the use of two novel modeling strategies: metric-based classification and loss re-weighting, leveraging distinct SSL-based embedding features. Extensive experimental results on the ICNALE benchmark dataset suggest that our approach can outperform existing strong baselines by a sizable margin, achieving a significant improvement of more than 10% in CEFR prediction accuracy.

pdf bib
GPT-Fathom: Benchmarking Large Language Models to Decipher the Evolutionary Path towards GPT-4 and Beyond
Shen Zheng | Yuyu Zhang | Yijie Zhu | Chenguang Xi | Pengyang Gao | Zhou Xun | Kevin Chang

With the rapid advancement of large language models (LLMs), there is a pressing need for a comprehensive evaluation suite to assess their capabilities and limitations. Existing LLM leaderboards often reference scores reported in other papers without consistent settings and prompts, which may inadvertently encourage cherry-picking favored settings and prompts for better results. In this work, we introduce GPT-Fathom, an open-source and reproducible LLM evaluation suite built on top of OpenAI Evals. We systematically evaluate 10+ leading LLMs as well as OpenAI’s legacy models on 20+ curated benchmarks across 7 capability categories, all under aligned settings. Our retrospective study on OpenAI’s earlier models offers valuable insights into the evolutionary path from GPT-3 to GPT-4. Currently, the community is eager to know how GPT-3 progressively improves to GPT-4, including technical details like whether adding code data improves LLM’s reasoning capability, which aspects of LLM capability can be improved by SFT and RLHF, how much is the alignment tax, etc. Our analysis sheds light on many of these questions, aiming to improve the transparency of advanced LLMs.

pdf bib
Subword Attention and Post-Processing for Rare and Unknown Contextualized Embeddings
Raj Patel | Carlotta Domeniconi

Word representations are an important aspect of Natural Language Processing (NLP). Representations are trained using large corpora, either as independent static embeddings or as part of a deep contextualized model. While word embeddings are useful, they struggle on rare and unknown words. As such, a large body of work has been done on estimating rare and unknown words. However, most of the methods focus on static embeddings, with few models focused on contextualized representations. In this work, we propose SPRUCE, a rare/unknown embedding architecture that focuses on contextualized representations. This architecture uses subword attention and embedding post-processing combined with the contextualized model to produce high quality embeddings. We then demonstrate these techniques lead to improved performance in most intrinsic and downstream tasks.

pdf bib
UGIF-DataSet: A New Dataset for Cross-lingual, Cross-modal Sequential actions on the UI
Sagar Gubbi Venkatesh | Partha Talukdar | Srini Narayanan

Help documents are supposed to aid smartphone users in resolving queries such as “How to block calls from unknown numbers?”. However, given a query, identifying the right help document, understanding instructions from the document, and using them to resolve the issue at hand is challenging. The user experience may be enhanced by converting the instructions in the help document to a step-by-step tutorial overlaid on the phone UI. Successful execution of this task requires overcoming research challenges in retrieval, parsing, and grounding in the multilingual-multimodal setting. For example, user queries in one language may have to be matched against instructions in another language, which in turn needs to be grounded in a multimodal UI in yet another language. Moreover, there isn’t any relevant dataset for such a task. In order to bridge this gap, we introduce UGIF-DataSet, a multi-lingual, multi-modal UI grounded dataset for step-by-step task completion on the smartphone, containing 4,184 tasks across 8 languages. The instruction steps in UGIF-DataSet are available only in English, so the challenge involves operations in the cross-modal, cross-lingual setting. We compare the performance of different large language models for this task and find that the end-to-end task completion rate drops from 48% in English to 32% for other languages, demonstrating significant overall headroom for improvement. We are hopeful that UGIF-DataSet and our analysis will aid further research on the important problem of sequential task completion in the multilingual and multimodal setting.

pdf bib
SimSCOOD: Systematic Analysis of Out-of-Distribution Generalization in Fine-tuned Source Code Models
Hossein Hajipour | Ning Yu | Cristian-Alexandru Staicu | Mario Fritz

Large code datasets have become increasingly accessible for pre-training source code models. However, for the fine-tuning phase, obtaining representative training data that fully covers the code distribution for specific downstream tasks remains challenging due to the task-specific nature and limited labeling resources. These lead to out-of-distribution (OOD) generalization issues with unexpected model inference behaviors that have not been systematically studied yet.In this paper, we contribute the first systematic approach that simulates various OOD scenarios along different dimensions of source code data properties and study the fine-tuned model behaviors in such scenarios. We investigate the behaviors of models under different fine-tuning methodologies, including full fine-tuning and Low-Rank Adaptation (LoRA) fine-tuning methods. Our comprehensive analysis, conducted on four state-of-the-art pretrained models and applied to two code generation tasks, exposes multiple failure modes attributed to OOD generalization issues.

pdf bib
Pruning as a Domain-specific LLM Extractor
Nan Zhang | Yanchi Liu | Xujiang Zhao | Wei Cheng | Runxue Bao | Rui Zhang | Prasenjit Mitra | Haifeng Chen

Large Language Models (LLMs) have exhibited remarkable proficiency across a wide array of NLP tasks. However, the escalation in model size also engenders substantial deployment costs. While few efforts have explored model pruning techniques to reduce the size of LLMs, they mainly center on general or task-specific weights. This leads to suboptimal performance due to lacking specificity on the target domain or generality on different tasks when applied to domain-specific challenges. This work introduces an innovative unstructured dual-pruning methodology, D-Pruner, for domain-specific compression on LLM. It extracts a compressed, domain-specific, and task- agnostic LLM by identifying LLM weights that are pivotal for general capabilities, like linguistic capability and multi-task solving, and domain-specific knowledge. More specifically, we first assess general weight importance by quantifying the error incurred upon their removal with the help of an open-domain calibration dataset. Then, we utilize this general weight importance to refine the training loss, so that it preserves generality when fitting into a specific domain. Moreover, by efficiently approximating weight importance with the refined training loss on a domain-specific calibration dataset, we obtain a pruned model emphasizing generality and specificity. Our comprehensive experiments across various tasks in healthcare and legal domains show the effectiveness of D-Pruner in domain-specific compression. Our code is available at https://github.com/psunlpgroup/D-Pruner.

pdf bib
LLMRefine: Pinpointing and Refining Large Language Models via Fine-Grained Actionable Feedback
Wenda Xu | Daniel Deutsch | Mara Finkelstein | Juraj Juraska | Biao Zhang | Zhongtao Liu | William Yang Wang | Lei Li | Markus Freitag

Recent large language models (LLM) areleveraging human feedback to improve theirgeneration quality. However, human feedbackis costly to obtain, especially during inference.In this work, we propose LLMRefine, aninference time optimization method to refineLLM’s output. The core idea is to usea learned fine-grained feedback model topinpoint defects and guide LLM to refinethem iteratively. Using original LLM as aproposal of edits, LLMRefine searches fordefect-less text via simulated annealing, tradingoff the exploration and exploitation. Weconduct experiments on three text generationtasks, including machine translation, long-form question answering (QA), and topicalsummarization. LLMRefine consistentlyoutperforms all baseline approaches, achievingimprovements up to 1.7 MetricX points ontranslation tasks, 8.1 ROUGE-L on ASQA, 2.2ROUGE-L on topical summarization.

pdf bib
Noisy Multi-Label Text Classification via Instance-Label Pair Correction
Pengyu Xu | Mingyang Song | Linkaida Liu | Bing Liu | Hongjian Sun | Liping Jing | Jian Yu

In noisy label learning, instance selection based on small-loss criteria has been proven to be highly effective. However, in the case of noisy multi-label text classification (NMLTC), the presence of noise is not limited to the instance-level but extends to the (instance-label) pair-level.This gives rise to two main challenges.(1) The loss information at the pair-level fails to capture the variations between instances. (2) There are two types of noise at the pair-level: false positives and false negatives. Identifying false negatives from a large pool of negative pairs presents an exceedingly difficult task. To tackle these issues, we propose a novel approach called instance-label pair correction (iLaCo), which aims to address the problem of noisy pair selection and correction in NMLTC tasks.Specifically, we first introduce a holistic selection metric that identifies noisy pairs by simultaneously considering global loss information and instance-specific ranking information.Secondly, we employ a filter guided by label correlation to focus exclusively on negative pairs with label relevance. This filter significantly reduces the difficulty of identifying false negatives.Experimental analysis indicates that our framework effectively corrects noisy pairs in NMLTC datasets, leading to a significant improvement in model performance.

pdf bib
Composite Backdoor Attacks Against Large Language Models
Hai Huang | Zhengyu Zhao | Michael Backes | Yun Shen | Yang Zhang

Large language models (LLMs) have demonstrated superior performance compared to previous methods on various tasks, and often serve as the foundation models for many researches and services. However, the untrustworthy third-party LLMs may covertly introduce vulnerabilities for downstream tasks. In this paper, we explore the vulnerability of LLMs through the lens of backdoor attacks. Different from existing backdoor attacks against LLMs, ours scatters multiple trigger keys in different prompt components. Such a Composite Backdoor Attack (CBA) is shown to be stealthier than implanting the same multiple trigger keys in only a single component. CBA ensures that the backdoor is activated only when all trigger keys appear. Our experiments demonstrate that CBA is effective in both natural language processing (NLP) and multimodal tasks. For instance, with 3% poisoning samples against the LLaMA-7B model on the Emotion dataset, our attack achieves a 100% Attack Success Rate (ASR) with a False Triggered Rate (FTR) below 2.06% and negligible model accuracy degradation. Our work highlights the necessity of increased security research on the trustworthiness of foundation LLMs.

pdf bib
Adapting Fake News Detection to the Era of Large Language Models
Jinyan Su | Claire Cardie | Preslav Nakov

In the age of large language models (LLMs) and the widespread adoption of AI-driven content creation, the landscape of information dissemination has witnessed a paradigm shift. With the proliferation of both human-written and machine-generated real and fake news, robustly and effectively discerning the veracity of news articles has become an intricate challenge. While substantial research has been dedicated to fake news detection, it has either assumed that all news articles are human-written or has abruptly assumed that all machine-generated news was fake. Thus, a significant gap exists in understanding the interplay between machine-paraphrased real news, machine-generated fake news, human-written fake news, and human-written real news. In this paper, we study this gap by conducting a comprehensive evaluation of fake news detectors trained in various scenarios. Our primary objectives revolve around the following pivotal question: How can we adapt fake news detectors to the era of LLMs?Our experiments reveal an interesting pattern that detectors trained exclusively on human-written articles can indeed perform well at detecting machine-generated fake news, but not vice versa. Moreover, due to the bias of detectors against machine-generated texts (CITATION), they should be trained on datasets with a lower machine-generated news ratio than the test set. Building on our findings, we provide a practical strategy for the development of robust fake news detectors.

pdf bib
MCAD: Multi-teacher Cross-modal Alignment Distillation for efficient image-text retrieval
Youbo Lei | Feifei He | Chen Chen | Yingbin Mo | Sijia Li | Defeng Xie | Haonan Lu

Due to the success of large-scale visual-language pretraining (VLP) models and the widespread use of image-text retrieval in industry areas, it is now critically necessary to reduce the model size and streamline their mobile-device deployment. Single- and dual-stream model structures are commonly used in image-text retrieval with the goal of closing the semantic gap between textual and visual modalities. While single-stream models use deep feature fusion to achieve more accurate cross-model alignment, dual-stream models are better at offline indexing and fast inference. We propose a Multi-teacher Cross-modality Alignment Distillation (MCAD) technique to integrate the advantages of single- and dual-stream models. By incorporating the fused single-stream features into the image and text features of the dual-stream model, we formulate new modified teacher similarity distributions and features. Then, we conduct both distribution and feature distillation to boost the capability of the student dual-stream model, achieving high retrieval performance without increasing inference complexity. Extensive experiments demonstrate the remarkable performance and high efficiency of MCAD on image-text retrieval tasks. Furthermore, we implement a lightweight CLIP model on Snapdragon/Dimensity chips with only ~100M running memory and ~8.0ms search latency, achieving the mobile-device application of VLP models.

pdf bib
Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting
Zhen Qin | Rolf Jagerman | Kai Hui | Honglei Zhuang | Junru Wu | Le Yan | Jiaming Shen | Tianqi Liu | Jialu Liu | Donald Metzler | Xuanhui Wang | Michael Bendersky

Ranking documents using Large Language Models (LLMs) by directly feeding the query and candidate documents into the prompt is an interesting and practical problem. However, researchers have found it difficult to outperform fine-tuned baseline rankers on benchmark datasets.We analyze pointwise and listwise ranking prompts used by existing methods and argue that off-the-shelf LLMs do not fully understand these challenging ranking formulations. In this paper, we propose to significantly reduce the burden on LLMs by using a new technique called Pairwise Ranking Prompting (PRP).Our results are the first in the literature to achieve state-of-the-art ranking performance on standard benchmarks using moderate-sized open-sourced LLMs. On TREC-DL 2019&2020, PRP based on the Flan-UL2 model with 20B parameters performs favorably with the previous best approach in the literature, which is based on the blackbox commercial GPT-4 that has 50x (estimated) model size, while outperforming other LLM-based solutions, such as InstructGPT which has 175B parameters, by over 10% for all ranking metrics. By using the same prompt template on seven BEIR tasks, PRP outperforms supervised baselines and outperforms the blackbox commercial ChatGPT solution by 4.2% and pointwise LLM-based solutions by more than 10% on average NDCG@10.Furthermore, we propose several variants of PRP to improve efficiency and show that it is possible to achieve competitive results even with linear complexity.

pdf bib
FedLFC: Towards Efficient Federated Multilingual Modeling with LoRA-based Language Family Clustering
Zhihan Guo | Yifei Zhang | Zhuo Zhang | Zenglin Xu | Irwin King

Federated Multilingual Modeling (FMM) plays a crucial role in the applications of natural language processing due to the increasing diversity of languages and the growing demand for data privacy. However, FMM faces limitations stemming from (1) the substantial communication costs in networking and (2) the conflicts arising from parameter interference between different languages. To address these challenges, we introduce a communication-efficient federated learning framework with low-rank adaptation and language family clustering for Multilingual Modeling (MM). In this framework, we maintain the weights of the base model, exclusively updating the lightweight Low-rank adaptation (LoRA) parameters to minimize communication costs. Additionally, we mitigate parameter conflicts by grouping languages based on their language family affiliations, as opposed to aggregating all LoRA parameters. Experiments demonstrate that our proposed model not only surpasses the baseline models in performance but also reduces the communication overhead. Our code is available at https://github.com/zhihan-guo/FedLFC.

pdf bib
Gaussian Process Optimization for Adaptable Multi-Objective Text Generation using Linearly-Weighted Language Models
Mohammad Mahdi Abdollah Pour | Ali Pesaranghader | Eldan Cohen | Scott Sanner

In multi-objective text generation, we aim to optimize over multiple weighted aspects (e.g., toxicity, semantic preservation, fluency) of the generated text. However, multi-objective weighting schemes may change dynamically in practice according to deployment requirements, evolving business needs, personalization requirements on edge devices, or the availability of new language models and/or objective requirements. Ideally, we need an efficient method to adapt to the dynamic requirements of the overall objective. To address these requirements, we propose a linear combination of objective-specific language models to efficiently adapt the decoding process and optimize for the desired objective without the significant computational overhead of retraining one or more language models. We show empirically that we can leverage Gaussian Process black box optimization to adapt the language model decoder weights to outperform other fixed weighting schemes and standard baselines of the task in only a few iterations of decoding. Overall this approach enables highly efficient adaptation of controllable language models via multi-objective weighting schemes that may evolve dynamically in practical deployment situations.

pdf bib
Groundedness in Retrieval-augmented Long-form Generation: An Empirical Study
Alessandro Stolfo

We present an empirical study of groundedness in long-form question answering (LFQA) by retrieval-augmented large language models (LLMs).In particular, we evaluate whether every generated sentence is grounded in the retrieved documents or the model’s pre-training data.Across 3 datasets and 4 model families, our findings reveal that a significant fraction of generated sentences are consistently ungrounded, even when those sentences contain correct ground-truth answers.Additionally, we examine the impacts of factors such as model size, decoding strategy, and instruction tuning on groundedness. Our results show that while larger models tend to ground their outputs more effectively, a significant portion of correct answers remains compromised by hallucinations. This study provides novel insights into the groundedness challenges in LFQA and underscores the necessity for more robust mechanisms in LLMs to mitigate the generation of ungrounded content.

pdf bib
TagDebias: Entity and Concept Tagging for Social Bias Mitigation in Pretrained Language Models
Mehrnaz Moslemi | Amal Zouaq

Pre-trained language models (PLMs) play a crucial role in various applications, including sensitive domains such as the hiring process. However, extensive research has unveiled that these models tend to replicate social biases present in their pre-training data, raising ethical concerns. In this study, we propose the TagDebias method, which proposes debiasing a dataset using type tags. It then proceeds to fine-tune PLMs on this debiased dataset. Experiments show that our proposed TagDebias model, when applied to a ranking task, exhibits significant improvements in bias scores.

pdf bib
Improving Absent Keyphrase Generation with Diversity Heads
Edwin Thomas | Sowmya Vajjala

Keyphrase Generation (KPG) is the task of automatically generating appropriate keyphrases for a given text, with a wide range of real-world applications such as document indexing and tagging, information retrieval, and text summarization. NLP research makes a distinction between present and absent keyphrases based on whether a keyphrase is directly present as a sequence of words in the document during evaluation. However, present and absent keyphrases are treated together in a text-to-text generation framework during training. We treat present keyphrase extraction as a sequence labeling problem and propose a new absent keyphrase generation model that uses a modified cross-attention layer with additional heads to capture diverse views for the same context encoding in this paper. Our experiments show improvements over the state-of-the-art for four datasets for present keyphrase extraction and five datasets for absent keyphrase generation among the six English datasets we explored, covering long and short documents.

pdf bib
mOthello: When Do Cross-Lingual Representation Alignment and Cross-Lingual Transfer Emerge in Multilingual Models?
Tianze Hua | Tian Yun | Ellie Pavlick

Many pretrained multilingual models exhibit cross-lingual transfer ability, which is often attributed to a learned language-neutral representation during pretraining. However, it remains unclear what factors contribute to the learning of a language-neutral representation, and whether the learned language-neutral representation suffices to facilitate cross-lingual transfer. We propose a synthetic task, Multilingual Othello (mOthello), as a testbed to delve into these two questions. We find that: (1) models trained with naive multilingual pretraining fail to learn a language-neutral representation across all input languages; (2) the introduction of “anchor tokens” (i.e., lexical items that are identical across languages) helps cross-lingual representation alignment; and (3) the learning of a language-neutral representation alone is not sufficient to facilitate cross-lingual transfer. Based on our findings, we propose a novel approach – multilingual pretraining with unified output space – that both induces the learning of language-neutral representation and facilitates cross-lingual transfer.

pdf bib
Discovering and Mitigating Indirect Bias in Attention-Based Model Explanations
Farsheed Haque | Depeng Xu | Shuhan Yuan

As the field of Natural Language Processing (NLP) increasingly adopts transformer-based models, the issue of bias becomes more pronounced. Such bias, manifesting through stereotypes and discriminatory practices, can disadvantage certain groups. Our study focuses on direct and indirect bias in the model explanations, where the model makes predictions relying heavily on identity tokens or associated contexts. We present a novel analysis of bias in model explanation, especially the subtle indirect bias, underlining the limitations of traditional fairness metrics. We first define direct and indirect bias in model explanations, which is complementary to fairness in predictions. We then develop an indirect bias discovery algorithm for quantitatively evaluating indirect bias in transformer models using their in-built self-attention matrix. We also propose an indirect bias mitigation algorithm to ensure fairness in transformer models by leveraging attention explanations. Our evaluation shows the significance of indirect bias and the effectiveness of our indirect bias discovery and mitigation.

pdf bib
i-Code V2: An Autoregressive Generation Framework over Vision, Language, and Speech Data
Ziyi Yang | Mahmoud Khademi | Yichong Xu | Reid Pryzant | Yuwei Fang | Chenguang Zhu | Dongdong Chen | Yao Qian | Xuemei Gao | Yi-Ling Chen | Robert Gmyr | Naoyuki Kanda | Noel Codella | Bin Xiao | Yu Shi | Lu Yuan | Takuya Yoshioka | Michael Zeng | Xuedong Huang

The convergence of text, visual, and audio data is crucial towards human-like artificial intelligence, however the current Vision-Language-Speech landscape is dominated by encoder-only models that lack generative abilities. We propose closing this gap with i-Code V2, one of the first models capable of generating natural language from any combination of Vision, Language, and Speech data. i-Code V2 leverages state-of-the-art single-modality encoders, combining their outputs with a new modality-fusing encoder to project combinations of modalities into a shared representational space. Language tokens are generated from these representations via an autoregressive decoder. i-Code V2 is pretrained end-to-end on a large collection of dual- and single-modality datasets with a novel text completion objective that can be generalized across arbitrary combinations of modalities. i-Code V2 matches or outperforms state-of-the-art single- and dual-modality baselines on 7 multimodal tasks, demonstrating the power of generative multimodal pretraining across a diversity of tasks and signals.

pdf bib
Think While You Write: Hypothesis Verification Promotes Faithful Knowledge-to-Text Generation
Yifu Qiu | Varun Embar | Shay Cohen | Benjamin Han

Knowledge-to-text generators often struggle to faithfully generate descriptions for the input facts: they may produce hallucinations that contradict the input, or describe facts not present in the input. To reduce hallucinations, we propose a decoding-only method, TWEAK (Think While Effectively Articulating Knowledge), which can be integrated with any generator without retraining. TWEAK treats the generated sequences at each decoding step and its future sequences as hypotheses, and ranks each generation candidate based on the extent to which their hypotheses are supported by the input facts using a Hypothesis Verification Model (HVM). We first demonstrate the effectiveness of TWEAK by using a Natural Language Inference (NLI) model as the HVM and report improved faithfulness with a minimal impact on the quality. We then replace the NLI model with a task-specific HVM trained with a first-of-a-kind dataset, FATE (Fact-Aligned Textual Entailment), which pairs input facts with their original and perturbed descriptions. We test TWEAK with two generators, and the best TWEAK variants improve on average for the two models by 2.24/7.17 points in faithfulness (FactKB) in in/out-of-distribution evaluations, respectively, and with only a 0.14/0.32-point decline in quality (BERTScore).

pdf bib
It’s All Relative! – A Synthetic Query Generation Approach for Improving Zero-Shot Relevance Prediction
Aditi Chaudhary | Karthik Raman | Michael Bendersky

Large language models (LLMs) have shown promising ability to generate synthetic query-document pairs by prompting with as few as 8 demonstrations. This has enabled building better IR models, especially for tasks with no training data. Typically, such synthetic query generation (QGen) approaches condition on an input context (e.g. a text document) and generate a query relevant to that context, or condition the QGen additionally on the relevance label (e.g. relevant vs irrelevant) to generate queries across relevance buckets. However, we find that such QGen approaches are sub-optimal as they require the model to reason about the desired label and the input from a handful of examples. In this work, we propose to reduce this burden of LLMs by generating queries simultaneously for different labels. We hypothesize that instead of asking the model to generate, say, an irrelevant query given an input context, asking the model to generate an irrelevant query relative to a relevant query is a much simpler task. Extensive experimentation across nine IR datasets shows that synthetic queries generated in such a fashion translates to better downstream performance.

pdf bib
RS-DPO: A Hybrid Rejection Sampling and Direct Preference Optimization Method for Alignment of Large Language Models
Saeed Khaki | JinJin Li | Lan Ma | Liu Yang | Prathap Ramachandra

Reinforcement learning from human feedback (RLHF) has been extensively employed to align large language models with user intent. However, proximal policy optimization (PPO) based RLHF is occasionally unstable requiring significant hyperparameter finetuning, and computationally expensive to maximize the estimated reward during alignment. Recently, direct preference optimization (DPO) is proposed to address those challenges. However, DPO often relies on contrastive responses generated from human annotator and alternative LLM, instead of the policy model, limiting the effectiveness of the RLHF. In this paper, we addresses both challenges by systematically combining rejection sampling (RS) and DPO. Our proposed method, RS-DPO, initiates with the development of a supervised fine-tuned policy model (SFT). A varied set of k responses per prompt are sampled directly from the SFT model. RS-DPO identifies pairs of contrastive samples based on their reward distribution. Finally, we apply DPO with the contrastive samples to align the model to human preference. Our experiments indicate that our proposed method effectively fine-tunes LLMs with limited resource environments, leading to improved alignment with user intent. Furthermore, it outperforms existing methods, including RS, PPO, and DPO.

pdf bib
Hypernetwork-Assisted Parameter-Efficient Fine-Tuning with Meta-Knowledge Distillation for Domain Knowledge Disentanglement
Changqun Li | Linlin Wang | Xin Lin | Shizhou Huang | Liang He

Domain adaptation from labeled source domains to the target domain is important in practical summarization scenarios. However, the key challenge is domain knowledge disentanglement. In this work, we explore how to disentangle domain-invariant knowledge from source domains while learning specific knowledge of the target domain. Specifically, we propose a hypernetwork-assisted encoder-decoder architecture with parameter-efficient fine-tuning. It leverages a hypernetwork instruction learning module to generate domain-specific parameters from the encoded inputs accompanied by task-related instruction. Further, to better disentangle and transfer knowledge from source domains to the target domain, we introduce a meta-knowledge distillation strategy to build a meta-teacher model that captures domain-invariant knowledge across multiple domains and use it to transfer knowledge to students. Experiments on three dialogue summarization datasets show the effectiveness of the proposed model. Human evaluations also show the superiority of our model with regard to the summary generation quality.

pdf bib
MICo: Preventative Detoxification of Large Language Models through Inhibition Control
Roy Siegelmann | Ninareh Mehrabi | Palash Goyal | Prasoon Goyal | Lisa Bauer | Jwala Dhamala | Aram Galstyan | Rahul Gupta | Reza Ghanadan

Large Language Models (LLMs) are powerful tools which have been both dominant and commonplace in the field of Artificial Intelligence. Yet, LLMs have a tendency to devolve into toxic degeneration, wherein otherwise safe and unproblematic models begin generating toxic content. For the sake of social responsibility and inspired by the biological mechanisms of inhibition control, we introduce the paradigm of Education for Societal Norms (ESN). By collecting and labeling examples as acceptable and unacceptable (in this case toxic and non-toxic), and including a corresponding acceptable rewrite with every unacceptable example, we introduce a new mechanism for LLM detoxification. We annotate a dataset of 2,850 entries and use it to fine-tune a model, which we call a Model with Inhibition Control (MICo). Evaluating this model on toxicity detection capability, rewrite detoxification, meaning preservation, and overall toxicity reduction, we discover significant improvements over the baseline model. In our experiments we show that overall toxicity of this model is more than 60% reduced, with over 75% reduction in severe toxicity.

pdf bib
Reinforcement Learning with Token-level Feedback for Controllable Text Generation
Wendi Li | Wei Wei | Kaihe Xu | Wenfeng Xie | Dangyang Chen | Yu Cheng

To meet the requirements of real-world applications, it is essential to control generations of large language models (LLMs). Prior research has tried to introduce reinforcement learning (RL) into controllable text generation while most existing methods suffer from overfitting issues (finetuning-based methods) or semantic collapse (post-processing methods). However, current RL methods are generally guided by coarse-grained (sentence/paragraph-level) feedback, which may lead to suboptimal performance owing to semantic twists or progressions within sentences. To tackle that, we propose a novel reinforcement learning algorithm named TOLE which formulates TOken-LEvel rewards for controllable text generation, and employs a “first-quantize-then-noise” paradigm to enhance the robustness of the RL algorithm. Furthermore, TOLE can be flexibly extended to multiple constraints with little computational expense. Experimental results show that our algorithm can achieve superior performance on both single-attribute and multi-attribute control tasks. We have released our codes at https://github.com/WindyLee0822/CTG.

pdf bib
CoMM: Collaborative Multi-Agent, Multi-Reasoning-Path Prompting for Complex Problem Solving
Pei Chen | Shuai Zhang | Boran Han

Large Language Models (LLMs) have shown great ability in solving traditional natural language tasks and elementary reasoning tasks with appropriate prompting techniques. However, their ability is still limited in solving complicated science problems. In this work, we aim to push the upper bound of the reasoning capability of LLMs by proposing a collaborative multi-agent, multi-reasoning-path (CoMM) prompting framework. Specifically, we prompt LLMs to play different roles in a problem-solving team, and encourage different role-play agents to collaboratively solve the target task. In particular, we discover that applying different reasoning paths for different roles is an effective strategy to implement few-shot prompting approaches in the multi-agent scenarios. Empirical results demonstrate the effectiveness of the proposed methods on two college-level science problems over competitive baselines. Our further analysis shows the necessity of prompting LLMs to play different roles or experts independently.

pdf bib
Tokenization Matters: Navigating Data-Scarce Tokenization for Gender Inclusive Language Technologies
Anaelia Ovalle | Ninareh Mehrabi | Palash Goyal | Jwala Dhamala | Kai-Wei Chang | Richard Zemel | Aram Galstyan | Yuval Pinter | Rahul Gupta

Gender-inclusive NLP research has documented the harmful limitations of gender binary-centric large language models (LLM), such as the inability to correctly use gender-diverse English neopronouns (e.g., xe, zir, fae). While data scarcity is a known culprit, the precise mechanisms through which scarcity affects this behavior remain underexplored. We discover LLM misgendering is significantly influenced by Byte-Pair Encoding (BPE) tokenization, the tokenizer powering many popular LLMs. Unlike binary pronouns, BPE overfragments neopronouns, a direct consequence of data scarcity during tokenizer training. This disparate tokenization mirrors tokenizer limitations observed in multilingual and low-resource NLP, unlocking new misgendering mitigation strategies. We propose two techniques: (1) pronoun tokenization parity, a method to enforce consistent tokenization across gendered pronouns, and (2) utilizing pre-existing LLM pronoun knowledge to improve neopronoun proficiency. Our proposed methods outperform finetuning with standard BPE, improving neopronoun accuracy from 14.1% to 58.4%. Our paper is the first to link LLM misgendering to tokenization and deficient neopronoun grammar, indicating that LLMs unable to correctly treat neopronouns as pronouns are more prone to misgender.

pdf bib
AdaPT: A Set of Guidelines for Hyperbolic Multimodal Multilingual NLP
Ramit Sawhney | Shrey Pandit | Vishwa Shah | Megh Thakkar | Shafiq Joty

The Euclidean space is the familiar space for training neural models and performing arithmetic operations.However, many data types inherently possess complex geometries, and model training methods involve operating over their latent representations, which cannot be effectively captured in the Euclidean space.The hyperbolic space provides a more generalized representative geometry to model the hierarchical complexities of the tree-like structure of natural language.We propose AdaPT a set of guidelines for initialization, parametrization, and training of neural networks, which adapts to the dataset and can be used with different manifolds. AdaPT can be generalized over any existing neural network training methodology and leads to more stable training without a substantial increase in training time.We apply AdaPT guidelines over two state-of-the-art deep learning approaches and empirically demonstrate its effectiveness through experiments on three tasks over 12 languages across speech and text.Through extensive qualitative analysis, we put forward the applicability of AdaPT as a set of guidelines optimally utilizing the manifold geometry, which can be extended to various downstream tasks across languages and modalities.

pdf bib
More Samples or More Prompts? Exploring Effective Few-Shot In-Context Learning for LLMs with In-Context Sampling
Bingsheng Yao | Guiming Chen | Ruishi Zou | Yuxuan Lu | Jiachen Li | Shao Zhang | Yisi Sang | Sijia Liu | James Hendler | Dakuo Wang

While most existing works on LLM prompting techniques focus only on how to select a better set of data samples inside one single prompt input (In-Context Learning or ICL), why can not we design and leverage multiple prompts together to further improve the LLM’s performance? In this work, we propose In-Context Sampling (ICS), a low-resource LLM prompting technique to produce confident predictions by optimizing the construction of multiple ICL prompt inputs. Extensive experiments with three open-source LLMs (FlanT5-XL, Mistral-7B, and Mixtral-8x7B) on four NLI datasets (e-SNLI, Multi-NLI, ANLI, and Contract-NLI) and one QA dataset (CommonsenseQA) illustrate that ICS can consistently enhance LLMs’ performance. An in-depth evaluation with three data similarity-based ICS strategies suggests that these strategies can further elevate LLM’s performance, which sheds light on a new yet promising future research direction.

pdf bib
ZSEE: A Dataset based on Zeolite Synthesis Event Extraction for Automated Synthesis Platform
Song He | Xin Peng | Yihan Cai | Xin Li | Zhiqing Yuan | WenLi Du | Weimin Yang

Automated synthesis of zeolite, one of the most important catalysts in chemical industries, holds great significance for attaining economic and environmental benefits. Structural synthesis data extracted through NLP technologies from zeolite experimental procedures can significantly expedite automated synthesis owing to its machine readability. However, the utilization of NLP technologies in information extraction of zeolite synthesis remains restricted due to the lack of annotated datasets. In this paper, we formulate an event extraction task to mine structural synthesis actions from experimental narratives for modular automated synthesis. Furthermore, we introduce ZSEE, a novel dataset containing fine-grained event annotations of zeolite synthesis actions. Our dataset features 16 event types and 13 argument roles which cover all the experimental operational steps of zeolite synthesis. We explore current state-of-the-art event extraction methods on ZSEE, perform error analysis based on the experimental results, and summarize the challenges and corresponding research directions to further facilitate the automated synthesis of zeolites. The code is publicly available at https://github.com/Hi-0317/ZSEE.

pdf bib
Mitigating Hallucination in Abstractive Summarization with Domain-Conditional Mutual Information
Kyubyung Chae | Jaepill Choi | Yohan Jo | Taesup Kim

A primary challenge in abstractive summarization is hallucination—the phenomenon where a model generates plausible text that is absent in the source text. We hypothesize that the domain (or topic) of the source text triggers the model to generate text that is highly probable in the domain, neglecting the details of the source text. To alleviate this model bias, we introduce a decoding strategy based on domain-conditional pointwise mutual information. This strategy adjusts the generation probability of each token by comparing it with the token’s marginal probability within the domain of the source text. According to evaluation on the XSUM dataset, our method demonstrates improvement in terms of faithfulness and source relevance.

pdf bib
Adversarial DPO: Harnessing Harmful Data for Reducing Toxicity with Minimal Impact on Coherence and Evasiveness in Dialogue Agents
San Kim | Gary Lee

Recent advancements in open-domain dialogue systems have been propelled by the emergence of high-quality large language models (LLMs) and various effective training methodologies. Nevertheless, the presence of toxicity within these models presents a significant challenge that can potentially diminish the user experience. In this study, we introduce an innovative training algorithm, an improvement upon direct preference optimization (DPO), called adversarial DPO (ADPO). The ADPO algorithm is designed to train models to assign higher probability distributions to preferred responses and lower distributions to unsafe responses, which are self-generated using the toxic control token. We demonstrate that ADPO enhances the model’s resilience against harmful conversations while minimizing performance degradation. Furthermore, we illustrate that ADPO offers a more stable training procedure compared to the traditional DPO. To the best of our knowledge, this is the first adaptation of the DPO algorithm that directly incorporates harmful data into the generative model, thereby reducing the need to artificially create safe dialogue data.

pdf bib
Prompt Space Optimizing Few-shot Reasoning Success with Large Language Models
Fobo Shi | Peijun Qing | Dong Yang | Nan Wang | Youbo Lei | Haonan Lu | Xiaodong Lin | Duantengchuan Li

Prompt engineering is an essential technique for enhancing the abilities of large language models (LLMs) by providing explicit and specific instructions. It enables LLMs to excel in various tasks, such as arithmetic reasoning, question answering, summarization, relation extraction, machine translation, and sentiment analysis. Researchers have been actively exploring different prompt engineering strategies, such as Chain of Thought (CoT), Zero-CoT, and In-context learning. However, an unresolved problem arises from the fact that current approaches lack a solid mathematical solution for determining optimal prompts. To address this issue in prompt engineering, we propose a new and effective approach called Prompt Space. Our methodology utilizes text embeddings to obtain basis vectors by matrix decomposition, and then constructs a space for representing all prompts. Prompt Space significantly outperforms state-of-the-art prompt paradigms on ten public reasoning benchmarks. Notably, without the help of the CoT method and the prompt “Let’s think step by step”, Prompt Space shows superior performance over the few-shot method. Overall, our approach provides a robust and effective mathematical framework for selecting simple and effective prompts. This advancement marks a significant step towards improving prompt engineering for a wide variety of applications in LLMs. Our code is publicly available at https://github.com/YouBLEI/Prompt-Space

pdf bib
DAGCN: Distance-based and Aspect-oriented Graph Convolutional Network for Aspect-based Sentiment Analysis
Zhihao Wang | Bo Zhang | Ru Yang | Chang Guo | Maozhen Li

Aspect-based sentiment analysis (ABSA) is a task that aims to determine the sentiment polarity of aspects by identifying opinion words. Recent advancements have predominantly been rooted either in semantic or syntactic methods. However, both of them tend to interference from local factors such as irrelevant words and edges, hindering the precise identification of opinion words. In this paper, we present Distance-based and Aspect-oriented Graph Convolutional Network (DAGCN) to address the aforementioned issue. Firstly, we introduce the Distance-based Syntactic Weight (DSW). It focuses on the local scope of aspects in the pruned dependency trees, thereby reducing the candidate pool of opinion words. Additionally, we propose Aspect-Fusion Attention (AF) to further filter opinion words within the local context and consider cases where opinion words are distant from the aspect. With the combination of DSW and AF, we achieve precise identification of corresponding opinion words. Extensive experiments on three public datasets demonstrate that the proposed model outperforms state-of-the-art models and verify the effectiveness of the proposed architecture.

pdf bib
Connecting the Dots: Inferring Patent Phrase Similarity with Retrieved Phrase Graphs
Zhuoyi Peng | Yi Yang

We study the patent phrase similarity inference task, which measures the semantic similarity between two patent phrases. As patent documents employ legal and highly technical language, existing semantic textual similarity methods that use localized contextual information do not perform satisfactorily in inferring patent phrase similarity. To address this, we introduce a graph-augmented approach to amplify the global contextual information of the patent phrases. For each patent phrase, we construct a phrase graph that links to its focal patents and a list of patents that are either cited by or cite these focal patents. The augmented phrase embedding is then derived from combining its localized contextual embedding with its global embedding within the phrase graph. We further propose a self-supervised learning objective that capitalizes on the retrieved topology to refine both the contextualized embedding and the graph parameters in an end-to-end manner. Experimental results from a unique patent phrase similarity dataset demonstrate that our approach significantly enhances the representation of patent phrases, resulting in marked improvements in similarity inference in a self-supervised fashion. Substantial improvements are also observed in the supervised setting, underscoring the potential benefits of leveraging retrieved phrase graph augmentation.

pdf bib
Self-Regulated Sample Diversity in Large Language Models
Mingyue Liu | Jonathan Frawley | Sarah Wyer | Hubert P. H. Shum | Sara Uckelman | Sue Black | Chris Willcocks

Sample diversity depends on the task; within mathematics, precision and determinism are paramount, while storytelling thrives on creativity and surprise. This paper presents a simple self-regulating approach where we adjust sample diversity inference parameters dynamically based on the input prompt—in contrast to existing methods that require expensive and inflexible setups, or maintain static values during inference. Capturing a broad spectrum of sample diversities can be formulated as a straightforward self-supervised inference task, which we find significantly improves the quality of responses generically without model retraining or fine-tuning. In particular, our method demonstrates significant improvement in all supercategories of the MMLU multitask benchmark (GPT-3.5: +4.4%, GPT-4: +1.5%), which captures a large variety of difficult tasks covering STEM, the humanities and social sciences.

pdf bib
Methods, Applications, and Directions of Learning-to-Rank in NLP Research
Justin Lee | Gabriel Bernier-Colborne | Tegan Maharaj | Sowmya Vajjala

Learning-to-rank (LTR) algorithms aim to order a set of items according to some criteria. They are at the core of applications such as web search and social media recommendations, and are an area of rapidly increasing interest, with the rise of large language models (LLMs) and the widespread impact of these technologies on society. In this paper, we survey the diverse use cases of LTR methods in natural language processing (NLP) research, looking at previously under-studied aspects such as multilingualism in LTR applications and statistical significance testing for LTR problems. We also consider how large language models are changing the LTR landscape. This survey is aimed at NLP researchers and practitioners interested in understanding the formalisms and best practices regarding the application of LTR approaches in their research.

pdf bib
When Quantization Affects Confidence of Large Language Models?
Irina Proskurina | Luc Brun | Guillaume Metzler | Julien Velcin

Recent studies introduced effective compression techniques for Large Language Models (LLMs) via post-training quantization or low-bit weight representation. Although quantized weights offer storage efficiency and allow for faster inference, existing works have indicated that quantization might compromise performance and exacerbate biases in LLMs.This study investigates the confidence and calibration of quantized models, considering factors such as language model type and scale as contributors to quantization loss.Firstly, we reveal that quantization with GPTQ to 4-bit results in a decrease in confidence regarding true labels, with varying impacts observed among different language models. Secondly, we observe fluctuations in the impact on confidence across different scales. Finally, we propose an explanation for quantization loss based on confidence levels, indicating that quantization disproportionately affects samples where the full model exhibited low confidence levels in the first place.We make our code and quantized models publicly available.

pdf bib
MedCycle: Unpaired Medical Report Generation via Cycle-Consistency
Elad Hirsch | Gefen Dawidowicz | Ayellet Tal

Generating medical reports for X-ray images presents a significant challenge, particularly in unpaired scenarios where access to paired image-report data for training is unavailable. Previous works have typically learned a joint embedding space for images and reports, necessitating a specific labeling schema for both. We introduce an innovative approach that eliminates the need for consistent labeling schemas, thereby enhancing data accessibility and enabling the use of incompatible datasets. This approach is based on cycle-consistent mapping functions that transform image embeddings into report embeddings, coupled with report auto encoding for medical report generation. Our model and objectives consider intricate local details and the overarching semantic context within images and reports. This approach facilitates the learning of effective mapping functions, resulting in the generation of coherent reports. It outperforms state-of-the-art results in unpaired chest X-ray report generation, demonstrating improvements in both language and clinical metrics.

pdf bib
Beta-LR: Interpretable Logical Reasoning based on Beta Distribution
Yizhuo Ma | Ke Qin | Shuang Liang

The logical information contained in text isof significant importance for logical reasoning.Previous approaches have relied on embeddingtext into a low-dimensional vector to capturelogical information and perform reasoning inEuclidean space. These methods involve constructing special graph architectures that matchlogical relations or designing data augmentation frameworks by extending texts based onsymbolic logic. However, it presents two obvious problems. 1) The logical informationreflected in the text exhibits uncertainty that isdifficult to represent using a vector. 2) Integrating logical information requires modeling logical operations (such as ∪, ∩, and ¬), while onlysimple arithmetic operations can be performedin Euclidean space. To address both the problems, we propose Beta-LR, a probabilistic embedding method to capture logical information.Specifically, we embed texts into beta distribution on each dimension to eliminate logical uncertainty. We also define neural operators thatenable interpretability and perform logical operations based on the characteristics of the betadistribution. We conduct experiments on twodatasets, ReClor and LogiQA, and our Beta-LRachieves competitive results. The experimentsdemonstrate that our method effectively captures the logical information in text for reasoning purposes. The source code is available athttps://github.com/myz12138/Beta-LR.

pdf bib
Applications of BERT Models Towards Automation of Clinical Coding in Icelandic
Haraldur Hauksson | Hafsteinn Einarsson

This study explores the potential of automating clinical coding in Icelandic, a language with limited digital resources, by leveraging over 25 years of electronic health records (EHR) from the Landspitali University Hospital. Traditionally a manual and error-prone task, clinical coding is essential for patient care, billing, and research. Our research delves into the effectiveness of Transformer-based models in automating this process. We investigate various model training strategies, including continued pretraining and model adaptation, under a constrained computational budget. Our findings reveal that the best-performing model achieves competitive results in both micro and macro F1 scores, with label attention contributing significantly to its success. The study also explores the possibility of training on unlabeled data. Our research provides valuable insights into the possibilities of using NLP for clinical coding in low-resource languages, demonstrating that small countries with unique languages and well-segmented healthcare records can achieve results comparable to those in higher-resourced languages.

pdf bib
“Tell me who you are and I tell you how you argue”: Predicting Stances and Arguments for Stakeholder Groups
Philipp Heinisch | Lorik Dumani | Philipp Cimiano | Ralf Schenkel

Argument mining has focused so far mainly on the identification, extraction, and formalization of arguments. An important yet unaddressedtask consists in the prediction of the argumentative behavior of stakeholders in a debate. Predicting the argumentative behavior in advance can support foreseeing issues in public policy making or help recognize potential disagreements early on and help to resolve them. In this paper, we consider the novel task of predicting the argumentative behavior of individual stakeholders. We present ARGENST, a framework that relies on a recommender-based architecture to predict the stance and the argumentative main point on a specific controversial topic for a given stakeholder, which is described in terms of a profile including properties related to demographic attributes, religious and political orientation, socio-economic background, etc. We evaluate our approach on the well-known debate.org dataset in terms of accuracy for predicting stance as well as in terms of similarity of the generated arguments to the ground truth arguments using BERTScore. As part of a case study, we show how juries of members representing different stakeholder groups and perspectives can be assembled to simulate the public opinion on a given topic.

pdf bib
Psychometric Predictive Power of Large Language Models
Tatsuki Kuribayashi | Yohei Oseki | Timothy Baldwin

Instruction tuning aligns the response of large language models (LLMs) with human preferences.Despite such efforts in human–LLM alignment, we find that instruction tuning does not always make LLMs human-like from a cognitive modeling perspective. More specifically, next-word probabilities estimated by instruction-tuned LLMs are often worse at simulating human reading behavior than those estimated by base LLMs.In addition, we explore prompting methodologies for simulating human reading behavior with LLMs. Our results show that prompts reflecting a particular linguistic hypothesis improve psychometric predictive power, but are still inferior to small base models.These findings highlight that recent advancements in LLMs, i.e., instruction tuning and prompting, do not offer better estimates than direct probability measurements from base LLMs in cognitive modeling. In other words, pure next-word probability remains a strong predictor for human reading behavior, even in the age of LLMs.

pdf bib
Large Language Models Sensitivity to The Order of Options in Multiple-Choice Questions
Pouya Pezeshkpour | Estevam Hruschka

Large Language Models (LLMs) have demonstrated remarkable capabilities in various NLP tasks. However, previous works have shown these models are sensitive towards prompt wording, and few-shot demonstrations and their order, posing challenges to fair assessment of these models. As these models become more powerful, it becomes imperative to understand and address these limitations. In this paper, we focus on LLMs robustness on the task of multiple-choice questions—commonly adopted task to study reasoning and fact-retrieving capability of LLMs. Investigating the sensitivity of LLMs towards the order of options in multiple-choice questions, we demonstrate a considerable performance gap of approximately 13% to 85% in LLMs on different benchmarks, when answer options are reordered, even when using demonstrations in a few-shot setting. Through a detailed analysis, we conjecture that this sensitivity arises when LLMs are uncertain about the prediction between the top-2/3 choices, and specific options placements may favor certain prediction between those top choices depending on the question caused by positional bias. We also identify patterns in top-2 choices that amplify or mitigate the model’s bias toward option placement. We found that for amplifying bias, the optimal strategy involves positioning the top two choices as the first and last options. Conversely, to mitigate bias, we recommend placing these choices among the adjacent options. To validate our conjecture, we conduct various experiments and adopt two approaches to calibrate LLMs’ predictions, leading to up to 8 percentage points improvement across different models and benchmarks.

pdf bib
PEEB: Part-based Image Classifiers with an Explainable and Editable Language Bottleneck
Thang Pham | Peijie Chen | Tin Nguyen | Seunghyun Yoon | Trung Bui | Anh Nguyen

CLIP-based classifiers rely on the prompt containing a class name that is known to the text encoder. Therefore, they perform poorly on new classes or the classes whose names rarely appear on the Internet (e.g., scientific names of birds). For fine-grained classification, we propose PEEB – an explainable and editable classifier to (1) express the class name into a set of text descriptors that describe the visual parts of that class; and (2) match the embeddings of the detected parts to their textual descriptors in each class to compute a logit score for classification. In a zero-shot setting where the class names are unknown, PEEB outperforms CLIP by a huge margin (∼10× in top-1 accuracy). Compared to part-based classifiers, PEEB is not only the state-of-the-art (SOTA) on the supervised-learning setting (88.80% and 92.20% accuracy on CUB-200 and Stanford Dogs-120, respectively) but also the first to enable users to edit the text descriptors to form a new classifier without any re-training. Compared to concept bottleneck models, PEEB is also the SOTA in both zero-shot and supervised-learning settings.

pdf bib
Ethos: Rectifying Language Models in Orthogonal Parameter Space
Lei Gao | Yue Niu | Tingting Tang | Salman Avestimehr | Murali Annavaram

Language models (LMs) have greatly propelled the research on natural language processing. However, LMs also raise concerns regarding the generation of biased or toxic content and the potential disclosure of private information from the training dataset. In this work, we present a new efficient approach, Ethos, that rectifies LMs to mitigate toxicity and bias in outputs and avoid privacy leakage. Ethos is built on task arithmetic. However, unlike current task arithmetic algorithms, Ethos distinguishes general beneficial and undesired knowledge when reconstructing task vectors. Specifically, Ethos first obtains a set of principal components from the pre-trained models using singular value decomposition. Then, by projecting the task vector onto principal components, Ethos separates the principal components that encode general from those associated with undesired knowledge. Ethos performs forgetting or unlearning by only negating the task vector with undesired knowledge, thereby minimizing collateral damage on general model utility. We demonstrate the efficacy of our approach on three different tasks: bias, toxicity, and memorization unlearning. Evaluations show Ethos is more effective in removing undesired knowledge while maintaining the overall model performance compared to current task arithmetic methods.

pdf bib
Crafting In-context Examples according to LMs’ Parametric Knowledge
Yoonsang Lee | Pranav Atreya | Xi Ye | Eunsol Choi

In-context learning can improve the performances of knowledge-rich tasks such as question answering. In such scenarios, in-context examples trigger a language model (LM) to surface information stored in its parametric knowledge. We study how to better construct in-context example sets, based on whether the model is aware of the in-context examples. We identify ‘known’ examples, where models can correctly answer from their parametric knowledge, and ‘unknown’ ones. Our experiments show that prompting with ‘unknown’ examples decreases the performance, potentially as it encourages hallucination rather than searching for its parametric knowledge. Constructing an in-context example set that presents both known and unknown information performs the best across diverse settings. We perform analysis on three multi-answer question answering datasets, which allows us to further study answer set ordering strategies based on the LM’s knowledge of each answer. Together, our study sheds light on how to best construct in-context example sets for knowledge-rich tasks.

pdf bib
ICXML: An In-Context Learning Framework for Zero-Shot Extreme Multi-Label Classification
Yaxin Zhu | Hamed Zamani

This paper focuses on the task of Extreme Multi-Label Classification (XMC) whose goal is to predict multiple labels for each instance from an extremely large label space. While existing research has primarily focused on fully supervised XMC, real-world scenarios often lack supervision signals, highlighting the importance of zero-shot settings. Given the large label space, utilizing in-context learning approaches is not trivial. We address this issue by introducing In-Context Extreme Multi-label Learning (ICXML), a two-stage framework that cuts down the search space by generating a set of candidate labels through in-context learning and then reranks them. Extensive experiments suggest that ICXML advances the state of the art on two diverse public benchmarks.

pdf bib
CLGSI: A Multimodal Sentiment Analysis Framework based on Contrastive Learning Guided by Sentiment Intensity
Yang Yang | Xunde Dong | Yupeng Qiang

Recently, contrastive learning has begun to gain popularity in multimodal sentiment analysis (MSA). However, most of existing MSA methods based on contrastive learning lacks more detailed learning of the distribution of sample pairs with different sentiment intensity differences in the contrastive learning representation space. In addition, limited research has been conducted on the fusion of each modality representation obtained by contrastive learning training.In this paper, we propose a novel framework for multimodal sentiment analysis based on Contrastive Learning Guided by Sentiment Intensity (CLGSI). Firstly, the proposed contrastive learning guided by sentiment intensity selects positive and negative sample pairs based on the difference in sentiment intensity and assigns corresponding weights accordingly.Subsequently, we propose a new multimodal representation fusion mechanism, called Global-Local-Fine-Knowledge (GLFK), which extracts common features between different modalities’ representations. At the same time, each unimodal encoder output is separately processed by a Multilayer Perceptron (MLP) to extract specific features of each modality. Finally, joint learning of the common and specific features is used to predict sentiment intensity. The effectiveness of CLGSI is assessed on two English datasets, MOSI and MOSEI, as well as one Chinese dataset, SIMS. We achieve competitive experimental results, which attest to the strong generalization performance of our approach. The code for our approach will be released in https://github.com/AZYoung233/CLGSI

pdf bib
Interpreting Answers to Yes-No Questions in Dialogues from Multiple Domains
Zijie Wang | Farzana Rashid | Eduardo Blanco

People often answer yes-no questions without explicitly saying yes, no, or similar polar key-words. Figuring out the meaning of indirectanswers is challenging, even for large language models. In this paper, we investigate this problem working with dialogues from multiple domains. We present new benchmarks in three diverse domains: movie scripts, tennis interviews, and airline customer service. We present an approach grounded on distant supervision and blended training to quickly adapt to a new dialogue domain. Experimental results show that our approach is never detrimental and yields F1 improvements as high as 11-34%.

pdf bib
Enhancing Perception: Refining Explanations of News Claims with LLM Conversations
Yi-Li Hsu | Jui-Ning Chen | Yang Fan Chiang | Shang-Chien Liu | Aiping Xiong | Lun-Wei Ku

We introduce Enhancing Perception, a framework for Large Language Models (LLMs) designed to streamline the time-intensive task typically undertaken by professional fact-checkers of crafting explanations for fake news. This study investigates the effectiveness of enhancing LLM explanations through conversational refinement. We compare various questioner agents, including state-of-the-art LLMs like GPT-4, Claude 2, PaLM 2, and 193 American participants acting as human questioners. Based on the histories of these refinement conversations, we further generate comprehensive summary explanations. We evaluated the effectiveness of these initial, refined, and summary explanations across 40 news claims by involving 2,797 American participants, measuring their self-reported belief change regarding both real and fake claims after receiving the explanations. Our findings reveal that, in the context of fake news, explanations that have undergone conversational refinement—whether by GPT-4 or human questioners, who ask more diverse and detail-oriented questions—were significantly more effective than both the initial unrefined explanations and the summary explanations. Moreover, these refined explanations achieved a level of effectiveness comparable to that of expert-written explanations. The results highlight the potential of automatic explanation refinement by LLMs in debunking fake news claims.

pdf bib
How Interpretable are Reasoning Explanations from Prompting Large Language Models?
Yeo Wei Jie | Ranjan Satapathy | Rick Goh | Erik Cambria

Prompt Engineering has garnered significant attention for enhancing the performance of large language models across a multitude of tasks. Techniques such as the Chain-of-Thought not only bolster task performance but also delineate a clear trajectory of reasoning steps, offering a tangible form of explanation for the audience. Prior works on interpretability assess the reasoning chains yielded by Chain-of-Thought solely along a singular axis, namely faithfulness. We present a comprehensive and multifaceted evaluation of interpretability, examining not only faithfulness but also robustness and utility across multiple commonsense reasoning benchmarks. Likewise, our investigation is not confined to a single prompting technique; it expansively covers a multitude of prevalent prompting techniques employed in large language models, thereby ensuring a wide-ranging and exhaustive evaluation. In addition, we introduce a simple interpretability alignment technique, termed Self-Entailment-Alignment Chain-of-thought, that yields more than 70% improvements across multiple dimensions of interpretability. Code is available at https://github.com/SenticNet/CoT_interpretability

pdf bib
Plug-in Language Model: Controlling Text Generation with a Simple Regression Model
Nai-Chi Yang | Wei-Yun Ma | Pu-Jen Cheng

Large-scale pre-trained language models have displayed unrivaled capacity in generating text that closely resembles human-written text. Nevertheless, generating texts adhering to specific conditions without fine-tuning or adding new parameters can be challenging. Contemporary approaches commonly rely on either prompts or auxiliary models to avoid modifying the language models. These auxiliary models are designed to assess whether a generated token contributes to meeting the desired requirements. These approaches adjust the distribution of the next token during the inference phase by leveraging the prediction score of the desired attribute to calculate gradients. However, these auxiliary models typically require the language model’s latent states. This prerequisite challenges integrating various existing black box attribute models or tools. We present the Plug-in Language Model (PiLM) as a solution to address the limitations. PiLM leverages reinforcement learning to utilize black box tools directly, adjusting the latent state to control text generation. However, performing backpropagation during the inference phase is time-consuming for PiLM. By replacing backpropagation with a simple regression model, PiLM can achieve an inference time comparable to that of the original LLM. Experiment results show that our approaches in this paper outperform existing state-of-the-art methods that rely on gradient-based, weighted decoding, or prompt-based methodologies.

pdf bib
Signer Diversity-driven Data Augmentation for Signer-Independent Sign Language Translation
Honghaofu Honghaofu | Liang Zhang | Biao Fu | Rui Zhao | Jinsong Su | Xiaodong Shi | Yidong Chen

The primary objective of sign language translation (SLT) is to transform sign language videos into natural sentences.A crucial challenge in this field is developing signer-independent SLT systems which requires models to generalize effectively to signers not encountered during training.This challenge is exacerbated by the limited diversity of signers in existing SLT datasets, which often results in suboptimal generalization capabilities of current models.Achieving robustness to unseen signers is essential for signer-independent SLT.However, most existing method relies on signer identity labels, which is often impractical and costly in real-world applications.To address this issue, we propose the Signer Diversity-driven Data Augmentation (SDDA) method that can achieve good generalization without relying on signer identity labels. SDDA comprises two data augmentation schemes. The first is data augmentation based on adversarial training, which aims to utilize the gradients of the model to generate adversarial examples. The second is data augmentation based on diffusion model, which focuses on using the advanced diffusion-based text guided image editing method to modify the appearances of the signer in images. The combination of the two strategies significantly enriches the diversity of signers in the training process.Moreover, we introduce a consistency loss and a discrimination loss to enhance the learning of signer-independent features.Our experimental results demonstrate our model significantly enhances the performance of SLT in the signer-independent setting, achieving state-of-the-art results without relying on signer identity labels.

pdf bib
A Systematic Analysis of Subwords and Cross-Lingual Transfer in Multilingual Translation
Francois Meyer | Jan Buys

Multilingual modelling can improve machine translation for low-resource languages, partly through shared subword representations. This paper studies the role of subword segmentation in cross-lingual transfer. We systematically compare the efficacy of several subword methods in promoting synergy and preventing interference across different linguistic typologies. Our findings show that subword regularisation boosts synergy in multilingual modelling, whereas BPE more effectively facilitates transfer during cross-lingual fine-tuning. Notably, our results suggest that differences in orthographic word boundary conventions (the morphological granularity of written words) may impede cross-lingual transfer more significantly than linguistic unrelatedness. Our study confirms that decisions around subword modelling can be key to optimising the benefits of multilingual modelling.

pdf bib
Multi-Granularity Guided Fusion-in-Decoder
Eunseong Choi | Hyeri Lee | Jongwuk Lee

In Open-domain Question Answering (ODQA), it is essential to discern relevant contexts as evidence and avoid spurious ones among retrieved results. The model architecture that uses concatenated multiple contexts in the decoding phase, *i.e.*, Fusion-in-Decoder, demonstrates promising performance but generates incorrect outputs from seemingly plausible contexts. To address this problem, we propose the ***M**ulti-**G**ranularity guided **F**usion-**i**n-**D**ecoder (**MGFiD**)*, discerning evidence across multiple levels of granularity. Based on multi-task learning, MGFiD harmonizes passage re-ranking with sentence classification. It aggregates evident sentences into an *anchor vector* that instructs the decoder. Additionally, it improves decoding efficiency by reusing the results of passage re-ranking for *passage pruning*. Through our experiments, MGFiD outperforms existing models on the Natural Questions (NQ) and TriviaQA (TQA) datasets, highlighting the benefits of its multi-granularity solution.

pdf bib
Group Fairness in Multilingual Speech Recognition Models
Anna Zee | Marc Zee | Anders Søgaard

We evaluate the performance disparity of the Whisper and MMS families of ASR models across the VoxPopuli and Common Voice multilingual datasets, with an eye toward intersectionality. Our two most important findings are that model size, surprisingly, correlates logarithmically with worst-case performance disparities, meaning that larger (and better) models are less fair. We also observe the importance of intersectionality. In particular, models often exhibit significant performance disparity across binary gender for adolescents.

pdf bib
Rethinking Machine Ethics – Can LLMs Perform Moral Reasoning through the Lens of Moral Theories?
Jingyan Zhou | Minda Hu | Junan Li | Xiaoying Zhang | Xixin Wu | Irwin King | Helen Meng

Making moral judgments is an essential step toward developing ethical AI systems. Prevalent approaches are mostly implemented in a bottom-up manner, which uses a large set of annotated data to train models based on crowd-sourced opinions about morality. These approaches have been criticized for potentially overgeneralizing a limited group of annotators’ moral stances and lacking explainability. This work proposes a flexible top-down framework to steer (Large) Language Models to perform moral reasoning with well-established moral theories from interdisciplinary research. The theory-guided top-down framework can incorporate various moral theories. Our experiments demonstrate the effectiveness of the proposed framework on datasets derived from moral theories. Furthermore, we show the alignment between different moral theories and existing morality datasets. Our analysis exhibits the potential and flaws in existing resources (models and datasets) in developing explainable moral judgment-making systems.

pdf bib
Role Prompting Guided Domain Adaptation with General Capability Preserve for Large Language Models
Rui Wang | Fei Mi | Yi Chen | Boyang Xue | Hongru Wang | Qi Zhu | Kam-Fai Wong | Ruifeng Xu

The growing interest in Large Language Models (LLMs) for specialized applications has revealed a significant challenge: when tailored to specific domains, LLMs tend to experience catastrophic forgetting, compromising their general capabilities and leading to a suboptimal user experience. Additionally, crafting a versatile model for multiple domains simultaneously often results in a decline in overall performance due to confusion between domains. In response to these issues, we present the RolE Prompting Guided Multi-Domain Adaptation (REGA) strategy. This novel approach effectively manages multi-domain LLM adaptation through three key components: 1) Self-Distillation constructs and replays general-domain exemplars to alleviate catastrophic forgetting. 2) Role Prompting assigns a central prompt to the general domain and a unique role prompt to each specific domain to minimize inter-domain confusion during training. 3) Role Integration reuses and integrates a small portion of domain-specific data to the general-domain data, which are trained under the guidance of the central prompt. The central prompt is used for a streamlined inference process, removing the necessity to switch prompts for different domains.Empirical results demonstrate that REGA effectively alleviates catastrophic forgetting and inter-domain confusion. This leads to improved domain-specific performance compared to standard fine-tuned models, while still preserving robust general capabilities.

pdf bib
BERTweet’s TACO Fiesta: Contrasting Flavors On The Path Of Inference And Information-Driven Argument Mining On Twitter
Marc Feger | Stefan Dietze

Argument mining, dealing with the classification of text based on inference and information, denotes a challenging analytical task in the rich context of Twitter (now 𝕏), a key platform for online discourse and exchange. Thereby, Twitter offers a diverse repository of short messages bearing on both of these elements. For text classification, transformer approaches, particularly BERT, offer state-of-the-art solutions. Our study delves into optimizing the embeddings of the understudied BERTweet transformer for argument mining on Twitter and broader generalization across topics.We explore the impact of pre-classification fine-tuning by aligning similar manifestations of inference and information while contrasting dissimilar instances. Using the TACO dataset, our approach augments tweets for optimizing BERTweet in a Siamese network, strongly improving classification and cross-topic generalization compared to standard methods.Overall, we contribute the transformer WRAPresentations and classifier WRAP, scoring 86.62% F1 for inference detection, 86.30% for information recognition, and 75.29% across four combinations of these elements, to enhance inference and information-driven argument mining on Twitter.

pdf bib
Testing the limits of logical reasoning in neural and hybrid models
Manuel Guzman | Jakub Szymanik | Maciej Malicki

We study the ability of neural and hybrid models to generalize logical reasoning patterns. We created a series of tests for analyzing various aspects of generalization in the context of language and reasoning, focusing on compositionality and recursiveness. We used them to study the syllogistic logic in hybrid models, where the network assists in premise selection. We analyzed feed-forward, recurrent, convolutional, and transformer architectures. Our experiments demonstrate that even though the models can capture elementary aspects of the meaning of logical terms, they learn to generalize logical reasoning only to a limited degree.

pdf bib
METAL: Towards Multilingual Meta-Evaluation
Rishav Hada | Varun Gumma | Mohamed Ahmed | Kalika Bali | Sunayana Sitaram

With the rising human-like precision of Large Language Models (LLMs) in numerous tasks, their utilization in a variety of real-world applications is becoming more prevalent. Several studies have shown that LLMs excel on many standard NLP benchmarks. However, it is challenging to evaluate LLMs due to test dataset contamination and the limitations of traditional metrics. Since human evaluations are difficult to collect, there is a growing interest in the community to use LLMs themselves as reference-free evaluators for subjective metrics. However, past work has shown that LLM-based evaluators can exhibit bias and have poor alignment with human judgments. In this study, we propose a framework for an end-to-end assessment of LLMs as evaluators in multilingual scenarios. We create a carefully curated dataset, covering 10 languages containing native speaker judgments for the task of summarization. This dataset is created specifically to evaluate LLM-based evaluators, which we refer to as meta-evaluation (METAL). We compare the performance of LLM-based evaluators created using GPT-3.5-Turbo, GPT-4, and PaLM2. Our results indicate that LLM-based evaluators based on GPT-4 perform the best across languages, while GPT-3.5-Turbo performs poorly. Additionally, we perform an analysis of the reasoning provided by LLM-based evaluators and find that it often does not match the reasoning provided by human judges.

pdf bib
AGIEval: A Human-Centric Benchmark for Evaluating Foundation Models
Wanjun Zhong | Ruixiang Cui | Yiduo Guo | Yaobo Liang | Shuai Lu | Yanlin Wang | Amin Saied | Weizhu Chen | Nan Duan

Assessing foundation models’ abilities for human-level tasks is crucial for Artificial General Intelligence (AGI) development.Traditional benchmarks, which rely on artificial datasets, may not accurately represent these capabilities. In this paper, we introduce AGIEval, a novel bilingual benchmark designed to assess foundation models in the context of human-centric standardized exams, such as college entrance exams, law school admission tests, math competitions, and lawyer qualification tests. We evaluate several state-of-the-art foundation models on our benchmark. Impressively, we show that GPT-4 exceeds the average human performance in SAT, LSAT, and math contests, with 95% accuracy on SAT Math and 92.5% on the Chinese college entrance English exam. This demonstrates the exceptional performance of contemporary foundation models. In contrast, we also find that GPT-4 is less proficient in tasks requiring complex reasoning or specific domain knowledge. Our comprehensive analyses of model capabilities (understanding, knowledge, reasoning, and calculation) reveal their strengths and limitations, providing valuable insights into future directions for enhancing general capabilities. By concentrating on tasks pertinent to human cognition and decision-making, our benchmark delivers a meaningful and robust evaluation of foundation models’ performance in real-world scenarios.

pdf bib
Product Description and QA Assisted Self-Supervised Opinion Summarization
Tejpalsingh Siledar | Rupasai Rangaraju | Sankara Muddu | Suman Banerjee | Amey Patil | Sudhanshu Singh | Muthusamy Chelliah | Nikesh Garera | Swaprava Nath | Pushpak Bhattacharyya

In e-commerce, opinion summarization is the process of summarizing the consensus opinions found in product reviews. However, the potential of additional sources such as product description and question-answers (QA) has been considered less often. Moreover, the absence of any supervised training data makes this task challenging. To address this, we propose a novel synthetic dataset creation (SDC) strategy that leverages information from reviews as well as additional sources for selecting one of the reviews as a pseudo-summary to enable supervised training. Our Multi-Encoder Decoder framework for Opinion Summarization (MEDOS) employs a separate encoder for each source, enabling effective selection of information while generating the summary. For evaluation, due to the unavailability of test sets with additional sources, we extend the Amazon, Oposum+, and Flipkart test sets and leverage ChatGPT to annotate summaries. Experiments across nine test sets demonstrate that the combination of our SDC approach and MEDOS model achieves on average a 14.5% improvement in ROUGE-1 F1 over the SOTA. Moreover, comparative analysis underlines the significance of incorporating additional sources for generating more informative summaries. Human evaluations further indicate that MEDOS scores relatively higher in coherence and fluency with 0.41 and 0.5 (−1 to 1) respectively, compared to existing models. To the best of our knowledge, we are the first to generate opinion summaries leveraging additional sources in a self-supervised setting.

pdf bib
COMEM: In-Context Retrieval-Augmented Mass-Editing Memory in Large Language Models
Shanbao Qiao | Xuebing Liu | Seung-Hoon Na

Noting that world knowledge continuously evolves over time, large language models (LLMs) need to be properly adjusted by performing the “knowledge editing”, which involves updating outdated information or correcting false information. To achieve reliable and “massive” editing capabilities in terms of generalization and specificity, this paper proposes a unified knowledge editing method called in-COntext retrieval-augmented Mass-Editing Memory (COMEM), which combines two types of editing approaches: parameter updating and in-context knowledge editing (IKE). In particular, COMEM incorporates retrieval-augmented IKE, a novel extension of IKE designed for massive editing tasks, based on an updating-aware demonstration construction.Experimental results on the zsRE and CounterFact datasets demonstrate that COMEM outperforms all existing methods, achieving state-of-the-art performance. Our code is available at https://github.com/JoveReCode/COMEM.git.

pdf bib
Content-Specific Humorous Image Captioning Using Incongruity Resolution Chain-of-Thought
Kohtaro Tanaka | Kohei Uehara | Lin Gu | Yusuke Mukuta | Tatsuya Harada

Although automated image captioning methods have benefited considerably from the development of large language models (LLMs), generating humorous captions is still a challenging task. Humorous captions generated by humans are unique to the image and reflect the content of the image. However, captions generated using previous captioning models tend to be generic. Therefore, we propose incongruity-resolution chain-of-thought (IRCoT) as a novel prompting framework that creates content-specific resolutions from fine details extracted from an image. Furthermore, we integrate logit bias and negative sampling to suppress the output of generic resolutions. The results of experiments with GPT4-V demonstrate that our proposed framework effectively generated humorous captions tailored to the content of specific input images.

pdf bib
Denoising Attention for Query-aware User Modeling
Elias Bassani | Pranav Kasela | Gabriella Pasi

Personalization of search results has gained increasing attention in the past few years, also thanks to the development of Neural Networks-based approaches for Information Retrieval. Recent works have proposed to build user models at query time by leveraging the Attention mechanism, which allows weighing the contribution of the user-related information w.r.t. the current query.This approach allows giving more importance to the user’s interests related to the current search performed by the user.In this paper, we discuss some shortcomings of the Attention mechanism when employed for personalization and introduce a novel Attention variant, the Denoising Attention, to solve them.Denoising Attention adopts a robust normalization scheme and introduces a filtering mechanism to better discern among the user-related data those helpful for personalization.Experimental evaluation shows improvements in MAP, MRR, and NDCG above 15% w.r.t. other Attention variants at the state-of-the-art.

pdf bib
A Lightweight Mixture-of-Experts Neural Machine Translation Model with Stage-wise Training Strategy
Fan Zhang | Mei Tu | Song Liu | Jinyao Yan

Dealing with language heterogeneity has always been one of the challenges in neural machine translation (NMT).The idea of using mixture-of-experts (MoE) naturally excels in addressing this issue by employing different experts to take responsibility for different problems.However, the parameter-inefficiency problem in MoE results in less performance improvement when boosting the number of parameters.Moreover, most of the MoE models are suffering from the training instability problem.This paper proposes MoA (Mixture-of-Adapters), a lightweight MoE-based NMT model that is trained via an elaborately designed stage-wise training strategy.With the standard Transformer as the backbone model, we introduce lightweight adapters as experts for easy expansion.To improve the parameter efficiency, we explicitly model and distill the language heterogeneity into the gating network with clustering.After freezing the gating network, we adopt the Gumbel-Max sampling as the routing scheme when training experts to balance the knowledge of generalization and specialization while preventing expert over-fitting.Empirical results show that MoA achieves stable improvements in different translation tasks by introducing much fewer extra parameters compared to other MoE baselines.Additionally, the performance evaluations on a multi-domain translation task illustrate the effectiveness of our training strategy.

pdf bib
BEAR: A Unified Framework for Evaluating Relational Knowledge in Causal and Masked Language Models
Jacek Wiland | Max Ploner | Alan Akbik

Knowledge probing assesses to which degree a language model (LM) has successfully learned relational knowledge during pre-training. Probing is an inexpensive way to compare LMs of different sizes and training configurations. However, previous approaches rely on the objective function used in pre-training LMs and are thus applicable only to masked or causal LMs. As a result, comparing different types of LMs becomes impossible. To address this, we propose an approach that uses an LM’s inherent ability to estimate the log-likelihood of any given textual statement. We carefully design an evaluation dataset of 7,731 instances (40,916 in a larger variant) from which we produce alternative statements for each relational fact, one of which is correct. We then evaluate whether an LM correctly assigns the highest log-likelihood to the correct statement. Our experimental evaluation of 22 common LMs shows that our proposed framework, BEAR, can effectively probe for knowledge across different LM types. We release the BEAR datasets and an open-source framework that implements the probing approach to the research community to facilitate the evaluation and development of LMs.

pdf bib
Conformal Intent Classification and Clarification for Fast and Accurate Intent Recognition
Floris Hengst | Ralf Wolter | Patrick Altmeyer | Arda Kaygan

We present Conformal Intent Classification and Clarification (CICC), a framework for fast and accurate intent classification for task-oriented dialogue systems. The framework turns heuristic uncertainty scores of any intent classifier into a clarification question that is guaranteed to contain the true intent at a pre-defined confidence level.By disambiguating between a small number of likely intents, the user query can be resolved quickly and accurately. Additionally, we propose to augment the framework for out-of-scope detection.In a comparative evaluation using seven intent recognition datasets we find that CICC generates small clarification questions and is capable of out-of-scope detection.CICC can help practitioners and researchers substantially in improving the user experience of dialogue agents with specific clarification questions.

pdf bib
Anonymity at Risk? Assessing Re-Identification Capabilities of Large Language Models in Court Decisions
Alex Nyffenegger | Matthias Stürmer | Joel Niklaus

Anonymity in court rulings is a critical aspect of privacy protection in the European Union and Switzerland but with the advent of LLMs, concerns about large-scale re-identification of anonymized persons are growing. In accordance with the Federal Supreme Court of Switzerland (FSCS), we study re-identification risks using actual legal data. Following the initial experiment, we constructed an anonymized Wikipedia dataset as a more rigorous testing ground to further investigate the findings. In addition to the datasets, we also introduce new metrics to measure performance. We systematically analyze the factors that influence successful re-identifications, identifying model size, input length, and instruction tuning among the most critical determinants. Despite high re-identification rates on Wikipedia, even the best LLMs struggled with court decisions. We demonstrate that for now, the risk of re-identifications using LLMs is minimal in the vast majority of cases. We hope that our system can help enhance the confidence in the security of anonymized decisions, thus leading the courts to publish more decisions.

pdf bib
X-LLaVA: Optimizing Bilingual Large Vision-Language Alignment
DongJae Shin | HyeonSeok Lim | Inho Won | ChangSu Choi | Minjun Kim | SeungWoo Song | HanGyeol Yoo | SangMin Kim | KyungTae Lim

The impressive development of large language models (LLMs) is expanding into the realm of large multimodal models (LMMs), which incorporate multiple types of data beyond text. However, the nature of multimodal models leads to significant expenses in the creation of training data. Furthermore, constructing multilingual data for LMMs presents its own set of challenges due to language diversity and complexity. Therefore, in this study, we propose two cost-effective methods to solve this problem: (1) vocabulary expansion and pretraining of multilingual LLM for specific languages, and (2) automatic and elaborate construction of multimodal datasets using GPT4-V. Based on these methods, we constructed a 91K English-Korean-Chinese multilingual, multimodal training dataset. Additionally, we developed a bilingual multimodal model that exhibits excellent performance in both Korean and English, surpassing existing approaches.

pdf bib
Why So Gullible? Enhancing the Robustness of Retrieval-Augmented Models against Counterfactual Noise
Giwon Hong | Jeonghwan Kim | Junmo Kang | Sung-Hyon Myaeng | Joyce Whang

Most existing retrieval-augmented language models (LMs) assume a naive dichotomy within a retrieved document set: query-relevance and irrelevance. Our work investigates a more challenging scenario in which even the “relevant” documents may contain misleading or incorrect information, causing conflict among the retrieved documents and thereby negatively influencing model decisions as noise. We observe that existing LMs are highly brittle to the presence of conflicting information in both the fine-tuning and in-context few-shot learning scenarios. We propose approaches for handling knowledge conflicts among retrieved documents by explicitly fine-tuning a discriminator or prompting GPT-3.5 to elicit its discriminative capability. Our empirical results on open-domain QA show that these approaches significantly enhance model robustness. We also provide our findings on incorporating the fine-tuned discriminator’s decision into the in-context learning process, proposing a way to exploit the benefits of two disparate learning schemes. Alongside our findings, we provide MacNoise, a machine-generated, conflict-induced dataset to further encourage research in this direction.

pdf bib
Heterogeneity over Homogeneity: Investigating Multilingual Speech Pre-Trained Models for Detecting Audio Deepfake
Orchid Chetia Phukan | Gautam Kashyap | Arun Balaji Buduru | Rajesh Sharma

In this work, we investigate multilingual speech Pre-Trained models (PTMs) for Audio deepfake detection (ADD). We hypothesize thatmultilingual PTMs trained on large-scale diverse multilingual data gain knowledge about diverse pitches, accents, and tones, during theirpre-training phase and making them more robust to variations. As a result, they will be more effective for detecting audio deepfakes. To validate our hypothesis, we extract representations from state-of-the-art (SOTA) PTMs including monolingual, multilingual as well as PTMs trained for speaker and emotion recognition, and evaluated them on ASVSpoof 2019 (ASV), In-the-Wild (ITW), and DECRO benchmark databases. We show that representations from multilingual PTMs, with simple downstream networks, attain the best performance for ADD compared to other PTM representations, which validates our hypothesis. We also explore the possibility of fusion of selected PTM representations for further improvements in ADD, and we propose a framework, MiO (Merge into One) for this purpose. With MiO, we achieve SOTA performance on ASV and ITW and comparable performance on DECRO with current SOTA works.

pdf bib
Identifying Self-Disclosures of Use, Misuse and Addiction in Community-based Social Media Posts
Chenghao Yang | Tuhin Chakrabarty | Karli Hochstatter | Melissa Slavin | Nabila El-Bassel | Smaranda Muresan

In the last decade, the United States has lost more than 500,000 people from an overdose involving prescription and illicit opioids making it a national public health emergency (USDHHS, 2017). Medical practitioners require robust and timely tools that can effectively identify at-risk patients. Community-based social media platforms such as Reddit allow self-disclosure for users to discuss otherwise sensitive drug-related behaviors. We present a moderate size corpus of 2500 opioid-related posts from various subreddits labeled with six different phases of opioid use: Medical Use, Misuse, Addiction, Recovery, Relapse, Not Using. For every post, we annotate span-level extractive explanations and crucially study their role both in annotation quality and model development. We evaluate several state-of-the-art models in a supervised, few-shot, or zero-shot setting. Experimental results and error analysis show that identifying the phases of opioid use disorder is highly contextual and challenging. However, we find that using explanations during modeling leads to a significant boost in classification accuracy demonstrating their beneficial role in a high-stakes domain such as studying the opioid use disorder continuum.

pdf bib
Self-Adaptive Sampling for Accurate Video Question Answering on Image Text Models
Wei Han | Hui Chen | Min-Yen Kan | Soujanya Poria

Image–text models (ITMs) is the prevalent architecture to solve video question–answering tasks, which requires only a few input frames to save huge computational cost compared to video–language models.However, we find existent ITM video question–answering solutions either 1) adopt simplistic and unintentional sampling strategies, which may miss key frames to offer the answer clues; or 2) sample a large number of frames into divided groups, which the computational sources can not accommodate. In this work, we aim at an efficient sampling method towards the few-frame situations.We first summarize a family of prior sampling methods based on question–frame correlation into a unified one, dubbed *Most Implied Frames* (MIF). Through some primary results and analysis, Through analysis, we form a hypothesis that question-aware sampling is not necessary, from which we further propose the other method *Most Dominant Frames* (MDF).Experimental results on four public datasets and three advanced ITMs demonstrate that our proposed strategies can boost the performance for image–text pretrained models, and have a wide application scenario in terms of model architectures and dataset types. Our code is available at https://github.com/declare-lab/Sealinghttps://github.com/declare-lab/Sealing.

pdf bib
Towards an On-device Agent for Text Rewriting
Yun Zhu | Yinxiao Liu | Felix Stahlberg | Shankar Kumar | Yu-Hui Chen | Liangchen Luo | Lei Shu | Renjie Liu | Jindong Chen | Lei Meng

Large Language Models (LLMs) have demonstrated impressive capabilities for text rewriting. However creating a smaller yet potent language model for text rewriting presents two formidable challenges: costly data collection and absence of emergent capabilities.In this paper we present solutions to address the above challenges.We propose an new instruction tuning method to develop a mo-bile text rewriting model that leverages LLM-generated data and heuristic reinforcement learning, eliminating the need for human data collection. Moreover, to bridge the performance gap from the constraint size, we pro-pose a cascading approach based on the confidence levels which are distilled from the large server model’s critiques. To evaluate the text rewriting tasks for mobile scenarios, we introduce MessageRewriteEval, a human-labeled benchmark that focuses on text rewriting of messages through natural language instructions. Through empirical experiments, we demonstrate that our on-device model surpasses the current state-of-the-art LLMs in text rewriting while maintaining a significantly reduced model size using public benchmark EditEval and our new benchmark. We also demonstrate that our proposed cascading approach improves model performance further.

pdf bib
Tailoring Vaccine Messaging with Common-Ground Opinions
Rickard Stureborg | Sanxing Chen | Roy Xie | Aayushi Patel | Christopher Li | Chloe Zhu | Tingnan Hu | Jun Yang | Bhuwan Dhingra

One way to personalize chatbot interactions is by establishing common ground with the intended reader. A domain where establishing mutual understanding could be particularly impactful is vaccine concerns and misinformation. Vaccine interventions are forms of messaging which aim to answer concerns expressed about vaccination. Tailoring responses in this domain is difficult, since opinions often have seemingly little ideological overlap. We define the task of tailoring vaccine interventions to a Common-Ground Opinion (CGO). Tailoring responses to a CGO involves meaningfully improving the answer by relating it to an opinion or belief the reader holds. In this paper we introduce Tailor-CGO, a dataset for evaluating how well responses are tailored to provided CGOs. We benchmark several major LLMs on this task; finding GPT-4-Turbo performs significantly better than others. We also build automatic evaluation metrics, including an efficient and accurate BERT model that outperforms finetuned LLMs, investigate how to successfully tailor vaccine messaging to CGOs, and provide actionable recommendations from this investigation.Tailor-CGO dataset and code available at: https://github.com/rickardstureborg/tailor-cgo

pdf bib
Best of Both Worlds: A Pliable and Generalizable Neuro-Symbolic Approach for Relation Classification
Robert Vacareanu | Fahmida Alam | Md Asiful Islam | Haris Riaz | Mihai Surdeanu

This paper introduces a novel neuro-symbolic architecture for relation classification (RC) that combines rule-based methods with contemporary deep learning techniques. This approach capitalizes on the strengths of both paradigms: the adaptability of rule-based systems and the generalization power of neural networks. Our architecture consists of two components: a declarative rule-based model for transparent classification and a neural component to enhance rule generalizability through semantic text matching.Notably, our semantic matcher is trained in an unsupervised domain-agnostic way, solely with synthetic data.Further, these components are loosely coupled, allowing for rule modifications without retraining the semantic matcher.In our evaluation, we focused on two few-shot relation classification datasets: Few-Shot TACRED and a Few-Shot version of NYT29. We show that our proposed method outperforms previous state-of-the-art models in three out of four settings, despite not seeing any human-annotated training data.Further, we show that our approach remains modular and pliable, i.e., the corresponding rules can be locally modified to improve the overall model. Human interventions to the rules for the TACRED relation org:parents boost the performance on that relation by as much as 26% relative improvement, without negatively impacting the other relations, and without retraining the semantic matching component.

pdf bib
Q-Tuning: Queue-based Prompt Tuning for Lifelong Few-shot Language Learning
Yanhui Guo | Shaoyuan Xu | Jinmiao Fu | Jia Liu | Chaosheng Dong | Bryan Wang

This paper introduces Q-tuning, a novel approach for continual prompt tuning that enables the lifelong learning of a pre-trained language model. When learning a new task, Q-tuning trains a task-specific prompt by adding it to a prompt queue consisting of the prompts from older tasks. To better transfer the knowledge of old tasks, we design an adaptive knowledge aggregation technique that reweighs previous prompts in the queue with a learnable low-rank matrix. Once the prompt queue reaches its maximum capacity, we leverage a PCA-based eviction rule to reduce the queue’s size, allowing the newly trained prompt to be added while preserving the primary knowledge of old tasks. In order to mitigate the accumulation of information loss caused by the eviction, we additionally propose a globally shared prefix prompt and a memory retention regularization based on information theory. Extensive experiments demonstrate that our approach outperforms the state-of-the-art methods substantially on continual prompt tuning benchmarks. Moreover, our approach enables lifelong learning on linearly growing task sequences while requiring constant complexity for training and inference.

pdf bib
In-Context Example Ordering Guided by Label Distributions
Zhichao Xu | Daniel Cohen | Bei Wang | Vivek Srikumar

By allowing models to predict without task-specific training, in-context learning (ICL) with pretrained LLMs has enormous potential in NLP. However, a number of problems persist in ICL. In particular, its performance is sensitive to the choice and order of in-context examples. Given the same set of in-context examples with different orderings, model performance may vary from near random to near state-of-the-art. In this work, we formulate in-context example ordering as an optimization problem. We examine three problem settings that differ in the assumptions they make about what is known about the task. Inspired by the idea of learning from label proportions, we propose two principles for in-context example ordering guided by model’s probability predictions. We apply our proposed principles to thirteen text classification datasets and nine different autoregressive LLMs with 700M to 13B parameters. We demonstrate that our approach outperforms the baselines by improving the classification accuracy, reducing model miscalibration, and also by selecting better in-context examples.

pdf bib
Beyond Surface Similarity: Detecting Subtle Semantic Shifts in Financial Narratives
Jiaxin Liu | Yi Yang | Kar Yan Tam

In this paper, we introduce the Financial-STS task, a financial domain-specific NLP task designed to measure the nuanced semantic similarity between pairs of financial narratives. These narratives originate from the financial statements of the same company but correspond to different periods, such as year-over-year comparisons. Measuring the subtle semantic differences between these paired narratives enables market stakeholders to gauge changes over time in the company’s financial and operational situations, which is critical for financial decision-making. We find that existing pretrained embedding models and LLM embeddings fall short in discerning these subtle financial narrative shifts. To address this gap, we propose an LLM-augmented pipeline specifically designed for the Financial-STS task. Evaluation on a human-annotated dataset demonstrates that our proposed method outperforms existing methods trained on classic STS tasks and generic LLM embeddings.

pdf bib
Laying Anchors: Semantically Priming Numerals in Language Modeling
Mandar Sharma | Rutuja Taware | Pravesh Koirala | Nikhil Muralidhar | Naren Ramakrishnan

Off-the-shelf pre-trained language models have become the de facto standard in NLP pipelines for a multitude of downstream tasks. However, the inability of these models to properly encode numerals limits their performance on tasks requiring numeric comprehension. We introduce strategies to semantically prime numerals in any corpus by generating anchors governed by the distribution of numerals in said corpus, thereby enabling mathematically grounded representations of these numeral tokens. We establish the superiority of our proposed techniques through evaluation on a range of numeracy tasks for both in-domain (seen) and out-domain (unseen) numerals. Further, we expand our empirical evaluations to numerals ranging from 1 to 10 billion, a significantly broader range compared to previous studies of the same nature, and we demonstrate significant improvements in the mathematical grounding of our learned embeddings.

pdf bib
UEGP: Unified Expert-Guided Pre-training for Knowledge Rekindle
Yutao Mou | Kexiang Wang | Jianhe Lin | Dehong Ma | Jun Fan | Daiting Shi | Zhicong Cheng | Gu Simiu | Dawei Yin | Weiran Xu

Pre-training and fine-tuning framework has become the standard training paradigm for NLP tasks and is also widely used in industrial-level applications. However, there are still a limitation with this paradigm: simply fine-tuning with task-specific objectives tends to converge to local minima, resulting in a sub-optimal performance. In this paper, we first propose a new paradigm: knowledge rekindle, which aims to re-incorporate the fine-tuned expert model into the training cycle and break through the performance upper bounds of experts without introducing additional annotated data. Then we further propose a unified expert-guided pre-training (UEGP) framework for knowledge rekindle. Specifically, we reuse fine-tuned expert models for various downstream tasks as knowledge sources and inject task-specific prior knowledge to pre-trained language models (PLMs) by means of knowledge distillation. In this process, we perform multi-task learning with knowledge distillation and masked language modeling (MLM) objectives. We also further explored whether mixture-of-expert guided pre-training (MoEGP) can further enhance the effect of knowledge rekindle. Experiments and analysis on eight datasets in GLUE benchmark and a industrial-level search re-ranking dataset show the effectiveness of our method.

pdf bib
LatticeGen: Hiding Generated Text in a Lattice for Privacy-Aware Large Language Model Generation on Cloud
Mengke Zhang | Tianxing He | Tianle Wang | Lu Mi | Niloofar Mireshghallah | Binyi Chen | Hao Wang | Yulia Tsvetkov

In the current user-server interaction paradigm of prompted generation with large language models (LLMs) on cloud, the server fully controls the generation process, which leaves zero options for users who want to keep the generated text private to themselves. For privacy-aware text generation on cloud, we propose LatticeGen, a cooperative protocol in which the server still handles most of the computation while the client controls the sampling operation. The key idea is that the true generated sequence is mixed with noise tokens by the client and hidden in a noised lattice. Only the client knows which tokens are the true ones. Considering potential attacks from a hypothetically malicious server and how the client can defend against it, we propose the repeated beam-search attack and the mixing noise scheme. In our experiments we apply LatticeGen to protect both prompt and generation. It is shown that while the noised lattice degrades generation quality, LatticeGen successfully protects the true generation to a remarkable degree under strong attacks (more than 50% of the semantic remains hidden as measured by BERTScore).

pdf bib
HateModerate: Testing Hate Speech Detectors against Content Moderation Policies
Jiangrui Zheng | Xueqing Liu | Mirazul Haque | Xing Qian | Guanqun Yang | Wei Yang

To protect users from massive hateful content, existing works studied automated hate speech detection. Despite the existing efforts, one question remains: Do automated hate speech detectors conform to social media content policies? A platform’s content policies are a checklist of content moderated by the social media platform. Because content moderation rules are often uniquely defined, existing hate speech datasets cannot directly answer this question. This work seeks to answer this question by creating HateModerate, a dataset for testing the behaviors of automated content moderators against content policies. First, we engage 28 annotators and GPT in a six-step annotation process, resulting in a list of hateful and non-hateful test suites matching each of Facebook’s 41 hate speech policies. Second, we test the performance of state-of-the-art hate speech detectors against HateModerate, revealing substantial failures these models have in their conformity to the policies. Third, using HateModerate, we augment the training data of a top-downloaded hate detector on HuggingFace. We observe significant improvement in the models’ conformity to content policies while having comparable scores on the original test data. Our dataset and code can be found on https://github.com/stevens-textmining/HateModerate.

pdf bib
Compensate Quantization Errors: Make Weights Hierarchical to Compensate Each Other
Yifei Gao | Jie Ou | Lei Wang | Yuting Xiao | Xiangzhiyuan Xiangzhiyuan | Ruiting Dai | Jun Cheng

Emergent Large Language Models (LLMs) use their extraordinary performance and powerful deduction capacity to discern from traditional language models. However, the expenses of computational resources and storage for these LLMs are stunning, quantization then arises as a trending conversation. To address accuracy decay caused by quantization, two streams of works in post-training quantization methods stand out. One uses other weights to compensate existing quantization error, while the other transfers the quantization difficulty to other parts in the model. Combining both merits, we introduce Learnable Singular value Increment (LSI) as an advanced solution. LSI uses Singular Value Decomposition to extract singular values of the weights and make them learnable to help weights compensate each other conditioned on activation. Incorporating LSI with existing techniques, we achieve state-of-the-art performance in diverse quantization settings, no matter in weight-only, weight-activation or extremely low bit scenarios. By unleashing the potential of LSI, efficient finetuning on quantized model is no longer a prohibitive problem.

pdf bib
Contrastive Preference Learning for Neural Machine Translation
Jianfei He | Shichao Sun | Sen Peng | Jie Xu | Xiaohua Jia | Wenjie Li

There exists a discrepancy between the token-level objective during training and the overall sequence-level quality that is expected from the model. This discrepancy leads to issues like exposure bias.To align the model with human expectations, sequence-level objectives are often used to fine-tune pre-trained models.In this paper, we introduce a contrastive preference model that enhances the traditional Plackett-Luce model by incorporating an indicator function. Building upon this novel preference model, we propose Contrastive Preference Learning (CPL), which uses offline samples with list-wise preferences to fine-tune a pre-trained model in Neural Machine Translation. Our experiments, conducted on three language pairs, demonstrate that CPL outperforms not only the vanilla Transformer model but also other token-level and sequence-level baselines. Furthermore, the ablation study highlights the essential role of the proposed indicator function in achieving this improvement.

pdf bib
SocREval: Large Language Models with the Socratic Method for Reference-free Reasoning Evaluation
Hangfeng He | Hongming Zhang | Dan Roth

To comprehensively gauge the capacity of current models for complex reasoning, it is crucial to assess their step-by-step reasoning in a scalable manner. Established reference-based evaluation metrics rely on human-annotated reasoning chains as references to assess the model-derived chains. However, such “gold-standard” human-written reasoning chains may not be unique and their acquisition is often labor-intensive. Existing reference-free reasoning evaluation metrics, while eliminating the need for human-crafted reasoning chains as references, often require fine-tuning with human-derived chains before evaluation, complicating the process and questioning their adaptability to other datasets. To address these challenges, we harness GPT-4 to automatically evaluate reasoning chain quality, thereby removing the dependency on human-written reasoning chains for both model fine-tuning and evaluative purposes. Leveraging the Socratic method, we develop SocREval (**Soc**ratic Method-Inspired **R**easoning **Eval**uation), a novel approach for prompt design in reference-free reasoning evaluation. Empirical results from four human annotated datasets reveal that SocREval significantly improves GPT-4’s performance, surpassing existing reference-free and reference-based reasoning evaluation metrics. Beyond its demonstrated efficacy, SocREval, proves to be both cost-efficient and robust to prompt writing and example selection, as substantiated by our in-depth analysis.

pdf bib
Multilingual Machine Translation with Large Language Models: Empirical Results and Analysis
Wenhao Zhu | Hongyi Liu | Qingxiu Dong | Jingjing Xu | Shujian Huang | Lingpeng Kong | Jiajun Chen | Lei Li

Large language models (LLMs) have demonstrated remarkable potential in handling multilingual machine translation (MMT). In this paper, we systematically investigate the advantages and challenges of LLMs for MMT by answering two questions: 1) How well do LLMs perform in translating massive languages? 2) Which factors affect LLMs’ performance in translation? We thoroughly evaluate eight popular LLMs, including ChatGPT and GPT-4. Our empirical results show that translation capabilities of LLMs are continually involving. GPT-4 has beat the strong supervised baseline NLLB in 40.91% of translation directions but still faces a large gap towards the commercial translation system like Google Translate, especially on low-resource languages. Through further analysis, we discover that LLMs exhibit new working patterns when used for MMT. First, LLM can acquire translation ability in a resource-efficient way and generate moderate translation even on zero-resource languages. Second, instruction semantics can surprisingly be ignored when given in-context exemplars. Third, cross-lingual exemplars can provide better task guidance for low-resource translation than exemplars in the same language pairs. Code will be released at: https://github.com/NJUNLP/MMT-LLM.

pdf bib
Unleashing the Power of LLMs in Court View Generation by Stimulating Internal Knowledge and Incorporating External Knowledge
Yifei Liu | Yiquan Wu | Ang Li | Yating Zhang | Changlong Sun | Weiming Lu | Fei Wu | Kun Kuang

Court View Generation (CVG) plays a vital role in the realm of legal artificial intelligence, which aims to support judges in crafting legal judgment documents. The court view consists of three essential judgment parts: the charge-related, law article-related, and prison term-related parts, each requiring specialized legal knowledge, rendering CVG a challenging task.Although Large Language Models (LLMs) have made remarkable strides in language generation, they encounter difficulties in the knowledge-intensive legal domain.Actually, there can be two types of knowledge: internal knowledge stored within LLMs’ parameters and external knowledge sourced from legal documents outside the models.In this paper, we decompose court views into different parts, stimulate internal knowledge, and incorporate external information to unleash the power of LLMs in the CVG task.To validate our method, we conduct a series of experiment results on two real-world datasets LAIC2021 and CJO2022. The experiments demonstrate that our method is capable of generating more accurate and reliable court views.

pdf bib
Prompting Vision-Language Models For Aspect-Controlled Generation of Referring Expressions
Danfeng Guo | Sanchit Agarwal | Arpit Gupta | Jiun-Yu Kao | Emre Barut | Tagyoung Chung | Jing Huang | Mohit Bansal

Referring Expression Generation (REG) is the task of generating a description that unambiguously identifies a given target in the scene. Different from Image Captioning (IC), REG requires learning fine-grained characteristics of not only the scene objects but also their surrounding context. Referring expressions are usually not singular; an object can often be uniquely referenced in numerous ways, for instance, by color, by location, or by relationship with other objects. Most prior works, however, have not explored this ‘aspect-based multiplicity’ of referring expressions. Hence, in this work, we focus on the Aspect-Controlled REG task, which requires generating a referring expression conditioned on the input aspect(s), where an aspect captures a style of reference. By changing the input aspect such as color, location, action etc., one can generate multiple distinct expressions per target region. To solve this new task, we first modify BLIP for aligning image-regions and text-expressions. We achieve this through a novel approach for feeding the input by drawing a bounding box around the target image-region and prompting the model to generate the referring expression. Our base REG model already beats all prior works in CIDEr score. To tackle Aspect-Controlled REG, we append ‘aspect tokens’ to the prompt and show that distinct expressions can be generated by just changing the prompt. Finally, to prove the high-quality and diversity of the data generated by our proposed aspect-controlled REG model, we also perform data-augmentation-based evaluation on the downstream Referring Expression Comprehension (REC) task. With just half of the real data augmented with the generated synthetic data, we achieve performance comparable to training with 100% of real data, using a SOTA REC model.

pdf bib
Task-Agnostic Detector for Insertion-Based Backdoor Attacks
Weimin Lyu | Xiao Lin | Songzhu Zheng | Lu Pang | Haibin Ling | Susmit Jha | Chao Chen

Textual backdoor attacks pose significant security threats. Current detection approaches, typically relying on intermediate feature representation or reconstructing potential triggers, are task-specific and less effective beyond sentence classification, struggling with tasks like question answering and named entity recognition. We introduce TABDet (Task-Agnostic Backdoor Detector), a pioneering task-agnostic method for backdoor detection. TABDet leverages final layer logits combined with an efficient pooling technique, enabling unified logit representation across three prominent NLP tasks. TABDet can jointly learn from diverse task-specific models, demonstrating superior detection efficacy over traditional task-specific methods.

pdf bib
Uncertainty Estimation on Sequential Labeling via Uncertainty Transmission
Jianfeng He | Linlin Yu | Shuo Lei | Chang-Tien Lu | Feng Chen

Sequential labeling is a task predicting labels for each token in a sequence, such as Named Entity Recognition (NER). NER tasks aim to extract entities and predict their labels given a text, which is important in information extraction. Although previous works have shown great progress in improving NER performance, uncertainty estimation on NER (UE-NER) is still underexplored but essential. This work focuses on UE-NER, which aims to estimate uncertainty scores for the NER predictions. Previous uncertainty estimation models often overlook two unique characteristics of NER: the connection between entities (i.e., one entity embedding is learned based on the other ones) and wrong span cases in the entity extraction subtask. Therefore, we propose a Sequential Labeling Posterior Network (SLPN) to estimate uncertainty scores for the extracted entities, considering uncertainty transmitted from other tokens. Moreover, we have defined an evaluation strategy to address the specificity of wrong-span cases. Our SLPN has achieved significant improvements on three datasets, such as a 5.54-point improvement in AUPR on the MIT-Restaurant dataset. Our code is available at https://github.com/he159ok/UncSeqLabeling_SLPN.

pdf bib
Exploring Language Model’s Code Generation Ability with Auxiliary Functions
Seonghyeon Lee | Sanghwan Jang | Seongbo Jang | Dongha Lee | Hwanjo Yu

Auxiliary function is a helpful component to improve language model’s code generation ability. However, a systematic exploration of how they affect has yet to be done. In this work, we comprehensively evaluate the ability to utilize auxiliary functions encoded in recent code-pretrained language models. First, we construct a human-crafted evaluation set, called HumanExtension, which contains examples of two functions where one function assists the other.With HumanExtension, we design several experiments to examine their ability in a multifaceted way. Our evaluation processes enable a comprehensive understanding of including auxiliary functions in the prompt in terms of effectiveness and robustness. An additional implementation style analysis captures the models’ various implementation patterns when they access the auxiliary function. Through this analysis, we discover the models’ promising ability to utilize auxiliary functions including their self-improving behavior by implementing the two functions step-by-step. However, our analysis also reveals the model’s underutilized behavior to call the auxiliary function, suggesting the future direction to enhance their implementation by eliciting the auxiliary function call ability encoded in the models. We release our code and dataset to facilitate this research direction.

pdf bib
Crossing Linguistic Horizons: Finetuning and Comprehensive Evaluation of Vietnamese Large Language Models
Sang Truong | Duc Nguyen | Toan Nguyen | Dong Le | Nhi Truong | Tho Quan | Sanmi Koyejo

Recent advancements in large language models (LLMs) have underscored their importance in the evolution of artificial intelligence. However, despite extensive pretraining on multilingual datasets, available open-sourced LLMs exhibit limited effectiveness in processing Vietnamese. The challenge is exacerbated by the absence of systematic benchmark datasets and metrics tailored for Vietnamese LLM evaluation. To mitigate these issues, we have finetuned LLMs specifically for Vietnamese and developed a comprehensive evaluation framework encompassing 10 tasks and 31 metrics. We observe that finetuning can help LLMs transfer knowledge across languages, serving as an efficient way to bolster their capabilities in non-English languages. Moreover, our analysis indicates that larger models can introduce more biases and uncalibrated outputs and the key factor influencing LLM performance is the quality of the training or finetuning datasets. These insights underscore the significance of meticulous finetuning with high-quality datasets in enhancing LLM performance.

pdf bib
GoT: Effective Graph-of-Thought Reasoning in Language Models
Yao Yao | Zuchao Li | Hai Zhao

With the widespread use of language models (LMs) in NLP tasks, researchers have discovered the potential of Chain-of-thought (CoT) to assist LMs in accomplishing complex reasoning tasks by generating intermediate steps. However, human thought processes are often non-linear, rather than simply sequential chains of thoughts. Therefore, we propose Graph-of-Thought (GoT) reasoning, which models human thought processes not only as a chain but also as a graph. By representing thought units as nodes and connections between them as edges, our approach captures the non-sequential nature of human thinking and allows for a more realistic modeling of thought processes. GoT adopts a two-stage framework with an additional GoT encoder for thought graph representation and fuses the graph representation with the original input representation through a gated fusion mechanism. We evaluate GoT’s performance on a text-only reasoning task (AQUA-RAT) and a multimodal reasoning task (ScienceQA). Our model achieves significant improvement over the strong CoT baseline on the AQUA-RAT test set and boosts accuracy from 85.19% to 87.59% using the T5-base model over the state-of-the-art Multimodal-CoT on the ScienceQA test set. Our code is publicly available at https://github.com/Zoeyyao27/Graph-of-Thought

pdf bib
Enhancing the General Agent Capabilities of Low-Paramter LLMs through Tuning and Multi-Branch Reasoning
Qinhao Zhou | Zihan Zhang | Xiang Xiang | Ke Wang | Yuchuan Wu | Yongbin Li

Open-source pre-trained Large Language Models (LLMs) exhibit strong language understanding and generation capabilities, making them highly successful in a variety of tasks. However, when used as agents for dealing with complex problems in the real world, their performance is far inferior to large commercial models such as ChatGPT and GPT-4. As intelligent agents, LLMs need to have the capabilities of task planning, long-term memory, and the ability to leverage external tools to achieve satisfactory performance. Various methods have been proposed to enhance the agent capabilities of LLMs. On the one hand, methods involve constructing agent-specific data and fine-tuning the models. On the other hand, some methods focus on designing prompts that effectively activate the reasoning abilities of the LLMs. We explore both strategies on the 7B and 13B models. We propose a comprehensive method for constructing agent-specific data using GPT-4. Through supervised fine-tuning with constructed data, we find that for these models with a relatively small number of parameters, supervised fine-tuning can significantly reduce hallucination outputs and formatting errors in agent tasks. Furthermore, techniques such as multi-path reasoning and task decomposition can effectively decrease problem complexity and enhance the performance of LLMs as agents. We evaluate our method on five agent tasks of AgentBench and achieve satisfactory results.

pdf bib
MuMath: Multi-perspective Data Augmentation for Mathematical Reasoning in Large Language Models
Weihao You | Shuo Yin | Xudong Zhao | Zhilong Ji | Guoqiang Zhong | Jinfeng Bai

Recently, the tool-use Large Language Models (LLMs) that integrate with external Python interpreters have significantly enhanced mathematical reasoning capabilities for open-source LLMs. However, these models fall short in demonstrating the calculation process, which compromises user-friendliness and understanding of problem-solving steps. Conversely, while tool-free methods offer a clear display of the problem-solving process, their accuracy leaves room for improvement.These tool-free methods typically employ a somewhat narrow range of augmentation techniques such as rephrasing and difficulty enhancement to boost performance. In response to this issue, we have amalgamated and further refined these strengths while broadening the scope of augmentation methods to construct a **mu**lti-perspective augmentation dataset for **math**ematics—termed **MuMath** (𝜇-Math) Dataset.Subsequently, we finetune LLaMA-2 on the MuMath dataset to derive the MuMath model. Our experiments indicate that our MuMath-70B model achieves new state-of-the-art performance among tool-free methods—achieving 88.3% on GSM8K and 34.5% on MATH .We release the MuMath dataset along with its corresponding models and code for public use.

pdf bib
Tram: A Token-level Retrieval-augmented Mechanism for Source Code Summarization
Tong Ye | Lingfei Wu | Tengfei Ma | Xuhong Zhang | Yangkai Du | Peiyu Liu | Shouling Ji | Wenhai Wang

Automatically generating human-readable text describing the functionality of a program is the intent of source code summarization. Although neural language models achieve significant performance in this field, they are limited by their inability to access external knowledge. To address this limitation, an emerging trend is combining neural models with external knowledge through retrieval methods. Previous methods have relied on the sentence-level retrieval paradigm on the encoder side. However, this paradigm is coarse-grained, noise-filled and cannot directly take advantage of the high-quality retrieved summary tokens on the decoder side. In this paper, we propose a fine-grained Token-level retrieval-augmented mechanism (Tram) on the decoder side rather than the encoder side to enhance the performance of neural models and produce more low-frequency tokens in generating summaries. Furthermore, to overcome the challenge of token-level retrieval in capturing contextual code semantics, we also propose integrating code semantics into individual summary tokens. The results of extensive experiments and human evaluation show that our token-level retrieval-augmented approach significantly improves performance and is more interpretable.

pdf bib
UNO-DST: Leveraging Unlabelled Data in Zero-Shot Dialogue State Tracking
Chuang Li | Yan Zhang | Min-Yen Kan | Haizhou Li

Previous zero-shot dialogue state tracking (DST) methods only apply transfer learning, but ignore unlabelled data in the target domain.We transform zero-shot DST into few-shot DST by utilising such unlabelled data via joint and self-training methods. Our method incorporates auxiliary tasks that generate slot types as inverse prompts for main tasks, creating slot values during joint training. Cycle consistency between these two tasks enables the generation and selection of quality samples in unknown target domains for subsequent fine-tuning. This approach also facilitates automatic label creation, thereby optimizing the training and fine-tuning of DST models. We demonstrate this method’s effectiveness on general language models in zero-shot scenarios, improving average joint goal accuracy by 8% across all domains in MultiWOZ.

pdf bib
Evaluating Step-by-Step Reasoning through Symbolic Verification
YiFan Zhang | Hanlin Zhang | Li Li | Eric Xing

Pre-trained language models (LMs) have shown remarkable reasoning performance using explanations or chain-of-thoughts (CoT)) for in-context learning. On the other hand, these reasoning tasks are usually presumed to be more approachable for symbolic programming. To understand the mechanism of reasoning of LMs, we curate synthetic datasets containing equivalent (natural, symbolic) data pairs, where symbolic examples contain first-order logic rules and predicates from non-parametric knowledge bases (KBs), supporting automated verification of intermediate reasoning results. Then we revisit neuro-symbolic approaches and propose to learn from demonstrations containing logic rules and corresponding examples to iteratively reason over KBs, recovering Prolog’s backward chaining algorithm and supporting automated verification of LMs’ outputs. Comprehensive experiments are included to systematically compare LMLP with CoT in deductive reasoning settings, showing that LMLP enjoys more than 25% higher accuracy than CoT on length generalization benchmarks even with smaller model sizes.

pdf bib
Multi-Review Fusion-in-Context
Aviv Slobodkin | Ori Shapira | Ran Levy | Ido Dagan

Grounded text generation, encompassing tasks such as long-form question-answering and summarization, necessitates both content selection and content consolidation. Current end-to-end methods are difficult to control and interpret due to their opaqueness.Accordingly, recent works have proposed a modular approach, with separate components for each step. Specifically, we focus on the second subtask, of generating coherent text given pre-selected content in a multi-document setting. Concretely, we formalize Fusion-in-Context (FiC) as a standalone task, whose input consists of source texts with highlighted spans of targeted content. A model then needs to generate a coherent passage that includes all and only the target information.Our work includes the development of a curated dataset of 1000 instances in the reviews domain, alongside a novel evaluation framework for assessing the faithfulness and coverage of highlights, which strongly correlate to human judgment. Several baseline models exhibit promising outcomes and provide insightful analyses.This study lays the groundwork for further exploration of modular text generation in the multi-document setting, offering potential improvements in the quality and reliability of generated content. Our benchmark, FuseReviews, including the dataset, evaluation framework, and designated leaderboard, can be found at https://fusereviews.github.io/.

pdf bib
Retrieving Examples from Memory for Retrieval Augmented Neural Machine Translation: A Systematic Comparison
Maxime Bouthors | Josep Crego | François Yvon

Retrieval-Augmented Neural Machine Translation (RAMT) architectures retrieve examples from memory to guide the generation process. While most works in this trend explore new ways to exploit the retrieved examples, the upstream retrieval step is mostly unexplored. In this paper, we study the effect of varying retrieval methods for several translation architectures to better understand the interplay between these two processes.We conduct experiments in two language pairs in a multi-domain setting and consider several downstream architectures based on a standard autoregressive model, an edit-based model, and a large language model with in-context learning. Our experiments show that the choice of the retrieval technique impacts the translation scores, with variance across architectures. We also discuss the effects of increasing the number and diversity of examples, which are mostly positive across the board.

pdf bib
Extending Input Contexts of Language Models through Training on Segmented Sequences
Petros Karypis | Julian McAuley | George Karypis

Effectively training language models on longinputs poses many technical challenges. As acost consideration, languages models are pre-trained on a fixed sequence length before beingadapted to longer sequences. We explore var-ious methods for adapting models to longerinputs by training on segmented sequences andan interpolation-based method for extendingabsolute positional embeddings. We developa training procedure to extend the input con-text size of pretrained models with no architec-tural changes and no additional memory coststhan training on the original input lengths. Bysub-sampling segments from long inputs whilemaintaining their original position the model isable to learn new positional interactions. Ourmethod benefits both models trained with abso-lute positional embeddings, by extending theirinput contexts, as well as popular relative posi-tional embedding methods showing a reducedperplexity on sequences longer than they weretrained on. We demonstrate our method canextend input contexts by a factor of 4× whileimproving perplexity.

pdf bib
Reason from Fallacy: Enhancing Large Language Models’ Logical Reasoning through Logical Fallacy Understanding
Yanda Li | Dixuan Wang | Jiaqing Liang | Guochao Jiang | Qianyu He | Yanghua Xiao | Deqing Yang

Large Language Models (LLMs) have demonstrated good performance in many reasoning tasks, but they still struggle with some complicated reasoning tasks including logical reasoning. One non-negligible reason for LLMs’ suboptimal performance on logical reasoning is their overlooking of understanding logical fallacies correctly. To evaluate LLMs’ capability of logical fallacy understanding (LFU), we propose five concrete tasks from three cognitive dimensions of WHAT, WHY, and HOW in this paper. Towards these LFU tasks, we have successfully constructed a new dataset LFUD based on GPT-4 accompanied by a little human effort. Our extensive experiments justify that our LFUD can be used not only to evaluate LLMs’ LFU capability, but also to fine-tune LLMs to obtain significantly enhanced performance on logical reasoning.

pdf bib
Exploring Automated Distractor Generation for Math Multiple-choice Questions via Large Language Models
Wanyong Feng | Jaewook Lee | Hunter McNichols | Alexander Scarlatos | Digory Smith | Simon Woodhead | Nancy Ornelas | Andrew Lan

Multiple-choice questions (MCQs) are ubiquitous in almost all levels of education since they are easy to administer, grade, and are a reliable format in assessments and practices. One of the most important aspects of MCQs is the distractors, i.e., incorrect options that are designed to target common errors or misconceptions among real students. To date, the task of crafting high-quality distractors largely remains a labor and time-intensive process for teachers and learning content designers, which has limited scalability. In this work, we study the task of automated distractor generation in the domain of math MCQs and explore a wide variety of large language model (LLM)-based approaches, from in-context learning to fine-tuning. We conduct extensive experiments using a real-world math MCQ dataset and find that although LLMs can generate some mathematically valid distractors, they are less adept at anticipating common errors or misconceptions among real students.

pdf bib
Aspect-based Sentiment Analysis with Context Denoising
Yuanhe Tian | Chang Liu | Yan Song | Fei Xia | Yongdong Zhang

Given a sentence and a particular aspect term, aspect-based sentiment analysis (ABSA) aims to predict the sentiment polarity towards this aspect term, which provides fine-grained analysis on sentiment understanding and it has attracted much attention in recent years. In order to achieve a good performance on ABSA, it is important for a model to appropriately encode contextual information, especially identifying salient features and eliminating noise in the context. To make incorrect predictions, most existing approaches employ powerful text encoders to locate important context features, as well as noises that mislead ABSA models. These approaches determine the noise in the text for ABSA by assigning low weights to context features or directly removing them from model input, which runs the risk of computing wrong weights or eliminating important context information. In this paper, we propose to improve ABSA with context denoising, where three types of word-level information are regarded as noise, namely, lexicographic noise, bag-of-words noise, and syntax noise. We utilize diffusion networks to perform the denoising process to gradually eliminate them so as to better predict sentiment polarities for given aspect terms. Our approach uses task-specific noise rather than the standard stochastic Gaussian noise in the diffusion networks. The experimental results on five widely used ABSA datasets demonstrate the validity and effectiveness of our approach.

pdf bib
IruMozhi: Automatically classifying diglossia in Tamil
Kabilan Prasanna | Aryaman Arora

Tamil, a Dravidian language of South Asia, is a highly diglossic language with two very different registers in everyday use: Literary Tamil (preferred in writing and formal communication) and Spoken Tamil (confined to speech and informal media). Spoken Tamil is under-studied in modern NLP systems compared to Literary Tamil written in the Tamil script, as evidenced by a lack of datasets explicitly targetting the Spoken variety. In this paper, we release IruMozhi, a human-translated dataset of parallel text in Literary and Spoken Tamil. Using IruMozhi, we train classifiers on the task of identifying which Tamil variety a text belongs to. We use these models to gauge the availability of pretraining data in Spoken Tamil, to audit the composition of existing labelled datasets for Tamil, and to encourage future work on the variety.

pdf bib
RENOVI: A Benchmark Towards Remediating Norm Violations in Socio-Cultural Conversations
Haolan Zhan | Zhuang Li | Xiaoxi Kang | Tao Feng | Yuncheng Hua | Lizhen Qu | Yi Ying | Mei Rianto Chandra | Kelly Rosalin | Jureynolds Jureynolds | Suraj Sharma | Shilin Qu | Linhao Luo | Ingrid Zukerman | Lay-Ki Soon | Zhaleh Semnani Azad | Reza Haf

Norm violations occur when individuals fail to conform to culturally accepted behaviors, which may lead to potential conflicts. Remediating norm violations requires social awareness and cultural sensitivity of the nuances at play. To equip interactive AI systems with a remediation ability, we offer ReNoVi — a large-scale corpus of 9,258 multi-turn dialogues annotated with social norms, as well as define a sequence of tasks to help understand and remediate norm violations step by step. ReNoVi consists of two parts: 512 human-authored dialogues (real data), and 8,746 synthetic conversations generated by ChatGPT through prompt learning. While collecting sufficient human-authored data is costly, synthetic conversations provide suitable amounts of data to help mitigate the scarcity of training data, as well as the chance to assess the alignment between LLMs and humans in the awareness of social norms. We thus harness the power of ChatGPT to generate synthetic training data for our task. To ensure the quality of both human-authored and synthetic data, we follow a quality control protocol during data collection. Our experimental results demonstrate the importance of remediating norm violations in socio-cultural conversations, as well as the improvement in performance obtained from synthetic data.

pdf bib
Human-in-the-Loop Synthetic Text Data Inspection with Provenance Tracking
Hong Jin Kang | Fabrice Harel-Canada | Muhammad Ali Gulzar | Nanyun Peng | Miryung Kim

Data augmentation techniques apply transformations to existing texts to generate additional data. The transformations may produce low-quality texts, where the meaning of the text is changed and the text may even be mangled beyond human comprehension. Analyzing the synthetically generated texts and their corresponding labels is slow and demanding. To winnow out texts with incorrect labels, we develop INSPECTOR, a human-in-the-loop data inspection technique. INSPECTOR combines the strengths of provenance tracking techniques with assistive labeling. INSPECTOR allows users to group related texts by their transformation provenance, i.e., the transformations applied to the original text, or feature provenance, the linguistic features of the original text. For assistive labeling, INSPECTOR computes metrics that approximate data quality, and allows users to compare the corresponding label of each text against the predictions of a large language model. In a user study, INSPECTOR increases the number of texts with correct labels identified by 3× on a sentiment analysis task and by 4× on a hate speech detection task. The participants found grouping the synthetically generated texts by their common transformation to be the most useful technique. Surprisingly, grouping texts by common linguistic features was perceived to be unhelpful. Contrary to prior work, our study finds that no single technique obviates the need for human inspection effort. This validates the design of INSPECTOR which combines both analysis of data provenance and assistive labeling to reduce human inspection effort.

pdf bib
COMMIT: Code-Mixing English-Centric Large Language Model for Multilingual Instruction Tuning
Jaeseong Lee | YeonJoon Jung | Seung-won Hwang

Recently, instruction-tuned large language models (LLMs) are showing prominent performance on various tasks, such as question answering. However, the majority of instruction-tuned LLMs are English-centric, which hinders their application to low-resource language QA. In this paper, we propose COde-Mixed Multilingual Instruction Tuning (COMMIT) to adapt English-centric LLM to low-resource language QA. We point out two main causes of English-centricness: imbalance of unlabeled data, and English-centric instruction tuning datasets. To deviate from English-centric instruction tuning, we propose to specialize code-mixing for instruction tuning, which blocks code-mixing in English templates, to leverage the potential of its superiority. To overcome data imbalance, we perform cross-lingual alignment. The majority of cross-lingual alignment works focused on making representations similar, which is not desirable to decoder-based LLMs, such as LLaMA. Therefore, we propose code-mixed continual causal language modeling to align the decoder. COMMIT improves the exact match score of low-resourced language QA by up to 32x. Code is publicly available.

pdf bib
DiLM: Distilling Dataset into Language Model for Text-level Dataset Distillation
Aru Maekawa | Satoshi Kosugi | Kotaro Funakoshi | Manabu Okumura

Dataset distillation aims to compress a training dataset by creating a small number of informative synthetic samples such that neural networks trained on them perform as well as those trained on the original training dataset. Current text dataset distillation methods create each synthetic sample as a sequence of word embeddings instead of a text to apply gradient-based optimization; however, such embedding-level distilled datasets cannot be used for training other models whose word embedding weights are different from the model used for distillation. To address this issue, we propose a novel text dataset distillation approach, called Distilling dataset into Language Model (DiLM), which trains a language model to generate informative synthetic training samples as text data, instead of directly optimizing synthetic samples. We evaluated DiLM on various text classification datasets and showed that distilled synthetic datasets from DiLM outperform those from current coreset selection methods. DiLM achieved remarkable generalization performance in training different types of models and in-context learning of large language models. Our code will be available at https://github.com/arumaekawa/DiLM.

pdf bib
MindAgent: Emergent Gaming Interaction
Ran Gong | Qiuyuan Huang | Xiaojian Ma | Yusuke Noda | Zane Durante | Zilong Zheng | Demetri Terzopoulos | Li Fei-Fei | Jianfeng Gao | Hoi Vo

Large Foundation Models (LFMs) can perform complex scheduling in a multi-agent system and can coordinate agents to complete sophisticated tasks that require extensive collaboration.However, despite the introduction of numerous gaming frameworks, the community lacks adequate benchmarks that support the implementation of a general multi-agent infrastructure encompassing collaboration between LFMs and human-NPCs. We propose a novel infrastructure—Mindagent—for evaluating planning and coordination capabilities in the context of gaming interaction. In particular, our infrastructure leverages an existing gaming framework to (i) act as the coordinator for a multi-agent system, (ii) collaborate with human players via instructions, and (iii) enable in-context learning based on few-shot prompting with feedback.Furthermore, we introduce “Cuisineworld”, a new gaming scenario and its related benchmark that supervises multiple agents playing the game simultaneously and measures multi-agent collaboration efficiency. We have conducted comprehensive evaluations with a new auto-metric Collaboration Score: CoS for assessing the collaboration efficiency. Finally, Mindagent can be deployed in real-world gaming scenarios in a customized VR version of Cuisineworld and adapted in the “Minecraft” domain. Our work involving LFMs within our new infrastructure for general-purpose scheduling and coordination can elucidate how such skills may be obtained by learning from large language corpora.

pdf bib
BotChat: Evaluating LLMs’ Capabilities of Having Multi-Turn Dialogues
Haodong Duan | Jueqi Wei | Chonghua Wang | Hongwei Liu | Yixiao Fang | Songyang Zhang | Dahua Lin | Kai Chen

In the realm of modern Large Language Models (LLMs), facilitating high-quality, multi-turn dialogues with humans represents a cornerstone feature. However, human-based evaluation of such a capability involves substantial manual effort. This study offers a formative assessment of current LLMs’ proficiency in emulating human-like, multi-turn conversations using an LLM-centric approach. The evaluation encompasses three key elements in the evaluation pipeline: utterance generation, evaluation protocol, and judgement, and we delve deeply into each aspect. GPT-4, both as an utterance generator and as a judge, exhibits exceptional performance. As a generator, GPT-4 crafts dialogues indistinguishable from human interactions in terms of style and flow. When judging, it shows a heightened alignment with human evaluative standards and consistency. Conversely, other LLMs face challenges in producing quality multi-turn dialogues, hindered by inadequate instruction-following abilities, a propensity for prolix utterances, and overall limited capabilities. Notably, generating extensive dialogues (e.g., spanning tens of turns) remains a formidable task for most LLMs, particularly in Chinese contexts. We hope that our work can serve as a valuable resource for evaluating the multi-turn chatting capabilities of LLMs. Related resources are available at https://github.com/open-compass/BotChat.

pdf bib
Learning Mutually Informed Representations for Characters and Subwords
Yilin Wang | Xinyi Hu | Matthew Gormley

Most pretrained language models rely on subword tokenization, which processes text as a sequence of subword tokens. However, different granularities of text, such as characters, subwords, and words, can contain different kinds of information. Previous studies have shown that incorporating multiple input granularities improves model generalization, yet very few of them outputs useful representations for each granularity. In this paper, we introduce the entanglement model, aiming to combine character and subword language models. Inspired by vision-language models, our model treats characters and subwords as separate modalities, and it generates mutually informed representations for both granularities as output. We evaluate our model on text classification, named entity recognition, POS-tagging, and character-level sequence labeling (intraword code-switching). Notably, the entanglement model outperforms its backbone language models, particularly in the presence of noisy texts and low-resource languages. Furthermore, the entanglement model even outperforms larger pre-trained models on all English sequence labeling tasks and classification tasks. We make our code publically available.

pdf bib
A Novel Two-step Fine-tuning Framework for Transfer Learning in Low-Resource Neural Machine Translation
Yuan Gao | Feng Hou | Ruili Wang

Existing transfer learning methods for neural machine translation typically use a well-trained translation model (i.e., a parent model) of a high-resource language pair to directly initialize a translation model (i.e., a child model) of a low-resource language pair, and the child model is then fine-tuned with corresponding datasets. In this paper, we propose a novel two-step fine-tuning (TSFT) framework for transfer learning in low-resource neural machine translation. In the first step, we adjust the parameters of the parent model to fit the child language by using the child source data. In the second step, we transfer the adjusted parameters to the child model and fine-tune it with a proposed distillation loss for efficient optimization. Our experimental results on five low-resource translations demonstrate that our framework yields significant improvements over various strong transfer learning baselines. Further analysis demonstrated the effectiveness of different components in our framework.

pdf bib
Enhancing Cross-lingual Sentence Embedding for Low-resource Languages with Word Alignment
Zhongtao Miao | Qiyu Wu | Kaiyan Zhao | Zilong Wu | Yoshimasa Tsuruoka

The field of cross-lingual sentence embeddings has recently experienced significant advancements, but research concerning low-resource languages has lagged due to the scarcity of parallel corpora. This paper shows that cross-lingual word representation in low-resource languages is notably under-aligned with that in high-resource languages in current models. To address this, we introduce a novel framework that explicitly aligns words between English and eight low-resource languages, utilizing off-the-shelf word alignment models. This framework incorporates three primary training objectives: aligned word prediction and word translation ranking, along with the widely used translation ranking. We evaluate our approach through experiments on the bitext retrieval task, which demonstrate substantial improvements on sentence embeddings in low-resource languages. In addition, the competitive performance of the proposed model across a broader range of tasks in high-resource languages underscores its practicality.

pdf bib
C3LPGCN:Integrating Contrastive Learning and Cooperative Learning with Prompt into Graph Convolutional Network for Aspect-based Sentiment Analysis
Ye He | Shihao Zou | YuzheChen YuzheChen | Xianying Huang

Aspect-based Sentiment Analysis (ABSA) is a fine-grained task. Recently, using graph convolutional networks (GCNs) to model syntactic information has become a popular topic. In addition, a growing consensus exists to enhance sentence representation using contrastive learning. However, when modeling syntactic information, incorrect syntactic structure may introduce additional noise. Meanwhile, we believe that contrastive learning implicitly introduce label information as priori. Therefore, we propose C3LPGCN, which integrates Contrastive Learning and Cooperative Learning with Prompt into GCN. Specifically, to alleviate the noise when modeling syntactic information, we propose mask-aware aspect information filter, which combines prompt information of template with aspect information to filter the syntactic information. Besides, we propose prompt-based contrastive learning and cooperative learning to utilise the label information further. On the one hand, we construct prompts containing labels for contrastive learning, by which the model can focus more on task-relevant features. On the other hand, cooperative learning further extracts label information by aligning input samples’ representation and output distribution with label samples. Extensive experiments on three datasets demonstrate that our method significantly improves the model’s performance compared to traditional contrastive learning methods. Moreover, our C3LPGCN outperforms state-of-the-art methods. Our source code and final models are publicly available at github

pdf bib
Visual Enhanced Entity-Level Interaction Network for Multimodal Summarization
Haolong Yan | Binghao Tang | Boda Lin | Gang Zhao | Si Li

MultiModal Summarization (MMS) aims to generate a concise summary based on multimodal data like texts and images and has wide application in multimodal fields.Previous works mainly focus on the coarse-level textual and visual features in which the overall features of the image interact with the whole sentence.However, the entities of the input text and the objects of the image may be underutilized, limiting the performance of current MMS models.In this paper, we propose a novel Visual Enhanced Entity-Level Interaction Network (VE-ELIN) to address the problem of underutilization of multimodal inputs at a fine-grained level in two ways.We first design a cross-modal entity interaction module to better fuse the entity information in text and the object information in vision.Then, we design an object-guided visual enhancement module to fully extract the visual features and enhance the focus of the image on the object area.We evaluate VE-ELIN on two MMS datasets and propose new metrics to measure the factual consistency of entities in the output.Finally, experimental results demonstrate that VE-ELIN is effective and outperforms previous methods under both traditional metrics and ours.The source code is available at https://github.com/summoneryhl/VE-ELIN.

pdf bib
Knowledgeable In-Context Tuning: Exploring and Exploiting Factual Knowledge for In-Context Learning
Jianing Wang | Chengyu Wang | Chuanqi Tan | Jun Huang | Ming Gao

Large language models (LLMs) enable in-context learning (ICL) by conditioning on a few labeled training examples as a text-based prompt, eliminating the need for parameter updates and achieving competitive performance. In this paper, we demonstrate that factual knowledge is imperative for the performance of ICL in three core facets: the inherent knowledge learned in LLMs, the factual knowledge derived from the selected in-context examples, and the knowledge biases in LLMs for output generation. To unleash the power of LLMs in few-shot learning scenarios, we introduce a novel Knowledgeable In-Context Tuning (KICT) framework to further improve the performance of ICL:1) injecting knowledge into LLMs during continual self-supervised pre-training, 2) judiciously selecting the examples for ICL with high knowledge relevance, and 3) calibrating the prediction results based on prior knowledge.We evaluate the proposed approaches on autoregressive models (e.g., GPT-style LLMs) over multiple text classification and question-answering tasks. Experimental results demonstrate that KICT substantially outperforms strong baselines and improves by more than 13% and 7% on text classification and question-answering tasks, respectively.

pdf bib
Time Machine GPT
Felix Drinkall | Eghbal Rahimikia | Janet Pierrehumbert | Stefan Zohren

Large language models (LLMs) are often trained on extensive, temporally indiscriminate text corpora, reflecting the lack of datasets with temporal metadata. This approach is not aligned with the evolving nature of language. Conventional methods for creating temporally adapted language models often depend on further pre-training static models on time-specific data. This paper presents a new approach: a series of point-in-time LLMs called TimeMachineGPT (TiMaGPT), specifically designed to be nonprognosticative. This ensures they remain uninformed about future factual information and linguistic changes. This strategy is beneficial for understanding language evolution and is of critical importance when applying models in dynamic contexts, such as time-series forecasting, where foresight of future information can prove problematic. We provide access to both the models and training datasets.

pdf bib
An End-to-End Submodular Framework for Data-Efficient In-Context Learning
Lilly Kumari | Shengjie Wang | Arnav Das | Tianyi Zhou | Jeff Bilmes

Recent advancements in natural language tasks leverage the emergent In-Context Learning (ICL) ability of pretrained Large Language Models (LLMs). ICL enables LLMs to perform new tasks by utilizing a limited number of input-output examples as prompts. While ICL circumvents the costly step of finetuning LLMs, its effectiveness is heavily dependent on the quality and ordering of provided examples (called exemplars). In this work, we propose a two-stage data-efficient framework Div-S3 for exemplar selection for ICL. The first stage focuses on data annotation and employs a pool-based active learning approach to select a set of Diverse and informative exemplars from the target tasks’ unlabeled pool. Given a test input/query, the second stage uses Submodular Span Summarization (S3) to select the most relevant and non-redundant exemplars from the annotated pool of a limited budget. On 7 different NLP datasets and 5 LLMs of varying complexities, we show Div-S3 outperforms (1) existing active learning-based methods for data annotation for ICL and (2) similarity-based methods for test query-specific exemplars retrieval.

pdf bib
Teaching Llama a New Language Through Cross-Lingual Knowledge Transfer
Hele-Andra Kuulmets | Taido Purason | Agnes Luhtaru | Mark Fishel

This paper explores cost-efficient methods to adapt pretrained Large Language Models (LLMs) to new lower-resource languages, with a specific focus on Estonian. Leveraging the Llama 2 model, we investigate the impact of combining cross-lingual instruction-tuning with additional monolingual pretraining. Our results demonstrate that even a relatively small amount of additional monolingual pretraining followed by cross-lingual instruction-tuning significantly enhances results on Estonian. Furthermore, we showcase cross-lingual knowledge transfer from high-quality English instructions to Estonian, resulting in improvements in commonsense reasoning and multi-turn conversation capabilities. Our best model, named Llammas, represents the first open-source instruction-following LLM for Estonian. Additionally, we publish Alpaca-est, the first general task instruction dataset for Estonia. These contributions mark the initial progress in the direction of developing open-source LLMs for Estonian.

pdf bib
Simulating Opinion Dynamics with Networks of LLM-based Agents
Yun-Shiuan Chuang | Agam Goyal | Nikunj Harlalka | Siddharth Suresh | Robert Hawkins | Sijia Yang | Dhavan Shah | Junjie Hu | Timothy Rogers

Accurately simulating human opinion dynamics is crucial for understanding a variety of societal phenomena, including polarization and the spread of misinformation. However, the agent-based models (ABMs) commonly used for such simulations often over-simplify human behavior. We propose a new approach to simulating opinion dynamics based on populations of Large Language Models (LLMs). Our findings reveal a strong inherent bias in LLM agents towards producing accurate information, leading simulated agents to consensus in line with scientific reality. This bias limits their utility for understanding resistance to consensus views on issues like climate change. After inducing confirmation bias through prompt engineering, however, we observed opinion fragmentation in line with existing agent-based modeling and opinion dynamics research. These insights highlight the promise and limitations of LLM agents in this domain and suggest a path forward: refining LLMs with real-world discourse to better simulate the evolution of human beliefs.

pdf bib
Probing the Category of Verbal Aspect in Transformer Language Models
Anisia Katinskaia | Roman Yangarber

We investigate how pretrained language models (PLM) encode the grammatical category of verbal aspect in Russian. Encoding of aspect in transformer LMs has not been studied previously in any language. A particular challenge is posed by ”alternative contexts”: where either the perfective or the imperfective aspect is suitable grammatically and semantically. We perform probing using BERT and RoBERTa on alternative and non-alternative contexts. First, we assess the models’ performance on aspect prediction, via behavioral probing. Next, we examine the models’ performance when their contextual representations are substituted with counterfactual representations, via causal probing. These counterfactuals alter the value of the “boundedness” feature—a semantic feature, which characterizes the action in the context. Experiments show that BERT and RoBERTa do encode aspect—mostly in their final layers. The counterfactual interventions affect perfective and imperfective in opposite ways, which is consistent with grammar: perfective is positively affected by adding the meaning of boundedness, and vice versa. The practical implications of our probing results are that fine-tuning only the last layers of BERT on predicting aspect is faster and more effective than fine-tuning the whole model. The model has high predictive uncertainty about aspect in alternative contexts, which tend to lack explicit hints about the boundedness of the described action.

pdf bib
A Measure for Transparent Comparison of Linguistic Diversity in Multilingual NLP Data Sets
Tanja Samardzic | Ximena Gutierrez | Christian Bentz | Steven Moran | Olga Pelloni

Typologically diverse benchmarks are increasingly created to track the progress achieved in multilingual NLP. Linguistic diversity of these data sets is typically measured as the number of languages or language families included in the sample, but such measures do not consider structural properties of the included languages. In this paper, we propose assessing linguistic diversity of a data set against a reference language sample as a means of maximising linguistic diversity in the long run. We represent languages as sets of features and apply a version of the Jaccard index suitable for comparing sets of measures. In addition to the features extracted from typological data bases, we propose an automatic text-based measure, which can be used as a means of overcoming the well-known problem of data sparsity in manually collected features. Our diversity score is interpretable in terms of linguistic features and can identify the types of languages that are not represented in a data set. Using our method, we analyse a range of popular multilingual data sets (UD, Bible100, mBERT, XTREME, XGLUE, XNLI, XCOPA, TyDiQA, XQuAD). In addition to ranking these data sets, we find, for example, that (poly)synthetic languages are missing in almost all of them.

pdf bib
Beyond Read-Only: Crafting a Comprehensive Chinese Text-to-SQL Dataset for Database Manipulation and Query
Xi Chen | Jinguo You | Likun Likun | Xiang Li

Text-to-SQL aims to convert natural language into structured query language, which is a challenging task. Current research focuses mainly on read operations and ignores other aspects of database operations such as create, update, and delete operations. The benchmark datasets as well as models that have been proposed also fail to cover these operations, limiting the development and practical applications in the field. To bridge this gap, we propose CRUDSQL, a large-scale cross-domain single-table CRUD operations Chinese Text-to-SQL dataset. The dataset contains 10,000 question/SQL pairs involving 625 tables from different domains. To support further research on this dataset, we also propose a baseline method, CRUDParser, which employs a two-phase approach based on BERT and T5 for SQL generation and incorporates two strategies, value matching, and value prompting, for interacting with databases to further improve the performance. The experimental results show that the new operation types bring different challenges for future research, and our approach achieves 67.08% and 83.8% exact set matching accuracy under both read and delete operations in the test set, but only 49.6% and 61.8% under create and update operations. We believe that the proposal of CRUDSQL as well as CRUDParser can provide new directions and possibilities for research and practical applications in the field of Text-to-SQL. The dataset is published at https://github.com/bizard-lab/CRUDSQL.

pdf bib
Normalizing without Modernizing: Keeping Historical Wordforms of Middle French while Reducing Spelling Variants
Raphael Rubino | Johanna Gerlach | Jonathan Mutal | Pierrette Bouillon

Conservation of historical documents benefits from computational methods by alleviating the manual labor related to digitization and modernization of textual content. Languages usually evolve over time and keeping historical wordforms is crucial for diachronic studies and digital humanities. However, spelling conventions did not necessarily exist when texts were originally written and orthographic variations are commonly observed depending on scribes and time periods. In this study, we propose to automatically normalize orthographic wordforms found in historical archives written in Middle French during the 16th century without fully modernizing textual content. We leverage pre-trained models in a low resource setting based on a manually curated parallel corpus and produce additional resources with artificial data generation approaches. Results show that causal language models and knowledge distillation improve over a strong baseline, thus validating the proposed methods.

pdf bib
Anti-LM Decoding for Zero-shot In-context Machine Translation
Suzanna Sia | Alexandra DeLucia | Kevin Duh

Zero-shot In-context learning is the phenomenon where models can perform a task given only the instructions. However, pre-trained large language models are known to be poorly calibrated for zero-shot tasks. One of the most effective approaches to handling this bias is to adopt a contrastive decoding objective, which accounts for the prior probability of generating the next token by conditioning on a context. This work introduces an Anti-Language Model objective with a decay factor designed to address the weaknesses of In-context Machine Translation. We conduct our experiments across 3 model types and sizes, 3 language directions, and for both greedy decoding and beam search. The proposed method outperforms other state-of-the-art decoding objectives, with up to 20 BLEU point improvement from the default objective in some settings.

pdf bib
Defending Against Weight-Poisoning Backdoor Attacks for Parameter-Efficient Fine-Tuning
Shuai Zhao | Leilei Gan | Anh Tuan Luu | Jie Fu | Lingjuan Lyu | Meihuizi Jia | Jinming Wen

Recently, various parameter-efficient fine-tuning (PEFT) strategies for application to language models have been proposed and successfully implemented. However, this raises the question of whether PEFT, which only updates a limited set of model parameters, constitutes security vulnerabilities when confronted with weight-poisoning backdoor attacks. In this study, we show that PEFT is more susceptible to weight-poisoning backdoor attacks compared to the full-parameter fine-tuning method, with pre-defined triggers remaining exploitable and pre-defined targets maintaining high confidence, even after fine-tuning. Motivated by this insight, we developed a Poisoned Sample Identification Module (PSIM) leveraging PEFT, which identifies poisoned samples through confidence, providing robust defense against weight-poisoning backdoor attacks. Specifically, we leverage PEFT to train the PSIM with randomly reset sample labels. During the inference process, extreme confidence serves as an indicator for poisoned samples, while others are clean. We conduct experiments on text classification tasks, five fine-tuning strategies, and three weight-poisoning backdoor attack methods. Experiments show near 100% success rates for weight-poisoning backdoor attacks when utilizing PEFT. Furthermore, our defensive approach exhibits overall competitive performance in mitigating weight-poisoning backdoor attacks.

pdf bib
Select and Summarize: Scene Saliency for Movie Script Summarization
Rohit Saxena | Frank Keller

Abstractive summarization for long-form narrative texts such as movie scripts is challenging due to the computational and memory constraints of current language models. A movie script typically comprises a large number of scenes; however, only a fraction of these scenes are salient, i.e., important for understanding the overall narrative. The salience of a scene can be operationalized by considering it as salient if it is mentioned in the summary. Automatically identifying salient scenes is difficult due to the lack of suitable datasets. In this work, we introduce a scene saliency dataset that consists of human-annotated salient scenes for 100 movies. We propose a two-stage abstractive summarization approach which first identifies the salient scenes in script and then generates a summary using only those scenes. Using QA-based evaluation, we show that our model outperforms previous state-of-the-art summarization methods and reflects the information content of a movie more accurately than a model that takes the whole movie script as input.

pdf bib
Don’t be a Fool: Pooling Strategies in Offensive Language Detection from User-Intended Adversarial Attacks
Seunguk Yu | Juhwan Choi | YoungBin Kim

Offensive language detection is an important task for filtering out abusive expressions and improving online user experiences. However, malicious users often attempt to avoid filtering systems through the involvement of textual noises. In this paper, we propose these evasions as user-intended adversarial attacks that insert special symbols or leverage the distinctive features of the Korean language. Furthermore, we introduce simple yet effective pooling strategies in a layer-wise manner to defend against the proposed attacks, focusing on the preceding layers not just the last layer to capture both offensiveness and token embeddings. We demonstrate that these pooling strategies are more robust to performance degradation even when the attack rate is increased, without directly training of such patterns. Notably, we found that models pre-trained on clean texts could achieve a comparable performance in detecting attacked offensive language, to models pre-trained on noisy texts by employing these pooling strategies.

pdf bib
Z-GMOT: Zero-shot Generic Multiple Object Tracking
Kim Tran | Anh Duy Le Dinh | Tien-Phat Nguyen | Thinh Phan | Pha Nguyen | Khoa Luu | Donald Adjeroh | Gianfranco Doretto | Ngan Le

Despite recent significant progress, Multi-Object Tracking (MOT) faces limitations such as reliance on prior knowledge and predefined categories and struggles with unseen objects. To address these issues, Generic Multiple Object Tracking (GMOT) has emerged as an alternative approach, requiring less prior information. However, current GMOT methods often rely on initial bounding boxes and struggle to handle variations in factors such as viewpoint, lighting, occlusion, and scale, among others. Our contributions commence with the introduction of the Referring GMOT dataset a collection of videos, each accompanied by detailed textual descriptions of their attributes. Subsequently, we propose Z-GMOT, a cutting-edge tracking solution capable of tracking objects from never-seen categories without the need of initial bounding boxes or predefined categories. Within our Z-GMOT framework, we introduce two novel components: (i) iGLIP, an improved Grounded language-image pretraining, for accurately detecting unseen objects with specific characteristics. (ii) MA-SORT, a novel object association approach that adeptly integrates motion and appearance-based matching strategies to tackle the complex task of tracking objects with high similarity. Our contributions are benchmarked through extensive experiments conducted on the Referring GMOT dataset for GMOT task. Additionally, to assess the generalizability of the proposed Z-GMOT, we conduct ablation studies on the DanceTrack and MOT20 datasets for the MOT task. Our dataset, code, and models are released at: https://fsoft-aic.github.io/Z-GMOT

pdf bib
NLP for Counterspeech against Hate: A Survey and How-To Guide
Helena Bonaldi | Yi-Ling Chung | Gavin Abercrombie | Marco Guerini

In recent years, counterspeech has emerged as one of the most promising strategies to fight online hate. These non-escalatory responses tackle online abuse while preserving the freedom of speech of the users, and can have a tangible impact in reducing online and offline violence. Recently, there has been growing interest from the Natural Language Processing (NLP) community in addressing the challenges of analysing, collecting, classifying, and automatically generating counterspeech, to reduce the huge burden of manually producing it. In particular, researchers have taken different directions in addressing these challenges, thus providing a variety of related tasks and resources. In this paper, we provide a guide for doing research on counterspeech, by describing - with detailed examples - the steps to undertake, and providing best practices that can be learnt from the NLP studies on this topic. Finally, we discuss open challenges and future directions of counterspeech research in NLP.

pdf bib
PRODIGy: a PROfile-based DIalogue Generation dataset
Daniela Occhipinti | Serra Tekiroglu | Marco Guerini

Providing dialogue agents with a profile representation can improve their consistency and coherence, leading to better conversations. However, current profile-based dialogue datasets for training such agents contain either explicit profile representations that are simple and dialogue-specific, or implicit representations that are difficult to collect. In this work, we introduce the PRODIGy (PROfile-based DIalogue Generation) dataset, which brings diverse representations together, providing a more comprehensive profile dimension set for each speaker. This resource comprises more than 20k dialogues, sourced from movie scripts, aligned with speaker representations such as communication style, biography, personality and gender. Initial experiments with diverse baselines show that providing generative language models with these aspects of a profile, both separately and jointly, enhances models’ performance. This improvement holds true in both in-domain and cross-domain settings, for both fine-tuned and instruction-based LLMs.

pdf bib
WaterJudge: Quality-Detection Trade-off when Watermarking Large Language Models
Piotr Molenda | Adian Liusie | Mark Gales

Watermarking generative-AI systems, such as LLMs, has gained considerable interest, driven by their enhanced capabilities across a wide range of tasks. Although current approaches have demonstrated that small, context-dependent shifts in the word distributions can be used to apply and detect watermarks, there has been little work in analyzing the impact that these perturbations have on the quality of generated texts. Balancing high detectability with minimal performance degradation is crucial in terms of selecting the appropriate watermarking setting; therefore this paper proposes a simple analysis framework where comparative assessment, a flexible NLG evaluation framework, is used to assess the quality degradation caused by a particular watermark setting. We demonstrate that our framework provides easy visualization of the quality-detection trade-off of watermark settings, enabling a simple solution to find an LLM watermark operating point that provides a well-balanced performance. This approach is applied to two different summarization systems and a translation system, enabling cross-model analysis for a task, and cross-task analysis.

pdf bib
Cognitive Overload: Jailbreaking Large Language Models with Overloaded Logical Thinking
Nan Xu | Fei Wang | Ben Zhou | Bangzheng Li | Chaowei Xiao | Muhao Chen

While large language models (LLMs) have demonstrated increasing power, they have also called upon studies on their vulnerabilities. As representatives, jailbreak attacks can provoke harmful or unethical responses from LLMs, even after safety alignment. In this paper, we investigate a novel category of jailbreak attacks specifically designed to target the cognitive structure and processes of LLMs. Specifically, we analyze the safety vulnerability of LLMs in the face of 1) multilingual cognitive overload, 2) veiled expression, and 3) effect-to- cause reasoning. Different from previous jailbreak attacks, our proposed cognitive overload is a black-box attack with no need for knowledge of model architecture or access to model weights. Experiments conducted on AdvBench and MasterKey reveal that various LLMs, including both popular open-source model Llama 2 and the proprietary model ChatGPT, can be compromised through cognitive overload. Motivated by cognitive psychology work on managing cognitive load, we further investigate defending cognitive overload attack from two perspectives. Empirical studies show that our cognitive overload from three perspectives can jailbreak all studied LLMs successfully, while existing defense strategies can hardly mitigate the caused malicious uses effectively.

pdf bib
PAELLA: Parameter-Efficient Lightweight Language-Agnostic Captioning Model
Rita Ramos | Emanuele Bugliarello | Bruno Martins | Desmond Elliott

We introduce PAELLA, a Parameter-Efficient Lightweight Language-Agnostic image captioning model designed to be both parameter and data-efficient using retrieval augmentation. The model is trained by learning a small mapping network with 34M parameters between a pre-trained visual model and a multilingual language model that is conditioned on two types of input: (i) the image itself, and (ii) a set of retrieved captions in the target language. The retrieved examples play a key role in guiding the model to generate captions across languages. Through retrieval, the model can be lightweight in terms of the number of trainable parameters, which only exist in its mapping network, and also in the amount of multilingual training data that is required. Experiments on the XM3600 dataset, featuring 36 languages, show that PAELLA can outperform or compete against some models with 3–77× more learned parameters and 35–863× more data, particularly in low-resource languages. We also find that PAELLA can be trained on only monolingual data and still show strong zero-shot abilities in other languages.

pdf bib
OSCaR: Object State Captioning and State Change Representation
Nguyen Nguyen | Jing Bi | Ali Vosoughi | Yapeng Tian | Pooyan Fazli | Chenliang Xu

The capability of intelligent models to extrapolate and comprehend changes in object states is a crucial yet demanding aspect of AI research, particularly through the lens of human interaction in real-world settings. This task involves describing complex visual environments, identifying active objects, and interpreting their changes as conveyed through language. Traditional methods, which isolate object captioning and state change detection, offer a limited view of dynamic environments. Moreover, relying on a small set of symbolic words to represent changes has restricted the expressiveness of language. To address these challenges, in this paper, we introduce the Object State Captioning and State Change Representation (OSCaR) dataset and benchmark. OSCaR consists of 14,084 annotated video segments with nearly 1,000 unique objects from various egocentric video collections. It sets a new testbed for evaluating Multimodal Large Language Models (MLLMs). Our experiments demonstrate that while MLLMs show some skill, they lack a full understanding of object state changes. The benchmark includes a fine-tuned model that, despite initial capabilities, requires significant improvements in accuracy and generalization ability for effective understanding of these changes. Our code and dataset are available at https://github.com/nguyennm1024/OSCaR.

pdf bib
SumCSE: Summary as a transformation for Contrastive Learning
Raghuveer Thirukovalluru | Xiaolan Wang | Jun Chen | Shuyang Li | Jie Lei | Rong Jin | Bhuwan Dhingra

Sentence embedding models are typically trained using contrastive learning (CL), either using human annotations directly or by repurposing other annotated datasets. In this work, we explore the recently introduced paradigm of generating CL data using generative language models (LM). In CL for computer vision (CV), compositional transformations (series of operations applied over an image. e.g. cropping + color distortion) which modify the input/image to retain minimal information were shown to be very effective. We show that composition of a ‘Summary’ transformation with diverse paraphrasing/contradicting transformations accomplishes the same and works very well in CL for sentence embeddings. Our final generated dataset (using Vicuna-13B) significantly outperforms the previous best unsupervised method (using ChatGPT) by 1.8 points, and SimCSE, a strong supervised baseline by 0.3 points on the semantic text similarity (STS) benchmark.

pdf bib
The Curious Decline of Linguistic Diversity: Training Language Models on Synthetic Text
Yanzhu Guo | Guokan Shang | Michalis Vazirgiannis | Chloé Clavel

This study investigates the consequences of training language models on synthetic data generated by their predecessors, an increasingly prevalent practice given the prominence of powerful generative models. Diverging from the usual emphasis on performance metrics, we focus on the impact of this training methodology on linguistic diversity, especially when conducted recursively over time. To assess this, we adapt and develop a set of novel metrics targeting lexical, syntactic, and semantic diversity, applying them in recursive finetuning experiments across various natural language generation tasks in English. Our findings reveal a consistent decrease in the diversity of the model outputs through successive iterations, especially remarkable for tasks demanding high levels of creativity. This trend underscores the potential risks of training language models on synthetic text, particularly concerning the preservation of linguistic richness. Our study highlights the need for careful consideration of the long-term effects of such training approaches on the linguistic capabilities of language models.

pdf bib
PersonaLLM: Investigating the Ability of Large Language Models to Express Personality Traits
Hang Jiang | Xiajie Zhang | Xubo Cao | Cynthia Breazeal | Deb Roy | Jad Kabbara

Despite the many use cases for large language models (LLMs) in creating personalized chatbots, there has been limited research on evaluating the extent to which the behaviors of personalized LLMs accurately and consistently reflect specific personality traits. We consider studying the behavior of LLM-based agents which we refer to as LLM personas and present a case study with GPT-3.5 and GPT-4 to investigate whether LLMs can generate content that aligns with their assigned personality profiles. To this end, we simulate distinct LLM personas based on the Big Five personality model, have them complete the 44-item Big Five Inventory (BFI) personality test and a story writing task, and then assess their essays with automatic and human evaluations. Results show that LLM personas’ self-reported BFI scores are consistent with their designated personality types, with large effect sizes observed across five traits. Additionally, LLM personas’ writings have emerging representative linguistic patterns for personality traits when compared with a human writing corpus. Furthermore, human evaluation shows that humans can perceive some personality traits with an accuracy of up to 80%. Interestingly, the accuracy drops significantly when the annotators were informed of AI authorship.

pdf bib
FIRE: A Dataset for Financial Relation Extraction
Hassan Hamad | Abhinav Kumar Thakur | Nijil Kolleri | Sujith Pulikodan | Keith Chugg

This paper introduces FIRE (**FI**nancial **R**elation **E**xtraction), a sentence-level dataset of named entities and relations within the financial sector. Comprising 3,025 instances, the dataset encapsulates 13 named entity types along with 18 relation types. Sourced from public financial reports and financial news articles, FIRE captures a wide array of financial information about a business including, but not limited to, corporate structure, business model, revenue streams, and market activities such as acquisitions. The full dataset was labeled by a single annotator to minimize labeling noise. The labeling time for each sentence was recorded during the labeling process. We show how this feature, along with curriculum learning techniques, can be used to improved a model’s performance. The FIRE dataset is designed to serve as a valuable resource for training and evaluating machine learning algorithms in the domain of financial information extraction. The dataset and the code to reproduce our experimental results are available at https://github.com/hmhamad/FIRE. The repository for the labeling tool can be found at https://github.com/abhinav-kumar-thakur/relation-extraction-annotator.

pdf bib
MusiLingo: Bridging Music and Text with Pre-trained Language Models for Music Captioning and Query Response
Zihao Deng | Yinghao Ma | Yudong Liu | Rongchen Guo | Ge Zhang | Wenhu Chen | Wenhao Huang | Emmanouil Benetos

Large Language Models (LLMs) have shown immense potential in multimodal applications, yet the convergence of textual and musical domains remains not well-explored. To address this gap, we present MusiLingo, a novel system for music caption generation and music-related query responses. MusiLingo employs a single projection layer to align music representations from the pre-trained frozen music audio model MERT (CITATION) with a frozen LLM, bridging the gap between music audio and textual contexts. We train it on an extensive music caption dataset and fine-tune it with instructional data. Due to the scarcity of high-quality music Q&A datasets, we created the MusicInstruct (MI) dataset from captions in the MusicCaps datasets, tailored for open-ended music inquiries. Empirical evaluations demonstrate its competitive performance in generating music captions and composing music-related Q&A pairs. Our introduced dataset enables notable advancements beyond previous ones.

pdf bib
Investigating Acceleration of LLaMA Inference by Enabling Intermediate Layer Decoding via Instruction Tuning with ‘LITE’
Neeraj Varshney | Agneet Chatterjee | Mihir Parmar | Chitta Baral

Large Language Models (LLMs) have achieved remarkable performance across a wide variety of tasks; however, their large size makes their inference slow and computationally expensive. Focusing on this problem, we study instruction tuning LLMs with additional explicit Losses from the Intermediate layers (LITE) and show that it enables these layers to acquire ‘good’ generation ability without affecting the generation ability of the final layer. We then perform ‘dynamic confidence-based early exiting’ at token level from the intermediate layers which improves the computational efficiency of text generation without sacrificing the quality of the generation. We conduct comprehensive experiments by instruction tuning LLaMA-2 models on the Alpaca dataset and evaluate on four different instruction test sets. We show that dynamic early exiting achieves consistent and considerable inference cost improvements (37.86% for 7B and 46.35% for 13B model) while maintaining the generation quality. We further conduct a thorough analysis of the results and dissect the efficiency improvements which reveals several important findings.

pdf bib
Instruction-following Evaluation through Verbalizer Manipulation
Shiyang Li | Jun Yan | Hai Wang | Zheng Tang | Xiang Ren | Vijay Srinivasan | Hongxia Jin

While instruction-tuned models have shown remarkable success in various natural language processing tasks, accurately evaluating their ability to follow instructions remains challenging. Existing benchmarks primarily focus on common instructions that align well with what the model learned during training. However, proficiency in responding to these instructions does not necessarily imply strong ability in instruction following. In this paper, we propose a novel instruction-following evaluation protocol called verbalizer manipulation. It instructs the model to verbalize the task label with words aligning with model priors to different extents, adopting verbalizers from highly aligned (e.g., outputting “positive” for positive sentiment), to minimally aligned (e.g., outputting “negative” for positive sentiment). Verbalizer manipulation can be seamlessly integrated with any classification benchmark to examine the model’s reliance on priors and its ability to override them to accurately follow the instructions. We conduct a comprehensive evaluation of four major model families across nine datasets, employing twelve sets of verbalizers for each of them. We observe that the instruction-following abilities of models, across different families and scales, are significantly distinguished by their performance on less natural verbalizers. Even the strongest GPT-4 model struggles to perform better than random guessing on the most challenging verbalizer, emphasizing the need for continued advancements to improve their instruction-following abilities.

pdf bib
WebWISE: Unlocking Web Interface Control for LLMs via Sequential Exploration
Heyi Tao | Sethuraman T V | Michal Shlapentokh-Rothman | Tanmay Gupta | Heng Ji | Derek Hoiem

This paper investigates using Large Language Models (LLMs) to automatically perform web software tasks using click, scroll, and text in- put operations. Previous approaches, such as reinforcement learning (RL) or imitation learning, are inefficient to train and task-specific. Our method uses filtered Document Object Model (DOM) elements as observations and performs tasks step-by-step, sequentially generating small programs based on the current observations. We use in-context learning, either benefiting from a single manually provided example, or an automatically generated example based on a successful zero-shot trial. We evaluate our proposed method on the MiniWob++ benchmark. With only one in-context example, our WebWISE method using gpt-3.5-turbo achieves similar or better performance than other methods that require many demonstrations or trials.

pdf bib
CodecLM: Aligning Language Models with Tailored Synthetic Data
Zifeng Wang | Chun-Liang Li | Vincent Perot | Long Le | Jin Miao | Zizhao Zhang | Chen-Yu Lee | Tomas Pfister

Instruction tuning has emerged as the key in aligning large language models (LLMs) with specific task instructions, thereby mitigating the discrepancy between the next-token prediction objective and users’ actual goals. To reduce the labor and time cost to collect or annotate data by humans, researchers start to explore the use of LLMs to generate instruction-aligned synthetic data. Recent works focus on generating diverse instructions and applying LLM to increase instruction complexity, often neglecting downstream use cases. It remains unclear how to tailor high-quality data to elicit better instruction-following abilities in different target instruction distributions and LLMs. To this end, we introduce CodecLM, a general framework for adaptively generating high-quality synthetic data for LLM alignment with different downstream instruction distributions and LLMs. Drawing on the Encode-Decode principles, we use LLMs as codecs to guide the data generation process. We first encode seed instructions into metadata, which are concise keywords generated on-the-fly to capture the target instruction distribution, and then decode metadata to create tailored instructions. We also introduce Self-Rubrics and Contrastive Filtering during decoding to tailor data-efficient samples. Extensive experiments on four open-domain instruction following benchmarks validate the effectiveness of CodecLM over the current state-of-the-arts.

pdf bib
Prompting Few-shot Multi-hop Question Generation via Comprehending Type-aware Semantics
Zefeng Lin | Weidong Chen | Yan Song | Yongdong Zhang

Given several documents, multi-hop question generation (MQG) is a task aims to generate complicated questions that require reasoning over multiple pieces of these documents to find the answer. To perform this task, existing studies focus on designing advanced architectures to locate essential keywords or sentences in multiple documents and then generate questions accordingly, where they normally do not note that question types could provide crucial hints for extracting key information from the documents for MQG. In general, supervised approaches are used that rely on large annotated data, which is not available in many low-resource scenarios and thus makes MQG hard in these domains. Consider the recent success of large language models (LLMs) on natural language processing tasks using limited labeled data under few-shot settings, in this paper, we propose an approach named type-aware semantics extraction-based chain-of-thought method (TASE-CoT) for few-shot MQG. Specifically, our approach firstly extracts question types and essential semantic phrases from the given documents and the answer. Then, we design a three-step CoT template to leverage the extracted question type and semantic phrases to predict multi-hop questions. Extensive experiments and the results demonstrate the effectiveness of our approach and the proposed modules.

pdf bib
When Hindsight is Not 20/20: Testing Limits on Reflective Thinking in Large Language Models
Yanhong Li | Chenghao Yang | Allyson Ettinger

Recent studies suggest that self-reflective prompting can significantly enhance the reasoning capabilities of Large Language Models (LLMs). However, the use of external feedback as a stop criterion raises doubts about the true extent of LLMs’ ability to emulate human-like self-reflection. In this paper, we set out to clarify these capabilities under a more stringent evaluation setting in which we disallow any kind of external feedback. Our findings under this setting show a split: while self-reflection enhances performance in TruthfulQA, it adversely affects results in HotpotQA.We conduct follow-up analyses to clarify the contributing factors in these patterns, and find that the influence of self-reflection is impacted both by reliability of accuracy in models’ initial responses, and by overall question difficulty: specifically, self-reflection shows the most benefit when models are less likely to be correct initially, and when overall question difficulty is higher. We also find that self-reflection reduces tendency toward majority voting. Based on our findings, we propose guidelines for decisions on when to implement self-reflection. We release the codebase for reproducing our experiments at https://github.com/yanhong-lbh/LLM-SelfReflection-Eval.

pdf bib
CoDa: Constrained Generation based Data Augmentation for Low-Resource NLP
Chandra Kiran Evuru | Sreyan Ghosh | Sonal Kumar | Ramaneswaran S | Utkarsh Tyagi | Dinesh Manocha

We present CoDa (**Co**nstrained Generation based **Da**ta Augmentation), a controllable, effective, and *training-free* data augmentation technique for low-resource (data-scarce) NLP. Our approach is based on prompting off-the-shelf instruction-following Large Language Models (LLMs) for generating text that satisfies a set of constraints. Precisely, we extract a set of simple constraints from every instance in the low-resource dataset and verbalize them to prompt an LLM to generate novel and diverse training instances. Our findings reveal that synthetic data that follows simple constraints in the downstream dataset act as highly effective augmentations, and CoDa can achieve this without intricate decoding-time constrained generation techniques or fine-tuning with complex algorithms that eventually make the model biased toward the small number of training instances. Additionally, CoDa is the first framework that provides users explicit control over the augmentation generation process, thereby also allowing easy adaptation to several domains. We demonstrate the effectiveness of CoDa across 11 datasets spanning 3 tasks and 3 low-resource settings. CoDa outperforms all our baselines, qualitatively and quantitatively, with improvements of 0.12%-7.19%. Code is available.

pdf bib
Synonym relations affect object detection learned on vision-language data
Giacomo Nebbia | Adriana Kovashka

We analyze whether object detectors trained on vision-language data learn effective visual representations for synonyms. Since many current vision-language models accept user-provided textual input, we highlight the need for such models to learn feature representations that are robust to changes in how such input is provided. Specifically, we analyze changes in synonyms used to refer to objects. Here, we study object detectors trained on vision-language data and investigate how to make their performance less dependent on whether synonyms are used to refer to an object. We propose two approaches to achieve this goal: data augmentation by back-translation and class embedding enrichment. We show the promise of such approaches, reporting improved performance on synonyms from mAP@0.5=33.87% to 37.93%.

pdf bib
CM-TTS: Enhancing Real Time Text-to-Speech Synthesis Efficiency through Weighted Samplers and Consistency Models
Xiang Li | FanBu FanBu | Ambuj Mehrish | Yingting Li | Jiale Han | Bo Cheng | Soujanya Poria

Neural Text-to-Speech (TTS) systems find broad applications in voice assistants, e-learning, and audiobook creation. The pursuit of modern models, like Diffusion Models (DMs), holds promise for achieving high-fidelity, real-time speech synthesis. Yet, the efficiency of multi-step sampling in Diffusion Models presents challenges. Efforts have been made to integrate GANs with DMs, speeding up inference by approximating denoising distributions, but this introduces issues with model convergence due to adversarial training. To overcome this, we introduce CM-TTS, a novel architecture grounded in consistency models (CMs). Drawing inspiration from continuous-time diffusion models, CM-TTS achieves top-quality speech synthesis in fewer steps without adversarial training or pre-trained model dependencies. We further design weighted samplers to incorporate different sampling positions into model training with dynamic probabilities, ensuring unbiased learning throughout the entire training process. We present a real-time mel-spectrogram generation consistency model, validated through comprehensive evaluations. Experimental results underscore CM-TTS’s superiority over existing single-step speech synthesis systems, representing a significant advancement in the field.

pdf bib
RobustSentEmbed: Robust Sentence Embeddings Using Adversarial Self-Supervised Contrastive Learning
Javad Rafiei Asl | Prajwal Panzade | Eduardo Blanco | Daniel Takabi | Zhipeng Cai

Pre-trained language models (PLMs) have consistently demonstrated outstanding performance across a diverse spectrum of natural language processing tasks. Nevertheless, despite their success with unseen data, current PLM-based representations often exhibit poor robustness in adversarial settings. In this paper, we introduce RobustSentEmbed, a self-supervised sentence embedding framework designed to improve both generalization and robustness in diverse text representation tasks and against a diverse set of adversarial attacks. Through the generation of high-risk adversarial perturbations and their utilization in a novel objective function, RobustSentEmbed adeptly learns high-quality and robust sentence embeddings. Our experiments confirm the superiority of RobustSentEmbed over state-of-the-art representations. Specifically, Our framework achieves a significant reduction in the success rate of various adversarial attacks, notably reducing the BERTAttack success rate by almost half (from 75.51% to 38.81%). The framework also yields improvements of 1.59% and 0.23% in semantic textual similarity tasks and various transfer tasks, respectively.

pdf bib
Characterizing Human and Zero-Shot GPT-3.5 Object-Similarity Judgments
D McKnight | Alona Fyshe

Recent advancements in large language models’ (LLMs) capabilities have yielded few-shot, human-comparable performance on a range of tasks. At the same time, researchers expend significant effort and resources gathering human annotations. At some point, LLMs may be able to perform some simple annotation tasks, but studies of LLM annotation accuracy and behavior are sparse. In this paper, we characterize OpenAI’s GPT-3.5’s judgment on a behavioral task for implicit object categorization. We characterize the embedding spaces of models trained on human vs. GPT responses and give similarities and differences between them, finding many similar dimensions. We also find that despite these similar dimensions, augmenting humans’ responses with GPT ones drives model divergence across the sizes of datasets tested.

pdf bib
Self-Demos: Eliciting Out-of-Demonstration Generalizability in Large Language Models
Wei He | Shichun Liu | Jun Zhao | Yiwen Ding | Yi Lu | Zhiheng Xi | Tao Gui | Qi Zhang | Xuanjing Huang

Large language models (LLMs) have shown promising abilities of in-context learning (ICL), adapting swiftly to new tasks with only few-shot demonstrations. However, current few-shot methods heavily depend on high-quality, query-specific demos, which are often lacking. When faced with out-of-demonstration (OOD) queries, methods that rely on hand-crafted demos or external retrievers might fail. To bridge the gap between limited demos and OOD queries, we propose Self-Demos, a novel prompting method that elicits the inherent generalizability in LLMs by query-aware demo generation. The generated demos strategically interpolate between existing demos and the given query, transforming the query from OOD to ID. To evaluate the effectiveness of our approach, we manually constructed OOD-Toolset, a dataset in the tool-using scenario with over 300 real-world APIs and 1000 instances, each consisting of three tool-use cases as demos and an OOD query. Thorough experiments on our dataset and two public math benchmarks have shown that our method can outperform state-of-the-art baselines in the OOD setting. Moreover, we conduct a range of analyses to validate Self-Demos’s generalization and provide more insights.

pdf bib
Getting Sick After Seeing a Doctor? Diagnosing and Mitigating Knowledge Conflicts in Event Temporal Reasoning
Tianqing Fang | Zhaowei Wang | Wenxuan Zhou | Hongming Zhang | Yangqiu Song | Muhao Chen

Event temporal reasoning aims at identifying the temporal relations between two or more events from narratives. However, knowledge conflicts arise when there is a mismatch between the actual temporal relations of events in the context and the prior knowledge or biases learned by the model. In this paper, we propose to detect knowledge-conflict examples in event temporal reasoning using bias indicators, which include event relation prior bias, tense bias, narrative bias, and dependency bias. We define conflict examples as those where event relations are opposite to biased or prior relations. To mitigate event-related knowledge conflicts, we introduce a Counterfactual Data Augmentation (CDA) based method that can be applied to both Pre-trained Language Models (PLMs) and Large Language Models (LLMs) either as additional training data or demonstrations for In- Context Learning. Experiments suggest both PLMs and LLMs suffer from knowledge conflicts in event temporal reasoning, and CDA has the potential for reducing hallucination and improving model performance.

pdf bib
MCECR: A Novel Dataset for Multilingual Cross-Document Event Coreference Resolution
Amir Pouran Ben Veyseh | Viet Lai | Chien Nguyen | Franck Dernoncourt | Thien Nguyen

Event coreference resolution (ECR) is a critical task in information extraction of natural language processing, aiming to identify and link event mentions across multiple documents. Despite recent progress, existing datasets for ECR primarily focus on within-document event coreference and English text, lacking cross-document ECR datasets for multiple languages beyond English. To address this issue, this work presents the first multiligual dataset for cross-document ECR, called MCECR (Multilingual Cross-Document Event Coreference Resolution), that manually annotates a diverse collection of documents for event mentions and coreference in five languages, i.e., English, Spanish, Hindi, Turkish, and Ukrainian. Using sampled articles from Wikinews over various topics as the seeds, our dataset fetches related news articles from the Google search engine to increase the number of non-singleton event clusters. In total, we annotate 5,802 news articles, providing a substantial and varied dataset for multilingual ECR in both within-document and cross-document scenarios. Extensive analysis of the proposed dataset reveals the challenging nature of multilingual event coreference resolution tasks, promoting MCECR as a strong benchmark dataset for future research in this area.

pdf bib
Sentiment Analysis in the Era of Large Language Models: A Reality Check
Wenxuan Zhang | Yue Deng | Bing Liu | Sinno Pan | Lidong Bing

Sentiment analysis (SA) has been a long-standing research area in natural language processing. With the recent advent of large language models (LLMs), there is great potential for their employment on SA problems. However, the extent to which current LLMs can be leveraged for different sentiment analysis tasks remains unclear. This paper aims to provide a comprehensive investigation into the capabilities of LLMs in performing various sentiment analysis tasks, from conventional sentiment classification to aspect-based sentiment analysis and multifaceted analysis of subjective texts. We evaluate performance across 13 tasks on 26 datasets and compare the results against small language models (SLMs) trained on domain-specific datasets. Our study reveals that while LLMs demonstrate satisfactory performance in simpler tasks, they lag behind in more complex tasks requiring a deeper understanding of specific sentiment phenomena or structured sentiment information. However, LLMs significantly outperform SLMs in few-shot learning settings, suggesting their potential when annotation resources are limited. We also highlight the limitations of current evaluation practices in assessing LLMs’ SA abilities and propose a novel benchmark, SentiEval, for a more comprehensive and realistic evaluation. Data and code are available at https://github.com/DAMO-NLP-SG/LLM-Sentiment.

pdf bib
Tokenizer Choice For LLM Training: Negligible or Crucial?
Mehdi Ali | Michael Fromm | Klaudia Thellmann | Richard Rutmann | Max Lübbering | Johannes Leveling | Katrin Klug | Jan Ebert | Niclas Doll | Jasper Buschhoff | Charvi Jain | Alexander Weber | Lena Jurkschat | Hammam Abdelwahab | Chelsea John | Pedro Ortiz Suarez | Malte Ostendorff | Samuel Weinbach | Rafet Sifa | Stefan Kesselheim | Nicolas Flores-Herr

The recent success of large language models (LLMs) has been predominantly driven by curating the training dataset composition, scaling of model architectures and dataset sizes and advancements in pretraining objectives, leaving tokenizer influence as a blind spot.Shedding light on this underexplored area, we conduct a comprehensive study on the influence of tokenizer choice on LLM downstream performance by training 24 mono- and multilingual LLMs at a 2.6B parameter scale, ablating different tokenizer algorithms and parameterizations. Our studies highlight that the tokenizer choice can significantly impact the model’s downstream performance and training costs. In particular, we find that the common tokenizer evaluation metrics fertility and parity are not always predictive of model downstream performance, rendering these metrics a questionable proxy for the model’s downstream performance. Furthermore, we show that multilingual tokenizers trained on the five most frequent European languages require vocabulary size increases of factor three in comparison to English. While English-centric tokenizers have been applied to the training of multi-lingual LLMs in the past, we find that this approach results in a severe downstream performance degradation and additional training costs of up to 68%, due to an inefficient tokenization vocabulary.

pdf bib
Think Before You Speak: Cultivating Communication Skills of Large Language Models via Inner Monologue
Junkai Zhou | Liang Pang | Huawei Shen | Xueqi Cheng

The emergence of large language models (LLMs) further improves the capabilities of open-domain dialogue systems and can generate fluent, coherent, and diverse responses. However, LLMs still lack a crucial ability: communication skills. This limitation renders them more like information seeking tools rather than anthropomorphic chatbots. Communication skills, such as topic transition, proactively asking questions, concept guidance, empathy, and summarising often should be taken into consideration, to make LLMs more anthropomorphic and proactive during the conversation, thereby increasing the interest of users and attracting them to chat for longer. However, enabling these communication skills in black-box LLMs remains a key challenge because they do not have the same utterance formation mode as real people: think before speaking. Inspired by linguistics and cognitive science, we empower LLMs with communication skills through inner monologues. To evaluate various communication skills, we construct a benchmark named Cskills, which can also more comprehensively evaluate the dialogue generation ability of the model. Experimental results show that the proposed CSIM strategy improves the backbone models and outperforms the baselines.

pdf bib
The Impact of Differential Privacy on Group Disparity Mitigation
Victor Hansen | Atula Neerkaje | Ramit Sawhney | Lucie Flek | Anders Søgaard

The performance cost of differential privacy has, for some applications, been shown to be higher for minority groups; fairness, conversely, has been shown to disproportionally compromise the privacy of members of such groups. Most work in this area has been restricted to computer vision and risk assessment. In response, we evaluate the impact of differential privacy on fairness across four diverse tasks, focusing on how attempts to mitigate privacy violations and between-group performance differences interact: Does privacy inhibit attempts to ensure fairness? To this end, we train (𝜀,𝛿)-differentially private models with empirical risk minimization and group distributionally robust training objectives. Consistent with previous findings, we find that differential privacy increases between-group performance differences in the baseline setting; more interestingly, differential privacy reduces between-group performance differences in the robust setting. We explain this by interpreting differential privacy as regularization.

pdf bib
Isometric Neural Machine Translation using Phoneme Count Ratio Reward-based Reinforcement Learning
Shivam Mhaskar | Nirmesh Shah | Mohammadi Zaki | Ashishkumar Gudmalwar | Pankaj Wasnik | Rajiv Shah

Traditional Automatic Video Dubbing (AVD) pipeline consists of three key modules, namely, Automatic Speech Recognition (ASR), Neural Machine Translation (NMT), and Text-to-Speech (TTS). Within AVD pipelines, isometric-NMT algorithms are employed to regulate the length of the synthesized output text. This is done to guarantee synchronization with respect to the alignment of video and audio subsequent to the dubbing process. Previous approaches have focused on aligning the number of characters and words in the source and target language texts of Machine Translation models. However, our approach aims to align the number of phonemes instead, as they are closely associated with speech duration. In this paper, we present the development of an isometric NMT system using Reinforcement Learning (RL), with a focus on optimizing the alignment of phoneme counts in the source and target language sentence pairs. To evaluate our models, we propose the Phoneme Count Compliance (PCC) score, which is a measure of length compliance. Our approach demonstrates a substantial improvement of approximately 36% in the PCC score compared to the state-of-the-art models when applied to English-Hindi language pairs. Moreover, we propose a student-teacher architecture within the framework of our RL approach to maintain a trade-off between the phoneme count and translation quality.

pdf bib
Read between the lines - Functionality Extraction From READMEs
Prince Kumar | Srikanth Tamilselvam | Dinesh Garg

While text summarization is a well-known NLP task, in this paper, we introduce a novel and useful variant of it called functionality extraction from Git README files. Though this task is a text2text generation at an abstract level, it involves its own peculiarities and challenges making existing text2text generation systems not very useful. The motivation behind this task stems from a recent surge in research and development activities around the use of large language models for code-related tasks, such as code refactoring, code summarization, etc. We also release a human-annotated dataset called FuncRead, and develop a battery of models for the task. Our exhaustive experimentation shows that small size fine-tuned models beat any baseline models that can be designed using popular black-box or white-box large language models (LLMs) such as ChatGPT and Bard. Our best fine-tuned 7 Billion CodeLlama model exhibit 70% and 20% gain on the F1 score against ChatGPT and Bard respectively.

pdf bib
AbsPyramid: Benchmarking the Abstraction Ability of Language Models with a Unified Entailment Graph
Zhaowei Wang | Haochen Shi | Weiqi Wang | Tianqing Fang | Hongming Zhang | Sehyun Choi | Xin Liu | Yangqiu Song

Cognitive research indicates that abstraction ability is essential in human intelligence, which remains under-explored in language models. In this paper, we present AbsPyramid, a unified entailment graph of 221K textual descriptions of abstraction knowledge. While existing resources only touch nouns or verbs within simplified events or specific domains, AbsPyramid collects abstract knowledge for three components of diverse events to comprehensively evaluate the abstraction ability of language models in the open domain. Experimental results demonstrate that current LLMs face challenges comprehending abstraction knowledge in zero-shot and few-shot settings. By training on our rich abstraction knowledge, we find LLMs can acquire basic abstraction abilities and generalize to unseen events. In the meantime, we empirically show that our benchmark is comprehensive to enhance LLMs across two previous abstraction tasks.

pdf bib
Few-TK: A Dataset for Few-shot Scientific Typed Keyphrase Recognition
Avishek Lahiri | Pratyay Sarkar | Medha Sen | Debarshi Sanyal | Imon Mukherjee

Scientific texts are distinctive from ordinary texts in quite a few aspects like their vocabulary and discourse structure. Consequently, Information Extraction (IE) tasks for scientific texts come with their own set of challenges. The classical definition of Named Entities restricts the inclusion of all scientific terms under its hood, which is why previous works have used the terms Named Entities and Keyphrases interchangeably. We suggest the rechristening of Named Entities for the scientific domain as Typed Keyphrases (TK), broadening their scope. We advocate for exploring this task in the few-shot domain due to the scarcity of labeled scientific IE data. Currently, no dataset exists for few-shot scientific Typed Keyphrase Recognition. To address this gap, we develop an annotation schema and present Few-TK, a dataset in the AI/ML field that includes scientific Typed Keyphrase annotations on abstracts of 500 research papers. To the best of our knowledge, this is the introductory few-shot Typed Keyphrase recognition dataset and only the second dataset structured specifically for few-shot NER, after Few-NERD. We report the results of several few-shot sequence-labelling models applied to our dataset. The data and code are available at https://github.com/AvishekLahiri/Few_TK.git

pdf bib
Language Models can be Deductive Solvers
Jiazhan Feng | Ruochen Xu | Junheng Hao | Hiteshi Sharma | Yelong Shen | Dongyan Zhao | Weizhu Chen

Logical reasoning is a fundamental aspect of human intelligence and a key component of tasks like problem-solving and decision-making. Recent advancements have enabled Large Language Models (LLMs) to potentially exhibit reasoning capabilities, but complex logical reasoning remains a challenge. The state-of-the-art, solver-augmented language models, use LLMs to parse natural language logical questions into symbolic representations first and then adopt external logical solvers to take in the symbolic representations and output the answers. Despite their impressive performance, any parsing errors will inevitably result in the failure of the execution of external logical solvers and no answer to the logical questions. In this paper, we introduce LoGiPT, a novel language model that directly internalizes and emulates the reasoning processes of logical solvers and avoids parsing errors by learning strict adherence to solver syntax and grammar. LoGiPT is fine-tuned on a newly constructed instruction-tuning dataset derived from revealing and refining the invisible reasoning process of deductive solvers. Experimental results on two public deductive reasoning benchmarks show that LoGiPT outperforms state-of-the-art solver-augmented LMs and few-shot prompting methods on competitive LLMs like GPT-4. This project is available in https://github.com/Cyril-JZ/LoGiPT.

pdf bib
Interpreting User Requests in the Context of Natural Language Standing Instructions
Nikita Moghe | Patrick Xia | Jacob Andreas | Jason Eisner | Benjamin Van Durme | Harsh Jhamtani

Users of natural language interfaces, frequently powered by Large Language Models (LLMs), must often repeat their full set of preferences each time they make a similar request. We describe an approach to LLM-based dialogue modeling in which persistent user constraints and preferences – collectively termed standing instructions – are provided as additional context for such interfaces. For example, when a user states “I’m hungry”, a previously expressed preference for Persian food can be automatically added to the LLM prompt, influencing the search for relevant restaurants.We develop NLSI, a language-to-program dataset consisting of over 2.4K English dialogues spanning 17 domains, in which each dialogue is paired with a user profile (a set of user-specific standing instructions) and corresponding structured representations (a sequence of API calls). A key challenge in NLSI is to identify which subset of the standing instructions is applicable to a given dialogue. NLSI contains diverse phenomena, from simple preferences to interdependent instructions such as triggering a hotel search whenever the user is booking tickets to an event. We conduct experiments on NLSI using prompting with large language models and various retrieval approaches, achieving a maximum of 46% exact match on API prediction. Our results demonstrate the challenges in identifying the relevant standing instructions and their interpretation into API calls

pdf bib
Secure Your Model: An Effective Key Prompt Protection Mechanism for Large Language Models
Ruixiang Tang | Yu-Neng Chuang | Xuanting Cai | Mengnan Du | Xia Hu

Large language models (LLMs) have notably revolutionized many domains within natural language processing due to their exceptional performance. Their security has become increasingly vital. This study is centered on protecting LLMs against unauthorized access and potential theft. We propose a simple yet effective protective measure wherein a unique key prompt is embedded within the LLM. This mechanism enables the model to respond only when presented with the correct key prompt; otherwise, LLMs will refuse to react to any input instructions. This key prompt protection offers a robust solution to prevent the unauthorized use of LLMs, as the model becomes unusable without the correct key. We evaluated the proposed protection on multiple LLMs and NLP tasks. Results demonstrate that our method can successfully protect the LLM without significantly impacting the model’s original function. Moreover, we demonstrate potential attacks that attempt to bypass the protection mechanism will adversely affect the model’s performance, further emphasizing the effectiveness of the proposed protection method.

pdf bib
Enhancing Chain-of-Thoughts Prompting with Iterative Bootstrapping in Large Language Models
Jiashuo Sun | Yi Luo | Yeyun Gong | Chen Lin | Yelong Shen | Jian Guo | Nan Duan

Large language models (LLMs) can achieve impressive performance on various reasoning tasks by incorporating chain-of-thought (CoT) prompting, where step-by-step reasoning is provided to guide LLMs to generate answers to questions, and the question-rationale-answer triplets are utilized as demonstration exemplars. However, the reasoning chains of demonstrations generated by LLMs are observed to be prone to errors, which can subsequently lead to incorrect reasoning during inference. Furthermore, inappropriate exemplars, e.g., overly simplistic or complex exemplars depending on the question’s difficulty level, can affect the LLM’s performance. To address these issues, we introduce Iter-CoT (Iterative bootstrapping in Chain-of-Thoughts prompting). Iter-CoT has two advantages: (1) it adopts iterative bootstrapping that enables LLMs to rectify errors autonomously, resulting in more precise and comprehensive reasoning chains. (2) it selects exemplars of challenging yet answerable (i.e., the LLM has the potential to answer correctly) questions, enhancing the LLMs’ generalizability to answer questions with varying difficulty levels. Experimental results exhibit Iter-CoT superior performance on three distinct reasoning tasks on ten datasets.

pdf bib
Do Prompt Positions Really Matter?
Junyu Mao | Stuart Middleton | Mahesan Niranjan

Prompt-based models have gathered a lot of attention from researchers due to their remarkable advancements in the fields of zero-shot and few-shot learning. Developing an effective prompt template plays a critical role. However, prior studies have mainly focused on prompt vocabulary searching or embedding initialization within a predefined template with the prompt position fixed. In this empirical study, we conduct the most comprehensive analysis to date of prompt position for diverse Natural Language Processing (NLP) tasks. Our findings quantify the substantial impact prompt position has on model performance. We observe that the prompt positions used in prior studies are often sub-optimal, and this observation is consistent even in widely used instruction-tuned models. These findings suggest prompt position optimisation as a valuable research direction to augment prompt engineering methodologies and prompt position-aware instruction tuning as a potential way to build more robust models in the future.

pdf bib
Natural Language Embedded Programs for Hybrid Language Symbolic Reasoning
Tianhua Zhang | Jiaxin Ge | Hongyin Luo | Yung-Sung Chuang | Mingye Gao | Yuan Gong | Yoon Kim | Xixin Wu | Helen Meng | James Glass

How can we perform computations over natural language representations to solve tasks that require symbolic and numeric reasoning? We propose natural language embedded programs (NLEP) as a unifying framework for addressing math/symbolic reasoning, natural language understanding, and instruction following tasks. Our approach prompts a language model to generate full Python programs that define functions over data structures which contain natural language representations of structured knowledge. A Python interpreter then executes the generated code and prints the output. Despite using a task-general prompt, we find that this approach can improve upon strong baselines across a range of different tasks including math and symbolic reasoning, text classification, question answering, and instruction following. We found that the generated programs are interpretable since they outline the exact reasoning process followed by the program interpreter.

pdf bib
A Study on Scaling Up Multilingual News Framing Analysis
Syeda Sabrina Akter | Antonios Anastasopoulos

Media framing is the study of strategically selecting and presenting specific aspects of political issues to shape public opinion. Despite its relevance to almost all societies around the world, research has been limited due to the lack of available datasets and other resources. This study explores the possibility of dataset creation through crowdsourcing, utilizing non-expert annotators to develop training corpora. We first extend framing analysis beyond English news to a multilingual context (12 typologically diverse languages) through automatic translation. We also present a novel benchmark in Bengali and Portuguese on the immigration and same-sex marriage domains.Additionally, we show that a system trained on our crowd-sourced dataset, combined with other existing ones, leads to a 5.32 percentage point increase from the baseline, showing that crowdsourcing is a viable option. Last, we study the performance of large language models (LLMs) for this task, finding that task-specific fine-tuning is a better approach than employing bigger non-specialized models.

pdf bib
ViGLUE: A Vietnamese General Language Understanding Benchmark and Analysis of Vietnamese Language Models
Minh-Nam Tran | Phu-Vinh Nguyen | Long Nguyen | Dien Dinh

As the number of language models has increased, various benchmarks have been suggested to assess the proficiency of the models in natural language understanding. However, there is a lack of such a benchmark in Vietnamese due to the difficulty in accessing natural language processing datasets or the scarcity of task-specific datasets. **ViGLUE**, the proposed dataset collection, is a **Vi**etnamese **G**eneral **L**anguage **U**nderstanding **E**valuation benchmark developed using three methods: translating an existing benchmark, generating new corpora, and collecting available datasets. ViGLUE contains twelve tasks and encompasses over ten areas and subjects, enabling it to evaluate models comprehensively over a broad spectrum of aspects. Baseline models utilizing multilingual language models are also provided for all tasks in the proposed benchmarks. In addition, the study of the available Vietnamese large language models is conducted to explore the language models’ ability in the few-shot learning framework, leading to the exploration of the relationship between specific tasks and the number of shots.

pdf bib
Exploring the Trade-off Between Model Performance and Explanation Plausibility of Text Classifiers Using Human Rationales
Lucas Resck | Marcos M. Raimundo | Jorge Poco

Saliency post-hoc explainability methods are important tools for understanding increasingly complex NLP models. While these methods can reflect the model’s reasoning, they may not align with human intuition, making the explanations not plausible. In this work, we present a methodology for incorporating rationales, which are text annotations explaining human decisions, into text classification models. This incorporation enhances the plausibility of post-hoc explanations while preserving their faithfulness. Our approach is agnostic to model architectures and explainability methods. We introduce the rationales during model training by augmenting the standard cross-entropy loss with a novel loss function inspired by contrastive learning. By leveraging a multi-objective optimization algorithm, we explore the trade-off between the two loss functions and generate a Pareto-optimal frontier of models that balance performance and plausibility. Through extensive experiments involving diverse models, datasets, and explainability methods, we demonstrate that our approach significantly enhances the quality of model explanations without causing substantial (sometimes negligible) degradation in the original model’s performance.

pdf bib
Unlocking Parameter-Efficient Fine-Tuning for Low-Resource Language Translation
Tong Su | Xin Peng | Sarubi Thillainathan | David Guzmán | Surangika Ranathunga | En-Shiun Lee

Parameter-efficient fine-tuning (PEFT) methods are increasingly vital in adapting large-scale pre-trained language models for diverse tasks, offering a balance between adaptability and computational efficiency. They are important in Low-Resource Language (LRL) Neural Machine Translation (NMT) to enhance translation accuracy with minimal resources. However, their practical effectiveness varies significantly across different languages. We conducted comprehensive empirical experiments with varying LRL domains and sizes to evaluate the performance of 8 PEFT methods with in total of 15 architectures using the SacreBLEU score. We showed that 6 PEFT architectures outperform the baseline for both in-domain and out-domain tests and the Houlsby+Inversion adapter has the best performance overall, proving the effectiveness of PEFT methods.

pdf bib
ADaPT: As-Needed Decomposition and Planning with Language Models
Archiki Prasad | Alexander Koller | Mareike Hartmann | Peter Clark | Ashish Sabharwal | Mohit Bansal | Tushar Khot

Large Language Models (LLMs) are increasingly being used for interactive decision-making tasks requiring planning and adapting to the environment. Recent works employ LLMs-as-agents in broadly two ways: iteratively determining the next action (iterative executors) or generating plans and executing sub-tasks using LLMs (plan-and-execute). However, these methods struggle with task complexity, as the inability to execute any sub-task may lead to task failure. To address these shortcomings, we introduce As-Needed Decomposition and Planning for complex Tasks (ADaPT), an approach that explicitly plans and decomposes complex sub-tasks as-needed, i.e., when the LLM is unable to execute them. ADaPT recursively decomposes sub-tasks to adapt to both task complexity and LLM capability. Our results demonstrate that ADaPT substantially outperforms established strong baselines, achieving success rates up to 28.3% higher in ALFWorld, 27% in WebShop, and 33% in TextCraft – a novel compositional dataset that we introduce. Through extensive analysis, we illustrate the importance of multilevel decomposition and establish that ADaPT dynamically adjusts to the capabilities of the executor LLM as well as to task complexity.

pdf bib
Guiding Large Language Models to Post-Edit Machine Translation with Error Annotations
Dayeon Ki | Marine Carpuat

Machine Translation (MT) remains one of the last NLP tasks where large language models (LLMs) have not yet replaced dedicated supervised systems. This work exploits the complementary strengths of LLMs and supervised MT by guiding LLMs to automatically post-edit MT with external feedback on its quality, derived from Multidimensional Quality Metric (MQM) annotations. Working with LLaMA-2 models, we consider prompting strategies varying the nature of feedback provided and then fine-tune the LLM to improve its ability to exploit the provided guidance. Through experiments on Chinese-English, English-German, and English-Russian MQM data, we demonstrate that prompting LLMs to post-edit MT improves TER, BLEU and COMET scores, although the benefits of fine-grained feedback are not clear. Fine-tuning helps integrate fine-grained feedback more effectively and further improves translation quality based on both automatic and human evaluation.

pdf bib
Non-contrastive sentence representations via self-supervision
Duccio Pappadopulo | Marco Farina

Sample contrastive methods, typically referred to simply as contrastive are the foundation of most unsupervised methods to learn text and sentence embeddings. On the other hand, a different class of self-supervised non-contrastive loss functions and methods have been considered in the computer vision community and referred to as dimension contrastive. In this paper, we thoroughly compare this class of methods with the standard baseline for contrastive sentence embeddings, SimCSE. We find that self-supervised embeddings trained using dimension contrastive objectives can outperform SimCSE on downstream tasks without needing auxiliary loss functions.

pdf bib
Semantically-Prompted Language Models Improve Visual Descriptions
Michael Ogezi | Bradley Hauer | Grzegorz Kondrak

Language-vision models like CLIP have made significant strides in vision tasks, such as zero-shot image classification (ZSIC). However, generating specific and expressive visual descriptions remains challenging; descriptions produced by current methods are often ambiguous and lacking in granularity. To tackle these issues, we propose V-GLOSS: Visual Glosses, a novel method built upon two key ideas. The first is Semantic Prompting, which conditions a language model on structured semantic knowledge. The second is a new contrastive algorithm that elicits fine-grained distinctions between similar concepts. With both ideas, we demonstrate that V-GLOSS improves visual descriptions and achieves strong results in the zero-shot setting on general and fine-grained image-classification datasets, including ImageNet, STL-10, FGVC Aircraft, and Flowers 102. Moreover, these descriptive capabilities contribute to enhancing image-generation performance. Finally, we introduce a quality-tested silver dataset with descriptions generated with V-GLOSS for all ImageNet classes.

pdf bib
GenTKG: Generative Forecasting on Temporal Knowledge Graph with Large Language Models
Ruotong Liao | Xu Jia | Yangzhe Li | Yunpu Ma | Volker Tresp

The rapid advancements in large language models (LLMs) have ignited interest in the temporal knowledge graph (tKG) domain, where conventional embedding-based and rule-based methods dominate. The question remains open of whether pre-trained LLMs can understand structured temporal relational data and replace them as the foundation model for temporal relational forecasting. Therefore, we bring temporal knowledge forecasting into the generative setting. However, challenges occur in the huge chasms between complex temporal graph data structure and sequential natural expressions LLMs can handle, and between the enormous data sizes of tKGs and heavy computation costs of finetuning LLMs. To address these challenges, we propose a novel retrieval-augmented generation framework named GenTKG combining a temporal logical rule-based retrieval strategy and few-shot parameter-efficient instruction tuning to solve the above challenges, respectively. Extensive experiments have shown that GenTKG outperforms conventional methods of temporal relational forecasting with low computation resources using extremely limited training data as few as 16 samples. GenTKG also highlights remarkable cross-domain generalizability with outperforming performance on unseen datasets without re-training, and in-domain generalizability regardless of time split in the same dataset. Our work reveals the huge potential of LLMs in the tKG domain and opens a new frontier for generative forecasting on tKGs. The code and data are released here: https://github.com/mayhugotong/GenTKG.

pdf bib
A Transformer with Stack Attention
Jiaoda Li | Jennifer White | Mrinmaya Sachan | Ryan Cotterell

Natural languages are believed to be (mildly) context-sensitive. Despite underpinning remarkably capable large language models, transformers are unable to model many context-free language tasks. In an attempt to address this limitation in the modeling power of transformer-based language models, we propose augmenting them with a differentiable, stack-based attention mechanism. Our stack-basedattention mechanism can be incorporated into any transformer-based language model and adds a level of interpretability to the model. We show that the addition of our stack-based attention mechanism enables the transformer to model some, but not all, deterministic context-freelanguages.

pdf bib
InstructEval: Systematic Evaluation of Instruction Selection Methods
Anirudh Ajith | Chris Pan | Mengzhou Xia | Ameet Deshpande | Karthik Narasimhan

In-context learning (ICL) performs tasks by prompting a large language model (LLM) using an instruction and a small set of annotated examples called demonstrations. Recent work has shown that precise details of the inputs used in the ICL prompt significantly impact performance, which has incentivized instruction selection algorithms. The effect of instruction-choice however is severely underexplored, with existing analyses restricted to shallow subsets of models and tasks, limiting the generalizability of their insights. We develop InstructEval, an ICL evaluation suite to conduct a thorough assessment of these techniques. The suite includes 13 open-sourced LLMs of varying scales from four model families, and covers nine tasks across three categories. Using the suite, we evaluate the relative performance of seven popular instruction selection methods over five metrics relevant to ICL. Our experiments reveal that using curated manually-written instructions or simple instructions without any task-specific descriptions often elicits superior ICL performance overall than that of automatic instruction-induction methods, pointing to a lack of generalizability among the latter. We release our evaluation suite (at https://github.com/princeton-nlp/InstructEval) for benchmarking instruction selection approaches and enabling more generalizable methods in this space.

pdf bib
RecMind: Large Language Model Powered Agent For Recommendation
Yancheng Wang | Ziyan Jiang | Zheng Chen | Fan Yang | Yingxue Zhou | Eunah Cho | Xing Fan | Yanbin Lu | Xiaojiang Huang | Yingzhen Yang

While the recommendation system (RS) has advanced significantly through deep learning, current RS approaches usually train and fine-tune models on task-specific datasets, limiting their generalizability to new recommendation tasks and their ability to leverage external knowledge due to model scale and data size constraints. Thus, we designed an LLM-powered autonomous recommender agent, RecMind, which is capable of leveraging external knowledge, utilizing tools with careful planning to provide zero-shot personalized recommendations. We propose a Self-Inspiring algorithm to improve the planning ability. At each intermediate step, the LLM “self-inspires” to consider all previously explored states to plan for the next step. This mechanism greatly improves the model’s ability to comprehend and utilize historical information in planning for recommendation. We evaluate RecMind’s performance in various recommendation scenarios. Our experiment shows that RecMind outperforms existing zero/few-shot LLM-based recommendation baseline methods in various tasks and achieves comparable performance to a fully trained recommendation model P5.

pdf bib
GOLD: Generalized Knowledge Distillation via Out-of-Distribution-Guided Language Data Generation
Mohsen Gholami | Mohammad Akbari | Tianxi Hu | Vaden Masrani | Z. Wang | Yong Zhang

Knowledge distillation from LLMs is essential for the efficient deployment of language models. Prior works have proposed data generation using LLMs for preparing distilled models. We argue that generating data with LLMs is prone to sampling mainly from the center of original content distribution. This limitation hinders the distilled model from learning the true underlying data distribution and to forget the tails of the distributions (samples with lower probability). To this end, we propose GOLD, a task-agnostic data generation and knowledge distillation framework, which employs an iterative out-of-distribution-guided feedback mechanism for the LLM. As a result, the generated data improves the generalizability of distilled models. An energy-based OOD evaluation approach is also introduced to deal with noisy generated data. Our extensive experiments on 10 different classification and sequence-to-sequence tasks in NLP show that GOLD respectively outperforms prior arts and the LLM with an average improvement of 5% and 14%. We will also show that the proposed method is applicable to less explored and novel tasks. Code is available in the Appendix.

pdf bib
How Lexical is Bilingual Lexicon Induction?
Harsh Kohli | Helian Feng | Nicholas Dronen | Calvin McCarter | Sina Moeini | Ali Kebarighotbi

In contemporary machine learning approaches to bilingual lexicon induction (BLI), a model learns a mapping between the embedding spaces of a language pair. Recently, retrieve-and-rank approach to BLI has achieved state of the art results on the task. However, the problem remains challenging in low-resource settings, due to the paucity of data. The task is complicated by factors such as lexical variation across languages. We argue that the incorporation of additional lexical information into the recent retrieve-and-rank approach should improve lexicon induction. We demonstrate the efficacy of our proposed approach on XLING, improving over the previous state of the art by an average of 2% across all language pairs.

pdf bib
Fumbling in Babel: An Investigation into ChatGPT’s Language Identification Ability
Wei-Rui Chen | Ife Adebara | Khai Doan | Qisheng Liao | Muhammad Abdul-Mageed

ChatGPT has recently emerged as a powerful NLP tool that can carry out a variety of tasks. However, the range of languages ChatGPT can handle remains largely a mystery. To uncover which languages ChatGPT ‘knows’, we investigate its language identification (LID) abilities. For this purpose, we compile Babel-670, a benchmark comprising 670 languages representing 23 language families spoken in five continents. Languages in Babel-670 run the gamut from the very high-resource to the very low-resource. We then study ChatGPT’s (both GPT-3.5 and GPT-4) ability to (i) identify language names and language codes (ii) under zero- and few-shot conditions (iii) with and without provision of a label set. When compared to smaller finetuned LID tools, we find that ChatGPT lags behind. For example, it has poor performance on African languages. We conclude that current large language models would benefit from further development before they can sufficiently serve diverse communities.

pdf bib
Targeted Augmentation for Low-Resource Event Extraction
Sijia Wang | Lifu Huang

Addressing the challenge of low-resource information extraction remains an ongoing issue due to the inherent information scarcity within limited training examples. Existing data augmentation methods, considered potential solutions, struggle to strike a balance between weak augmentation (e.g., synonym augmentation) and drastic augmentation (e.g., conditional generation without proper guidance). This paper introduces a novel paradigm that employs targeted augmentation and back validation to produce augmented examples with enhanced diversity, polarity, accuracy, and coherence. Extensive experimental results demonstrate the effectiveness of the proposed paradigm. Furthermore, identified limitations are discussed, shedding light on areas for future improvement.

pdf bib
Asking More Informative Questions for Grounded Retrieval
Sedrick Keh | Justin Chiu | Daniel Fried

When a model is trying to gather information in an interactive setting, it benefits from asking informative questions. However, in the case of a grounded multi-turn image identification task, previous studies have been constrained to polar yes/no questions (White et al., 2021), limiting how much information the model can gain in a single turn. We present an approach that formulates more informative, open-ended questions. In doing so, we discover that off-the-shelf visual question answering (VQA) models often make presupposition errors, which standard information gain question selection methods fail to account for. To address this issue, we propose a method that can incorporate presupposition handling into both question selection and belief updates. Specifically, we use a two-stage process, where the model first filters out images which are irrelevant to a given question, then updates its beliefs about which image the user intends. Through self-play and human evaluations, we show that our method is successful in asking informative open-ended questions, increasing accuracy over the past state-of-the-art by 14%, while resulting in 48% more efficient games in human evaluations.

pdf bib
Efficient Citer: Tuning Large Language Models for Enhanced Answer Quality and Verification
Marzieh Tahaei | Aref Jafari | Ahmad Rashid | David Alfonso-Hermelo | Khalil Bibi | Yimeng Wu | Ali Ghodsi | Boxing Chen | Mehdi Rezagholizadeh

In recent years, there has been a growing interest in utilizing external knowledge to reduce hallucinations in large language models (LLMs) and provide them with updated information. Despite this improvement, a major challenge lies in the lack of explicit citations, which hampers the ability to verify the information generated by these models.This paper focuses on providing models with citation capabilities efficiently. By constructing a dataset of citations, we train two model architectures: an FID-style FLAN-T5 model for efficient answer composition and a 13B model known for its success in instruction following after tuning. Evaluation on fluency, correctness, and citation quality is conducted through human assessment and the newly introduced Automatic LLMs’ Citation Evaluation (ALCE) benchmark.Results demonstrate significant improvements in answer quality and efficiency, surpassing the performance of the popular ChatGPT on some of the metrics. The models exhibit exceptional out-of-domain generalization in both human and automatic evaluation. Notably, the FID-style FLAN-T5 model with only 3B parameters performs impressively compared to the 13B model.

pdf bib
Addressing Healthcare-related Racial and LGBTQ+ Biases in Pretrained Language Models
Sean Xie | Saeed Hassanpour | Soroush Vosoughi

Recent studies have highlighted the issue of Pretrained Language Models (PLMs) inadvertently propagating social stigmas and stereotypes, a critical concern given their widespread use. This is particularly problematic in sensitive areas like healthcare, where such biases could lead to detrimental outcomes. Our research addresses this by adapting two intrinsic bias benchmarks to quantify racial and LGBTQ+ biases in prevalent PLMs. We also empirically evaluate the effectiveness of various debiasing methods in mitigating these biases. Furthermore, we assess the impact of debiasing on both Natural Language Understanding and specific biomedical applications. Our findings reveal that while PLMs commonly exhibit healthcare-related racial and LGBTQ+ biases, the applied debiasing techniques successfully reduce these biases without compromising the models’ performance in downstream tasks.

pdf bib
ATG: Benchmarking Automated Theorem Generation for Generative Language Models
Xiaohan Lin | Qingxing Cao | Yinya Huang | Zhicheng Yang | Zhengying Liu | Zhenguo Li | Xiaodan Liang

Humans can develop new theorems to explore broader and more complex mathematical results.While current generative language models (LMs) have achieved significant improvement in automatically proving theorems, their ability to generate new or reusable theorems is still under-explored. Without the new theorems, current LMs struggle to prove harder theorems that are distant from the given hypotheses with the exponentially growing search space.More advanced theorem proving is if an agent (for instance, a generative LM) can leverage its creativity to generate new but also reasonable theorems that properly substitute part of a proof and also be saved as reusable knowledge for future theorem proving.Therefore, this paper proposes an Automated Theorem Generation (ATG) benchmark that evaluates whether an agent can automatically generate valuable (and possibly brand new) theorems that are applicable for downstream theorem proving as reusable knowledge. Specifically, we construct the ATG benchmark by splitting the Metamath library into three sets: axioms, library, and problem based on their proving depth.We conduct extensive experiments to investigate whether current LMs can generate theorems in the library and benefit the problem theorems proving. The results demonstrate that high-quality ATG data facilitates models’ performances on downstream ATP. However, there is still room for current LMs to develop better ATG and generate more advanced and human-like theorems. We hope the new ATG challenge can shed some light on advanced complex theorem proving.

pdf bib
Benchmarking Generation and Evaluation Capabilities of Large Language Models for Instruction Controllable Summarization
Yixin Liu | Alexander Fabbri | Jiawen Chen | Yilun Zhao | Simeng Han | Shafiq Joty | Pengfei Liu | Dragomir Radev | Chien-Sheng Wu | Arman Cohan

While large language models (LLMs) can already achieve strong performance on standard generic summarization benchmarks, their performance on more complex summarization task settings is less studied. Therefore, we benchmark LLMs on instruction controllable text summarization, where the model input consists of both a source article and a natural language requirement for desired summary characteristics. To this end, we curate an evaluation-only dataset for this task setting and conduct human evaluations of five LLM-based systems to assess their instruction-following capabilities in controllable summarization. We then benchmark LLM-based automatic evaluation for this task with 4 different evaluation protocols and 11 LLMs, resulting in 40 evaluation methods. Our study reveals that instruction controllable text summarization remains a challenging task for LLMs, since (1) all LLMs evaluated still make factual and other types of errors in their summaries; (2) no LLM-based evaluation methods can achieve a strong alignment with human annotators when judging the quality of candidate summaries; (3) different LLMs show large performance gaps in summary generation and evaluation capabilities. We make our collected benchmark InstruSum publicly available to facilitate future research in this direction.

pdf bib
NeuroComparatives: Neuro-Symbolic Distillation of Comparative Knowledge
Phillip Howard | Junlin Wang | Vasudev Lal | Gadi Singer | Yejin Choi | Swabha Swayamdipta

Comparative knowledge (e.g., steel is stronger and heavier than styrofoam) is an essential component of our world knowledge, yet understudied in prior literature. In this paper, we harvest the dramatic improvements in knowledge capabilities of language models into a large-scale comparative knowledge base. While the ease of acquisition of such comparative knowledge is much higher from extreme-scale models like GPT-4, compared to their considerably smaller and weaker counterparts such as GPT-2, not even the most powerful models are exempt from making errors. We thus ask: to what extent are models at different scales able to generate valid and diverse comparative knowledge?We introduce NeuroComparatives, a novel framework for comparative knowledge distillation overgenerated from language models such as GPT-variants and LLaMA, followed by stringent filtering of the generated knowledge. Our framework acquires comparative knowledge between everyday objects, producing a corpus of up to 8.8M comparisons over 1.74M entity pairs - 10X larger and 30% more diverse than existing resources. Moreover, human evaluations show that NeuroComparatives outperform existing resources in terms of validity (up to 32% absolute improvement). Our acquired NeuroComparatives leads to performance improvements on five downstream tasks.We find that neuro-symbolic manipulation of smaller models offers complementary benefits to the currently dominant practice of prompting extreme-scale language models for knowledge distillation.

pdf bib
Emotion-Anchored Contrastive Learning Framework for Emotion Recognition in Conversation
Fangxu Yu | Junjie Guo | Zhen Wu | Xinyu Dai

Emotion Recognition in Conversation (ERC) involves detecting the underlying emotion behind each utterance within a conversation. Effectively generating representations for utterances remains a significant challenge in this task. Recent works propose various models to address this issue, but they still struggle with differentiating similar emotions such as excitement and happiness. To alleviate this problem, We propose an Emotion-Anchored Contrastive Learning (EACL) framework that can generate more distinguishable utterance representations for similar emotions. To achieve this, we utilize label encodings as anchors to guide the learning of utterance representations and design an auxiliary loss to ensure the effective separation of anchors for similar emotions. Moreover, an additional adaptation process is proposed to adapt anchors to serve as effective classifiers to improve classification performance. Across extensive experiments, our proposed EACL achieves state-of-the-art emotion recognition performance and exhibits superior performance on similar emotions. Our code is available at https://github.com/Yu-Fangxu/EACL.

pdf bib
SUQL: Conversational Search over Structured and Unstructured Data with Large Language Models
Shicheng Liu | Jialiang Xu | Wesley Tjangnaka | Sina Semnani | Chen Yu | Monica Lam

While most conversational agents are grounded on either free-text or structured knowledge, many knowledge corpora consist of hybrid sources.This paper presents the first conversational agent that supports the full generality of hybrid data access for large knowledge corpora, through a language we developed called SUQL (Structured and Unstructured Query Language). Specifically, SUQL extends SQL with free-text primitives (\smallSUMMARY and \smallANSWER), so information retrieval can be composed with structured data accesses arbitrarily in a formal, succinct, precise, and interpretable notation. With SUQL, we propose the first semantic parser, an LLM with in-context learning, that can handle hybrid data sources.Our in-context learning-based approach, when applied to the HybridQA dataset, comes within 8.9% Exact Match and 7.1% F1 of the SOTA, which was trained on 62K data samples. More significantly, unlike previous approaches, our technique is applicable to large databases and free-text corpora. We introduce a dataset consisting of crowdsourced questions and conversations on Yelp, a large, real restaurant knowledge base with structured and unstructured data. We show that our few-shot conversational agent based on SUQL finds an entity satisfying all user requirements 90.3% of the time, compared to 63.4% for a baseline based on linearization.

pdf bib
On Evaluating the Integration of Reasoning and Action in LLM Agents with Database Question Answering
Linyong Nan | Ellen Zhang | Weijin Zou | Yilun Zhao | Wenfei Zhou | Arman Cohan

This study introduces a new long-form database question answering dataset designed to evaluate how Large Language Models (LLMs) interact with a SQL interpreter. The task necessitates LLMs to strategically generate multiple SQL queries to retrieve sufficient data from a database, to reason with the acquired context, and to synthesize them into a comprehensive analytical narrative. Our findings highlight that this task poses great challenges even for the state-of-the-art **GPT-4** model. We propose and evaluate two interaction strategies, and provide a fine-grained analysis of the individual stages within the interaction. A key discovery is the identification of two primary bottlenecks hindering effective interaction: the capacity for planning and the ability to generate multiple SQL queries. To address the challenge of accurately assessing answer quality, we introduce a multi-agent evaluation framework that simulates the academic peer-review process, enhancing the precision and reliability of our evaluations. This framework allows for a more nuanced understanding of the strengths and limitations of current LLMs in complex retrieval and reasoning tasks.

pdf bib
CARE: Extracting Experimental Findings From Clinical Literature
Aakanksha Naik | Bailey Kuehl | Erin Bransom | Doug Downey | Tom Hope

Extracting fine-grained experimental findings from literature can provide dramatic utility for scientific applications. Prior work has developed annotation schemas and datasets for limited aspects of this problem, failing to capture the real-world complexity and nuance required. Focusing on biomedicine, this work presents CARE—a new IE dataset for the task of extracting clinical findings. We develop a new annotation schema capturing fine-grained findings as n-ary relations between entities and attributes, which unifies phenomena challenging for current IE systems such as discontinuous entity spans, nested relations, variable arity n-ary relations and numeric results in a single schema. We collect extensive annotations for 700 abstracts from two sources: clinical trials and case reports. We also demonstrate the generalizability of our schema to the computer science and materials science domains. We benchmark state-of-the-art IE systems on CARE, showing that even models such as GPT4 struggle. We release our resources to advance research on extracting and aggregating literature findings.

pdf bib
Personalized Federated Learning for Text Classification with Gradient-Free Prompt Tuning
Rui Wang | Tong Yu | Ruiyi Zhang | Sungchul Kim | Ryan Rossi | Handong Zhao | Junda Wu | Subrata Mitra | Lina Yao | Ricardo Henao

In this paper, we study personalized federated learning for text classification with Pretrained Language Models (PLMs). We identify two challenges in efficiently leveraging PLMs for personalized federated learning: 1) Communication. PLMs are usually large in size, e.g., with hundreds of millions of parameters, inducing huge communication cost in a federated setting. 2) Local Training. Training with PLMs generally requires back-propagation, during which memory consumption can be several times that of the forward-propagation. This may not be affordable when the PLMs are trained locally on the clients that are resource constrained, e.g., mobile devices with limited access to memory resources. Additionally, the proprietary PLMs can be provided as concealed APIs, for which the back-propagation operations may not be available. In solving these, we propose a training framework that includes an approach of discrete local search for gradient-free local training, along with a compression mechanism inspired from the linear word analogy that allows communicating with discretely indexed tokens, thus significantly reducing the communication cost. Experiments show that our gradient-free framework achieves superior performance compared with baselines.

pdf bib
SGSH: Stimulate Large Language Models with Skeleton Heuristics for Knowledge Base Question Generation
Shasha Guo | Lizi Liao | Jing Zhang | Yanling Wang | Cuiping Li | Hong Chen

Knowledge base question generation (KBQG) aims to generate natural language questions from a set of triplet facts extracted from KB. Existing methods have significantly boosted the performance of KBQG via pre-trained language models (PLMs) thanks to the richly endowed semantic knowledge. With the advance of pre-training techniques, large language models (LLMs) (e.g., GPT-3.5) undoubtedly possess much more semantic knowledge. Therefore, how to effectively organize and exploit the abundant knowledge for KBQG becomes the focus of our study. In this work, we propose SGSH — a simple and effective framework to Stimulate GPT-3.5 with Skeleton Heuristics to enhance KBQG. The framework incorporates “skeleton heuristics”, which provides more fine-grained guidance associated with each input to stimulate LLMs to generate optimal questions, encompassing essential elements like the question phrase and the auxiliary verb.More specifically, we devise an automatic data construction strategy leveraging ChatGPT to construct a skeleton training dataset, based on which we employ a soft prompting approach to train a BART model dedicated to generating the skeleton associated with each input.Subsequently, skeleton heuristics are encoded into the prompt to incentivize GPT-3.5 to generate desired questions. Extensive experiments demonstrate that SGSH derives the new state-of-the-art performance on the KBQG tasks.

pdf bib
Biomedical Entity Representation with Graph-Augmented Multi-Objective Transformer
Andrey Sakhovskiy | Natalia Semenova | Artur Kadurin | Elena Tutubalina

Modern biomedical concept representations are mostly trained on synonymous concept names from a biomedical knowledge base, ignoring the inter-concept interactions and a concept’s local neighborhood in a knowledge base graph. In this paper, we introduce Biomedical Entity Representation with a Graph-Augmented Multi-Objective Transformer (BERGAMOT), which adopts the power of pre-trained language models (LMs) and graph neural networks to capture both inter-concept and intra-concept interactions from the multilingual UMLS graph. To obtain fine-grained graph representations, we introduce two additional graph-based objectives: (i) a node-level contrastive objective and (ii) the Deep Graph Infomax (DGI) loss, which maximizes the mutual information between a local subgraph and a high-level graph summary. We apply contrastive loss on textual and graph representations to make them less sensitive to surface forms and enable intermodal knowledge exchange. BERGAMOT achieves state-of-the-art results in zero-shot entity linking without task-specific supervision on 4 of 5 languages of the Mantra corpus and on 8 of 10 languages of the XL-BEL benchmark.

pdf bib
Cross-Lingual Summarization with Pseudo-Label Regularization
Thang Le

Cross-Lingual Summarization (XLS) aims to summarize a document in the source language into a condensed version in the target language, effectively removing language barriers for non-native readers. Previous approaches, however, have the same limitation that only a single reference (gold summary) is exploited during model training, making the base model exposed to an underrepresented hypothesis space since the actual number of possible hypotheses is exponentially large. To alleviate this problem, we present a study adopting pseudo-labels in regularizing standard cross-lingual summarization training. We investigate several components leading to the gains in regularization training with verified experiments involving 8 diverse languages from different families. Conclusively, we show that pseudo-labeling is a simple and effective approach that significantly improves over standard gold reference training in XLS.

pdf bib
On the Way to Gentle AI Counselor: Politeness Cause Elicitation and Intensity Tagging in Code-mixed Hinglish Conversations for Social Good
Priyanshu Priya | Gopendra Singh | Mauajama Firdaus | Jyotsna Agrawal | Asif Ekbal

Politeness is a multifaceted concept influenced by individual perceptions of what is considered polite or impolite. With this objective, we introduce a novel task - Politeness Cause Elicitation and Intensity Tagging (PCEIT). This task focuses on conversations and aims to identify the underlying reasons behind the use of politeness and gauge the degree of politeness conveyed. To address this objective, we create HING-POEM, a new conversational dataset in Hinglish (a blend of Hindi and English) for mental health and legal counseling of crime victims. The rationale for the domain selection lies in the paramount importance of politeness in mental health and legal counseling of crime victims to ensure a compassionate and cordial atmosphere for them. We enrich the HING-POEM dataset by annotating it with politeness labels, politeness causal spans, and intensity values at the level of individual utterances. In the context of the introduced PCEIT task, we present PAANTH (Politeness CAuse ElicitAion and INtensity Tagging in Hinglish), a comprehensive framework based on Contextual Enhanced Attentive Convolution Transformer. We conduct extensive quantitative and qualitative evaluations to establish the effectiveness of our proposed approach using the newly constructed dataset. Our approach is compared against state-of-the-art baselines, and these analyses help demonstrate the superiority of our method.

pdf bib
Leveraging Summarization for Unsupervised Dialogue Topic Segmentation
Aleksei Artemiev | Daniil Parinov | Alexey Grishanov | Ivan Borisov | Alexey Vasilev | Daniil Muravetskii | Aleksey Rezvykh | Aleksei Goncharov | Andrey Savchenko

Traditional approaches to dialogue segmentation perform reasonably well on synthetic or written dialogues but suffer when dealing with spoken, noisy dialogs. In addition, such methods require careful tuning of hyperparameters. We propose to leverage a novel approach that is based on dialogue summaries. Experiments on different datasets showed that the new approach outperforms popular state-of-the-art algorithms in unsupervised topic segmentation and requires less setup.

pdf bib
LLaMA-Rider: Spurring Large Language Models to Explore the Open World
Yicheng Feng | Yuxuan Wang | Jiazheng Liu | Sipeng Zheng | Zongqing Lu

Recently, various studies have leveraged Large Language Models (LLMs) to help decision-making and planning in environments and try to align the LLMs’ knowledge with the world conditions. Nonetheless, the capacity of LLMs to continuously acquire environmental knowledge and adapt in an open world remains uncertain. In this paper, we propose an approach to spur LLMs to explore the open world, gather experiences, and learn to improve their task-solving capabilities. In this approach, a multi-round feedback-revision mechanism is utilized to encourage LLMs to actively select appropriate revision actions guided by feedback information from the environment. This facilitates exploration and enhances the model’s performance. Besides, we integrate sub-task relabeling to assist LLMs in maintaining consistency in sub-task planning and help the model learn the combinatorial nature between tasks, enabling it to complete a wider range of tasks through training based on the acquired exploration experiences. By evaluation in Minecraft, an open-ended sandbox world, we demonstrate that our approach LLaMA-Rider enhances the efficiency of the LLM in exploring the environment, and effectively improves the LLM’s ability to accomplish more tasks through fine-tuning with merely 1.3k instances of collected data, showing minimal training costs compared to the baseline using reinforcement learning. The code is available at https://github.com/PKU-RL/LLaMA-Rider.

pdf bib
Contrastive Learning as a Polarizer: Mitigating Gender Bias by Fair and Biased sentences
Kyungmin Park | Sihyun Oh | Daehyun Kim | Juae Kim

Recently, language models have accelerated the improvement in natural language processing. However, recent studies have highlighted a significant issue: social biases inherent in training data can lead models to learn and propagate these biases. In this study, we propose a contrastive learning method for bias mitigation, utilizing anchor points to push further negatives and pull closer positives within the representation space. This approach employs stereotypical data as negatives and stereotype-free data as positives, enhancing debiasing performance. Our model attained state-of-the-art performance in the ICAT score on the StereoSet, a benchmark for measuring bias in models. In addition, we observed that effective debiasing is achieved through an awareness of biases, as evidenced by improved hate speech detection scores. The implementation code and trained models are available at https://github.com/HUFS-NLP/CL_Polarizer.git.

pdf bib
PoLLMgraph: Unraveling Hallucinations in Large Language Models via State Transition Dynamics
Derui Zhu | Dingfan Chen | Qing Li | Zongxiong Chen | Lei Ma | Jens Grossklags | Mario Fritz

Despite tremendous advancements in large language models (LLMs) over recent years, a notably urgent challenge for their practical deployment is the phenomenon of "hallucination”, where the model fabricates facts and produces non-factual statements. In response, we propose PoLLMgraph—a Polygraph for LLMs—as an effective model-based white-box detection and forecasting approach. PoLLMgraph distinctly differs from the large body of existing research that concentrates on addressing such challenges through black-box evaluations. In particular, we demonstrate that hallucination can be effectively detected by analyzing the LLM’s internal state transition dynamics during generation via tractable probabilistic models. Experimental results on various open-source LLMs confirm the efficacy of PoLLMgraph, outperforming state-of-the-art methods by a considerable margin, evidenced by over 20% improvement in AUC-ROC on common benchmarking datasets like TruthfulQA. Our work paves a new way for model-based white-box analysis of LLMs, motivating the research community to further explore, understand, and refine the intricate dynamics of LLM behaviors.

pdf bib
Improving Health Question Answering with Reliable and Time-Aware Evidence Retrieval
Juraj Vladika | Florian Matthes

In today’s digital world, seeking answers to health questions on the Internet is a common practice. However, existing question answering (QA) systems often rely on using pre-selected and annotated evidence documents, thus making them inadequate for addressing novel questions. Our study focuses on the open-domain QA setting, where the key challenge is to first uncover relevant evidence in large knowledge bases. By utilizing the common retrieve-then-read QA pipeline and PubMed as a trustworthy collection of medical research documents, we answer health questions from three diverse datasets. We modify different retrieval settings to observe their influence on the QA pipeline’s performance, including the number of retrieved documents, sentence selection process, the publication year of articles, and their number of citations. Our results reveal that cutting down on the amount of retrieved documents and favoring more recent and highly cited documents can improve the final macro F1 score up to 10%. We discuss the results, highlight interesting examples, and outline challenges for future research, like managing evidence disagreement and crafting user-friendly explanations.

pdf bib
DecoderLens: Layerwise Interpretation of Encoder-Decoder Transformers
Anna Langedijk | Hosein Mohebbi | Gabriele Sarti | Willem Zuidema | Jaap Jumelet

In recent years, several interpretability methods have been proposed to interpret the inner workings of Transformer models at different levels of precision and complexity.In this work, we propose a simple but effective technique to analyze encoder-decoder Transformers. Our method, which we name DecoderLens, allows the decoder to cross-attend representations of intermediate encoder activations instead of using the default final encoder output.The method thus maps uninterpretable intermediate vector representations to human-interpretable sequences of words or symbols, shedding new light on the information flow in this popular but understudied class of models.We apply DecoderLens to question answering, logical reasoning, speech recognition and machine translation models, finding that simpler subtasks are solved with high precision by low and intermediate encoder layers.
up

pdf (full)
bib (full) Proceedings of the 4th Workshop on Natural Language Processing for Indigenous Languages of the Americas (AmericasNLP 2024)

pdf bib
Proceedings of the 4th Workshop on Natural Language Processing for Indigenous Languages of the Americas (AmericasNLP 2024)
Manuel Mager | Abteen Ebrahimi | Shruti Rijhwani | Arturo Oncevay | Luis Chiruzzo | Robert Pugh | Katharina von der Wense

pdf bib
NLP for Language Documentation: Two Reasons for the Gap between Theory and Practice
Luke Gessler | Katharina von der Wense

Both NLP researchers and linguists have expressed a desire to use language technologies in language documentation, but most documentary work still proceeds without them, presenting a lost opportunity to hasten the preservation of the world’s endangered languages, such as those spoken in Latin America. In this work, we empirically measure two factors that have previously been identified as explanations of this low utilization: curricular offerings in graduate programs, and rates of interdisciplinary collaboration in publications related to NLP in language documentation. Our findings verify the claim that interdisciplinary training and collaborations are scarce and support the view that interdisciplinary curricular offerings facilitate interdisciplinary collaborations.

pdf bib
Translation systems for low-resource Colombian Indigenous languages, a first step towards cultural preservation
Juan Prieto | Cristian Martinez | Melissa Robles | Alberto Moreno | Sara Palacios | Rubén Manrique

The use of machine learning and Natural Language Processing (NLP) technologies can assist in the preservation and revitalization of indigenous languages, particularly those classified as “low-resource.” Given the increasing digitization of information, the development of translation tools for these languages is of significant importance. These tools not only facilitate better access to digital resources for indigenous communities but also stimulate language preservation efforts and potentially foster more inclusive, equitable societies, as demonstrated by the AmericasNLP workshop since 2021. The focus of this paper is Colombia, a country home to 65 distinct indigenous languages, presenting a vast spectrum of linguistic characteristics. This cultural and linguistic diversity is an inherent pillar of the nation’s identity, and safeguarding it has been increasingly challenging given the dwindling number of native speakers and the communities’ inclination towards oral traditions. Considering this context, scattered initiatives exist to develop translation systems for these languages. However, these endeavors suffer from a lack of consolidated, comparable data. This paper consolidates a dataset of parallel data in four Colombian indigenous languages - Wayuunaiki, Arhuaco, Inga, and Nasa - gathered from existing digital resources. It also presents the creation of baseline models for future translation and comparison, ultimately serving as a catalyst for incorporating more digital resources progressively.

pdf bib
Word-level prediction in Plains Cree: First steps
Olga Kriukova | Antti Arppe

Plains Cree (nêhiyawêwin) is a morphologically complex and predominantly prefixing language. The combinatory potential of inflectional and derivational/lexical prefixes and verb stems in Plains Cree makes it challenging for traditional auto-completion (or word suggestion) approaches to handle. The lack of a large corpus of Plains Cree also complicates the situation. This study attempts to investigate how well a BiLSTM model trained on a small Cree corpus can handle a word suggestion task. Moreover, this study evaluates whether the use of semantically and morphosyntactically refined Word2Vec embeddings can improve the overall accuracy and quality of BiLSTM suggestions. The results show that some models trained with the refined vectors provide semantically and morphosyntactically better suggestions. They are also more accurate in predictions of content words. The model trained with the non-refined vectors, in contrast, was better at predicting conjunctions, particles, and other non-inflecting words. The models trained with different refined vector combinations provide the expected next word among top-10 predictions in 36.73 to 37.88% of cases (depending on the model).

pdf bib
Mapping ‘when’-clauses in Latin American and Caribbean languages: an experiment in subtoken-based typology
Nilo Pedrazzini

Languages can encode temporal subordination lexically, via subordinating conjunctions, and morphologically, by marking the relation on the predicate. Systematic cross-linguistic variation among the former can be studied using well-established token-based typological approaches to token-aligned parallel corpora. Variation among different morphological means is instead much harder to tackle and therefore more poorly understood, despite being predominant in several language groups. This paper explores variation in the expression of generic temporal subordination (‘when’-clauses) among the languages of Latin America and the Caribbean, where morphological marking is particularly common. It presents probabilistic semantic maps computed on the basis of the languages of the region, thus avoiding bias towards the many world’s languages that exclusively use lexified connectors, incorporating associations between character in/i-grams and English iwhen/i. The approach allows capturing morphological clause-linkage devices in addition to lexified connectors, paving the way for larger-scale, strategy-agnostic analyses of typological variation in temporal subordination.

pdf bib
Comparing LLM prompting with Cross-lingual transfer performance on Indigenous and Low-resource Brazilian Languages
David Ifeoluwa Adelani | A. Seza Doğruöz | André Coneglian | Atul Kr. Ojha

Large Language Models are transforming NLP for a lot of tasks. However, how LLMs perform NLP tasks for LRLs is less explored. In alliance with the theme track of the NAACL’24, we focus on 12 low-resource languages (LRLs) from Brazil, 2 LRLs from Africa and 2 high-resource languages (HRLs) (e.g., English and Brazilian Portuguese). Our results indicate that the LLMs perform worse for the labeling of LRLs in comparison to HRLs in general. We explain the reasons behind this failure and provide an error analyses through examples from 2 Brazilian LRLs.

pdf bib
Analyzing Finetuned Vision Models for Mixtec Codex Interpretation
Alexander Webber | Zachary Sayers | Amy Wu | Elizabeth Thorner | Justin Witter | Gabriel Ayoubi | Christan Grant

Throughout history, pictorial record-keeping has been used to document events, stories, and concepts. A popular example of this is the Tzolk’in Maya Calendar. The pre-Columbian Mixtec society also recorded many works through graphical media called codices that depict both stories and real events. Mixtec codices are unique because the depicted scenes are highly structured within and across documents. As a first effort toward translation, we created two binary classification tasks over Mixtec codices, namely, gender and pose. The composition of figures within a codex is essential for understanding the codex’s narrative. We labeled a dataset with around 1300 figures drawn from three codices of varying qualities. We finetuned the Visual Geometry Group 16 (VGG-16) and Vision Transformer 16 (ViT-16) models, measured their performance, and compared learned features with expert opinions found in literature. The results show that when finetuned, both VGG and ViT perform well, with the transformer-based architecture (ViT) outperforming the CNN-based architecture (VGG) at higher learning rates. We are releasing this work to allow collaboration with the Mixtec community and domain scientists.

pdf bib
A New Benchmark for Kalaallisut-Danish Neural Machine Translation
Ross Kristensen-Mclachlan | Johanne Nedergård

Kalaallisut, also known as (West) Greenlandic, poses a number of unique challenges to contemporary natural language processing (NLP). In particular, the language has historically lacked benchmarking datasets and robust evaluation of specific NLP tasks, such as neural machine translation (NMT). In this paper, we present a new benchmark dataset for Greenlandic to Danish NMT comprising over 1.2m words of Greenlandic and 2.1m words of parallel Danish translations. We provide initial metrics for models trained on this dataset and conclude by suggesting how these findings can be taken forward to other NLP tasks for the Greenlandic language.

pdf bib
Morphological Tagging in Bribri Using Universal Dependency Features
Jessica Karson | Rolando Coto-Solano

This paper outlines the Universal Features tagging of a dependency treebank for Bribri, an Indigenous language of Costa Rica. Universal Features are a morphosyntactic tagging component of Universal Dependencies, which is a framework that aims to provide an annotation system inclusive of all languages and their diverse structures (Nivre et al., 2016; de Marneffe et al., 2021). We used a rule-based system to do a first-pass tagging of a treebank of 1572 words. After manual corrections, the treebank contained 3051 morphological features. We then used this morphologically-tagged treebank to train a UDPipe 2 parsing and tagging model. This model has a UFEATS precision of 80.5 ± 3.6, which is a statistically significant improvement upon the previously available FOMA-based morphological tagger for Bribri. An error analysis suggests that missing TAM and case markers are the most common problem for the model. We hope to use this model to expand upon existing treebanks and facilitate the construction of linguistically-annotated corpora for the language.

pdf bib
LLM-Assisted Rule Based Machine Translation for Low/No-Resource Languages
Jared Coleman | Bhaskar Krishnamachari | Ruben Rosales | Khalil Iskarous

We propose a new paradigm for machine translation that is particularly useful for no-resource languages (those without any publicly available bilingual or monolingual corpora): LLM-RBMT (LLM-Assisted Rule Based Machine Translation). Using the LLM-RBMT paradigm, we design the first language education/revitalization-oriented machine translator for Owens Valley Paiute (OVP), a critically endangered Indigenous American language for which there is virtually no publicly available data. We present a detailed evaluation of the translator’s components: a rule-based sentence builder, an OVP to English translator, and an English to OVP translator. We also discuss the potential of the paradigm, its limitations, and the many avenues for future research that it opens up.

pdf bib
A Concise Survey of OCR for Low-Resource Languages
Milind Agarwal | Antonios Anastasopoulos

Modern natural language processing (NLP) techniques increasingly require substantial amounts of data to train robust algorithms. Building such technologies for low-resource languages requires focusing on data creation efforts and data-efficient algorithms. For a large number of low-resource languages, especially Indigenous languages of the Americas, this data exists in image-based non-machine-readable documents. This includes scanned copies of comprehensive dictionaries, linguistic field notes, children’s stories, and other textual material. To digitize these resources, Optical Character Recognition (OCR) has played a major role but it comes with certain challenges in low-resource settings. In this paper, we share the first survey of OCR techniques specific to low-resource data creation settings and outline several open challenges, with a special focus on Indigenous Languages of the Americas. Based on experiences and results from previous research, we conclude with recommendations on utilizing and improving OCR for the benefit of computational researchers, linguists, and language communities.

pdf bib
Unlocking Knowledge with OCR-Driven Document Digitization for Peruvian Indigenous Languages
Shadya Sanchez Carrera | Roberto Zariquiey | Arturo Oncevay

The current focus on resource-rich languages poses a challenge to linguistic diversity, affecting minority languages with limited digital presence and relatively old published and unpublished resources. In addressing this issue, this study targets the digitalization of old scanned textbooks written in four Peruvian indigenous languages (Asháninka, Shipibo-Konibo, Yanesha, and Yine) using Optical Character Recognition (OCR) technology. This is complemented with text correction methods to minimize extraction errors. Contributions include the creation of an annotated dataset with 454 scanned page images, for a rigorous evaluation, and the development of a module to correct OCR-generated transcription alignments.

pdf bib
Awajun-OP: Multi-domain dataset for Spanish–Awajun Machine Translation
Oscar Moreno | Yanua Atamain | Arturo Oncevay

We introduce a Spanish-Awajun parallel dataset of 22k high-quality sentence pairs with the help of the journalistic organization Company C. This dataset consists of parallel data obtained from various web sources such as poems, stories, laws, protocols, guidelines, handbooks, the Bible, and news published by Company C. The study also includes an analysis of the dataset’s performance for Spanish-Awajun translation using a Transformer architecture with transfer learning from a parent model, utilizing Spanish-English and Spanish-Finnish as high-resource language-pairs. As far as we know, this is the first Spanish-Awajun machine translation study, and we hope that this work will serve as a starting point for future research on this neglected Peruvian language.

pdf bib
Wav2pos: Exploring syntactic analysis from audio for Highland Puebla Nahuatl
Robert Pugh | Varun Sreedhar | Francis Tyers

We describe an approach to part-of-speech tagging from audio with very little human-annotated data, for Highland Puebla Nahuatl, a low-resource language of Mexico. While automatic morphosyntactic analysis is typically trained on annotated textual data, large amounts of text is rarely available for low-resource, marginalized, and/or minority languages, and morphosyntactically-annotated data is even harder to come by. Much of the data from these languages may exist in the form of recordings, often only partially-transcribed or analyzed by field linguists working on language documentation projects. Given this relatively low-availability of text in the low-resource language scenario, we explore end-to-end automated morphosyntactic analysis directly from audio. The experiments described in this paper focus on one piece of morphosyntax, part-of-speech tagging, and builds on existing work in a high-resource setting. We use weak supervision to increase training volume, and explore a few techniques for generating word-level predictions from the acoustic features. Our experiments show promising results, despite less than 400 sentences of audio-aligned, manually-labeled text.

pdf bib
From Field Linguistics to NLP: Creating a curated dataset in Amuzgo language
Antonio Reyes | Hamlet Antonio García

This article presents an ongoing research on one of the several native languages of the Americas: Amuzgo or jny’on3 nda3 . This language is spoken in Southern Mexico and belongs to the Otomanguean family. Although Amuzgo vitality is stable and there are some available resources, such as grammars, dictionaries, or literature, its digital inclusion is emerging (cf. Eberhard et al. (2024)). In this respect, here is described the creation of a curated dataset in Amuzgo. This resource is intended to contribute the development of tools for scarce resources languages by providing fine-grained linguistic information in different layers: From data collection with native speakers to data annotation. The dataset was built according to the following method: i) data collection in Amuzgo by means of linguistic fieldwork; ii) acoustic data processing; iii) data transcription; iv) glossing and translating data into Spanish; v) semiautomatic alignment of translations; and vi) data systematization. This resource is released as an open access dataset to foster the academic community to explore the richness of this language.

pdf bib
Enenlhet as a case-study to investigate ASR model generalizability for language documentation
Éric Le Ferrand | Raina Heaton | Emily Prud’hommeaux

Although both linguists and language community members recognize the potential utility of automatic speech recognition (ASR) for documentation, one of the obstacles to using these technologies is the scarcity of data necessary to train effective systems. Recent advances in ASR, particularly the ability to fine-tune large multilingual acoustic models to small amounts of data from a new language, have demonstrated the potential of ASR for transcription. However, many proof-of-concept demonstrations of ASR in low-resource settings rely on a single data collection project, which may yield models that are biased toward that particular data scenario, whether in content, recording quality, transcription conventions, or speaker population. In this paper, we investigate the performance of two state-of-the art ASR architectures for fine-tuning acoustic models to small speech datasets with the goal of transcribing recordings of Enenlhet, an endangered Indigenous language spoken in South America. Our results suggest that while ASR offers utility for generating first-pass transcriptions of speech collected in the course of linguistic fieldwork, individual vocabulary diversity and data quality have an outsized impact on ASR accuracy.

pdf bib
Advancing NMT for Indigenous Languages: A Case Study on Yucatec Mayan and Chol
Julio Rangel | Norio Kobayashi

This study leverages Spanish-trained large language models (LLMs) to develop neural machine translation (NMT) systems for Mayan languages. For this, we first compile and process a low-resource dataset of 28,135 translation pairs of Chol and Yucatec Mayan extracted from documents in the CPLM Corpus (Martínez et al.). Then, we implement a prompt-based approach to train one-to-many and many-to-many models. By comparing several training strategies for two LLMs, we found that, on average, training multilingual models is better, as shown by the ChrF++ reaching 50 on the test set in the best case. This study reinforces the viability of using LLMs to improve accessibility and preservation for languages with limited digital resources. We share our code, datasets, and models to promote collaboration and progress in this field: https://github.com/RIKEN-DKO/iikim_translator.

pdf bib
BSC Submission to the AmericasNLP 2024 Shared Task
Javier Garcia Gilabert | Aleix Sant | Carlos Escolano | Francesca De Luca Fornaciari | Audrey Mash | Maite Melero

This paper describes the BSC’s submission to the AmericasNLP 2024 Shared Task. We participated in the Spanish to Quechua and Spanish to Guarani tasks. In this paper we show that by using LoRA adapters we can achieve similar performance as a full parameter fine-tuning by only training 14.2% of the total number of parameters. Our systems achieved the highest ChrF++ scores and ranked first for both directions in the final results outperforming strong baseline systems in the provided development and test datasets.

pdf bib
System Description of the NordicsAlps Submission to the AmericasNLP 2024 Machine Translation Shared Task
Joseph Attieh | Zachary Hopton | Yves Scherrer | Tanja Samardžić

This paper presents the system description of the NordicsAlps team for the AmericasNLP 2024 Machine Translation Shared Task 1. We investigate the effect of tokenization on translation quality by exploring two different tokenization schemes: byte-level and redundancy-driven tokenization. We submitted three runs per language pair. The redundancy-driven tokenization ranked first among all submissions, scoring the highest average chrF2++, chrF, and BLEU metrics (averaged across all languages). These findings demonstrate the importance of carefully tailoring the tokenization strategies of machine translation systems, particularly in resource-constrained scenarios.

pdf bib
On the Robustness of Neural Models for Full Sentence Transformation
Michael Ginn | Ali Marashian | Bhargav Shandilya | Claire Post | Enora Rice | Juan Vásquez | Marie Mcgregor | Matthew Buchholz | Mans Hulden | Alexis Palmer

This paper describes the LECS Lab submission to the AmericasNLP 2024 Shared Task on the Creation of Educational Materials for Indigenous Languages. The task requires transforming a base sentence with regards to one or more linguistic properties (such as negation or tense). We observe that this task shares many similarities with the well-studied task of word-level morphological inflection, and we explore whether the findings from inflection research are applicable to this task. In particular, we experiment with a number of augmentation strategies, finding that they can significantly benefit performance, but that not all augmented data is necessarily beneficial. Furthermore, we find that our character-level neural models show high variability with regards to performance on unseen data, and may not be the best choice when training data is limited.

pdf bib
The unreasonable effectiveness of large language models for low-resource clause-level morphology: In-context generalization or prior exposure?
Coleman Haley

This paper describes the submission of Team “Giving it a Shot” to the AmericasNLP 2024 Shared Task on Creation of Educational Materials for Indigenous Languages. We use a simple few-shot prompting approach with several state of the art large language models, achieving competitive performance on the shared task, with our best system placing third overall. We perform a preliminary analysis to determine to what degree the performance of our model is due to prior exposure to the task languages, finding that generally our performance is better explained as being derived from in-context learning capabilities.

pdf bib
A Comparison of Fine-Tuning and In-Context Learning for Clause-Level Morphosyntactic Alternation
Jim Su | Justin Ho | George Broadwell | Sarah Moeller | Bonnie Dorr

This paper presents our submission to the AmericasNLP 2024 Shared Task on the Creation of Educational Materials for Indigenous Languages. We frame this task as one of morphological inflection generation, treating each sentence as a single word. We investigate and compare two distinct approaches: fine-tuning neural encoder-decoder models such as NLLB- 200, and in-context learning with proprietary large language models (LLMs). Our findings demonstrate that for this task, no one approach is perfect. Anthropic’s Claude 3 Opus, when supplied with grammatical description entries, achieves the highest performance on Bribri among the evaluated models. This outcome corroborates and extends previous research exploring the efficacy of in-context learning in low- resource settings. For Maya, fine-tuning NLLB- 200-3.3B using StemCorrupt augmented data yielded the best performance.

pdf bib
Experiments in Mamba Sequence Modeling and NLLB-200 Fine-Tuning for Low Resource Multilingual Machine Translation
Dan Degenaro | Tom Lupicki

This paper presents DC_DMV’s submission to the AmericasNLP 2024 Shared Task 1: Machine Translation Systems for Indigenous Languages. Our submission consists of two multilingual approaches to building machine translation systems from Spanish to eleven Indigenous languages: fine-tuning the 600M distilled variant of NLLB-200, and an experiment in training from scratch a neural network using the Mamba State Space Modeling architecture. We achieve the best results on the test set for a total of 4 of the language pairs between two checkpoints by fine-tuning NLLB-200, and outperform the baseline score on the test set for 2 languages.

pdf bib
JGU Mainz’s Submission to the AmericasNLP 2024 Shared Task on the Creation of Educational Materials for Indigenous Languages
Minh Duc Bui | Katharina von der Wense

In this paper, we present the four systems developed by the Meenzer team from JGU for the AmericasNLP 2024 shared task on the creation of educational materials for Indigenous languages. The task involves accurately applying specific grammatical modifications to given source sentences across three low-resource Indigenous languages: Bribri, Guarani, and Maya. We train two types of model architectures: finetuning a sequence-to-sequence pointer-generator LSTM and finetuning the Mixtral 8x7B model by incorporating in-context examples into the training phase. System 1, an ensemble combining finetuned LSTMs, finetuned Mixtral models, and GPT-4, achieves the best performance on Guarani. Meanwhile, system 4, another ensemble consisting solely of fine-tuned Mixtral models, outperforms all other teams on Maya and secures the second place overall. Additionally, we conduct an ablation study to understand the performance of our system 4.

pdf bib
Applying Linguistic Expertise to LLMs for Educational Material Development in Indigenous Languages
Justin Vasselli | Arturo Martínez Peguero | Junehwan Sung | Taro Watanabe

This paper presents our approach to the AmericasNLP 2024 Shared Task 2 as the JAJ (/dʒæz/) team. The task aimed at creating educational materials for indigenous languages, and we focused on Maya and Bribri. Given the unique linguistic features and challenges of these languages, and the limited size of the training datasets, we developed a hybrid methodology combining rule-based NLP methods with prompt-based techniques. This approach leverages the meta-linguistic capabilities of large language models, enabling us to blend broad, language-agnostic processing with customized solutions. Our approach lays a foundational framework that can be expanded to other indigenous languages languages in future work.

pdf bib
Exploring Very Low-Resource Translation with LLMs: The University of Edinburgh’s Submission to AmericasNLP 2024 Translation Task
Vivek Iyer | Bhavitvya Malik | Wenhao Zhu | Pavel Stepachev | Pinzhen Chen | Barry Haddow | Alexandra Birch

This paper describes the University of Edinburgh’s submission to the AmericasNLP 2024 shared task on the translation of Spanish into 11 indigenous American languages. We explore the ability of multilingual Large Language Models (LLMs) to model low-resource languages by continued pre-training with LoRA, and conduct instruction fine-tuning using a variety of datasets, demonstrating that this improves LLM performance. Furthermore, we demonstrate the efficacy of checkpoint averaging alongside decoding techniques like beam search and sampling, resulting in further improvements. We participate in all 11 translation directions.

pdf bib
The role of morphosyntactic similarity in generating related sentences
Michael Hammond

In this paper we describe our work on Task~2: Creation of Educational Materials. We tried three approaches, but only the third approach yielded improvement over the baseline system. The first system was a fairly generic transformer model. The second system was our own implementation of the edit tree approach from the baseline system. Our final attempt was a version of the baseline system where if no transformation succeeded, we applied transformations from similar morphosyntactic relations. We describe all three here, but, in the end, we only submitted the third system.

pdf bib
Findings of the AmericasNLP 2024 Shared Task on the Creation of Educational Materials for Indigenous Languages
Luis Chiruzzo | Pavel Denisov | Alejandro Molina-Villegas | Silvia Fernandez-Sabido | Rolando Coto-Solano | Marvin Agüero-Torales | Aldo Alvarez | Samuel Canul-Yah | Lorena Hau-Ucán | Abteen Ebrahimi | Robert Pugh | Arturo Oncevay | Shruti Rijhwani | Katharina von der Wense | Manuel Mager

This paper presents the results of the first shared task about the creation of educational materials for three indigenous languages of the Americas.The task proposes to automatically generate variations of sentences according to linguistic features that could be used for grammar exercises.The languages involved in this task are Bribri, Maya, and Guarani.Seven teams took part in the challenge, submitting a total of 22 systems, obtaining very promising results.

pdf bib
Findings of the AmericasNLP 2024 Shared Task on Machine Translation into Indigenous Languages
Abteen Ebrahimi | Ona de Gibert | Raul Vazquez | Rolando Coto-Solano | Pavel Denisov | Robert Pugh | Manuel Mager | Arturo Oncevay | Luis Chiruzzo | Katharina von der Wense | Shruti Rijhwani

This paper presents the findings of the third iteration of the AmericasNLP Shared Task on Machine Translation. This year’s competition features eleven Indigenous languages found across North, Central, and South America. A total of six teams participate with a total of 157 submissions across all languages and models. Two baselines – the Sheffield and Helsinki systems from 2023 – are provided and represent hard-to-beat starting points for the competition. In addition to the baselines, teams are given access to a new repository of training data which consists of data collected by teams in prior shared tasks. Using ChrF++ as the main competition metric, we see improvements over the baseline for 4 languages: Chatino, Guarani, Quechua, and Rarámuri, with performance increases over the best baseline of 4.2 ChrF++. In this work, we present a summary of the submitted systems, results, and a human evaluation of system outputs for Bribri, which consists of both (1) a rating of meaning and fluency and (2) a qualitative error analysis of outputs from the best submitted system.
up

pdf (full)
bib (full) Proceedings of the 19th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2024)

pdf bib
Proceedings of the 19th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2024)
Ekaterina Kochmar | Marie Bexte | Jill Burstein | Andrea Horbach | Ronja Laarmann-Quante | Anaïs Tack | Victoria Yaneva | Zheng Yuan

pdf bib
How Good are Modern LLMs in Generating Relevant and High-Quality Questions at Different Bloom’s Skill Levels for Indian High School Social Science Curriculum?
Nicy Scaria | Suma Chenna | Deepak Subramani

The creation of pedagogically effective questions is a challenge for teachers and requires significant time and meticulous planning, especially in resource-constrained economies. For example, in India, assessments for social science in high schools are characterized by rote memorization without regard to higher-order skill levels. Automated educational question generation (AEQG) using large language models (LLMs) has the potential to help teachers develop assessments at scale. However, it is important to evaluate the quality and relevance of these questions. In this study, we examine the ability of different LLMs (Falcon 40B, Llama2 70B, Palm 2, GPT 3.5, and GPT 4) to generate relevant and high-quality questions of different cognitive levels, as defined by Bloom’s taxonomy. We prompt each model with the same instructions and different contexts to generate 510 questions in the social science curriculum of a state educational board in India. Two human experts used a nine-item rubric to assess linguistic correctness, pedagogical relevance and quality, and adherence to Bloom’s skill levels. Our results showed that 91.56% of the LLM-generated questions were relevant and of high quality. This suggests that LLMs can generate relevant and high-quality questions at different cognitive levels, making them useful for creating assessments for scaling education in resource-constrained economies.

pdf bib
Synthetic Data Generation for Low-resource Grammatical Error Correction with Tagged Corruption Models
Felix Stahlberg | Shankar Kumar

Tagged corruption models provide precise control over the introduction of grammatical errors into clean text. This capability has made them a powerful tool for generating pre-training data for grammatical error correction (GEC) in English. In this work, we demonstrate their application to four languages with substantially fewer GEC resources than English: German, Romanian, Russian, and Spanish. We release a new tagged-corruption dataset consisting of 2.5M examples per language that was generated by a fine-tuned PaLM 2 foundation model. Pre-training on tagged corruptions yields consistent gains across all four languages, especially for small model sizes and languages with limited human-labelled data.

pdf bib
Pillars of Grammatical Error Correction: Comprehensive Inspection Of Contemporary Approaches In The Era of Large Language Models
Kostiantyn Omelianchuk | Andrii Liubonko | Oleksandr Skurzhanskyi | Artem Chernodub | Oleksandr Korniienko | Igor Samokhin

In this paper, we carry out experimental research on Grammatical Error Correction, delving into the nuances of single-model systems, comparing the efficiency of ensembling and ranking methods, and exploring the application of large language models to GEC as single-model systems, as parts of ensembles, and as ranking methods. We set new state-of-the-art records with F_0.5 scores of 72.8 on CoNLL-2014-test and 81.4 on BEA-test, respectively. To support further advancements in GEC and ensure the reproducibility of our research, we make our code, trained models, and systems’ outputs publicly available, facilitating future findings.

pdf bib
Using Adaptive Empathetic Responses for Teaching English
Li Siyan | Teresa Shao | Julia Hirschberg | Zhou Yu

Existing English-teaching chatbots rarely incorporate empathy explicitly in their feedback, but empathetic feedback could help keep students engaged and reduce learner anxiety. Toward this end, we propose the task of negative emotion detection via audio, for recognizing empathetic feedback opportunities in language learning. We then build the first spoken English-teaching chatbot with adaptive, empathetic feedback. This feedback is synthesized through automatic prompt optimization of ChatGPT and is evaluated with English learners. We demonstrate the effectiveness of our system through a preliminary user study.

pdf bib
Beyond Flesch-Kincaid: Prompt-based Metrics Improve Difficulty Classification of Educational Texts
Donya Rooein | Paul Röttger | Anastassia Shaitarova | Dirk Hovy

Using large language models (LLMs) for educational applications like dialogue-based teaching is a hot topic. Effective teaching, however, requires teachers to adapt the difficulty of content and explanations to the education level of their students. Even the best LLMs today struggle to do this well. If we want to improve LLMs on this adaptation task, we need to be able to measure adaptation success reliably. However, current Static metrics for text difficulty, like the Flesch-Kincaid Reading Ease score, are known to be crude and brittle. We, therefore, introduce and evaluate a new set of Prompt-based metrics for text difficulty. Based on a user study, we create Prompt-based metrics as inputs for LLMs. They leverage LLM’s general language understanding capabilities to capture more abstract and complex features than Static metrics. Regression experiments show that adding our Prompt-based metrics significantly improves text difficulty classification over Static metrics alone. Our results demonstrate the promise of using LLMs to evaluate text adaptation to different education levels.

pdf bib
Large Language Models Are State-of-the-Art Evaluator for Grammatical Error Correction
Masamune Kobayashi | Masato Mita | Mamoru Komachi

Large Language Models (LLMs) have been reported to outperform existing automatic evaluation metrics in some tasks, such as text summarization and machine translation. However, there has been a lack of research on LLMs as evaluators in grammatical error correction (GEC). In this study, we investigate the performance of LLMs in GEC evaluation by employing prompts designed to incorporate various evaluation criteria inspired by previous research. Our extensive experimental results demonstrate that GPT-4 achieved Kendall’s rank correlation of 0.662 with human judgments, surpassing all existing methods. Furthermore, in recent GEC evaluations, we have underscored the significance of the LLMs scale and particularly emphasized the importance of fluency among evaluation criteria.

pdf bib
Can Language Models Guess Your Identity? Analyzing Demographic Biases in AI Essay Scoring
Alexander Kwako | Christopher Ormerod

Large language models (LLMs) are increasingly used for automated scoring of student essays. However, these models may perpetuate societal biases if not carefully monitored. This study analyzes potential biases in an LLM (XLNet) trained to score persuasive student essays, based on data from the PERSUADE corpus. XLNet achieved strong performance based on quadratic weighted kappa, standardized mean difference, and exact agreement with human scores. Using available metadata, we performed analyses of scoring differences across gender, race/ethnicity, English language learning status, socioeconomic status, and disability status. Automated scores exhibited small magnifications of marginal differences in human scoring, favoring female students over males and White students over Black students. To further probe potential biases, we found that separate XLNet classifiers and XLNet hidden states weakly predicted demographic membership. Overall, results reinforce the need for continued fairness analyses as use of LLMs expands in education.

pdf bib
Automated Scoring of Clinical Patient Notes: Findings From the Kaggle Competition and Their Translation into Practice
Victoria Yaneva | King Yiu Suen | Le An Ha | Janet Mee | Milton Quranda | Polina Harik

Scoring clinical patient notes (PNs) written by medical students is a necessary but resource-intensive task in medical education. This paper describes the organization and key lessons from a Kaggle competition on automated scoring of such notes. 1,471 teams took part in the competition and developed an extensive, publicly available code repository of varying solutions evaluated over the first public dataset for this task. The most successful approaches from this community effort are described and utilized in the development of a PN scoring system. We discuss the choice of models and system architecture with a view to operational use and scalability, and evaluate its performance on both the public Kaggle data (10 clinical cases, 43,985 PNs) and an extended internal dataset (178 clinical cases, 6,940 PNs). The results show that the system significantly outperforms a state-of-the-art existing tool for PN scoring and that task-adaptive pretraining using masked language modeling can be an effective approach even for small training samples.

pdf bib
A World CLASSE Student Summary Corpus
Scott Crossley | Perpetual Baffour | Mihai Dascalu | Stefan Ruseti

This paper introduces the Common Lit Augmented Student Summary Evaluation (CLASSE) corpus. The corpus comprises 11,213 summaries written over six prompts by students in grades 3-12 while using the CommonLit website. Each summary was scored by expert human raters on analytic features related to main points, details, organization, voice, paraphrasing, and language beyond the source text. The human scores were aggregated into two component scores related to content and wording. The final corpus was the focus of a Kaggle competition hosted in late 2022 and completed in 2023 in which over 2,000 teams participated. The paper includes a baseline scoring model for the corpus based on a Large Language Model (Longformer model). The paper also provides an overview of the winning models from the Kaggle competition.

pdf bib
Improving Socratic Question Generation using Data Augmentation and Preference Optimization
Nischal Ashok Kumar | Andrew Lan

The Socratic method is a way of guiding students toward solving a problem independently without directly revealing the solution to the problem by asking incremental questions. Although this method has been shown to significantly improve student learning outcomes, it remains a complex labor-intensive task for instructors. Large language models (LLMs) can be used to augment human effort by automatically generating Socratic questions for students. However, existing methods that involve prompting these LLMs sometimes produce invalid outputs, e.g., those that directly reveal the solution to the problem or provide irrelevant or premature questions. To alleviate this problem, inspired by reinforcement learning with AI feedback (RLAIF), we first propose a data augmentation method to enrich existing Socratic questioning datasets with questions that are invalid in specific ways. Also, we propose a method to optimize open-source LLMs such as LLama 2 to prefer ground-truth questions over generated invalid ones, using direct preference optimization (DPO). Our experiments on a Socratic questions dataset for student code debugging show that a DPO-optimized LLama 2-7B model can effectively avoid generating invalid questions, and as a result, outperforms existing state-of-the-art prompting methods.

pdf bib
Scoring with Confidence? – Exploring High-confidence Scoring for Saving Manual Grading Effort
Marie Bexte | Andrea Horbach | Lena Schützler | Oliver Christ | Torsten Zesch

A possible way to save manual grading effort in short answer scoring is to automatically score answers for which the classifier is highly confident. We explore the feasibility of this approach in a high-stakes exam setting, evaluating three different similarity-based scoring methods, where the similarity score is a direct proxy for model confidence. The decision on an appropriate level of confidence should ideally be made before scoring a new prompt. We thus probe to what extent confidence thresholds are consistent across different datasets and prompts. We find that high-confidence thresholds vary on a prompt-to-prompt basis, and that the overall potential of increased performance at a reasonable cost of additional manual effort is limited.

pdf bib
Predicting Initial Essay Quality Scores to Increase the Efficiency of Comparative Judgment Assessments
Michiel De Vrindt | Anaïs Tack | Renske Bouwer | Wim Van Den Noortgate | Marije Lesterhuis

Comparative judgment (CJ) is a method that can be used to assess the writing quality of student essays based on repeated pairwise comparisons by multiple assessors. Although the assessment method is known to have high validity and reliability, it can be particularly inefficient, as assessors must make many judgments before the scores become reliable. Prior research has investigated methods to improve the efficiency of CJ, yet these methods introduce additional challenges, notably stemming from the initial lack of information at the start of the assessment, which is known as a cold-start problem. This paper reports on a study in which we predict the initial quality scores of essays to establish a warm start for CJ. To achieve this, we construct informative prior distributions for the quality scores based on the predicted initial quality scores. Through simulation studies, we demonstrate that our approach increases the efficiency of CJ: On average, assessors need to make 30% fewer judgments for each essay to reach an overall reliability level of 0.70.

pdf bib
Improving Transfer Learning for Early Forecasting of Academic Performance by Contextualizing Language Models
Ahatsham Hayat | Bilal Khan | Mohammad Hasan

This paper presents a cutting-edge method that harnesses contextualized language models (LMs) to significantly enhance the prediction of early academic performance in STEM fields. Our approach uniquely tackles the challenge of transfer learning with limited-domain data. Specifically, we overcome this challenge by contextualizing students’ cognitive trajectory data through the integration of both distal background factors (comprising academic information, demographic details, and socioeconomic indicators) and proximal non-cognitive factors (such as emotional engagement). By tapping into the rich prior knowledge encoded within pre-trained LMs, we effectively reframe academic performance forecasting as a task ideally suited for natural language processing.Our research rigorously examines three key aspects: the impact of data contextualization on prediction improvement, the effectiveness of our approach compared to traditional numeric-based models, and the influence of LM capacity on prediction accuracy. The results underscore the significant advantages of utilizing larger LMs with contextualized inputs, representing a notable advancement in the precision of early performance forecasts. These findings emphasize the importance of employing contextualized LMs to enhance artificial intelligence-driven educational support systems and overcome data scarcity challenges.

pdf bib
Can GPT-4 do L2 analytic assessment?
Stefano Banno | Hari Krishna Vydana | Kate Knill | Mark Gales

Automated essay scoring (AES) to evaluate second language (L2) proficiency has been a firmly established technology used in educational contexts for decades. Although holistic scoring has seen advancements in AES that match or even exceed human performance, analytic scoring still encounters issues as it inherits flaws and shortcomings from the human scoring process. The recent introduction of large language models presents new opportunities for automating the evaluation of specific aspects of L2 writing proficiency. In this paper, we perform a series of experiments using GPT-4 in a zero-shot fashion on a publicly available dataset annotated with holistic scores based on the Common European Framework of Reference and aim to extract detailed information about their underlying analytic components. We observe significant correlations between the automatically predicted analytic scores and multiple features associated with the individual proficiency components.

pdf bib
Using Program Repair as a Proxy for Language Models’ Feedback Ability in Programming Education
Charles Koutcheme | Nicola Dainese | Arto Hellas

One of the key challenges in programming education is being able to provide high-quality feedback to learners. Such feedback often includes explanations of the issues in students’ programs coupled with suggestions on how to fix these issues. Large language models (LLMs) have recently emerged as valuable tools that can help in this effort. In this article, we explore the relationship between the program repair ability of LLMs and their proficiency in providing natural language explanations of coding mistakes. We outline a benchmarking study that evaluates leading LLMs (including open-source ones) on program repair and explanation tasks. Our experiments study the capabilities of LLMs both on a course level and on a programming concept level, allowing us to assess whether the programming concepts practised in exercises with faulty student programs relate to the performance of the models. Our results highlight that LLMs proficient in repairing student programs tend to provide more complete and accurate natural language explanations of code issues. Overall, these results enhance our understanding of the role and capabilities of LLMs in programming education. Using program repair as a proxy for explanation evaluation opens the door for cost-effective assessment methods.

pdf bib
Automated Evaluation of Teacher Encouragement of Student-to-Student Interactions in a Simulated Classroom Discussion
Michael Ilagan | Beata Beigman Klebanov | Jamie Mikeska

Leading students to engage in argumentation-focused discussions is a challenge for elementary school teachers, as doing so requires facilitating group discussions with student-to-student interaction. The Mystery Powder (MP) Task was designed to be used in online simulated classrooms to develop teachers’ skill in facilitating small group science discussions. In order to provide timely and scaleable feedback to teachers facilitating a discussion in the simulated classroom, we employ a hybrid modeling approach that successfully combines fine-tuned large language models with features capturing important elements of the discourse dynamic to evaluate MP discussion transcripts. To our knowledge, this is the first application of a hybrid model to automate evaluation of teacher discourse.

pdf bib
Explainable AI in Language Learning: Linking Empirical Evidence and Theoretical Concepts in Proficiency and Readability Modeling of Portuguese
Luisa Ribeiro-Flucht | Xiaobin Chen | Detmar Meurers

While machine learning methods have supported significantly improved results in education research, a common deficiency lies in the explainability of the result. Explainable AI (XAI) aims to fill that gap by providing transparent, conceptually understandable explanations for the classification decisions, enhancing human comprehension and trust in the outcomes. This paper explores an XAI approach to proficiency and readability assessment employing a comprehensive set of 465 linguistic complexity measures. We identify theoretical descriptions associating such measures with varying levels of proficiency and readability and validate them using cross-corpus experiments employing supervised machine learning and Shapley Additive Explanations. The results not only highlight the utility of a diverse set of complexity measures in effectively modeling proficiency and readability in Portuguese, achieving a state-of-the-art accuracy of 0.70 in the proficiency classification task and of 0.84 in the readability classification task, but they largely corroborate the theoretical research assumptions, especially in the lexical domain.

pdf bib
Fairness in Automated Essay Scoring: A Comparative Analysis of Algorithms on German Learner Essays from Secondary Education
Nils-Jonathan Schaller | Yuning Ding | Andrea Horbach | Jennifer Meyer | Thorben Jansen

Pursuing educational equity, particularly in writing instruction, requires that all students receive fair (i.e., accurate and unbiased) assessment and feedback on their texts. Automated Essay Scoring (AES) algorithms have so far focused on optimizing the mean accuracy of their scores and paid less attention to fair scores for all subgroups, although research shows that students receive unfair scores on their essays in relation to demographic variables, which in turn are related to their writing competence. We add to the literature arguing that AES should also optimize for fairness by presenting insights on the fairness of scoring algorithms on a corpus of learner texts in the German language and introduce the novelty of examining fairness on psychological and demographic differences in addition to demographic differences. We compare shallow learning, deep learning, and large language models with full and skewed subsets of training data to investigate what is needed for fair scoring. The results show that training on a skewed subset of higher and lower cognitive ability students shows no bias but very low accuracy for students outside the training set. Our results highlight the need for specific training data on all relevant user groups, not only for demographic background variables but also for cognitive abilities as psychological student characteristics.

pdf bib
Improving Automated Distractor Generation for Math Multiple-choice Questions with Overgenerate-and-rank
Alexander Scarlatos | Wanyong Feng | Andrew Lan | Simon Woodhead | Digory Smith

Multiple-choice questions (MCQs) are commonly used across all levels of math education since they can be deployed and graded at a large scale. A critical component of MCQs is the distractors, i.e., incorrect answers crafted to reflect student errors or misconceptions. Automatically generating them in math MCQs, e.g., with large language models, has been challenging. In this work, we propose a novel method to enhance the quality of generated distractors through overgenerate-and-rank, training a ranking model to predict how likely distractors are to be selected by real students. Experimental results on a real-world dataset and human evaluation with math teachers show that our ranking model increases alignment with human-authored distractors, although human-authored ones are still preferred over generated ones.

pdf bib
Identifying Fairness Issues in Automatically Generated Testing Content
Kevin Stowe

Natural language generation tools are powerful and effective for generating content. However, language models are known to display bias and fairness issues, making them impractical to deploy for many use cases. We here focus on how fairness issues impact automatically generated test content, which can have stringent requirements to ensure the test measures only what it was intended to measure. Specifically, we review test content generated for a large-scale standardized English proficiency test with the goal of identifying content that only pertains to a certain subset of the test population as well as content that has the potential to be upsetting or distracting to some test takers. Issues like these could inadvertently impact a test taker’s score and thus should be avoided. This kind of content does not reflect the more commonly-acknowledged biases, making it challenging even for modern models that contain safeguards. We build a dataset of 601 generated texts annotated for fairness and explore a variety of methods for classification: fine-tuning, topic-based classification, and prompting, including few-shot and self-correcting prompts. We find that combining prompt self-correction and few-shot learning performs best, yielding an F1 score of 0.79 on our held-out test set, while much smaller BERT- and topic-based models have competitive performance on out-of-domain data.

pdf bib
Towards Automated Document Revision: Grammatical Error Correction, Fluency Edits, and Beyond
Masato Mita | Keisuke Sakaguchi | Masato Hagiwara | Tomoya Mizumoto | Jun Suzuki | Kentaro Inui

Natural language processing (NLP) technology has rapidly improved automated grammatical error correction (GEC) tasks, and the GEC community has begun to explore document-level revision. However, there are two major obstacles to going beyond automated sentence-level GEC to NLP-based document-level revision support: (1) there are few public corpora with document-level revisions annotated by professional editors, and (2) it is infeasible to obtain all possible references and evaluate revision quality using such references because there are infinite revision possibilities. To address these challenges, this paper proposes a new document revision corpus, Text Revision of ACL papers (TETRA), in which multiple professional editors have revised academic papers sampled from the ACL anthology. This corpus enables us to focus on document-level and paragraph-level edits, such as edits related to coherence and consistency. Additionally, as a case study using the TETRA corpus, we investigate reference-less and interpretable methods for meta-evaluation to detect quality improvements according to document revisions. We show the uniqueness of TETRA compared with existing document revision corpora and demonstrate that a fine-tuned pre-trained language model can discriminate the quality of documents after revision even when the difference is subtle.

pdf bib
Evaluating Vocabulary Usage in LLMs
Matthew Durward | Christopher Thomson

The paper focuses on investigating vocabulary usage for AI and human-generated text. We define vocabulary usage in two ways: structural differences and keyword differences. Structural differences are evaluated by converting text into Vocabulary-Managment Profiles, initially used for discourse analysis. Through VMPs, we can treat the text data as a time series, allowing an evaluation by implementing Dynamic time-warping distance measures and subsequently deriving similarity scores to provide an indication of whether the structural dynamics in AI texts resemble human texts. To analyze keywords, we use a measure that emphasizes frequency and dispersion to source ‘key’ keywords. A qualitative approach is then applied, noting thematic differences between human and AI writing.

pdf bib
Exploring LLM Prompting Strategies for Joint Essay Scoring and Feedback Generation
Maja Stahl | Leon Biermann | Andreas Nehring | Henning Wachsmuth

Individual feedback can help students improve their essay writing skills. However, the manual effort required to provide such feedback limits individualization in practice. Automatically-generated essay feedback may serve as an alternative to guide students at their own pace, convenience, and desired frequency. Large language models (LLMs) have demonstrated strong performance in generating coherent and contextually relevant text. Yet, their ability to provide helpful essay feedback is unclear. This work explores several prompting strategies for LLM-based zero-shot and few-shot generation of essay feedback. Inspired by Chain-of-Thought prompting, we study how and to what extent automated essay scoring (AES) can benefit the quality of generated feedback. We evaluate both the AES performance that LLMs can achieve with prompting only and the helpfulness of the generated essay feedback. Our results suggest that tackling AES and feedback generation jointly improves AES performance. However, while our manual evaluation emphasizes the quality of the generated essay feedback, the impact of essay scoring on the generated feedback remains low ultimately.

pdf bib
Towards Fine-Grained Pedagogical Control over English Grammar Complexity in Educational Text Generation
Dominik Glandorf | Detmar Meurers

Teaching foreign languages and fostering language awareness in subject matter teaching requires a profound knowledge of grammar structures. Yet, while Large Language Models can act as tutors, it is unclear how effectively they can control grammar in generated text and adapt to learner needs. In this study, we investigate the ability of these models to exemplify pedagogically relevant grammar patterns, detect instances of grammar in a given text, and constrain text generation to grammar characteristic of a proficiency level. Concretely, we (1) evaluate the ability of GPT3.5 and GPT4 to generate example sentences for the standard English Grammar Profile CEFR taxonomy using few-shot in-context learning, (2) train BERT-based detectors with these generated examples of grammatical patterns, and (3) control the grammatical complexity of text generated by the open Mistral model by ranking sentence candidates with these detectors. We show that the grammar pattern instantiation quality is accurate but too homogeneous, and our classifiers successfully detect these patterns. A GPT-generated dataset of almost 1 million positive and negative examples for the English Grammar Profile is released with this work. With our method, Mistral’s output significantly increases the number of characteristic grammar constructions on the desired level, outperforming GPT4. This showcases how language domain knowledge can enhance Large Language Models for specific education needs, facilitating their effective use for intelligent tutor development and AI-generated materials. Code, models, and data are available at https://github.com/dominikglandorf/LLM-grammar.

pdf bib
LLMs in Short Answer Scoring: Limitations and Promise of Zero-Shot and Few-Shot Approaches
Imran Chamieh | Torsten Zesch | Klaus Giebermann

In this work, we investigate the potential of Large Language Models (LLMs) for automated short answer scoring. We test zero-shot and few-shot settings, and compare with fine-tuned models and a supervised upper-bound, across three diverse datasets. Our results, in zero-shot and few-shot settings, show that LLMs perform poorly in these settings: LLMs have difficulty with tasks that require complex reasoning or domain-specific knowledge. While the models show promise on general knowledge tasks. The fine-tuned model come close to the supervised results but are still not feasible for application, highlighting potential overfitting issues. Overall, our study highlights the challenges and limitations of LLMs in short answer scoring and indicates that there currently seems to be no basis for applying LLMs for short answer scoring.

pdf bib
Automated Essay Scoring Using Grammatical Variety and Errors with Multi-Task Learning and Item Response Theory
Kosuke Doi | Katsuhito Sudoh | Satoshi Nakamura

This study examines the effect of grammatical features in automatic essay scoring (AES). We use two kinds of grammatical features as input to an AES model: (1) grammatical items that writers used correctly in essays, and (2) the number of grammatical errors. Experimental results show that grammatical features improve the performance of AES models that predict the holistic scores of essays. Multi-task learning with the holistic and grammar scores, alongside using grammatical features, resulted in a larger improvement in model performance. We also show that a model using grammar abilities estimated using Item Response Theory (IRT) as the labels for the auxiliary task achieved comparable performance to when we used grammar scores assigned by human raters. In addition, we weight the grammatical features using IRT to consider the difficulty of grammatical items and writers’ grammar abilities. We found that weighting grammatical features with the difficulty led to further improvement in performance.

pdf bib
Error Tracing in Programming: A Path to Personalised Feedback
Martha Shaka | Diego Carraro | Kenneth Brown

Knowledge tracing, the process of estimating students’ mastery over concepts from their past performance and predicting future outcomes, often relies on binary pass/fail predictions. This hinders the provision of specific feedback by failing to diagnose precise errors. We present an error-tracing model for learning programming that advances traditional knowledge tracing by employing multi-label classification to forecast exact errors students may generate. Through experiments on a real student dataset, we validate our approach and compare it to two baseline knowledge-tracing methods. We demonstrate an improved ability to predict specific errors, for first attempts and for subsequent attempts at individual problems.

pdf bib
Improving Readability Assessment with Ordinal Log-Loss
Ho Hung Lim | John Lee

Automatic Readability Assessment (ARA) predicts the level of difficulty of a text, e.g. at Grade 1 to Grade 12. ARA is an ordinal classification task since the predicted levels follow an underlying order, from easy to difficult. However, most neural ARA models ignore the distance between the gold level and predicted level, treating all levels as independent labels. This paper investigates whether distance-sensitive loss functions can improve ARA performance. We evaluate a variety of loss functions on neural ARA models, and show that ordinal log-loss can produce statistically significant improvement over the standard cross-entropy loss in terms of adjacent accuracy in a majority of our datasets.

pdf bib
Automated Sentence Generation for a Spaced Repetition Software
Benjamin Paddags | Daniel Hershcovich | Valkyrie Savage

This paper presents and tests AllAI, an app that utilizes state-of-the-art NLP technology to assist second language acquisition through a novel method of sentence-based spaced repetition. Diverging from current single word or fixed sentence repetition, AllAI dynamically combines words due for repetition into sentences, enabling learning words in context while scheduling them independently. This research explores various suitable NLP paradigms and finds a few-shot prompting approach and retrieval of existing sentences from a corpus to yield the best correctness and scheduling accuracy. Subsequently, it evaluates these methods on 26 learners of Danish, finding a four-fold increase in the speed at which new words are learned, compared to conventional spaced repetition. Users of the retrieval method also reported significantly higher enjoyment, hinting at a higher user engagement.

pdf bib
Using Large Language Models to Assess Young Students’ Writing Revisions
Tianwen Li | Zhexiong Liu | Lindsay Matsumura | Elaine Wang | Diane Litman | Richard Correnti

Although effective revision is the crucial component of writing instruction, few automated writing evaluation (AWE) systems specifically focus on the quality of the revisions students undertake. In this study we investigate the use of a large language model (GPT-4) with Chain-of-Thought (CoT) prompting for assessing the quality of young students’ essay revisions aligned with the automated feedback messages they received. Results indicate that GPT-4 has significant potential for evaluating revision quality, particularly when detailed rubrics are included that describe common revision patterns shown by young writers. However, the addition of CoT prompting did not significantly improve performance. Further examination of GPT-4’s scoring performance across various levels of student writing proficiency revealed variable agreement with human ratings. The implications for improving AWE systems focusing on young students are discussed.

pdf bib
Automatic Crossword Clues Extraction for Language Learning
Santiago Berruti | Arturo Collazo | Diego Sellanes | Aiala Rosá | Luis Chiruzzo

Crosswords are a powerful tool that could be used in educational contexts, but they are not that easy to build. In this work, we present experiments on automatically extracting clues from simple texts that could be used to create crosswords, with the aim of using them in the context of teaching English at the beginner level. We present a series of heuristic patterns based on NLP tools for extracting clues, and use them to create a set of 2209 clues from a collection of 400 simple texts. Human annotators labeled the clues, and this dataset is used to evaluate the performance of our heuristics, and also to create a classifier that predicts if an extracted clue is correct. Our best classifier achieves an accuracy of 84%.

pdf bib
Anna Karenina Strikes Again: Pre-Trained LLM Embeddings May Favor High-Performing Learners
Abigail Gurin Schleifer | Beata Beigman Klebanov | Moriah Ariely | Giora Alexandron

Unsupervised clustering of student responses to open-ended questions into behavioral and cognitive profiles using pre-trained LLM embeddings is an emerging technique, but little is known about how well this captures pedagogically meaningful information. We investigate this in the context of student responses to open-ended questions in biology, which were previously analyzed and clustered by experts into theory-driven Knowledge Profiles (KPs).Comparing these KPs to ones discovered by purely data-driven clustering techniques, we report poor discoverability of most KPs, except for the ones including the correct answers. We trace this ‘discoverability bias’ to the representations of KPs in the pre-trained LLM embeddings space.

pdf bib
Assessing Student Explanations with Large Language Models Using Fine-Tuning and Few-Shot Learning
Dan Carpenter | Wookhee Min | Seung Lee | Gamze Ozogul | Xiaoying Zheng | James Lester

The practice of soliciting self-explanations from students is widely recognized for its pedagogical benefits. However, the labor-intensive effort required to manually assess students’ explanations makes it impractical for classroom settings. As a result, many current solutions to gauge students’ understanding during class are often limited to multiple choice or fill-in-the-blank questions, which are less effective at exposing misconceptions or helping students to understand and integrate new concepts. Recent advances in large language models (LLMs) present an opportunity to assess student explanations in real-time, making explanation-based classroom response systems feasible for implementation. In this work, we investigate LLM-based approaches for assessing the correctness of students’ explanations in response to undergraduate computer science questions. We investigate alternative prompting approaches for multiple LLMs (i.e., Llama 2, GPT-3.5, and GPT-4) and compare their performance to FLAN-T5 models trained in a fine-tuning manner. The results suggest that the highest accuracy and weighted F1 score were achieved by fine-tuning FLAN-T5, while an in-context learning approach with GPT-4 attains the highest macro F1 score.

pdf bib
Harnessing GPT to Study Second Language Learner Essays: Can We Use Perplexity to Determine Linguistic Competence?
Ricardo Muñoz Sánchez | Simon Dobnik | Elena Volodina

Generative language models have been used to study a wide variety of phenomena in NLP. This allows us to better understand the linguistic capabilities of those models and to better analyse the texts that we are working with. However, these studies have mainly focused on text generated by L1 speakers of English. In this paper we study whether linguistic competence of L2 learners of Swedish (through their performance on essay tasks) correlates with the perplexity of a decoder-only model (GPT-SW3). We run two sets of experiments, doing both quantitative and qualitative analyses for each of them. In the first one, we analyse the perplexities of the essays and compare them with the CEFR level of the essays, both from an essay-wide level and from a token level. In our second experiment, we compare the perplexity of an L2 learner essay with a normalised version of it. We find that the perplexity of essays tends to be lower for higher CEFR levels and that normalised essays have a lower perplexity than the original versions. Moreover, we find that different factors can lead to spikes in perplexity, not all of them being related to L2 learner language.

pdf bib
BERT-IRT: Accelerating Item Piloting with BERT Embeddings and Explainable IRT Models
Kevin P. Yancey | Andrew Runge | Geoffrey LaFlair | Phoebe Mulcaire

Estimating item parameters (e.g., the difficulty of a question) is an important part of modern high-stakes tests. Conventional methods require lengthy pilots to collect response data from a representative population of test-takers. The need for these pilots limit item bank size and how often those item banks can be refreshed, impacting test security, while increasing costs needed to support the test and taking up the test-taker’s valuable time. Our paper presents a novel explanatory item response theory (IRT) model, BERT-IRT, that has been used on the Duolingo English Test (DET), a high-stakes test of English, to reduce the length of pilots by a factor of 10. Our evaluation shows how the model uses BERT embeddings and engineered NLP features to accelerate item piloting without sacrificing criterion validity or reliability.

pdf bib
Transfer Learning of Argument Mining in Student Essays
Yuning Ding | Julian Lohmann | Nils-Jonathan Schaller | Thorben Jansen | Andrea Horbach

This paper explores the transferability of a cross-prompt argument mining model trained on argumentative essays authored by native English-speaking learners (EN-L1) across educational contexts and languages. Specifically, the adaptability of a multilingual transformer model is assessed through its application to comparable argumentative essays authored by English-as-a-foreign-language learners (EN-L2) for context transfer, and a dataset composed of essays written by native German learners (DE) for both language and task transfer. To separate language effects from educational context effects, we also perform experiments on a machine-translated version of the German dataset (DE-MT). Our findings demonstrate that, even under zero-shot conditions, a model trained on native English speakers exhibits satisfactory performance on the EN-L2/DE datasets. Machine translation does not substantially enhance this performance, suggesting that distinct writing styles across educational contexts impact performance more than language differences.

pdf bib
Building Robust Content Scoring Models for Student Explanations of Social Justice Science Issues
Allison Bradford | Kenneth Steimel | Brian Riordan | Marcia Linn

With increased attention to connecting science topics to real-world contexts, like issues of social justice, teachers need support to assess student progress in explaining such issues. In this work, we explore the robustness of NLP-based automatic content scoring models that provide insight into student ability to integrate their science and social justice ideas in two different environmental science contexts. We leverage encoder-only transformer models to capture the degree to which students explain a science phenomenon, understand the intersecting justice issues, and integrate their understanding of science and social justice. We developed models training on data from each of the contexts as well as from a combined dataset. We found that the models developed in one context generate educationally useful scores in the other context. The model trained on the combined dataset performed as well as or better than the models trained on separate datasets in most cases. Quadratic weighted kappas demonstrate that these models are above threshold for use in classrooms.

pdf bib
From Miscue to Evidence of Difficulty: Analysis of Automatically Detected Miscues in Oral Reading for Feedback Potential
Beata Beigman Klebanov | Michael Suhan | Tenaha O’Reilly | Zuowei Wang

This research is situated in the space between an existing NLP capability and its use(s) in an educational context. We analyze oral reading data collected with a deployed automated speech analysis software and consider how the results of automated speech analysis can be interpreted and used to inform the ideation and design of a new feature – feedback to learners and teachers. Our analysis shows how the details of the system’s performance and the details of the context of use both significantly impact the ideation process.

pdf bib
Findings from the First Shared Task on Automated Prediction of Difficulty and Response Time for Multiple-Choice Questions
Victoria Yaneva | Kai North | Peter Baldwin | Le An Ha | Saed Rezayi | Yiyun Zhou | Sagnik Ray Choudhury | Polina Harik | Brian Clauser

This paper reports findings from the First Shared Task on Automated Prediction of Difficulty and Response Time for Multiple-Choice Questions. The task was organized as part of the 19th Workshop on Innovative Use of NLP for Building Educational Applications (BEA’24), held in conjunction with NAACL 2024, and called upon the research community to contribute solutions to the problem of modeling difficulty and response time for clinical multiple-choice questions (MCQs). A set of 667 previously used and now retired MCQs from the United States Medical Licensing Examination (USMLE®) and their corresponding difficulties and mean response times were made available for experimentation. A total of 17 teams submitted solutions and 12 teams submitted system report papers describing their approaches. This paper summarizes the findings from the shared task and analyzes the main approaches proposed by the participants.

pdf bib
Predicting Item Difficulty and Item Response Time with Scalar-mixed Transformer Encoder Models and Rational Network Regression Heads
Sebastian Gombert | Lukas Menzel | Daniele Di Mitri | Hendrik Drachsler

This paper describes a contribution to the BEA 2024 Shared Task on Automated Prediction of Item Difficulty and Response Time. The participants in this shared task are to develop models for predicting the difficulty and response time of multiple-choice items in the medical field. These items were taken from the United States Medical Licensing Examination® (USMLE®), a high-stakes medical exam. For this purpose, we evaluated multiple BERT-like pre-trained transformer encoder models, which we combined with Scalar Mixing and two custom 2-layer classification heads using learnable Rational Activations as an activation function, each for predicting one of the two variables of interest in a multi-task setup. Our best models placed first out of 43 for predicting item difficulty and fifth out of 34 for predicting Item Response Time.

pdf bib
UnibucLLM: Harnessing LLMs for Automated Prediction of Item Difficulty and Response Time for Multiple-Choice Questions
Ana-Cristina Rogoz | Radu Tudor Ionescu

This work explores a novel data augmentation method based on Large Language Models (LLMs) for predicting item difficulty and response time of retired USMLE Multiple-Choice Questions (MCQs) in the BEA 2024 Shared Task. Our approach is based on augmenting the dataset with answers from zero-shot LLMs (Falcon, Meditron, Mistral) and employing transformer-based models based on six alternative feature combinations. The results suggest that predicting the difficulty of questions is more challenging. Notably, our top performing methods consistently include the question text, and benefit from the variability of LLM answers, highlighting the potential of LLMs for improving automated assessment in medical licensing exams. We make our code available at: https://github.com/ana-rogoz/BEA-2024.

pdf bib
The British Council submission to the BEA 2024 shared task
Mariano Felice | Zeynep Duran Karaoz

This paper describes our submission to the item difficulty prediction track of the BEA 2024 shared task. Our submission included the output of three systems: 1) a feature-based linear regression model, 2) a RoBERTa-based model and 3) a linear regression ensemble built on the predictions of the two previous models. Our systems ranked 7th, 8th and 5th respectively, demonstrating that simple models can achieve optimal results. A closer look at the results shows that predictions are more accurate for items in the middle of the difficulty range, with no other obvious relationships between difficulty and the accuracy of predictions.

pdf bib
ITEC at BEA 2024 Shared Task: Predicting Difficulty and Response Time of Medical Exam Questions with Statistical, Machine Learning, and Language Models
Anaïs Tack | Siem Buseyne | Changsheng Chen | Robbe D’hondt | Michiel De Vrindt | Alireza Gharahighehi | Sameh Metwaly | Felipe Kenji Nakano | Ann-Sophie Noreillie

This paper presents the results of our participation in the BEA 2024 shared task on the automated prediction of item difficulty and item response time (APIDIRT), hosted by the NBME (National Board of Medical Examiners). During this task, practice multiple-choice questions from the United States Medical Licensing Examination® (USMLE®) were shared, and research teams were tasked with devising systems capable of predicting the difficulty and average response time for new exam questions.Our team, part of the interdisciplinary itec research group, participated in the task. We extracted linguistic features and clinical embeddings from question items and tested various modeling techniques, including statistical regression, machine learning, language models, and ensemble methods. Surprisingly, simplermodels such as Lasso and random forest regression, utilizing principal component features from linguistic and clinical embeddings, outperformed more complex models. In the competition, our random forest model ranked 4th out of 43 submissions for difficulty prediction, while the Lasso model secured the 2nd position out of 34 submissions for response time prediction. Further analysis suggests that had we submitted the Lasso model for difficulty prediction, we would have achieved an even higher ranking. We also observed that predicting response time is easier than predicting difficulty, with features such as item length, type, exam step, and analytical thinking influencing response time prediction more significantly.

pdf bib
Item Difficulty and Response Time Prediction with Large Language Models: An Empirical Analysis of USMLE Items
Okan Bulut | Guher Gorgun | Bin Tan

This paper summarizes our methodology and results for the BEA 2024 Shared Task. This competition focused on predicting item difficulty and response time for retired multiple-choice items from the United States Medical Licensing Examination® (USMLE®). We extracted linguistic features from the item stem and response options using multiple methods, including the BiomedBERT model, FastText embeddings, and Coh-Metrix. The extracted features were combined with additional features available in item metadata (e.g., item type) to predict item difficulty and average response time. The results showed that the BiomedBERT model was the most effective in predicting item difficulty, while the fine-tuned model based on FastText word embeddings was the best model for predicting response time.

pdf bib
Utilizing Machine Learning to Predict Question Difficulty and Response Time for Enhanced Test Construction
Rishikesh Fulari | Jonathan Rusert

In this paper, we present the details of ourcontribution to the BEA Shared Task on Automated Prediction of Item Difficulty and Response Time. Participants in this collaborativeeffort are tasked with developing models to predict the difficulty and response time of multiplechoice items within the medical domain. Theseitems are sourced from the United States Medical Licensing Examination® (USMLE®), asignificant medical assessment. In order toachieve this, we experimented with two featurization techniques, one using lingusitic features and the other using embeddings generated by BERT fine-tuned over MS-MARCOdataset. Further, we tried several different machine learning models such as Linear Regression, Decision Trees, KNN and Boosting models such as XGBoost and GBDT. We found thatout of all the models we experimented withRandom Forest Regressor trained on Linguisticfeatures gave the least root mean squared error.

pdf bib
Leveraging Physical and Semantic Features of text item for Difficulty and Response Time Prediction of USMLE Questions
Gummuluri Venkata Ravi Ram | Ashinee Kesanam | Anand Kumar M

This paper presents our system developed for the Shared Task on Automated Prediction of Item Difficulty and Item Response Time for USMLE questions, organized by the Association for Computational Linguistics (ACL) Special Interest Group for building Educational Applications (BEA SIGEDU). The Shared Task, held as a workshop at the North American Chapter of the Association for Computational Linguistics (NAACL) 2024 conference, aimed to advance the state-of-the-art in predicting item characteristics directly from item text, with implications for the fairness and validity of standardized exams. We compared various methods ranging from BERT for regression to Random forest, Gradient Boosting(GB), Linear Regression, Support Vector Regressor (SVR), k-nearest neighbours (KNN) Regressor, MultiLayer Perceptron(MLP) to custom-ANN using BioBERT and Word2Vec embeddings and provided inferences on which performed better. This paper also explains the importance of data augmentation to balance the data in order to get better results. We also proposed five hypotheses regarding factors impacting difficulty and response time for a question and also verified it thereby helping researchers to derive meaningful numerical attributes for accurate prediction. We achieved a RSME score of 0.315 for Difficulty prediction and 26.945 for Response Time.

pdf bib
UPN-ICC at BEA 2024 Shared Task: Leveraging LLMs for Multiple-Choice Questions Difficulty Prediction
George Duenas | Sergio Jimenez | Geral Mateus Ferro

We describe the second-best run for the shared task on predicting the difficulty of Multi-Choice Questions (MCQs) in the medical domain. Our approach leverages prompting Large Language Models (LLMs). Rather than straightforwardly querying difficulty, we simulate medical candidate’s responses to questions across various scenarios. For this, more than 10,000 prompts were required for the 467 training questions and the 200 test questions. From the answers to these prompts, we extracted a set of features which we combined with a Ridge Regression to which we only adjusted the regularization parameter using the training set. Our motivation stems from the belief that MCQ difficulty is influenced more by the respondent population than by item-specific content features. We conclude that the approach is promising and has the potential to improve other item-based systems on this task, which turned out to be extremely challenging and has ample room for future improvement.

pdf bib
Using Machine Learning to Predict Item Difficulty and Response Time in Medical Tests
Mehrdad Yousefpoori-Naeim | Shayan Zargari | Zahra Hatami

Prior knowledge of item characteristics, suchas difficulty and response time, without pretestingitems can substantially save time and costin high-standard test development. Using a varietyof machine learning (ML) algorithms, thepresent study explored several (non-)linguisticfeatures (such as Coh-Metrix indices) alongwith MPNet word embeddings to predict thedifficulty and response time of a sample of medicaltest items. In both prediction tasks, thecontribution of embeddings to models alreadycontaining other features was found to be extremelylimited. Moreover, a comparison offeature importance scores across the two predictiontasks revealed that cohesion-based featureswere the strongest predictors of difficulty, whilethe prediction of response time was primarilydependent on length-related features.

pdf bib
Large Language Model-based Pipeline for Item Difficulty and Response Time Estimation for Educational Assessments
Hariram Veeramani | Surendrabikram Thapa | Natarajan Balaji Shankar | Abeer Alwan

This work presents a novel framework for the automated prediction of item difficulty and response time within educational assessments. Utilizing data from the BEA 2024 Shared Task, we integrate Named Entity Recognition, Semantic Role Labeling, and linguistic features to prompt a Large Language Model (LLM). Our best approach achieves an RMSE of 0.308 for item difficulty and 27.474 for response time prediction, improving on the provided baseline. The framework’s adaptability is demonstrated on audio recordings of 3rd-8th graders from the Atlanta, Georgia area responding to the Test of Narrative Language. These results highlight the framework’s potential to enhance test development efficiency.

pdf bib
UNED team at BEA 2024 Shared Task: Testing different Input Formats for predicting Item Difficulty and Response Time in Medical Exams
Alvaro Rodrigo | Sergio Moreno-Álvarez | Anselmo Peñas

This paper presents the description and primary outcomes of our team’s participation in the BEA 2024 shared task. Our primary exploration involved employing transformer-based systems, particularly BERT models, due to their suitability for Natural Language Processing tasks and efficiency with computational resources. We experimented with various input formats, including concatenating all text elements and incorporating only the clinical case. Surprisingly, our results revealed different impacts on predicting difficulty versus response time, with the former favoring clinical text only and the latter benefiting from including the correct answer. Despite moderate performance in difficulty prediction, our models excelled in response time prediction, ranking highest among all participants. This study lays the groundwork for future investigations into more complex approaches and configurations, aiming to advance the automatic prediction of exam difficulty and response time.

pdf bib
The BEA 2024 Shared Task on the Multilingual Lexical Simplification Pipeline
Matthew Shardlow | Fernando Alva-Manchego | Riza Batista-Navarro | Stefan Bott | Saul Calderon Ramirez | Rémi Cardon | Thomas François | Akio Hayakawa | Andrea Horbach | Anna Huelsing

We report the findings of the 2024 Multilingual Lexical Simplification Pipeline shared task. We released a new dataset comprising 5,927 instances of lexical complexity prediction and lexical simplification on common contexts across 10 languages, split into trial (300) and test (5,627). 10 teams participated across 2 tracks and 10 languages with 233 runs evaluated across all systems. Five teams participated in all languages for the lexical complexity prediction task and 4 teams participated in all languages for the lexical simplification task. Teams employed a range of strategies, making use of open and closed source large language models for lexical simplification, as well as feature-based approaches for lexical complexity prediction. The highest scoring team on the combined multilingual data was able to obtain a Pearson’s correlation of 0.6241 and an ACC@1@Top1 of 0.3772, both demonstrating that there is still room for improvement on two difficult sub-tasks of the lexical simplification pipeline.

pdf bib
TMU-HIT at MLSP 2024: How Well Can GPT-4 Tackle Multilingual Lexical Simplification?
Taisei Enomoto | Hwichan Kim | Tosho Hirasawa | Yoshinari Nagai | Ayako Sato | Kyotaro Nakajima | Mamoru Komachi

Lexical simplification (LS) is a process of replacing complex words with simpler alternatives to help readers understand sentences seamlessly. This process is divided into two primary subtasks: assessing word complexities and replacing high-complexity words with simpler alternatives. Employing task-specific supervised data to train models is a prevalent strategy for addressing these subtasks. However, such approach cannot be employed for low-resource languages. Therefore, this paper introduces a multilingual LS pipeline system that does not rely on supervised data. Specifically, we have developed systems based on GPT-4 for each subtask. Our systems demonstrated top-class performance on both tasks in many languages. The results indicate that GPT-4 can effectively assess lexical complexity and simplify complex words in a multilingual context with high quality.

pdf bib
ANU at MLSP-2024: Prompt-based Lexical Simplification for English and Sinhala
Sandaru Seneviratne | Hanna Suominen

Lexical simplification, the process of simplifying complex content in text without any modifications to the syntactical structure of text, plays a crucial role in enhancing comprehension and accessibility. This paper presents an approach to lexical simplification that relies on the capabilities of generative Artificial Intelligence (AI) models to predict the complexity of words and substitute complex words with simpler alternatives. Early lexical simplification methods predominantly relied on rule-based approaches, transitioning gradually to machine learning and deep learning techniques, leveraging contextual embeddings from large language models. However, the the emergence of generative AI models revolutionized the landscape of natural language processing, including lexical simplification. In this study, we proposed a straightforward yet effective method that employs generative AI models for both predicting lexical complexity and generating appropriate substitutions. To predict lexical complexity, we adopted three distinct types of prompt templates, while for lexical substitution, we employed three prompt templates alongside an ensemble approach. Extending our experimentation to include both English and Sinhala data, our approach demonstrated comparable performance across both languages, with particular strengths in lexical substitution.

pdf bib
ISEP_Presidency_University at MLSP 2024 Shared Task: Using GPT-3.5 to Generate Substitutes for Lexical Simplification
Benjamin Dutilleul | Mathis Debaillon | Sandeep Mathias

Lexical substitute generation is a task where we generate substitutes for a given word to fit in the required context. It is one of the main steps for automatic lexical simplifcation. In this paper, we introduce an automatic lexical simplification system using the GPT-3 large language model. The system generates simplified candidate substitutions for complex words to aid readability and comprehension for the reader. The paper describes the system that we submitted for the Multilingual Lexical Simplification Pipeline Shared Task at the 2024 BEA Workshop. During the shared task, we experimented with Catalan, English, French, Italian, Portuguese, and German for the Lexical Simplification Shared Task. We achieved the best results in Catalan and Portuguese, and were runners-up in English, French and Italian. To further research in this domain, we also release our code upon acceptance of the paper.

pdf bib
Machine Translation for Lexical Complexity Prediction and Lexical Simplification
Petru Cristea | Sergiu Nisioi

We present the submissions of team Archaeology for the Lexical Simplification and Lexical Complexity Prediction Shared Tasks at BEA2024. Our approach for this shared task consists in creating two pipelines for generating lexical substitutions and estimating the complexity: one using machine translation texts into English and one using the original language.For the LCP subtask, our xgb regressor is trained with engineered features (based primarily on English language resources) and shallow word structure features. For the LS subtask we use a locally-executed quantized LLM to generate candidates and sort them by complexity score computed using the pipeline designed for LCP.These pipelines provide distinct perspectives on the lexical simplification process, offering insights into the efficacy and limitations of employing Machine Translation versus direct processing on the original language data.

pdf bib
RETUYT-INCO at MLSP 2024: Experiments on Language Simplification using Embeddings, Classifiers and Large Language Models
Ignacio Sastre | Leandro Alfonso | Facundo Fleitas | Federico Gil | Andrés Lucas | Tomás Spoturno | Santiago Góngora | Aiala Rosá | Luis Chiruzzo

In this paper we present the participation of the RETUYT-INCO team at the BEA-MLSP 2024 shared task. We followed different approaches, from Multilayer Perceptron models with word embeddings to Large Language Models fine-tuned on different datasets: already existing, crowd-annotated, and synthetic.Our best models are based on fine-tuning Mistral-7B, either with a manually annotated dataset or with synthetic data.

pdf bib
GMU at MLSP 2024: Multilingual Lexical Simplification with Transformer Models
Dhiman Goswami | Kai North | Marcos Zampieri

This paper presents GMU’s submission to the Multilingual Lexical Simplification Pipeline (MLSP) shared task at the BEA workshop 2024. The task includes Lexical Complexity Prediction (LCP) and Lexical Simplification (LS) sub-tasks across 10 languages. Our submissions achieved rankings ranging from 1st to 5th in LCP and from 1st to 3rd in LS. Our best performing approach for LCP is a weighted ensemble based on Pearson correlation of language specific transformer models trained on all languages combined. For LS, GPT4-turbo zero-shot prompting achieved the best performance.

pdf bib
ITEC at MLSP 2024: Transferring Predictions of Lexical Difficulty from Non-Native Readers
Anaïs Tack

This paper presents the results of our team’s participation in the BEA 2024 shared task on the multilingual lexical simplification pipeline (MLSP; Shardlow et al., 2024). During the task, organizers supplied data that combined two components of the simplification pipeline: lexical complexity prediction and lexical substitution. This dataset encompassed ten languages, including French. Given the absence of dedicated training data, teams were challenged with employing systems trained on pre-existing resources and evaluating their performance on unexplored test data.Our team contributed to the task using previously developed models for predicting lexical difficulty in French (Tack, 2021). These models were built on deep learning architectures, adding to our participation in the CWI 2018 shared task (De Hertog and Tack, 2018). The training dataset comprised 262,054 binary decision annotations, capturing perceived lexical difficulty, collected from a sample of 56 non-native French readers. Two pre-trained neural logistic models were used: (1) a model for predicting difficulty for words within their sentence context, and (2) a model for predicting difficulty for isolated words.The findings revealed that despite being trained for a distinct prediction task (as indicated by a negative R2 fit), transferring the logistic predictions of lexical difficulty to continuous scores of lexical complexity exhibited a positive correlation. Specifically, the results indicated that isolated predictions exhibited a higher correlation (r = .36) compared to contextualized predictions (r = .33). Moreover, isolated predictions demonstrated a remarkably higher Spearman rank correlation (ρ = .50) than contextualized predictions (ρ = .35). These results align with earlier observations by Tack (2021), suggesting that the ground truth primarily captures more lexical access difficulties than word-to-context integration problems.
up

pdf (full)
bib (full) Proceedings of the 6th Clinical Natural Language Processing Workshop

pdf bib
Proceedings of the 6th Clinical Natural Language Processing Workshop
Tristan Naumann | Asma Ben Abacha | Steven Bethard | Kirk Roberts | Danielle Bitterman

pdf bib
Exploring Robustness in Doctor-Patient Conversation Summarization: An Analysis of Out-of-Domain SOAP Notes
Yu-Wen Chen | Julia Hirschberg

Summarizing medical conversations poses unique challenges due to the specialized domain and the difficulty of collecting in-domain training data. In this study, we investigate the performance of state-of-the-art doctor-patient conversation generative summarization models on the out-of-domain data. We divide the summarization model of doctor-patient conversation into two configurations: (1) a general model, without specifying subjective (S), objective (O), and assessment (A) and plan (P) notes; (2) a SOAP-oriented model that generates a summary with SOAP sections. We analyzed the limitations and strengths of the fine-tuning language model-based methods and GPTs on both configurations. We also conducted a Linguistic Inquiry and Word Count analysis to compare the SOAP notes from different datasets. The results exhibit a strong correlation for reference notes across different datasets, indicating that format mismatch (i.e., discrepancies in word distribution) is not the main cause of performance decline on out-of-domain data. Lastly, a detailed analysis of SOAP notes is included to provide insights into missing information and hallucinations introduced by the models.

pdf bib
Efficient Medical Question Answering with Knowledge-Augmented Question Generation
Julien Khlaut | Corentin Dancette | Elodie Ferreres | Benani Alaedine | Herent Herent | Pierre Manceron

In the expanding field of language model applications, medical knowledge representation remains a significant challenge due to the specialized nature of the domain. Large language models, such as GPT-4, obtain reasonable scores on medical question-answering tasks, but smaller models are far behind.In this work, we introduce a method to improve the proficiency of a small language model in the medical domain by employing a two-fold approach. We first fine-tune the model on a corpus of medical textbooks. Then, we use GPT-4 to generate questions similar to the downstream task, prompted with textbook knowledge, and use them to fine-tune the model. Additionally, we introduce ECN-QA, a novel Medical QA dataset containing “progressive questions” composed of related sequential questions. We show the benefits of our training strategy on this dataset.The study’s findings highlight the potential of small language models in the medical domain when appropriately fine-tuned.

pdf bib
Gemini Goes to Med School: Exploring the Capabilities of Multimodal Large Language Models on Medical Challenge Problems & Hallucinations
Ankit Pal | Malaikannan Sankarasubbu

Large language models have the potential to be valuable in the healthcare industry, but it’s crucial to verify their safety and effectiveness through rigorous evaluation. In our study, we evaluated LLMs, including Google’s Gemini, across various medical tasks. Despite Gemini’s capabilities, it underperformed compared to leading models like MedPaLM 2 and GPT-4, particularly in medical visual question answering (VQA), with a notable accuracy gap (Gemini at 61.45% vs. GPT-4V at 88%). Our analysis revealed that Gemini is highly susceptible to hallucinations, overconfidence, and knowledge gaps, which indicate risks if deployed uncritically. We also performed a detailed analysis by medical subject and test type, providing actionable feedback for developers and clinicians. To mitigate risks, we implemented effective prompting strategies, improving performance, and contributed to the field by releasing a Python module for medical LLM evaluation and establishing a leaderboard on Hugging Face for ongoing research and development. Python module can be found at https://github.com/promptslab/RosettaEval

pdf bib
Retrieval augmented text-to-SQL generation for epidemiological question answering using electronic health records
Angelo Ziletti | Leonardo DAmbrosi

Electronic health records (EHR) and claims data are rich sources of real-world data that reflect patient health status and healthcare utilization. Querying these databases to answer epidemiological questions is challenging due to the intricacy of medical terminology and the need for complex SQL queries. Here, we introduce an end-to-end methodology that combines text-to-SQL generation with retrieval augmented generation (RAG) to answer epidemiological questions using EHR and claims data. We show that our approach, which integrates a medical coding step into the text-to-SQL process, significantly improves the performance over simple prompting. Our findings indicate that although current language models are not yet sufficiently accurate for unsupervised use, RAG offers a promising direction for improving their capabilities, as shown in a realistic industry setting.

pdf bib
ClinicalMamba: A Generative Clinical Language Model on Longitudinal Clinical Notes
Zhichao Yang | Avijit Mitra | Sunjae Kwon | Hong Yu

The advancement of natural language processing (NLP) systems in healthcare hinges on language models’ ability to interpret the intricate information contained within clinical notes. This process often requires integrating information from various time points in a patient’s medical history. However, most earlier clinical language models were pretrained with a context length limited to roughly one clinical document. In this study, We introduce ClinicalMamba, a specialized version of the Mamba language model, pretrained on a vast corpus of longitudinal clinical notes to address the unique linguistic characteristics and information processing needs of the medical domain. ClinicalMamba models, with 130 million and 2.8 billion parameters, demonstrate superior performance in modeling clinical language across extended text lengths compared to Mamba and other clinical models based on longformer and Llama. With few-shot learning, ClinicalMamba achieves notable benchmarks in speed and performance, outperforming existing clinical language models and large language models like GPT-4 in longitudinal clinical tasks.

pdf bib
Working Alliance Transformer for Psychotherapy Dialogue Classification
Baihan Lin | Guillermo Cecchi | Djallel Bouneffouf

As a predictive measure of the treatment outcome in psychotherapy, the working alliance measures the agreement of the patient and the therapist in terms of their bond, task and goal. Long been a clinical quantity estimated by the patients’ and therapists’ self-evaluative reports, we believe that the working alliance can be better characterized using natural language processing technique directly in the dialogue transcribed in each therapy session. In this work, we propose the Working Alliance Transformer (WAT), a Transformer-based classification model that has a psychological state encoder which infers the working alliance scores by projecting the embedding of the dialogues turns onto the embedding space of the clinical inventory for working alliance. We evaluate our method in a real-world dataset with over 950 therapy sessions with anxiety, depression, schizophrenia and suicidal patients and demonstrate an empirical advantage of using information about therapeutic states in the sequence classification task of psychotherapy dialogues.

pdf bib
Building A German Clinical Named Entity Recognition System without In-domain Training Data
Siting Liang | Daniel Sonntag

Clinical Named Entity Recognition (NER) is essential for extracting important medical insights from clinical narratives. Given the challenges in obtaining expert training datasets for real-world clinical applications related to data protection regulations and the lack of standardised entity types, this work represents a collaborative initiative aimed at building a German clinical NER system with a focus on addressing these obstacles effectively. In response to the challenge of training data scarcity, we propose a Conditional Relevance Learning (CRL) approach in low-resource transfer learning scenarios. CRL effectively leverages a pre-trained language model and domain-specific open resources, enabling the acquisition of a robust base model tailored for clinical NER tasks, particularly in the face of changing label sets. This flexibility empowers the implementation of a Multilayered Semantic Annotation (MSA) schema in our NER system, capable of organizing a diverse array of entity types, thus significantly boosting the NER system’s adaptability and utility across various clinical domains. In the case study, we demonstrate how our NER system can be applied to overcome resource constraints and comply with data privacy regulations. Lacking prior training on in-domain data, feedback from expert users in respective domains is essential in identifying areas for system refinement. Future work will focus on the integration of expert feedback to improve system performance in specific clinical contexts.

pdf bib
DAIC-WOZ: On the Validity of Using the Therapist’s prompts in Automatic Depression Detection from Clinical Interviews
Sergio Burdisso | Ernesto Reyes-Ramírez | Esaú Villatoro-tello | Fernando Sánchez-Vega | Adrian Lopez Monroy | Petr Motlicek

Automatic depression detection from conversational data has gained significant interest in recent years.The DAIC-WOZ dataset, interviews conducted by a human-controlled virtual agent, has been widely used for this task.Recent studies have reported enhanced performance when incorporating interviewer’s prompts into the model.In this work, we hypothesize that this improvement might be mainly due to a bias present in these prompts, rather than the proposed architectures and methods.Through ablation experiments and qualitative analysis, we discover that models using interviewer’s prompts learn to focus on a specific region of the interviews, where questions about past experiences with mental health issues are asked, and use them as discriminative shortcuts to detect depressed participants. In contrast, models using participant responses gather evidence from across the entire interview.Finally, to highlight the magnitude of this bias, we achieve a 0.90 F1 score by intentionally exploiting it, the highest result reported to date on this dataset using only textual information.Our findings underline the need for caution when incorporating interviewers’ prompts into models, as they may inadvertently learn to exploit targeted prompts, rather than learning to characterize the language and behavior that are genuinely indicative of the patient’s mental health condition.

pdf bib
Parameter-Efficient Fine-Tuning of LLaMA for the Clinical Domain
Aryo Gema | Pasquale Minervini | Luke Daines | Tom Hope | Beatrice Alex

Adapting pretrained language models to novel domains, such as clinical applications, traditionally involves retraining their entire set of parameters. Parameter-Efficient Fine-Tuning (PEFT) techniques for fine-tuning language models significantly reduce computational requirements by selectively fine-tuning small subsets of parameters. In this study, we propose a two-step PEFT framework and evaluate it in the clinical domain. Our approach combines a specialised PEFT adapter layer designed for clinical domain adaptation with another adapter specialised for downstream tasks. We evaluate the framework on multiple clinical outcome prediction datasets, comparing it to clinically trained language models. Our framework achieves a better AUROC score averaged across all clinical downstream tasks compared to clinical language models. In particular, we observe large improvements of 4-5% AUROC in large-scale multilabel classification tasks, such as diagnoses and procedures classification. To our knowledge, this study is the first to provide an extensive empirical analysis of the interplay between PEFT techniques and domain adaptation in an important real-world domain of clinical applications.

pdf bib
A Multilevel Analysis of PubMed-only BERT-based Biomedical Models
Vicente Sanchez Carmona | Shanshan Jiang | Bin Dong

Biomedical NLP models play a big role in the automatic extraction of information from biomedical documents, such as COVID research papers. Three landmark models have led the way in this area: BioBERT, MSR BiomedBERT, and BioLinkBERT. However, their shallow evaluation –a single mean score– forbid us to better understand how the contributions proposed in each model advance the Biomedical NLP field. We show through a Multilevel Analysis how we can assess these contributions. Our analyses across 5000 fine-tuned models show that, actually, BiomedBERT’s true effect is bigger than BioLinkBERT’s effect, and the success of BioLinkBERT does not seem to be due to its contribution –the Link function– but due to an unknown factor.

pdf bib
A Privacy-Preserving Corpus for Occupational Health in Spanish: Evaluation for NER and Classification Tasks
Claudio Aracena | Luis Miranda | Thomas Vakili | Fabián Villena | Tamara Quiroga | Fredy Núñez-Torres | Victor Rocco | Jocelyn Dunstan

Annotated corpora are essential to reliable natural language processing. While they are expensive to create, they are essential for building and evaluating systems. This study introduces a new corpus of 2,869 medical and admission reports collected by an occupational insurance and health provider. The corpus has been carefully annotated for personally identifiable information (PII) and is shared, masking this information. Two annotators adhered to annotation guidelines during the annotation process, and a referee later resolved annotation conflicts in a consolidation process to build a gold standard subcorpus. The inter-annotator agreement values, measured in F1, range between 0.86 and 0.93 depending on the selected subcorpus. The value of the corpus is demonstrated by evaluating its use for NER of PII and a classification task. The evaluations find that fine-tuned models and GPT-3.5 reach F1 of 0.911 and 0.720 in NER of PII, respectively. In the case of the insurance coverage classification task, using the original or de-identified corpus results in similar performance. The annotated data are released in de-identified form.

pdf bib
DERA: Enhancing Large Language Model Completions with Dialog-Enabled Resolving Agents
Varun Nair | Elliot Schumacher | Geoffrey Tso | Anitha Kannan

Large language models (LLMs) have emerged as valuable tools for many natural language understanding tasks. In safety-critical applications such as healthcare, the utility of these models is governed by their ability to generate factually accurate and complete outputs. In this work, we present dialog-enabled resolving agents (DERA). DERA is a paradigm made possible by the increased conversational abilities of LLMs. It provides a simple, interpretable forum for models to communicate feedback and iteratively improve output. We frame our dialog as a discussion between two agent types – a Researcher, who processes information and identifies crucial problem components, and a Decider, who has the autonomy to integrate the Researcher’s information and makes judgments on the final output.We test DERA against three clinically-focused tasks, with GPT-4 serving as our LLM. DERA shows significant improvement over the base GPT-4 performance in both human expert preference evaluations and quantitative metrics for medical conversation summarization and care plan generation. In a new finding, we also show that GPT-4’s performance (70%) on an open-ended version of the MedQA question-answering (QA) dataset (Jin 2021; USMLE) is well above the passing level (60%), with DERA showing similar performance. We will release the open-ended MedQA dataset.

pdf bib
LlamaMTS: Optimizing Metastasis Detection with Llama Instruction Tuning and BERT-Based Ensemble in Italian Clinical Reports
Livia Lilli | Stefano Patarnello | Carlotta Masciocchi | Valeria Masiello | Fabio Marazzi | Tagliaferri Luca | Nikola Capocchiano

Information extraction from Electronic Health Records (EHRs) is a crucial task in healthcare, and the lack of resources and language specificity pose significant challenges. This study addresses the limited availability of Italian Natural Language Processing (NLP) tools for clinical applications and the computational demand of large language models (LLMs) for training. We present LlamaMTS, an instruction-tuned Llama for the Italian language, leveraging the LoRA technique. It is ensembled with a BERT-based model to classify EHRs based on the presence or absence of metastasis in patients affected by Breast cancer. Through our evaluation analysis, we discovered that LlamaMTS exhibits superior performance compared to both zero-shot LLMs and other Italian BERT-based models specifically fine-tuned on the same metastatic task. LlamaMTS demonstrates promising results in resource-constrained environments, offering a practical solution for information extraction from Italian EHRs in oncology, potentially improving patient care and outcomes.

pdf bib
Using Structured Health Information for Controlled Generation of Clinical Cases in French
Hugo Boulanger | Nicolas Hiebel | Olivier Ferret | Karën Fort | Aurélie Névéol

Text generation opens up new prospects for overcoming the lack of open corpora in fields such as healthcare, where data sharing is bound by confidentiality. In this study, we compare the performance of encoder-decoder and decoder-only language models for the controlled generation of clinical cases in French. To do so, we fine-tuned several pre-trained models on French clinical cases for each architecture and generate clinical cases conditioned by patient demographic information (gender and age) and clinical features.Our results suggest that encoder-decoder models are easier to control than decoder-only models, but more costly to train.

pdf bib
Large Language Models Provide Human-Level Medical Text Snippet Labeling
Ibtihel Amara | Haiyang Yu | Fan Zhang | Yuchen Liu | Benny Li | Chang Liu | Rupesh Kartha | Akshay Goel

This study evaluates the proficiency of Large Language Models (LLMs) in accurately labeling clinical document excerpts. Our focus is on the assignment of potential or confirmed diagnoses and medical procedures to snippets of medical text sourced from unstructured clinical patient records. We explore how the performance of LLMs compare against human annotators in classifying these excerpts. Employing a few-shot, chain-of-thought prompting approach with the MIMIC-III dataset, Med-PaLM 2 showcases annotation accuracy comparable to human annotators, achieving a notable precision rate of approximately 92% relative to the gold standard labels established by human experts.

pdf bib
Conversational Topic Recommendation in Counseling and Psychotherapy with Decision Transformer and Large Language Models
Aylin Gunal | Baihan Lin | Djallel Bouneffouf

Given the increasing demand for mental health assistance, artificial intelligence (AI), particularly large language models (LLMs), may be valuable for integration into automated clinical support systems. In this work, we leverage a decision transformer architecture for topic recommendation in counseling conversations between patients and mental health professionals. The architecture is utilized for offline reinforcement learning, and we extract states (dialogue turn embeddings), actions (conversation topics), and rewards (scores measuring the alignment between patient and therapist) from previous turns within a conversation to train a decision transformer model. We demonstrate an improvement over baseline reinforcement learning methods, and propose a novel system of utilizing our model’s output as synthetic labels for fine-tuning a large language model for the same task. Although our implementation based on LLaMA-2 7B has mixed results, future work can undoubtedly build on the design.

pdf bib
Leveraging Wikidata for Biomedical Entity Linking in a Low-Resource Setting: A Case Study for German
Faizan Mustafa | Corina Dima | Juan Ochoa | Steffen Staab

Biomedical Entity Linking (BEL) is a challenging task for low-resource languages, due to the lack of appropriate resources: datasets, knowledge bases (KBs), and pre-trained models. In this paper, we propose an approach to create a biomedical knowledge base for German BEL using UMLS information from Wikidata, that provides good coverage and can be easily extended to further languages. As a further contribution, we adapt several existing approaches for use in the German BEL setup, and report on their results. The chosen methods include a sparse model using character n-grams, a multilingual biomedical entity linker, and two general-purpose text retrieval models. Our results show that a language-specific KB that provides good coverage leads to most improvement in entity linking performance, irrespective of the used model. The finetuned German BEL model, newly created UMLSWikidata KB as well as the code to reproduce our results are publicly available.

pdf bib
Revisiting Clinical Outcome Prediction for MIMIC-IV
Tom Röhr | Alexei Figueroa | Jens-Michalis Papaioannou | Conor Fallon | Keno Bressem | Wolfgang Nejdl | Alexander Löser

Clinical Decision Support Systems assist medical professionals in providing optimal care for patients.A prominent data source used for creating tasks for such systems is the Medical Information Mart for Intensive Care (MIMIC).MIMIC contains electronic health records (EHR) gathered in a tertiary hospital in the United States.The majority of past work is based on the third version of MIMIC, although the fourth is the most recent version.This new version, not only introduces more data into MIMIC, but also increases the variety of patients.While MIMIC-III is limited to intensive care units, MIMIC-IV also offers EHRs from the emergency department.In this work, we investigate how to adapt previous work to update clinical outcome prediction for MIMIC-IV.We revisit several established tasks, including prediction of diagnoses, procedures, length-of-stay, and also introduce a novel task: patient routing prediction.Furthermore, we quantitatively and qualitatively evaluate all tasks on several bio-medical transformer encoder models.Finally, we provide narratives for future research directions in the clinical outcome prediction domain. We make our source code publicly available to reproduce our experiments, data, and tasks.

pdf bib
Can LLMs Correct Physicians, Yet? Investigating Effective Interaction Methods in the Medical Domain
Burcu Sayin | Pasquale Minervini | Jacopo Staiano | Andrea Passerini

We explore the potential of Large Language Models (LLMs) to assist and potentially correct physicians in medical decision-making tasks. We evaluate several LLMs, including Meditron, Llama2, and Mistral, to analyze the ability of these models to interact effectively with physicians across different scenarios. We consider questions from PubMedQA and several tasks, ranging from binary (yes/no) responses to long answer generation, where the answer of the model is produced after an interaction with a physician. Our findings suggest that prompt design significantly influences the downstream accuracy of LLMs and that LLMs can provide valuable feedback to physicians, challenging incorrect diagnoses and contributing to more accurate decision-making. For example, when the physician is accurate 38% of the time, Mistral can produce the correct answer, improving accuracy up to 74% depending on the prompt being used, while Llama2 and Meditron models exhibit greater sensitivity to prompt choice. Our analysis also uncovers the challenges of ensuring that LLM-generated suggestions are pertinent and useful, emphasizing the need for further research in this area.

pdf bib
Leveraging pre-trained large language models for aphasia detection in English and Chinese speakers
Yan Cong | Jiyeon Lee | Arianna LaCroix

We explore the utility of pre-trained Large Language Models (LLMs) in detecting the presence, subtypes, and severity of aphasia across English and Mandarin Chinese speakers. Our investigation suggests that even without fine-tuning or domain-specific training, pre-trained LLMs can offer some insights on language disorders, regardless of speakers’ first language. Our analysis also reveals noticeable differences between English and Chinese LLMs. While the English LLMs exhibit near-chance level accuracy in subtyping aphasia, the Chinese counterparts demonstrate less than satisfactory performance in distinguishing between individuals with and without aphasia. This research advocates for the importance of linguistically tailored and specified approaches in leveraging LLMs for clinical applications, especially in the context of multilingual populations.

pdf bib
Fusion of Domain-Adapted Vision and Language Models for Medical Visual Question Answering
Cuong Ha | Shima Asaadi | Sanjeev Kumar Karn | Oladimeji Farri | Tobias Heimann | Thomas Runkler

Vision-language models, while effective in general domains and showing strong performance in diverse multi-modal applications like visual question-answering (VQA), struggle to maintain the same level of effectiveness in more specialized domains, e.g., medical. We propose a medical vision-language model that integrates large vision and language models adapted for the medical domain. This model goes through three stages of parameter-efficient training using three separate biomedical and radiology multi-modal visual and text datasets. The proposed model achieves state-of-the-art performance on the SLAKE 1.0 medical VQA (MedVQA) dataset with an overall accuracy of 87.5% and demonstrates strong performance on another MedVQA dataset, VQA-RAD, achieving an overall accuracy of 73.2%.

pdf bib
LLM-Based Section Identifiers Excel on Open Source but Stumble in Real World Applications
Saranya Krishnamoorthy | Ayush Singh | Shabnam Tafreshi

Electronic health records (EHR) even though a boon for healthcare practitioners, are grow- ing convoluted and longer every day. Sifting around these lengthy EHRs is taxing and be- comes a cumbersome part of physician-patient interaction. Several approaches have been pro- posed to help alleviate this prevalent issue ei- ther via summarization or sectioning, however, only a few approaches have truly been helpful in the past. With the rise of automated methods, machine learning (ML) has shown promise in solving the task of identifying relevant sections in EHR. However, most ML methods rely on labeled data which is difficult to get in health- care. Large language models (LLMs) on the other hand, have performed impressive feats in natural language processing (NLP), that too in a zero-shot manner, i.e. without any labeled data. To that end, we propose using LLMs to identify relevant section headers. We find that GPT-4 can effectively solve the task on both zero and few-shot settings as well as segment dramatically better than state-of-the-art meth- ods. Additionally, we also annotate a much harder real world dataset and find that GPT-4 struggles to perform well, alluding to further research and harder benchmarks.

pdf bib
Adapting Abstract Meaning Representation Parsing to the Clinical Narrative – the SPRING THYME parser
Jon Cai | Kristin Wright-Bettner | Martha Palmer | Guergana Savova | James Martin

This paper is dedicated to the design and evaluation of the first AMR parser tailored for clinical notes. Our objective was to facilitate the precise transformation of the clinical notes into structured AMR expressions, thereby enhancing the interpretability and usability of clinical text data at scale. Leveraging the colon cancer dataset from the Temporal Histories of Your Medical Events (THYME) corpus, we adapted a state-of-the-art AMR parser utilizing continuous training. Our approach incorporates data augmentation techniques to enhance the accuracy of AMR structure predictions. Notably, through this learning strategy, our parser achieved an impressive F1 score of 88% on the THYME corpus’s colon cancer dataset. Moreover, our research delved into the efficacy of data required for domain adaptation within the realm of clinical notes, presenting domain adaptation data requirements for AMR parsing. This exploration not only underscores the parser’s robust performance but also highlights its potential in facilitating a deeper understanding of clinical narratives through structured semantic representations.

pdf bib
SERPENT-VLM : Self-Refining Radiology Report Generation Using Vision Language Models
Manav Kapadnis | Sohan Patnaik | Abhilash Nandy | Sourjyadip Ray | Pawan Goyal | Debdoot Sheet

Radiology Report Generation (R2Gen) demonstrates how Multi-modal Large Language Models (MLLMs) can automate the creation of accurate and coherent radiological reports. Existing methods often hallucinate details in text-based reports that don’t accurately reflect the image content. To mitigate this, we introduce a novel strategy, SERPENT-VLM (SElf Refining Radiology RePort GENeraTion using Vision Language Models), which improves the R2Gen task by integrating a self-refining mechanism into the MLLM framework. We employ a unique self-supervised loss that leverages similarity between pooled image representations and the contextual representations of the generated radiological text, alongside the standard Causal Language Modeling objective, to refine image-text representations. This allows the model to scrutinize and align the generated text through dynamic interaction between a given image and the generated text, therefore reducing hallucination and continuously enhancing nuanced report generation. SERPENT-VLM outperforms existing baselines such as LlaVA-Med, BiomedGPT, etc., achieving SoTA performance on the IU X-ray and Radiology Objects in COntext (ROCO) datasets, and also proves to be robust against noisy images. A qualitative case study emphasizes the significant advancements towards more sophisticated MLLM frameworks for R2Gen, opening paths for further research into self-supervised refinement in the medical imaging domain.

pdf bib
ERD: A Framework for Improving LLM Reasoning for Cognitive Distortion Classification
Sehee Lim | Yejin Kim | Chi-Hyun Choi | Jy-yong Sohn | Byung-Hoon Kim

Improving the accessibility of psychotherapy with the aid of Large Language Models (LLMs) is garnering a significant attention in recent years. Recognizing cognitive distortions from the interviewee’s utterances can be an essential part of psychotherapy, especially for cognitive behavioral therapy. In this paper, we propose ERD, which improves LLM-based cognitive distortion classification performance with the aid of additional modules of (1) extracting the parts related to cognitive distortion, and (2) debating the reasoning steps by multiple agents. Our experimental results on a public dataset show that ERD improves the multi-class F1 score as well as binary specificity score. Regarding the latter score, it turns out that our method is effective in debiasing the baseline method which has high false positive rate, especially when the summary of multi-agent debate is provided to LLMs.

pdf bib
Leveraging Prompt-Learning for Structured Information Extraction from Crohn’s Disease Radiology Reports in a Low-Resource Language
Liam Hazan | Naama Gavrielov | Roi Reichart | Talar Hagopian | Mary-Louise Greer | Ruth Cytter-Kuint | Gili Focht | Dan Turner | Moti Freiman

Automatic conversion of free-text radiology reports into structured data using Natural Language Processing (NLP) techniques is crucial for analyzing diseases on a large scale. While effective for tasks in widely spoken languages like English, generative large language models (LLMs) typically underperform with less common languages and can pose potential risks to patient privacy. Fine-tuning local NLP models is hindered by the skewed nature of real-world medical datasets, where rare findings represent a significant data imbalance. We introduce SMP-BERT, a novel prompt learning method that leverages the structured nature of reports to overcome these challenges. In our studies involving a substantial collection of Crohn’s disease radiology reports in Hebrew (over 8,000 patients and 10,000 reports), SMP-BERT greatly surpassed traditional fine-tuning methods in performance, notably in detecting infrequent conditions (AUC: 0.99 vs 0.94, F1: 0.84 vs 0.34). SMP-BERT empowers more accurate AI diagnostics available for low-resource languages.

pdf bib
Context Aggregation with Topic-focused Summarization for Personalized Medical Dialogue Generation
Zhengyuan Liu | Siti Salleh | Pavitra Krishnaswamy | Nancy Chen

In the realm of dialogue systems, generated responses often lack personalization. This is particularly true in the medical domain, where research is limited by scarce available domain-specific data and the complexities of modeling medical context and persona information. In this work, we investigate the potential of harnessing large language models for personalized medical dialogue generation. In particular, to better aggregate the long conversational context, we adopt topic-focused summarization to distill core information from the dialogue history, and use such information to guide the conversation flow and generated content. Drawing inspiration from real-world telehealth conversations, we outline a comprehensive pipeline encompassing data processing, profile construction, and domain adaptation. This work not only highlights our technical approach but also shares distilled insights from the data preparation and model construction phases.

pdf bib
Evaluating Lexicon Incorporation for Depression Symptom Estimation
Kirill Milintsevich | Gaël Dias | Kairit Sirts

This paper explores the impact of incorporating sentiment, emotion, and domain-specific lexicons into a transformer-based model for depression symptom estimation. Lexicon information is added by marking the words in the input transcripts of patient-therapist conversations as well as in social media posts. Overall results show that the introduction of external knowledge within pre-trained language models can be beneficial for prediction performance, while different lexicons show distinct behaviours depending on the targeted task. Additionally, new state-of-the-art results are obtained for the estimation of depression level over patient-therapist interviews.

pdf bib
Semi-automatic Construction of a Word Complexity Lexicon for Japanese Medical Terminology
Soichiro Sugihara | Tomoyuki Kajiwara | Takashi Ninomiya | Shoko Wakamiya | Eiji Aramaki

We construct a word complexity lexicon for medical terms in Japanese.To facilitate communication between medical practitioners and patients, medical text simplification is being studied.Medical text simplification is a natural language processing task that paraphrases complex technical terms into expressions that patients can understand.However, in contrast to English, where this task is being actively studied, there are insufficient language resources in Japanese.As a first step in advancing research on medical text simplification in Japanese, we annotate the 370,000 words from a large-scale medical terminology lexicon with a five-point scale of complexity for patients.

pdf bib
TEAM MIPAL at MEDIQA-M3G 2024: Large VQA Models for Dermatological Diagnosis
Hyeonjin Kim | Min Kim | Jae Jang | KiYoon Yoo | Nojun Kwak

This paper describes the methods used for the NAACL 2024 workshop MEDIQA-M3G shared task for generating medical answers from image and query data for skin diseases. MedVInT-Decoder, LLaVA, and LLaVA-Med are chosen as base models. Finetuned with the task dataset on the dermatological domain, MedVInT-Decoder achieved a BLEU score of 3.82 during competition, while LLaVA and LLaVA-Med reached 6.98 and 4.62 afterward, respectively.

pdf bib
MediFact at MEDIQA-M3G 2024: Medical Question Answering in Dermatology with Multimodal Learning
Nadia Saeed

The MEDIQA-M3G 2024 challenge necessitates novel solutions for Multilingual & Multimodal Medical Answer Generation in dermatology (wai Yim et al., 2024a). This paper addresses the limitations of traditional methods by proposing a weakly supervised learning approach for open-ended medical question-answering (QA). Our system leverages readily available MEDIQA-M3G images via a VGG16-CNN-SVM model, enabling multilingual (English, Chinese, Spanish) learning of informative skin condition representations. Using pre-trained QA models, we further bridge the gap between visual and textual information through multimodal fusion. This approach tackles complex, open-ended questions even without predefined answer choices. We empower the generation of comprehensive answers by feeding the ViT-CLIP model with multiple responses alongside images. This work advances medical QA research, paving the way for clinical decision support systems and ultimately improving healthcare delivery.

pdf bib
MediFact at MEDIQA-CORR 2024: Why AI Needs a Human Touch
Nadia Saeed

Accurate representation of medical information is crucial for patient safety, yet artificial intelligence (AI) systems, such as Large Language Models (LLMs), encounter challenges in error-free clinical text interpretation. This paper presents a novel approach submitted to the MEDIQA-CORR 2024 shared task k (Ben Abacha et al., 2024a), focusing on the automatic correction of single-word errors in clinical notes. Unlike LLMs that rely on extensive generic data, our method emphasizes extracting contextually relevant information from available clinical text data. Leveraging an ensemble of extractive and abstractive question-answering approaches, we construct a supervised learning framework with domain-specific feature engineering. Our methodology incorporates domain expertise to enhance error correction accuracy. By integrating domain expertise and prioritizing meaningful information extraction, our approach underscores the significance of a human-centric strategy in adapting AI for healthcare.

pdf bib
KnowLab_AIMed at MEDIQA-CORR 2024: Chain-of-Though (CoT) prompting strategies for medical error detection and correction
Zhaolong Wu | Abul Hasan | Jinge Wu | Yunsoo Kim | Jason Cheung | Teng Zhang | Honghan Wu

This paper describes our submission to the MEDIQA-CORR 2024 shared task for automatically detecting and correcting medical errors in clinical notes. We report results for three methods of few-shot In-Context Learning (ICL) augmented with Chain-of-Thought (CoT) and reason prompts using a large language model (LLM). In the first method, we manually analyse a subset of train and validation dataset to infer three CoT prompts by examining error types in the clinical notes. In the second method, we utilise the training dataset to prompt the LLM to deduce reasons about their correctness or incorrectness. The constructed CoTs and reasons are then augmented with ICL examples to solve the tasks of error detection, span identification, and error correction. Finally, we combine the two methods using a rule-based ensemble method. Across the three sub-tasks, our ensemble method achieves a ranking of 3rd for both sub-task 1 and 2, while securing 7th place in sub-task 3 among all submissions.

pdf bib
PromptMind Team at EHRSQL-2024: Improving Reliability of SQL Generation using Ensemble LLMs
Satya Gundabathula | Sriram Kolar

This paper presents our approach to the EHRSQL-2024 shared task, which aims to develop a reliable Text-to-SQL system for electronic health records. We propose two approaches that leverage large language models (LLMs) for prompting and fine-tuning to generate EHRSQL queries. In both techniques, we concentrate on bridging the gap between the real-world knowledge on which LLMs are trained and the domain-specific knowledge required for the task. The paper provides the results of each approach individually, demonstrating that they achieve high execution accuracy. Additionally, we show that an ensemble approach further enhances generation reliability by reducing errors. This approach secured us 2nd place in the shared task competition. The methodologies outlined in this paper are designed to be transferable to domain-specific Text-to-SQL problems that emphasize both accuracy and reliability.

pdf bib
PromptMind Team at MEDIQA-CORR 2024: Improving Clinical Text Correction with Error Categorization and LLM Ensembles
Satya Gundabathula | Sriram Kolar

This paper describes our approach to the MEDIQA-CORR shared task, which involves error detection and correction in clinical notes curated by medical professionals. This task involves handling three subtasks: detecting the presence of errors, identifying the specific sentence containing the error, and correcting it. Through our work, we aim to assess the capabilities of Large Language Models (LLMs) trained on a vast corpora of internet data that contain both factual and unreliable information. We propose to comprehensively address all subtasks together, and suggest employing a unique prompt-based in-context learning strategy. We will evaluate its efficacy in this specialized task demanding a combination of general reasoning and medical knowledge. In medical systems where prediction errors can have grave consequences, we propose leveraging self-consistency and ensemble methods to enhance error correction and error detection performance.

pdf bib
Maven at MEDIQA-CORR 2024: Leveraging RAG and Medical LLM for Error Detection and Correction in Medical Notes
Suramya Jadhav | Abhay Shanbhag | Sumedh Joshi | Atharva Date | Sheetal Sonawane

Addressing the critical challenge of identifying and rectifying medical errors in clinical notes, we present a novel approach tailored for the MEDIQA-CORR task @ NAACL-ClinicalNLP 2024, which comprises three subtasks: binary classification, span identification, and natural language generation for error detection and correction. Binary classification involves detecting whether the text contains a medical error; span identification entails identifying the text span associated with any detected error; and natural language generation focuses on providing a free text correction if a medical error exists. Our proposed architecture leverages Named Entity Recognition (NER) for identifying disease-related terms, Retrieval-Augmented Generation (RAG) for contextual understanding from external datasets, and a quantized and fine-tuned Palmyra model for error correction. Our model achieved a global rank of 5 with an aggregate score of 0.73298, calculated as the mean of ROUGE-1-F, BERTScore, and BLEURT scores.

pdf bib
LAILab at Chemotimelines 2024: Finetuning sequence-to-sequence language models for temporal relation extraction towards cancer patient undergoing chemotherapy treatment
Shohreh Haddadan | Tuan-Dung Le | Thanh Duong | Thanh Thieu

In this paper, we report our effort to tackle the challenge of extracting chemotimelines from EHR notes across a dataset of three cancer types. We focus on the two subtasks: 1) detection and classification of temporal relations given the annotated chemotherapy events and time expressions and 2) directly extracting patient chemotherapy timelines from EHR notes. We address both subtasks using Large Language Models. Our best-performing methods in both subtasks use Flan-T5, an instruction-tuned language model. Our proposed system achieves the highest average score in both subtasks. Our results underscore the effectiveness of finetuning general-domain large language models in domain-specific and unseen tasks.

pdf bib
Lexicans at Chemotimelines 2024: Chemotimeline Chronicles - Leveraging Large Language Models (LLMs) for Temporal Relations Extraction in Oncological Electronic Health Records
Vishakha Sharma | Andres Fernandez | Andrei Ioanovici | David Talby | Frederik Buijs

Automatic generation of chemotherapy treatment timelines from electronic health records (EHRs) notes not only streamlines clinical workflows but also promotes better coordination and improvements in cancer treatment and quality of care. This paper describes the submission to the Chemotimelines 2024 shared task that aims to automatically build a chemotherapy treatment timeline for each patient using their complete set of EHR notes, spanning various sources such as primary care provider, oncology, discharge summaries, emergency department, pathology, radiology, and more. We report results from two large language models (LLMs), namely Llama 2 and Mistral 7B, applied to the shared task data using zero-shot prompting.

pdf bib
Team NLPeers at Chemotimelines 2024: Evaluation of two timeline extraction methods, can generative LLM do it all or is smaller model fine-tuning still relevant ?
Nesrine Bannour | Judith Jeyafreeda Andrew | Marc Vincent

This paper presents our two deep learning-based approaches to participate in subtask 1 of the Chemotimelines 2024 Shared task. The first uses a fine-tuning strategy on a relatively small general domain Masked Language Model (MLM) model, with additional normalization steps obtained using a simple Large Language Model (LLM) prompting technique. The second is an LLM-based approach combining advanced automated prompt search with few-shot in-context learning using the DSPy framework.Our results confirm the continued relevance of the smaller MLM fine-tuned model. It also suggests that the automated few-shot LLM approach can perform close to the fine-tuning-based method without extra LLM normalization and be advantageous under scarce data access conditions. We finally hint at the possibility to choose between lower training examples or lower computing resources requirements when considering both methods.

pdf bib
KCLab at Chemotimelines 2024: End-to-end system for chemotherapy timeline extraction – Subtask2
Yukun Tan | Merve Dede | Ken Chen

This paper presents our participation in the Chemotimelines 2024 subtask2, focusing on the development of an end-to-end system for chemotherapy timeline extraction. We initially adopt a basic framework from subtask2, utilizing Apache cTAKES for entity recognition and a BERT-based model for classifying the temporal relationship between chemotherapy events and associated times. Subsequently, we enhance this pipeline through two key directions: first, by expanding the exploration of the system, achieved by extending the search dictionary of cTAKES with the UMLS database; second, by reducing false positives through preprocessing of clinical notes and implementing filters to reduce the potential errors from the BERT-based model. To validate the effectiveness of our framework, we conduct extensive experiments using clinical notes from breast, ovarian, and melanoma cancer cases. Our results demonstrate improvements over the previous approach.

pdf bib
Project PRIMUS at EHRSQL 2024 : Text-to-SQL Generation using Large Language Model for EHR Analysis
Sourav Joy | Rohan Ahmed | Argha Saha | Minhaj Habil | Utsho Das | Partha Bhowmik

This paper explores the application of the sqlcoders model, a pre-trained neural network, for automatic SQL query generation from natural language questions. We focus on the model’s internal functionality and demonstrate its effectiveness on a domain-specific validation dataset provided by EHRSQL. The sqlcoders model, based on transformers with attention mechanisms, has been trained on paired examples of natural language questions and corresponding SQL queries. It takes advantage of a carefully crafted prompt that incorporates the database schema alongside the question to guide the model towards the desired output format.

pdf bib
NYULangone at Chemotimelines 2024: Utilizing Open-Weights Large Language Models for Chemotherapy Event Extraction
Jeff Zhang | Yin Aphinyanaphongs | Anthony Cardillo

The extraction of chemotherapy treatment timelines from clinical narratives poses significant challenges due to the complexity of medical language and patient-specific treatment regimens. This paper describes the NYULangone team’s approach to Subtask 2 of the Chemotimelines 2024 shared task, focusing on leveraging a locally hosted Large Language Model (LLM), Mixtral 8x7B (Mistral AI, France), to interpret and extract relevant events from clinical notes without relying on domain-specific training data. Despite facing challenges due to the task’s complexity and the current capacity of open-source AI, our methodology highlights the future potential of local foundational LLMs in specialized domains like biomedical data processing.

pdf bib
AIRI NLP Team at EHRSQL 2024 Shared Task: T5 and Logistic Regression to the Rescue
Oleg Somov | Alexey Dontsov | Elena Tutubalina

This paper presents a system developed for the Clinical NLP 2024 Shared Task, focusing on reliable text-to-SQL modeling on Electronic Health Records (EHRs). The goal is to create a model that accurately generates SQL queries for answerable questions while avoiding incorrect responses and handling unanswerable queries. Our approach comprises three main components: a query correspondence model, a text-to-SQL model, and an SQL verifier.For the query correspondence model, we trained a logistic regression model using hand-crafted features to distinguish between answerable and unanswerable queries. As for the text-to-SQL model, we utilized T5-3B as a pretrained language model, further fine-tuned on pairs of natural language questions and corresponding SQL queries. Finally, we applied the SQL verifier to inspect the resulting SQL queries.During the evaluation stage of the shared task, our system achieved an accuracy of 68.9 % (metric version without penalty), positioning it at the fifth-place ranking. While our approach did not surpass solutions based on large language models (LMMs) like ChatGPT, it demonstrates the promising potential of domain-specific specialized models that are more resource-efficient. The code is publicly available at https://github.com/runnerup96/EHRSQL-text2sql-solution.

pdf bib
IKIM at MEDIQA-M3G 2024: Multilingual Visual Question-Answering for Dermatology through VLM Fine-tuning and LLM Translations
Marie Bauer | Amin Dada | Constantin Seibold | Jens Kleesiek

This paper presents our solution to the MEDIQA-M3G Challenge at NAACL-ClinicalNLP 2024. We participated in all three languages, ranking first in Chinese and Spanish and third in English. Our approach utilizes LLaVA-med, an open-source, medical vision-language model (VLM) for visual question-answering in Chinese, and Mixtral-8x7B-instruct, a Large Language Model (LLM) for a subsequent translation into English and Spanish. In addition to our final method, we experiment with alternative approaches: Training three different models for each language instead of translating the results from one model, using different combinations and numbers of input images, and additional training on publicly available data that was not part of the original challenge training set.

pdf bib
NEUI at MEDIQA-M3G 2024: Medical VQA through consensus
Ricardo García | Oscar Lithgow-Serrano

This document describes our solution to the MEDIQA-M3G: Multilingual & Multimodal Medical Answer Generation. To build our solution, we leveraged two pre-trained models, a Visual Language Model (VLM) and a Large Language Model (LLM). We fine-tuned both models using the MEDIQA-M3G and MEDIQA-CORR training datasets, respectively. In the first stage, the VLM provides singular responses for each pair of image & text inputs in a case. In the second stage, the LLM consolidates the VLM responses using it as context among the original text input. By changing the original English case content field in the context component of the second stage to the one in Spanish, we adapt the pipeline to generate submissions in English and Spanish. We performed an ablation study to explore the impact of the different models’ capabilities, such as multimodality and reasoning, on the MEDIQA-M3G task. Our approach favored privacy and feasibility by adopting open-source and self-hosted small models and ranked 4th in English and 2nd in Spanish.

pdf bib
VerbaNexAI at MEDIQA-CORR: Efficacy of GRU with BioWordVec and ClinicalBERT in Error Correction in Clinical Notes
Juan Pajaro | Edwin Puertas | David Villate | Laura Estrada | Laura Tinjaca

The automatic identification of medical errors in clinical notes is crucial for improving the quality of healthcare services.LLMs emerge as a powerful artificial intelligence tool for automating this task. However, LLMs present vulnerabilities, high costs, and sometimes a lack of transparency. This article addresses the detection of medical errors through the fine-tuning approach, conducting a comprehensive comparison between various models and exploring in depth the components of the machine learning pipeline. The results obtained with the fine-tuned ClinicalBert and Gated recurrent units (Gru) models show an accuracy of 0.56 and 0.55, respectively. This approach not only mitigates the problems associated with the use of LLMs but also demonstrates how exhaustive iteration in critical phases of the pipeline, especially in feature selection, can facilitate the automation of clinical record analysis.

pdf bib
HSE NLP Team at MEDIQA-CORR 2024 Task: In-Prompt Ensemble with Entities and Knowledge Graph for Medical Error Correction
Airat Valiev | Elena Tutubalina

This paper presents our LLM-based system designed for the MEDIQA-CORR @ NAACL-ClinicalNLP 2024 Shared Task 3, focusing on medical error detection and correction in medical records. Our approach consists of three key components: entity extraction, prompt engineering, and ensemble. First, we automatically extract biomedical entities such as therapies, diagnoses, and biological species. Next, we explore few-shot learning techniques and incorporate graph information from the MeSH database for the identified entities. Finally, we investigate two methods for ensembling: (i) combining the predictions of three previous LLMs using an AND strategy within a prompt and (ii) integrating the previous predictions into the prompt as separate ‘expert’ solutions, accompanied by trust scores representing their performance. The latter system ranked second with a BERTScore score of 0.8059 and third with an aggregated score of 0.7806 out of the 15 teams’ solutions in the shared task.

pdf bib
Wonder at Chemotimelines 2024: MedTimeline: An End-to-End NLP System for Timeline Extraction from Clinical Narratives
Liwei Wang | Qiuhao Lu | Rui Li | Sunyang Fu | Hongfang Liu

Extracting timeline information from clinical narratives is critical for cancer research and practice using electronic health records (EHRs). In this study, we apply MedTimeline, our end-to-end hybrid NLP system combining large language model, deep learning with knowledge engineering, to the ChemoTimeLine challenge subtasks. Our experiment results in 0.83, 0.90, 0.84, and 0.53, 0.63, 0.39, respectively, for subtask1 and subtask2 in breast, melanoma and ovarian cancer.

pdf bib
Edinburgh Clinical NLP at MEDIQA-CORR 2024: Guiding Large Language Models with Hints
Aryo Gema | Chaeeun Lee | Pasquale Minervini | Luke Daines | T. Simpson | Beatrice Alex

The MEDIQA-CORR 2024 shared task aims to assess the ability of Large Language Models (LLMs) to identify and correct medical errors in clinical notes. In this study, we evaluate the capability of general LLMs, specifically GPT-3.5 and GPT-4, to identify and correct medical errors with multiple prompting strategies. Recognising the limitation of LLMs in generating accurate corrections only via prompting strategies, we propose incorporating error-span predictions from a smaller, fine-tuned model in two ways: 1) by presenting it as a hint in the prompt and 2) by framing it as multiple-choice questions from which the LLM can choose the best correction. We found that our proposed prompting strategies significantly improve the LLM’s ability to generate corrections. Our best-performing solution with 8-shot + CoT + hints ranked sixth in the shared task leaderboard. Additionally, our comprehensive analyses show the impact of the location of the error sentence, the prompted role, and the position of the multiple-choice option on the accuracy of the LLM. This prompts further questions about the readiness of LLM to be implemented in real-world clinical settings.

pdf bib
UMass-BioNLP at MEDIQA-M3G 2024: DermPrompt - A Systematic Exploration of Prompt Engineering with GPT-4V for Dermatological Diagnosis
Parth Vashisht | Abhilasha Lodha | Mukta Maddipatla | Zonghai Yao | Avijit Mitra | Zhichao Yang | Sunjae Kwon | Junda Wang | Hong Yu

This paper presents our team’s participation in the MEDIQA-ClinicalNLP 2024 shared task B. We present a novel approach to diagnosing clinical dermatology cases by integrating large multimodal models, specifically leveraging the capabilities of GPT-4V under a retriever and a re-ranker framework. Our investigation reveals that GPT-4V, when used as a retrieval agent, can accurately retrieve the correct skin condition 85% of the time using dermatological images and brief patient histories. Additionally, we empirically show that Naive Chain-of-Thought (CoT) works well for retrieval while Medical Guidelines Grounded CoT is required for accurate dermatological diagnosis. Further, we introduce a Multi-Agent Conversation (MAC) framework and show it’s superior performance and potential over the best CoT strategy. The experiments suggest that using naive CoT for retrieval and multi-agent conversation for critique-based diagnosis, GPT-4V can lead to an early and accurate diagnosis of dermatological conditions. The implications of this work extend to improving diagnostic workflows, supporting dermatological education, and enhancing patient care by providing a scalable, accessible, and accurate diagnostic tool.

pdf bib
KU-DMIS at MEDIQA-CORR 2024: Exploring the Reasoning Capabilities of Small Language Models in Medical Error Correction
Hyeon Hwang | Taewhoo Lee | Hyunjae Kim | Jaewoo Kang

Recent advancements in large language models (LM) like OpenAI’s GPT-4 have shown promise in healthcare, particularly in medical question answering and clinical applications. However, their deployment raises privacy concerns and their size limits use in resource-constrained environments.Smaller open-source LMs have emerged as alternatives, but their reliability in medicine remains underexplored.This study evaluates small LMs in the medical field using the MEDIQA-CORR 2024 task, which assesses the ability of models to identify and correct errors in clinical notes. Initially, zero-shot inference and simple fine-tuning of small models resulted in poor performance. When fine-tuning with chain-of-thought (CoT) reasoning using synthetic data generated by GPT-4, their performance significantly improved. Meerkat-7B, a small LM trained with medical CoT reasoning, demonstrated notable performance gains. Our model outperforms other small non-commercial LMs and some larger models, achieving a 73.36 aggregate score on MEDIQA-CORR 2024.

pdf bib
CLD-MEC at MEDIQA- CORR 2024 Task: GPT-4 Multi-Stage Clinical Chain of Thought Prompting for Medical Errors Detection and Correction
Renad Alzghoul | Ayaabdelhaq Ayaabdelhaq | Abdulrahman Tabaza | Ahmad Altamimi

This paper demonstrates CLD-MEC team submission to the MEDIQA-CORR 2024 shared task for identifying and correcting medical errors from clinical notes. We developed a framework to track two main types of medical errors: diagnostics and medical management-related errors. The tracking framework is implied utilizing a GPT-4 multi-stage prompting-based pipeline that ends with the three downstream tasks: classification of medical error existence (Task 1), identification of error location (Task 2), and correction error (Task 3). Throughout the pipeline, we employed clinical Chain of Thought (CoT) and Chain-of-Verification (CoVe) techniques to mitigate the hallucination and enforce the clinical context learning. The model performance is acceptable, given it is based on zero-shot learning. In addition, we developed a RAG system injected with clinical practice guidelines as an external knowledge datastore. Our RAG is based on the Bio_ClinicalBERT as a vector embedding model. However, our RAG system failed to get the desired results. We proposed recommendations to be investigated in future research work to overcome the limitations of our approach.

pdf bib
Overview of the 2024 Shared Task on Chemotherapy Treatment Timeline Extraction
Jiarui Yao | Harry Hochheiser | WonJin Yoon | Eli Goldner | Guergana Savova

The 2024 Shared Task on Chemotherapy Treatment Timeline Extraction aims to advance the state of the art of clinical event timeline extraction from the Electronic Health Records (EHRs). Specifically, this edition focuses on chemotherapy event timelines from EHRs of patients with breast, ovarian and skin cancers. These patient-level timelines present a novel challenge which involves tasks such as the extraction of relevant events, time expressions and temporal relations from each document and then summarizing over the documents. De-identified EHRs for 57,530 patients with breast and ovarian cancer spanning 2004-2020, and approximately 15,946 patients with melanoma spanning 2010-2020 were made available to participants after executing a Data Use Agreement. A subset of patients is annotated for gold entities, time expressions, temporal relations and patient-level timelines. The rest is considered unlabeled data. In Subtask1, gold chemotherapy event mentions and time expressions are provided (along with the EHR notes). Participants are asked to build the patient-level timelines using gold annotations as input. Thus, the subtask seeks to explore the topics of temporal relations extraction and timeline creation if event and time expression input is perfect. In Subtask2, which is the realistic real-world setting, only EHR notes are provided. Thus, the subtask aims at developing an end-to-end system for chemotherapy treatment timeline extraction from patient’s EHR notes. There were 18 submissions for Subtask 1 and 9 submissions for Subtask 2. The organizers provided a baseline system. The teams employed a variety of methods including Logistic Regression, TF-IDF, n-grams, transformer models, zero-shot prompting with Large Language Models (LLMs), and instruction tuning. The gap in performance between prompting LLMs and finetuning smaller-sized LMs indicates that for a challenging task such as patient-level chemotherapy timeline extraction, more sophisticated LLMs or prompting techniques are necessary in order to achieve optimal results as finetuing smaller-sized LMs outperforms by a wide margin.

pdf bib
IryoNLP at MEDIQA-CORR 2024: Tackling the Medical Error Detection & Correction Task on the Shoulders of Medical Agents
Jean-Philippe Corbeil

In natural language processing applied to the clinical domain, utilizing large language models has emerged as a promising avenue for error detection and correction on clinical notes, a knowledge-intensive task for which annotated data is scarce. This paper presents MedReAct’N’MedReFlex, which leverages a suite of four LLM-based medical agents. The MedReAct agent initiates the process by observing, analyzing, and taking action, generating trajectories to guide the search to target a potential error in the clinical notes. Subsequently, the MedEval agent employs five evaluators to assess the targeted error and the proposed correction. In cases where MedReAct’s actions prove insufficient, the MedReFlex agent intervenes, engaging in reflective analysis and proposing alternative strategies. Finally, the MedFinalParser agent formats the final output, preserving the original style while ensuring the integrity of the error correction process. One core component of our method is our RAG pipeline based on our ClinicalCorp corpora. Among other well-known sources containing clinical guidelines and information, we preprocess and release the open-source MedWiki dataset for clinical RAG application. Our results demonstrate the central role of our RAG approach with ClinicalCorp leveraged through the MedReAct’N’MedReFlex framework. It achieved the ninth rank on the MEDIQA-CORR 2024 final leaderboard.

pdf bib
Overview of the MEDIQA-M3G 2024 Shared Task on Multilingual Multimodal Medical Answer Generation
Wen-wai Yim | Asma Ben Abacha | Yujuan Fu | Zhaoyi Sun | Fei Xia | Meliha Yetisgen | Martin Krallinger

Remote patient care provides opportunities for expanding medical access, saving healthcare costs, and offering on-demand convenient services. In the MEDIQA-M3G 2024 Shared Task, researchers explored solutions for the specific task of dermatological consumer health visual question answering, where user generated queries and images are used as input and a free-text answer response is generated as output. In this novel challenge, eight teams with a total of 48 submissions were evaluated across three language test sets. In this work, we provide a summary of the dataset, as well as results and approaches. We hope that the insights learned here will inspire future research directions that can lead to technology that deburdens clinical workload and improves care.

pdf bib
EM_Mixers at MEDIQA-CORR 2024: Knowledge-Enhanced Few-Shot In-Context Learning for Medical Error Detection and Correction
Swati Rajwal | Eugene Agichtein | Abeed Sarker

This paper describes our submission to MEDIQA-CORR 2024 shared task for automatic identification and correction of medical errors in a given clinical text. We report results from two approaches: the first uses a few-shot in-context learning (ICL) with a Large Language Model (LLM) and the second approach extends the idea by using a knowledge-enhanced few-shot ICL approach. We used Azure OpenAI GPT-4 API as the LLM and Wikipedia as the external knowledge source. We report evaluation metrics (accuracy, ROUGE, BERTScore, BLEURT) across both approaches for validation and test datasets. Of the two approaches implemented, our experimental results show that the knowledge-enhanced few-shot ICL approach with GPT-4 performed better with error flag (subtask A) and error sentence detection (subtask B) with accuracies of 68% and 64%, respectively on the test dataset. These results positioned us fourth in subtask A and second in subtask B, respectively in the shared task.

pdf bib
Overview of the MEDIQA-CORR 2024 Shared Task on Medical Error Detection and Correction
Asma Ben Abacha | Wen-wai Yim | Yujuan Fu | Zhaoyi Sun | Fei Xia | Meliha Yetisgen

Automatic detection and correction of medical errors enables a more rigorous validation of medical documentation as well as clinical notes generated by large language models. Such solutions can ensure the accuracy and medical coherence of clinical texts and enhance patient care and health outcomes. The MEDIQA-CORR 2024 shared task focused on detecting and correcting different types of medical errors in clinical texts. Seventeen teams participated in the shared task and experimented with a broad range of approaches and models. In this paper, we describe the MEDIQA-CORR task, datasets, and the participants’ results and methods.

pdf bib
UTSA-NLP at ChemoTimelines 2024: Evaluating Instruction-Tuned Language Models for Temporal Relation Extraction
Xingmeng Zhao | Anthony Rios

This paper presents our approach for the 2024 ChemoTimelines shared task. Specifically, we explored using Large Language Models (LLMs) for temporal relation extraction. We evaluate multiple model variations based on how the training data is used. For instance, we transform the task into a question-answering problem and use QA pairs to extract chemo-related events and their temporal relations. Next, we add all the documents to each question-answer pair as examples in our training dataset. Finally, we explore adding unlabeled data for continued pretraining. Each addition is done iteratively. Our results show that adding the document helps, but unlabeled data does not yield performance improvements, possibly because we used only 1% of the available data. Moreover, we find that instruction-tuned models still substantially underperform more traditional systems (e.g., EntityBERT).

pdf bib
WangLab at MEDIQA-CORR 2024: Optimized LLM-based Programs for Medical Error Detection and Correction
Augustin Toma | Ronald Xie | Steven Palayew | Patrick Lawler | Bo Wang

Medical errors in clinical text pose significant risks to patient safety. The MEDIQA-CORR 2024 shared task focuses on detecting and correcting these errors across three subtasks: identifying the presence of an error, extracting the erroneous sentence, and generating a corrected sentence. In this paper, we present our approach that achieved top performance in all three subtasks. For the MS dataset, which contains subtle errors, we developed a retrieval-based system leveraging external medical question-answering datasets. For the UW dataset, reflecting more realistic clinical notes, we created a pipeline of modules to detect, localize, and correct errors. Both approaches utilized the DSPy framework for optimizing prompts and few-shot examples in large language model (LLM) based programs. Our results demonstrate the effectiveness of LLM based programs for medical error correction. However, our approach has limitations in addressing the full diversity of potential errors in medical documentation. We discuss the implications of our work and highlight future research directions to advance the robustness and applicability of medical error detection and correction systems.

pdf bib
WangLab at MEDIQA-M3G 2024: Multimodal Medical Answer Generation using Large Language Models
Augustin Toma | Ronald Xie | Steven Palayew | Gary Bader | Bo Wang

This paper outlines our submission to the MEDIQA2024 Multilingual and Multimodal Medical Answer Generation (M3G) shared task. We report results for two standalone solutions under the English category of the task, the first involving two consecutive API calls to the Claude 3 Opus API and the second involving training an image-disease label joint embedding in the style of CLIP for image classification. These two solutions scored 1st and 2nd place respectively on the competition leaderboard, substantially outperforming the next best solution. Additionally, we discuss insights gained from post-competition experiments. While the performance of these two described solutions have significant room for improvement due to the difficulty of the shared task and the challenging nature of medical visual question answering in general, we identify the multi-stage LLM approach and the CLIP image classification approach as promising avenues for further investigation.

pdf bib
LG AI Research & KAIST at EHRSQL 2024: Self-Training Large Language Models with Pseudo-Labeled Unanswerable Questions for a Reliable Text-to-SQL System on EHRs
Yongrae Jo | Seongyun Lee | Minju Seo | Sung Ju Hwang | Moontae Lee

Text-to-SQL models are pivotal for making Electronic Health Records (EHRs) accessible to healthcare professionals without SQL knowledge. With the advancements in large language models, these systems have become more adept at translating complex questions into SQL queries. Nonetheless, the critical need for reliability in healthcare necessitates these models to accurately identify unanswerable questions or uncertain predictions, preventing misinformation. To address this problem, we present a self-training strategy using pseudo-labeled unanswerable questions to enhance the reliability of text-to-SQL models for EHRs. This approach includes a two-stage training process followed by a filtering method based on the token entropy and query execution. Our methodology’s effectiveness is validated by our top performance in the EHRSQL 2024 shared task, showcasing the potential to improve healthcare decision-making through more reliable text-to-SQL systems.

pdf bib
Overview of the EHRSQL 2024 Shared Task on Reliable Text-to-SQL Modeling on Electronic Health Records
Gyubok Lee | Sunjun Kweon | Seongsu Bae | Edward Choi

Electronic Health Records (EHRs) are relational databases that store the entire medical histories of patients within hospitals. They record numerous aspects of patients’ medical care, from hospital admission and diagnosis to treatment and discharge. While EHRs are vital sources of clinical data, exploring them beyond a predefined set of queries requires skills in query languages like SQL. To make information retrieval more accessible, one strategy is to build a question-answering system, possibly leveraging text-to-SQL models that can automatically translate natural language questions into corresponding SQL queries and use these queries to retrieve the answers. The EHRSQL 2024 shared task aims to advance and promote research in developing a question-answering system for EHRs using text-to-SQL modeling, capable of reliably providing requested answers to various healthcare professionals to improve their clinical work processes and satisfy their needs. Among more than 100 participants who applied to the shared task, eight teams completed the entire shared task processes and demonstrated a wide range of methods to effectively solve this task. In this paper, we describe the task of reliable text-to-SQL modeling, the dataset, and the methods and results of the participants. We hope this shared task will spur further research and insights into developing reliable question-answering systems for EHRs.

pdf bib
Saama Technologies at EHRSQL 2024: SQL Generation through Classification Answer Selector by LLM
Mohammed Jabir | Kamal Kanakarajan | Malaikannan Sankarasubbu

The EHRSQL task aims to develop a dependable text-to-SQL model for Electronic Health Records (EHR) databases, which are crucial sources of clinical data that store patients’ medical histories in hospitals. Large language models (LLM) have been proven to exhibit state-of-the-art performance for text-to-SQL tasks across various domains. To this end, we have developed a framework, SQL Generation through Classification Answer Selector by LLM (SCAS), which comprises two modules. The CAS module determines the answerability of the question, while the SG model generates the SQL query exclusively for answerable questions. Our system ranked 7th on the leaderboard with a Reliability Score of 53.21 on the official test set.

pdf bib
KU-DMIS at EHRSQL 2024 : Generating SQL query via question templatization in EHR
Hajung Kim | Chanhwi Kim | Hoonick Lee | Kyochul Jang | Jiwoo Lee | Kyungjae Lee | Gangwoo Kim | Jaewoo Kang

Transforming natural language questions into SQL queries is crucial for precise data retrieval from electronic health record (EHR) databases. A significant challenge in this process is detecting and rejecting unanswerable questions that request information outside the database’s scope or exceed the system’s capabilities. In this paper, we introduce a novel text-to-SQL framework that focuses on standardizing the structure of questions into a templated format. Our framework begins by fine-tuning GPT-3.5-turbo, a powerful large language model (LLM), with detailed prompts involving the table schemas of the EHR database system. Our approach shows promising results on the EHRSQL-2024 benchmark dataset, part of the ClinicalNLP shared task. Although fine-tuning GPT achieves third place on the development set, it struggled with the diverse questions in the test set. With our framework, we improve our system’s adaptability and achieve fourth position in the official leaderboard of the EHRSQL-2024 challenge.

pdf bib
ProbGate at EHRSQL 2024: Enhancing SQL Query Generation Accuracy through Probabilistic Threshold Filtering and Error Handling
Sangryul Kim | Donghee Han | Sehyun Kim

Recently, deep learning-based language models have significantly enhanced text-to-SQL tasks, with promising applications in retrieving patient records within the medical domain. One notable challenge in such applications is discerning unanswerable queries. Through fine-tuning model, we demonstrate the feasibility of converting medical record inquiries into SQL queries. Additionally, we introduce an entropy-based method to identify and filter out unanswerable results. We further enhance result quality by filtering low-confidence SQL through log probability-based distribution, while grammatical and schema errors are mitigated by executing queries on the actual database.We experimentally verified that our method can filter unanswerable questions, which can be widely utilized even when the parameters of the model are not accessible, and that it can be effectively utilized in practice.

pdf bib
LTRC-IIITH at EHRSQL 2024: Enhancing Reliability of Text-to-SQL Systems through Abstention and Confidence Thresholding
Jerrin Thomas | Pruthwik Mishra | Dipti Sharma | Parameswari Krishnamurthy

In this paper, we present our work in the EHRSQL 2024 shared task which tackles reliable text-to-SQL modeling on Electronic Health Records. Our proposed system tackles the task with three modules - abstention module, text-to-SQL generation module, and reliability module. The abstention module identifies whether the question is answerable given the database schema. If the question is answerable, the text-to-SQL generation module generates the SQL query and associated confidence score. The reliability module has two key components - confidence score thresholding, which rejects generations with confidence below a pre-defined level, and error filtering, which identifies and excludes SQL queries that result in execution errors. In the official leaderboard for the task, our system ranks 6th. We have also made the source code public.

pdf bib
LTRC-IIITH at MEDIQA-M3G 2024: Medical Visual Question Answering with Vision-Language Models
Jerrin Thomas | Sushvin Marimuthu | Parameswari Krishnamurthy

In this paper, we present our work to the MEDIQA-M3G 2024 shared task, which tackles multilingual and multimodal medical answer generation. Our system consists of a lightweight Vision-and-Language Transformer (ViLT) model which is fine-tuned for the clinical dermatology visual question-answering task. In the official leaderboard for the task, our system ranks 6th. After the challenge, we experiment with training the ViLT model on more data. We also explore the capabilities of large Vision-Language Models (VLMs) such as Gemini and LLaVA.
up

pdf (full)
bib (full) Proceedings of the Fifth Workshop on Data Science with Human-in-the-Loop (DaSH 2024)

pdf bib
Proceedings of the Fifth Workshop on Data Science with Human-in-the-Loop (DaSH 2024)
Eduard Dragut | Yunyao Li | Lucian Popa | Slobodan Vucetic | Shashank Srivastava

pdf bib
APE: Active Learning-based Tooling for Finding Informative Few-shot Examples for LLM-based Entity Matching
Kun Qian | Yisi Sang | Farima Bayat† | Anton Belyi | Xianqi Chu | Yash Govind | Samira Khorshidi | Rahul Khot | Katherine Luna | Azadeh Nikfarjam | Xiaoguang Qi | Fei Wu | Xianhan Zhang | Yunyao Li

Prompt engineering is an iterative procedure that often requires extensive manual effort to formulate suitable instructions for effectively directing large language models (LLMs) in specific tasks. Incorporating few-shot examples is a vital and effective approach to provide LLMs with precise instructions, leading to improved LLM performance. Nonetheless, identifying the most informative demonstrations for LLMs is labor-intensive, frequently entailing sifting through an extensive search space. In this demonstration, we showcase a human-in-the-loop tool called ool (Active Prompt Engineering) designed for refining prompts through active learning. Drawing inspiration from active learning, ool iteratively selects the most ambiguous examples for human feedback, which will be transformed into few-shot examples within the prompt.

pdf bib
Towards Optimizing and Evaluating a Retrieval Augmented QA Chatbot using LLMs with Human-in-the-Loop
Anum Afzal | Alexander Kowsik | Rajna Fani | Florian Matthes

Large Language Models have found application in various mundane and repetitive tasks including Human Resource (HR) support. We worked with the domain experts of a large multinational company to develop an HR support chatbot as an efficient and effective tool for addressing employee inquiries. We inserted a human-in-the-loop in various parts of the development cycles such as dataset collection, prompt optimization, and evaluation of generated output. By enhancing the LLM-driven chatbot’s response quality and exploring alternative retrieval methods, we have created an efficient, scalable, and flexible tool for HR professionals to address employee inquiries effectively. Our experiments and evaluation conclude that GPT-4 outperforms other models and can overcome inconsistencies in data through internal reasoning capabilities. Additionally, through expert analysis, we infer that reference-free evaluation metrics such as G-Eval and Prometheus demonstrate reliability closely aligned with that of human evaluation.

pdf bib
Evaluation and Continual Improvement for an Enterprise AI Assistant
Akash Maharaj | Kun Qian | Uttaran Bhattacharya | Sally Fang | Horia Galatanu | Manas Garg | Rachel Hanessian | Nishant Kapoor | Ken Russell | Shivakumar Vaithyanathan | Yunyao Li

The development of conversational AI assistants is an iterative process with many components involved. As such, the evaluation and continual improvement of these assistants is a complex and multifaceted problem. This paper introduces the challenges in evaluating and improving a generative AI assistant for enterprise that is under active development and how we address these challenges. We also share preliminary results and discuss lessons learned.

pdf bib
Mini-DA: Improving Your Model Performance through Minimal Data Augmentation using LLM
Shuangtao Yang | Xiaoyi Liu | Xiaozheng Dong | Bo Fu

When performing data augmentation using large language models (LLMs), the common approach is to directly generate a large number of new samples based on the original dataset, and then model is trained on the integration of augmented dataset and the original dataset. However, data generation demands extensive computational resources. In this study, we propose Mini-DA, a minimized data augmentation method that leverages the feedback from the target model during the training process to select only the most challenging samples from the validation set for augmentation. Our experimental results show in text classification task, by using as little as 13 percent of the original augmentation volume, Mini-DA can achieve performance comparable to full data augmentation for intent detection task, significantly improving data and computational resource utilization efficiency.

pdf bib
CURATRON: Complete and Robust Preference Data for Rigorous Alignment of Large Language Models
Son The Nguyen | Niranjan Uma Naresh | Theja Tulabandhula

This paper addresses the challenges of aligning large language models (LLMs) with human values via preference learning (PL), focusing on incomplete and corrupted data in preference datasets. We propose a novel method for robustly and completely recalibrating values within these datasets to enhance LLMs’ resilience against the issues. In particular, we devise a guaranteed polynomial time ranking algorithm that robustifies several existing models, such as the classic Bradley–Terry–Luce (BTL) model and certain generalizations of it. To the best of our knowledge, our present work is the first to propose an algorithm that provably recovers an 𝜖-optimal ranking with high probability while allowing as large as O(n) perturbed pairwise comparison results per model response. Furthermore, we show robust recovery results in the partially observed setting. Our experiments confirm that our algorithms handle adversarial noise and unobserved comparisons well in LLM preference dataset settings. This work contributes to the development and scaling of more reliable and ethically aligned AI models by equipping the dataset curation pipeline with the ability to handle missing and maliciously manipulated inputs.
up

pdf (full)
bib (full) Proceedings of the 4th Workshop on Figurative Language Processing (FigLang 2024)

pdf bib
Proceedings of the 4th Workshop on Figurative Language Processing (FigLang 2024)
Debanjan Ghosh | Smaranda Muresan | Anna Feldman | Tuhin Chakrabarty | Emmy Liu

pdf bib
Context vs. Human Disagreement in Sarcasm Detection
Hyewon Jang | Moritz Jakob | Diego Frassinelli

Prior work has highlighted the importance of context in the identification of sarcasm by humans and language models. This work examines how much context is required for a better identification of sarcasm by both parties. We collect textual responses to dialogical prompts and sarcasm judgment to the responses placed after long contexts, short contexts, and no contexts. We find that both for humans and language models, the presence of context is generally important in identifying sarcasm in the response. But increasing the amount of context provides no added benefit to humans (long = short > none). This is the same for language models, but only on easily agreed-upon sentences; for sentences with disagreement among human evaluators, different models show different behavior. We also show how sarcasm detection patterns stay consistent as the amount of context is manipulated despite the low agreement in human evaluation.

pdf bib
Optimizing Multilingual Euphemism Detection using Low-Rank Adaption Within and Across Languages
Nicholas Hankins

This short paper presents an investigation into the effectiveness of various classification methods as a submission in the Multilingual Euphemism Detection Shared Task for the Fourth Workshop on Figurative Language Processing co-located with NAACL 2024. The process used by the participant utilizes pre-trained large language models combined with parameter efficient fine-tuning methods, specifically Low-Rank Adaptation (LoRA), in classifying euphemisms across four different languages - Mandarin Chinese, American English, Spanish, and Yorùbá. The study is comprised of three main components that aim to explore heuristic methods to navigate how base models can most efficiently be fine-tuned into classifiers to learn figurative language. Multilingual labeled training data was utilized to fine-tune classifiers for each language, and later combined for one large classifier, while unseen test data was finally used to evaluate the accuracy of the best performing classifiers. In addition, cross-lingual tests were conducted by applying each language’s data on each of the other language’s classifiers. All of the results provide insights into the potential of pre-trained base models combined with LoRA fine-tuning methods in accurately classifying euphemisms across and within different languages.

pdf bib
Comparison of Image Generation Models for Abstract and Concrete Event Descriptions
Mohammed Khaliq | Diego Frassinelli | Sabine Schulte Im Walde

With the advent of diffusion-based image generation models such as DALL-E, Stable Diffusion and Midjourney, high quality images can be easily generated using textual inputs. It is unclear, however, to what extent the generated images resemble human mental representations, especially regarding abstract event knowledge. We analyse the capability of four state-of-the-art models in generating images of verb-object event pairs when we systematically manipulate the degrees of abstractness of both the verbs and the object nouns. Human judgements assess the generated images and demonstrate that DALL-E is strongest for event pairs with concrete nouns (e.g., “pour water”; “believe person”), while Midjourney is preferred for event pairs with abstract nouns (e.g., “raise awareness”; “remain mystery”), irrespective of the concreteness of the verb. Across models, humans were most unsatisfied with images of events pairs that combined concrete verbs with abstract direct-object nouns (e.g., “speak truth”), and an additional ad-hoc annotation contributes this to its potential for figurative language.

pdf bib
Cross-Lingual Metaphor Detection for Low-Resource Languages
Anna Hülsing | Sabine Schulte Im Walde

Research on metaphor detection (MD) in a multilingual setup has recently gained momentum. As for many tasks, it is however unclear how the amount of data used to pretrain large language models affects the performance, and whether non-neural models might provide a reasonable alternative, especially for MD in low-resource languages. This paper compares neural and non-neural cross-lingual models for English as the source language and Russian, German and Latin as target languages. In a series of experiments we show that the neural cross-lingual adapter architecture MAD-X performs best across target languages. Zero-shot classification with mBERT achieves decent results above the majority baseline, while few-shot classification with mBERT heavily depends on shot-selection, which is inconvenient in a cross-lingual setup where no validation data for the target language exists. The non-neural model, a random forest classifier with conceptual features, is outperformed by the neural models. Overall, we recommend MAD-X for metaphor detection not only in high-resource but also in low-resource scenarios regarding the amounts of pretraining data for mBERT.

pdf bib
A Hard Nut to Crack: Idiom Detection with Conversational Large Language Models
Francesca De Luca Fornaciari | Begoña Altuna | Itziar Gonzalez-Dios | Maite Melero

In this work, we explore idiomatic language processing with Large Language Models (LLMs). We introduce the Idiomatic language Test Suite IdioTS, a dataset of difficult examples specifically designed by language experts to assess the capabilities of LLMs to process figurative language at sentence level. We propose a comprehensive evaluation methodology based on an idiom detection task, where LLMs are prompted with detecting an idiomatic expression in a given English sentence. We present a thorough automatic and manual evaluation of the results and a comprehensive error analysis.

pdf bib
The Elephant in the Room: Ten Challenges of Computational Detection of Rhetorical Figures
Ramona Kühn | Jelena Mitrović

Computational detection of rhetorical figures focuses mostly on figures such as metaphor, irony, or analogy. However, there exist many more figures that are neither less important nor less prevalent. We wanted to pinpoint the reasons why researchers often avoid other figures and to shed light on the challenges they struggle with when investigating those figures. In this comprehensive survey, we analyzed over 40 papers dealing with the computational detection of rhetorical figures other than metaphor, simile, sarcasm, and irony. We encountered recurrent challenges from which we compiled a ten point list. Furthermore, we suggest solutions for each challenge to encourage researchers to investigate a greater variety of rhetorical figures.

pdf bib
Guidelines for the Annotation of Intentional Linguistic Metaphor
Stefanie Dipper | Adam Roussel | Alexandra Wiemann | Won Kim | Tra-my Nguyen

This paper presents guidelines for the annotation of intentional (i.e. non-conventionalized) linguistic metaphors. Expressions that contribute to the same metaphorical image are annotated as a chain, additionally a semantically contrasting expression of the target domain is marked as an anchor. So far, a corpus of ten TEDx talks with a total of 20k tokens has been annotated according to these guidelines. 1.25% of the tokens are intentional metaphorical expressions.

pdf bib
Evaluating the Development of Linguistic Metaphor Annotation in Mexican Spanish Popular Science Tweets
Alec Montero | Gemma Bel-Enguix | Sergio-Luis Ojeda-Trueba | Marisela Colín Rodea

Following previous work on metaphor annotation and automatic metaphor processing, this study presents the evaluation of an initial phase in the novel area of linguistic metaphor detection in Mexican Spanish popular science tweets. Specifically, we examine the challenges posed by the annotation process stemming from disagreement among annotators. During this phase of our work, we conducted the annotation of a corpus comprising 3733 Mexican Spanish popular science tweets. This corpus was divided into two halves and each half was then assigned to two different pairs of native Mexican Spanish-speaking annotators. Despite rigorous methodology and continuous training, inter-annotator agreement as measured by Cohen’s kappa was found to be low, slightly above chance levels, although the concordance percentage exceeded 60%. By elucidating the inherent complexity of metaphor annotation tasks, our evaluation emphasizes the implications of these findings and offers insights for future research in this field, with the aim of creating a robust dataset for machine learning.

pdf bib
Can GPT4 Detect Euphemisms across Multiple Languages?
Todd Firsich | Anthony Rios

A euphemism is a word or phrase used in place of another word or phrase that might be considered harsh, blunt, unpleasant, or offensive. Euphemisms generally soften the impact of what is being said, making it more palatable or appropriate for the context or audience. Euphemisms can vary significantly between languages, reflecting cultural sensitivities and taboos, and what might be a mild expression in one language could carry a stronger connotation or be completely misunderstood in another. This paper uses prompting techniques to evaluate OpenAI’s GPT4 for detecting euphemisms across multiple languages as part of the 2024 FigLang shared task. We evaluate both zero-shot and few-shot approaches. Our method achieved an average macro F1 of .732, ranking first in the competition. Moreover, we found that GPT4 does not perform uniformly across all languages, with a difference of .233 between the best (English .831) and the worst (Spanish .598) languages.

pdf bib
Ensemble-based Multilingual Euphemism Detection: a Behavior-Guided Approach
Fedor Vitiugin | Henna Paakki

This paper describes the system submitted by our team to the Multilingual Euphemism Detection Shared Task for the Fourth Workshop on Figurative Language Processing (FigLang 2024). We propose a novel model for multilingual euphemism detection, combining contextual and behavior-related features. The system classifies texts that potentially contain euphemistic terms with an ensemble classifier based on outputs from behavior-related fine-tuned models. Our results show that, for this kind of task, our model outperforms baselines and state-of-the-art euphemism detection methods. As for the leader-board, our classification model achieved a macro averaged F1 score of [anonymized], reaching the [anonymized] place.

pdf bib
An Expectation-Realization Model for Metaphor Detection: Within Distribution, Out of Distribution, and Out of Pretraining
Oseremen Uduehi | Razvan Bunescu

We propose a new model for metaphor detection in which an expectation component estimates representations of expected word meanings in a given context, whereas a realization component computes representations of target word meanings in context. We also introduce a systematic evaluation methodology that estimates generalization performance in three settings: within distribution, a new strong out of distribution setting, and a novel out-of-pretraining setting. Across all settings, the expectation-realization model obtains results that are competitive with or better than previous metaphor detection models.

pdf bib
A Textual Modal Supplement Framework for Understanding Multi-Modal Figurative Language
Jiale Chen | Qihao Yang | Xuelian Dong | Xiaoling Mao | Tianyong Hao

Figurative language in media such as memes, art, or comics has gained dramatic interest recently. However, the challenge remains in accurately justifying and explaining whether an image caption complements or contradicts the image it accompanies. To tackle this problem, we design a modal-supplement framework MAPPER consisting of a describer and thinker. The describer based on a frozen large vision model is designed to describe an image in detail to capture entailed semantic information. The thinker based on a finetuned large multi-modal model is designed to utilize description, claim and image to make prediction and explanation. Experiment results on a publicly available benchmark dataset from FigLang2024 Task 2 show that our method ranks at top 1 in overall evaluation, the performance exceeds the second place by 28.57%. This indicates that MAPPER is highly effective in understanding, judging and explaining of the figurative language. The source code is available at https://github.com/Libv-Team/figlang2024.

pdf bib
FigCLIP: A Generative Multimodal Model with Bidirectional Cross-attention for Understanding Figurative Language via Visual Entailment
Qihao Yang | Xuelin Wang

This is a system paper for the FigLang-2024 Multimodal Figurative Language Shared Task. Figurative language is generally represented through multiple modalities, facilitating the expression of complex and abstract ideas. With the popularity of various text-to-image tools, a large number of images containing metaphors or ironies are created. Traditional recognizing textual entailment has been extended to the task of understanding figurative language via visual entailment. However, existing pre-trained multimodal models in open domains often struggle with this task due to the intertwining of counterfactuals, human culture, and imagination. To bridge this gap, we propose FigCLIP, an end-to-end model based on CLIP and GPT-2, to identify multimodal figurative semantics and generate explanations. It employs a bidirectional fusion module with cross-attention and leverages explanations to promote the alignment of figurative image-text representations. Experimental results on the benchmark demonstrate the effectiveness of our method, achieving 70% F1-score, 67% F1@50-score and 50% F1@60-score. It outperforms GPT-4V, which has robust visual reasoning capabilities.

pdf bib
The Register-specific Distribution of Personification in Hungarian: A Corpus-driven Analysis
Gabor Simon

The aim of the paper is twofold: (i) to present an extended version of the PerSE corpus, the language resource for investigating personification in Hungarian; (ii) to explore the semantic and lexicogrammatical patterns of Hungarian personification in a corpus-driven analysis, based on the current version of the research corpus. PerSE corpus is compiled from online available Hungarian texts in different registers including journalistic (car reviews and reports on interstate relations) and academic discourse (original research papers from different fields). The paper provides the reader with the infrastructure and the protocol of the semi-automatic and manual annotation in the corpus. Then it gives an overview of the register-specific distribution of personifications and focuses on some of its lexicogrammatical patterns.

pdf bib
Report on the Multilingual Euphemism Detection Task
Patrick Lee | Anna Feldman

This paper presents the Multilingual Euphemism Detection Shared Task for the Fourth Workshop on Figurative Language Processing (FigLang 2024) held in conjunction with NAACL 2024. Participants were invited to attempt the euphemism detection task on four different languages (American English, global Spanish, Yorùbá, and Mandarin Chinese): given input text containing a potentially euphemistic term (PET), determine if its use is euphemistic or not. We present the expanded datasets used for the shared task, summarize each team’s methods and findings, and analyze potential implications for future research.

pdf bib
A Report on the FigLang 2024 Shared Task on Multimodal Figurative Language
Shreyas Kulkarni | Arkadiy Saakyan | Tuhin Chakrabarty | Smaranda Muresan

We present the outcomes of the Multimodal Figurative Language Shared Task held at the 4th Workshop on Figurative Language Processing (FigLang 2024) co-located at NAACL 2024. The task utilized the V-FLUTE dataset which is comprised of <image, text> pairs that use figurative language and includes detailed textual explanations for the entailment or contradiction relationship of each pair. The challenge for participants was to develop models capable of accurately identifying the visual entailment relationship in these multimodal instances and generating persuasive free-text explanations. The results showed that the participants’ models significantly outperformed the initial baselines in both automated and human evaluations. We also provide an overview of the systems submitted and analyze the results of the evaluations. All participating systems outperformed the LLaVA-ZS baseline, provided by us in F1-score.
up

pdf (full)
bib (full) Proceedings of the Third Workshop on Bridging Human--Computer Interaction and Natural Language Processing

pdf bib
Proceedings of the Third Workshop on Bridging Human--Computer Interaction and Natural Language Processing
Su Lin Blodgett | Amanda Cercas Curry | Sunipa Dey | Michael Madaio | Ani Nenkova | Diyi Yang | Ziang Xiao

pdf bib
Examining Prosody in Spoken Navigation Instructions for People with Disabilities
Cathy Jiao | Aaron Steinfeld | Maxine Eskenazi

The introduction of conversational systems have made synthesized speech technologies common tools for daily activities. However, not all synthetic speech systems are designed with the needs of people with disabilities in mind. This paper describes a study in which 198 people – 80 participants with self-reported disabilities and 118 participants without – were recruited to listen to navigation instructions from a spoken dialogue system with different prosodic features. Results showed that slowing down speech rate aids in participants’ number recall, but not in noun recall. From our results, we provide suggestions for developers for building accessible synthetic speech systems.

pdf bib
Properties and Challenges of LLM-Generated Explanations
Jenny Kunz | Marco Kuhlmann

The self-rationalising capabilities of large language models (LLMs) have been explored in restricted settings, using task-specific data sets.However, current LLMs do not (only) rely on specifically annotated data; nonetheless, they frequently explain their outputs.The properties of the generated explanations are influenced by the pre-training corpus and by the target data used for instruction fine-tuning.As the pre-training corpus includes a large amount of human-written explanations “in the wild”, we hypothesise that LLMs adopt common properties of human explanations.By analysing the outputs for a multi-domain instruction fine-tuning data set, we find that generated explanations show selectivity and contain illustrative elements, but less frequently are subjective or misleading.We discuss reasons and consequences of the properties’ presence or absence. In particular, we outline positive and negative implications depending on the goals and user groups of the self-rationalising system.

pdf bib
This Reference Does Not Exist: An Exploration of LLM Citation Accuracy and Relevance
Courtni Byun | Piper Vasicek | Kevin Seppi

Citations are a fundamental and indispensable part of research writing. They provide support and lend credibility to research findings. Recent GPT-fueled interest in large language models (LLMs) has shone a spotlight on the capabilities and limitations of these models when generating relevant citations for a document. Recent work has focused largely on title and author accuracy. We underline this effort and expand on it with a preliminary exploration in relevance of model-recommended citations. We define three citation-recommendation tasks. We also collect and annotate a dataset of model-recommended citations for those tasks. We find that GPT-4 largely outperforms earlier models on both author and title accuracy in two markedly different CS venues, but may not recommend references that are more relevant than those recommended by the earlier models. The two venues we compare are CHI and EMNLP. All models appear to perform better at recommending EMNLP papers than CHI papers.

pdf bib
Combining Multiple Metrics for Evaluating Retrieval-Augmented Conversations
Jason Ingyu Choi | Marcus Collins | Eugene Agichtein | Oleg Rokhlenko | Shervin Malmasi

Conversational AI is a subtype of Human Computer Interaction that has gained wide adoption. These systems are typically powered by Large Language Models (LLMs) that use Retrieval Augmented Generation (RAG) to infuse external knowledge, which is effective against issues like hallucination. However, automatically evaluating retrieval augmented conversations with minimal human effort remains challenging, particularly in online settings. We address this challenge by proposing a lexical metric, and a novel method for combining it with other metrics, including semantic models. Our approach involves: (1) Conversational Information Utility (CIU), a new automated metric inspired by prior user studies on web search evaluation, to compute information overlap between conversation context and grounded information in an unsupervised, purely lexical way; and (2) a generalized reward model through Mixture-of-Experts (MoE-CIU) that dynamically ensembles CIU with other metrics, including learned ones, into a single reward. Evaluation against human ratings on two public datasets (Topical Chat and Persona Chat) shows that CIU improves correlation against human judgments by 2.0% and 0.9% respectively compared to the second best metric. When MoE is applied to combine lexical and learned semantic metrics, correlations further improve by 9.9% and 5.0%, suggesting that unified reward models are a promising approach.

pdf bib
How Much Annotation is Needed to Compare Summarization Models?
Chantal Shaib | Joe Barrow | Alexa Siu | Byron Wallace | Ani Nenkova

Modern instruction-tuned models have become highly capable in text generation tasks such as summarization, and are expected to be released at a steady pace. In practice one may now wish to choose confidently, but with minimal effort, the best performing summarization model when applied to a new domain or purpose. In this work, we empirically investigate the test sample size necessary to select a preferred model in the context of news summarization. Empirical results reveal that comparative evaluation converges quickly for both automatic and human evaluation, with clear preferences for a system emerging from under 100 examples. The human preference data allows us to quantify how well automatic scores can reproduce preference rankings across a variety of downstream summarization tasks. We find that, while automatic metrics are stable at smaller sample sizes, only some automatic metrics are able to moderately predict model win rates according to human preference.

pdf bib
An Interactive Co-Pilot for Accelerated Research Ideation
Harshit Nigam | Manasi Patwardhan | Lovekesh Vig | Gautam Shroff

In the realm of research support tools, there exists a notable void in resources tailored specifically for aiding researchers during the crucial ideation phase of the research life-cycle. We address this gap by introducing ‘Acceleron’, a ‘Co-Pilot’ for researchers, designed specifically to accelerate the ideation phase of the research life-cycle. Leveraging the reasoning and domain-specific skills of Large Language Models (LLMs) within an agent-based architecture with distinct personas, Acceleron aids researchers through the formulation of a comprehensive research proposals. It emulates the ideation process, engaging researchers in an interactive fashion to validate the novelty of the proposal and generate plausible set-of hypotheses. Notably, it addresses challenges inherent in LLMs, such as hallucinations, implements a two-stage aspect-based retrieval to manage precision-recall trade-offs, and tackles issues of unanswerability. Our observations and end-user evaluations illustrate the efficacy of Acceleron as an enhancer of researcher’s productivity.

pdf bib
Sensemaking of Socially-Mediated Crisis Information
Vrushali Koli | Jun Yuan | Aritra Dasgupta

In times of crisis, the human mind is often a voracious information forager. It might not be immediately apparent what one wants or needs, and people frequently look for answers to their most pressing questions and worst fears. In that context, the pandemic has demonstrated that social media sources, like erstwhile Twitter, are a rich medium for data-driven communication between experts and the public.However, as lay users, we must find needles in a haystack to distinguish credible and actionable information signals from the noise. In this work, we leverage the literature on crisis communication to propose an AI-driven sensemaking model that bridges the gap between what people seek and what they need during a crisis. Our model learns to contrast social media messages concerning expert guidance with subjective opinion and enables semantic interpretation of message characteristics based on the communicative intent of the message author. We provide examples from our tweet collection and present a hypothetical social media usage scenario to demonstrate the efficacy of our proposed model.

pdf bib
Blind Spots and Biases: Exploring the Role of Annotator Cognitive Biases in NLP
Sanjana Gautam | Mukund Srinath

With the rapid proliferation of artificial intelligence, there is growing concern over its potential to exacerbate existing biases and societal disparities and introduce novel ones. This issue has prompted widespread attention from academia, policymakers, industry, and civil society. While evidence suggests that integrating human perspectives can mitigate bias-related issues in AI systems, it also introduces challenges associated with cognitive biases inherent in human decision-making. Our research focuses on reviewing existing methodologies and ongoing investigations aimed at understanding annotation attributes that contribute to bias.

pdf bib
LLMCheckup: Conversational Examination of Large Language Models via Interpretability Tools and Self-Explanations
Qianli Wang | Tatiana Anikina | Nils Feldhus | Josef Genabith | Leonhard Hennig | Sebastian Möller

Interpretability tools that offer explanations in the form of a dialogue have demonstrated their efficacy in enhancing users’ understanding (Slack et al., 2023; Shen et al., 2023), as one-off explanations may fall short in providing sufficient information to the user. Current solutions for dialogue-based explanations, however, often require external tools and modules and are not easily transferable to tasks they were not designed for. With LLMCheckup, we present an easily accessible tool that allows users to chat with any state-of-the-art large language model (LLM) about its behavior. We enable LLMs to generate explanations and perform user intent recognition without fine-tuning, by connecting them with a broad spectrum of Explainable AI (XAI) methods, including white-box explainability tools such as feature attributions, and self-explanations (e.g., for rationale generation). LLM-based (self-)explanations are presented as an interactive dialogue that supports follow-up questions and generates suggestions. LLMCheckup provides tutorials for operations available in the system, catering to individuals with varying levels of expertise in XAI and supporting multiple input modalities. We introduce a new parsing strategy that substantially enhances the user intent recognition accuracy of the LLM. Finally, we showcase LLMCheckup for the tasks of fact checking and commonsense question answering. Our code repository: https://github.com/DFKI-NLP/LLMCheckup
up

pdf (full)
bib (full) Proceedings of the Fifth Workshop on Insights from Negative Results in NLP

pdf bib
Proceedings of the Fifth Workshop on Insights from Negative Results in NLP
Shabnam Tafreshi | Arjun Akula | João Sedoc | Aleksandr Drozd | Anna Rogers | Anna Rumshisky

pdf bib
MoSECroT: Model Stitching with Static Word Embeddings for Crosslingual Zero-shot Transfer
Haotian Ye | Yihong Liu | Chunlan Ma | Hinrich Schütze

Transformer-based pre-trained language models (PLMs) have achieved remarkable performance in various natural language processing (NLP) tasks. However, pre-training such models can take considerable resources that are almost only available to high-resource languages. On the contrary, static word embeddings are easier to train in terms of computing resources and the amount of data required. In this paper, we introduce MoSECroT (Model Stitching with Static Word Embeddings for Crosslingual Zero-shot Transfer, a novel and challenging task that is especially relevant to low-resource languages for which static word embeddings are available. To tackle the task, we present the first framework that leverages relative representations to construct a common space for the embeddings of a source language PLM and the static word embeddings of a target language. In this way, we can train the PLM on source-language training data and perform zero-shot transfer to the target language by simply swapping the embedding layer. However, through extensive experiments on two classification datasets, we show that although our proposed framework is competitive with weak baselines when addressing MoSECroT, it fails to achieve competitive results compared with some strong baselines. In this paper, we attempt to explain this negative result and provide several thoughts on possible improvement.

pdf bib
What explains the success of cross-modal fine-tuning with ORCA?
Paloma Garcia De Herreros | Vagrant Gautam | Philipp Slusallek | Dietrich Klakow | Marius Mosbach

ORCA (Shen et al., 2023) is a recent technique for cross-modal fine-tuning, i.e., applying pre-trained transformer models to modalities beyond their training data. The technique consists primarily of training an embedder and fine-tuning the embedder and model. Despite its high performance on a variety of downstream tasks, we do not understand precisely how each of these components contribute to ORCA’s success. Therefore, we run a series of ablations and find that embedder training does not help 2D tasks at all, contrary to what the original paper posits. In 1D tasks, some amount of embedder training is necessary but more is not better. In 4 out of 6 datasets we experiment with, it is model fine-tuning that makes the biggest difference. Through our ablations and baselines, we contribute a better understanding of the individual components of ORCA.

pdf bib
Does Fine-tuning a Classifier Help in Low-budget Scenarios? Not Much
Cesar Gonzalez - Gutierrez | Audi Primadhanty | Francesco Cazzaro | Ariadna Quattoni

In recent years, the two-step approach for text classification based on pre-training plus fine-tuning has led to significant improvements in classification performance. In this paper, we study the low-budget scenario, and we ask whether it is justified to allocate the additional resources needed for fine-tuning complex models. To do so, we isolate the gains obtained from pre-training from those obtained from fine-tuning. We find out that, when the gains from pre-training are factored out, the performance attained by using complex transformer models leads to marginal improvements over simpler models. Therefore, in this scenario, utilizing simpler classifiers on top of pre-trained representations proves to be a viable alternative.

pdf bib
How Well Can a Genetic Algorithm Fine-tune Transformer Encoders? A First Approach
Vicente Ivan Sanchez Carmona | Shanshan Jiang | Bin Dong

Genetic Algorithms (GAs) have been studied across different fields such as engineering or medicine to optimize diverse problems such as network routing, or medical image segmentation. Moreover, they have been used to automatically find optimal architectures for deep neural networks. However, to our knowledge, they have not been applied as a weight optimizer for the Transformer model. While gradient descent has been the main paradigm for this task, we believe that GAs have advantages to bring to the table. In this paper, we will show that even though GAs are capable of fine-tuning Transformer encoders, their generalization ability is considerably poorer than that from Adam; however, on a closer look, GAs ability to exploit knowledge from 2 different pretraining datasets surpasses Adam’s ability to do so.

pdf bib
I Have an Attention Bridge to Sell You: Generalization Capabilities of Modular Translation Architectures
Timothee Mickus | Raul Vazquez | Joseph Attieh

Modularity is a paradigm of machine translation with the potential of bringing forth models that are large at training time and small during inference. Within this field of study, modular approaches, and in particular attention bridges, have been argued to improve the generalization capabilities of models by fostering language-independent representations. In the present paper, we study whether modularity affects translation quality; as well as how well modular architectures generalize across different evaluation scenarios. For a given computational budget, we find non-modular architectures to be always comparable or preferable to all modular designs we study.

pdf bib
Knowledge Distillation vs. Pretraining from Scratch under a Fixed (Computation) Budget
Minh Duc Bui | Fabian Schmidt | Goran Glavaš | Katharina Von Der Wense

Compared to standard language model (LM) pretraining (i.e., from scratch), Knowledge Distillation (KD) entails an additional forward pass through a teacher model that is typically substantially larger than the target student model. As such, KD in LM pretraining materially slows down throughput of pretraining instances vis-a-vis pretraining from scratch. Scaling laws of LM pretraining suggest that smaller models can close the gap to larger counterparts if trained on more data (i.e., processing more tokens)—and under a fixed computation budget, smaller models are able to process more data than larger models. We thus hypothesize that KD might, in fact, be suboptimal to pretraining from scratch for obtaining smaller LMs, when appropriately accounting for the compute budget. To test this, we compare pretraining from scratch against several KD strategies for masked language modeling (MLM) in a fair experimental setup, with respect to amount of computation as well as pretraining data. Downstream results on GLUE, however, do not confirm our hypothesis: while pretraining from scratch performs comparably to ordinary KD under a fixed computation budget, more sophisticated KD strategies, namely TinyBERT and MiniLM, outperform it by a notable margin. We further find that KD yields larger gains over pretraining from scratch when the data can be repeated under the fixed computation budget.

pdf bib
An Analysis of BPE Vocabulary Trimming in Neural Machine Translation
Marco Cognetta | Tatsuya Hiraoka | Rico Sennrich | Yuval Pinter | Naoaki Okazaki

We explore threshold vocabulary trimming in Byte-Pair Encoding subword tokenization, a tokenization postprocessing step that replaces rare subwords with their component subwords. The technique is available in popular tokenization libraries but has not been subjected to rigorous scientific scrutiny. While the removal of rare subwords is suggested as best practice in model implementations, both as a means to reduce model size and for improving model performance through robustness, our experiments indicate that, across a large space of hyperparameter settings, vocabulary trimming fails to consistently improve model performance, and is even prone to incurring heavy degradation.

pdf bib
On the Limits of Multi-modal Meta-Learning with Auxiliary Task Modulation Using Conditional Batch Normalization
Jordi Armengol - Estape | Vincent Michalski | Ramnath Kumar | Pierre - Luc St-Charles | Doina Precup | Samira Ebrahimi Kahou

Few-shot learning aims to learn representations that can tackle novel tasks given a small number of examples. Recent studies show that cross-modal learning can improve representations for few-shot classification. More specifically, language is a rich modality that can be used to guide visual learning. In this work, we experiment with a multi-modal architecture for few-shot learning that consists of three components: a classifier, an auxiliary network, and a bridge network. While the classifier performs the main classification task, the auxiliary network learns to predict language representations from the same input, and the bridge network transforms high-level features of the auxiliary network into modulation parameters for layers of the few-shot classifier using conditional batch normalization. The bridge should encourage a form of lightweight semantic alignment between language and vision which could be useful for the classifier. However, after evaluating the proposed approach on two popular few-shot classification benchmarks we find that a) the improvements do not reproduce across benchmarks, and b) when they do, the improvements are due to the additional compute and parameters introduced by the bridge network. We contribute insights and recommendations for future work in multi-modal meta-learning, especially when using language representations.

pdf bib
Pointer-Generator Networks for Low-Resource Machine Translation: Don’t Copy That!
Niyati Bafna | Philipp Koehn | David Yarowsky

While Transformer-based neural machine translation (NMT) is very effective in high-resource settings, many languages lack the necessary large parallel corpora to benefit from it. In the context of low-resource (LR) MT between two closely-related languages, a natural intuition is to seek benefits from structural “shortcuts”, such as copying subwords from the source to the target, given that such language pairs often share a considerable number of identical words, cognates, and borrowings. We test Pointer-Generator Networks for this purpose for six language pairs over a variety of resource ranges, and find weak improvements for most settings. However, analysis shows that the model does not show greater improvements for closely-related vs. more distant language pairs, or for lower resource ranges, and that the models do not exhibit the expected usage of the mechanism for shared subwords. Our discussion of the reasons for this behaviour highlights several general challenges for LR NMT, such as modern tokenization strategies, noisy real-world conditions, and linguistic complexities. We call for better scrutiny of linguistically motivated improvements to NMT given the blackbox nature of Transformer models, as well as for a focus on the above problems in the field.

pdf bib
Imaginary Numbers! Evaluating Numerical Referring Expressions by Neural End-to-End Surface Realization Systems
Rossana Cunha | Osuji Chinonso | João Campos | Brian Timoney | Brian Davis | Fabio Cozman | Adriana Pagano | Thiago Castro Ferreira

Neural end-to-end surface realizers output more fluent texts than classical architectures. However, they tend to suffer from adequacy problems, in particular hallucinations in numerical referring expression generation. This poses a problem to language generation in sensitive domains, as is the case of robot journalism covering COVID-19 and Amazon deforestation. We propose an approach whereby numerical referring expressions are converted from digits to plain word form descriptions prior to being fed to state-of-the-art Large Language Models. We conduct automatic and human evaluations to report the best strategy to numerical superficial realization. Code and data are publicly available.

pdf bib
Using Locally Learnt Word Representations for better Textual Anomaly Detection
Alicia Breidenstein | Matthieu Labeau

The literature on general purpose textual Anomaly Detection is quite sparse, as most textual anomaly detection methods are implemented as out of domain detection in the context of pre-established classification tasks. Notably, in a field where pre-trained representations and models are of common use, the impact of the pre-training data on a task that lacks supervision has not been studied. In this paper, we use the simple setting of k-classes out anomaly detection and search for the best pairing of representation and classifier. We show that well-chosen embeddings allow a simple anomaly detection baseline such as OC-SVM to achieve similar results and even outperform deep state-of-the-art models.

pdf bib
Can probing classifiers reveal the learning by contact center large language models?: No, it doesn’t!
Varun Nathan | Ayush Kumar | Digvijay Ingle

Fine-tuning large language models (LLMs) with domain-specific instruction dataset has emerged as an effective method to enhance their domain-specific understanding. Yet, there is limited work that examines the core characteristics acquired during this process. In this study, we benchmark the fundamental characteristics learned by contact-center (CC) domain specific instruction fine-tuned LLMs with out-of-the-box (OOB) LLMs via probing tasks encompassing conversational, channel, and automatic speech recognition (ASR) properties. We explore different LLM architectures (Flan-T5 and Llama) and sizes (3B, 7B, 11B, 13B). Our findings reveal remarkable effectiveness of CC-LLMs on the in-domain downstream tasks, with improvement in response acceptability by over 48% compared to OOB-LLMs. However, we observe that the performance of probing classifiers are relatively similar and does not reflect the performance of in-domain downstream tasks. A similar observation is also noted on SentEval dataset that assess capabilities of models in terms of surface, syntactic, and semantic information through probing tasks. Our study challenges the premise that probing classifiers can reveal the fundamental characteristics learned by large language models and is reflective of the downstream task performance, via a case-study of LLMs tuned for contact center domain.

pdf bib
Can Abstract Meaning Representation Facilitate Fair Legal Judgement Predictions?
Supriti Vijay | Daniel Hershcovich

Legal judgment prediction encompasses the automated prediction of case outcomes by leveraging historical facts and opinions. While this approach holds the potential to enhance the efficiency of the legal system, it also raises critical concerns regarding the perpetuation of biases. Abstract Meaning Representation has shown promise as an intermediate text representation in various downstream NLP tasks due to its ability to capture semantically meaningful information in a graph-like structure. In this paper, we employ this ability of AMR in the legal judgement prediction task and assess to what extent it encodes biases, or conversely, abstracts away from them. Our study reveals that while AMR-based models exhibit worse overall performance than transformer-based models, they are less biased for attributes like age and defendant state compared to gender. By shedding light on these findings, this paper contributes to a more nuanced understanding of AMR’s potential benefits and limitations in legal NLP.

pdf bib
WINOVIZ: Probing Visual Properties of Objects Under Different States
Woojeong Jin | Tejas Srinivasan | Jesse Thomason | Xiang Ren

Humans interpret visual aspects of objects based on contexts. For example, a banana appears brown when rotten and green when unripe. Previous studies focused on language models’ grasp of typical object properties. We introduce WINOVIZ, a text-only dataset with 1,380 examples of probing language models’ reasoning about diverse visual properties under different contexts. Our task demands pragmatic and visual knowledge reasoning. We also present multi-hop data, a more challenging version requiring multi-step reasoning chains. Experimental findings include: a) GPT-4 excels overall but struggles with multi-hop data. b) Large models perform well in pragmatic reasoning but struggle with visual knowledge reasoning. c) Vision-language models outperform language-only models.

pdf bib
Harnessing the Power of Multiple Minds: Lessons Learned from LLM Routing
Kv Aditya Srivatsa | Kaushal Maurya | Ekaterina Kochmar

With the rapid development of LLMs, it is natural to ask how to harness their capabilities efficiently. In this paper, we explore whether it is feasible to direct each input query to a single most suitable LLM. To this end, we propose LLM routing for challenging reasoning tasks. Our extensive experiments suggest that such routing shows promise but is not feasible in all scenarios, so more robust approaches should be investigated to fill this gap.

pdf bib
The Paradox of Preference: A Study on LLM Alignment Algorithms and Data Acquisition Methods
Rishikesh Devanathan | Varun Nathan | Ayush Kumar

This research investigates the impact of preference annotation acquisition methods on the performance of LLM alignment algorithms, including Direct Preference Optimization (DPO), Identity Preference Optimization (IPO), and Conservative DPO (cDPO), compared to Supervised Fine-Tuning (SFT) in NLP tasks. We analyze the influence of LLM and human-based preferences on algorithm performance, considering data volume and quality. Additionally, we assess DPO’s vulnerability to overfitting and IPO’s resilience against it, addressing four main research questions. Using the GAIR dataset and Zephyr-7b as the SFT model, we reveal unexpected negative outcomes. Specifically, DPO trained on LLM preferences outperforms human preferences, contrary to expectations. Moreover, there’s no correlation between preference data volume or quality and algorithm performance. Contrary to expectations, DPO shows no overfitting in both human and LLM preference datasets. Surprisingly, cDPO doesn’t fare better than DPO under flip noise. Our findings highlight the complexities of preference annotation methods and underscore the importance of scrutinizing negative results in NLP algorithm research.

pdf bib
The Ups and Downs of Large Language Model Inference with Vocabulary Trimming by Language Heuristics
Nikolay Bogoychev | Pinzhen Chen | Barry Haddow | Alexandra Birch

Deploying large language models (LLMs) encounters challenges due to intensive computational and memory requirements. Our research examines vocabulary trimming (VT) inspired by restricting embedding entries to the language of interest to bolster time and memory efficiency. While such modifications have been proven effective in tasks like machine translation, tailoring them to LLMs demands specific modifications given the diverse nature of LLM applications. We apply two language heuristics to trim the full vocabulary—Unicode-based script filtering and corpus-based selection—to different LLM families and sizes. The methods are straightforward, interpretable, and easy to implement. It is found that VT reduces the memory usage of small models by nearly 50% and has an upper bound of 25% improvement in generation speed. Yet, we reveal the limitations of these methods in that they do not perform consistently well for each language with diminishing returns in larger models.

pdf bib
Multi-Task Learning with Adapters for Plausibility Prediction: Bridging the Gap or Falling into the Trenches?
Annerose Eichel | Sabine Schulte Im Walde

We present a multi-task learning approach to predicting semantic plausibility by leveraging 50+ adapters categorized into 17 tasks within an efficient training framework. Across four plausibility datasets in English of varying size and linguistic constructions, we compare how models provided with knowledge from a range of NLP tasks perform in contrast to models without external information. Our results show that plausibility prediction benefits from complementary knowledge (e.g., provided by syntactic tasks) are significant but non-substantial, while performance may be hurt when injecting knowledge from an unsuitable task. Similarly important, we find that knowledge transfer may be hindered by class imbalance, and demonstrate the positive yet minor effect of balancing training data, even at the expense of size.

pdf bib
Investigating Multi-Pivot Ensembling with Massively Multilingual Machine Translation Models
Alireza Mohammadshahi | Jannis Vamvas | Rico Sennrich

Massively multilingual machine translation models allow for the translation of a large number of languages with a single model, but have limited performance on low- and very-low-resource translation directions. Pivoting via high-resource languages remains a strong strategy for low-resource directions, and in this paper we revisit ways of pivoting through multiple languages. Previous work has used a simple averaging of probability distributions from multiple paths, but we find that this performs worse than using a single pivot, and exacerbates the hallucination problem because the same hallucinations can be probable across different paths. We also propose MaxEns, a novel combination strategy that makes the output biased towards the most confident predictions, hypothesising that confident predictions are less prone to be hallucinations. We evaluate different strategies on the FLORES benchmark for 20 low-resource language directions, demonstrating that MaxEns improves translation quality for low-resource languages while reducing hallucination in translations, compared to both direct translation and an averaging approach. On average, multi-pivot strategies still lag behind using English as a single pivot language, raising the question of how to identify the best pivoting strategy for a given translation direction.
up

pdf (full)
bib (full) Proceedings of the Sixth Workshop on Natural Language Processing and Computational Social Science (NLP+CSS 2024)

pdf bib
Proceedings of the Sixth Workshop on Natural Language Processing and Computational Social Science (NLP+CSS 2024)
Dallas Card | Anjalie Field | Dirk Hovy | Katherine Keith

pdf bib
Detecting Perspective-Getting in Wikipedia Discussions
Evgeny Vasilets | Tijs Broek | Anna Wegmann | David Abadi | Dong Nguyen

Perspective-getting (i.e., the effort to obtain information about the other person’s perspective) can lead to more accurate interpersonal understanding. In this paper, we develop an approach to measure perspective-getting and apply it to English Wikipedia discussions. First, we develop a codebook based on perspective-getting theory to operationalize perspective-getting into two categories: asking questions about and attending the other’s perspective. Second, we use the codebook to annotate perspective-getting in Wikipedia discussion pages. Third, we fine-tune a RoBERTa model that achieves an average F-1 score of 0.76 on the two perspective-getting categories. Last, we test whether perspective-getting is associated with discussion outcomes. Perspective-getting was not higher in non-escalated discussions. However, discussions starting with a post attending the other’s perspective are followed by responses that are more likely to also attend the other’s perspective. Future research may use our model to study the influence of perspective-getting on the dynamics and outcomes of online discussions.

pdf bib
Connecting the Dots in News Analysis: Bridging the Cross-Disciplinary Disparities in Media Bias and Framing
Gisela Vallejo | Timothy Baldwin | Lea Frermann

The manifestation and effect of bias in news reporting have been central topics in the social sciences for decades, and have received increasing attention in the NLP community recently. While NLP can help to scale up analyses or contribute automatic procedures to investigate the impact of biased news in society, we argue that methodologies that are currently dominant fall short of capturing the complex questions and effects addressed in theoretical media studies. This is problematic because it diminishes the validity and safety of the resulting tools and applications. Here, we review and critically compare task formulations, methods and evaluation schemes in the social sciences and NLP. We discuss open questions and suggest possible directions to close identified gaps between theory and predictive models, and their evaluation. These include model transparency, considering document-external information, and cross-document reasoning.

pdf bib
The Crime of Being Poor: Associations between Crime and Poverty on Social Media in Eight Countries
Georgina Curto | Svetlana Kiritchenko | Kathleen Fraser | Isar Nejadgholi

Negative public perceptions of people living in poverty can hamper policies and programs that aim to help the poor. One prominent example of social bias and discrimination against people in need is the persistent association of poverty with criminality. The phenomenon has two facets: first, the belief that poor people are more likely to engage in crime (e.g., stealing, mugging, violence) and second, the view that certain behaviors directly resulting from poverty (e.g., living outside, panhandling) warrant criminal punishment. In this paper, we use large language models (LLMs) to identify examples of crime–poverty association (CPA) in English social media texts. We analyze the online discourse on CPA across eight geographically-diverse countries, and find evidence that the CPA rates are higher within the sample obtained from the U.S. and Canada, as compared to the other countries such as South Africa, despite the latter having higher poverty, criminality, and inequality indexes. We further uncover and analyze the most common themes in CPA posts and find more negative and biased attitudes toward people living in poverty in posts from the U.S. and Canada. These results could partially be explained by cultural factors related to the tendency to overestimate the equality of opportunities and social mobility in the U.S. and Canada. These findings have consequences for policy-making and open a new path of research for poverty mitigation with the focus not only on the redistribution of wealth but also on the mitigation of bias and discrimination against people in need.

pdf bib
Discovering Implicit Meanings of Cultural Motifs from Text
Anurag Acharya | Diego Estrada | Shreeja Dahal | W. Victor H. Yarlott | Diana Gomez | Mark Finlayson

Motifs are distinctive, recurring, widely used idiom-like words or phrases, often originating in folklore and usually strongly anchored to a particular cultural or national group. Motifs are significant communicative devices across a wide range of media—including news, literature, and propaganda—because they can concisely imply a large set of culturally relevant associations. One difficulty of understanding motifs is that their meaning is usually implicit, so for an out-group person the meaning is inaccessible. We present the Motif Implicit Meaning Extractor (MIME), a proof-of-concept system designed to automatically identify a motif’s implicit meaning, as evidenced by textual uses of the motif across a large set data. MIME uses several sources (including motif indices, Wikipedia pages on the motifs, explicit explanations of motifs from in-group informants, and news/social media posts where the motif is used) and can generate a structured report of information about a motif understandable to an out-group person. In addition to a variety of examples and information drawn from structured sources, the report includes implicit information about a motif such as the type of reference (e.g., a person, an organization, etc.), it’s general connotation (strongly negative, slightly negative, neutral, etc.), and it’s associations (typically adjectives). We describe how MIME works and demonstrate its operation on a small set of manually curated motifs. We perform a qualitative evaluation of the output, and assess the difficulty of the problem, showing that explicit motif information provided by cultural informants is critical to high quality output, although mining motif usages in news and social media provides useful additional depth. A system such as MIME, appropriately scaled up, would potentially be quite useful to an out-group person trying to understand in-group usages of motifs, and has wide potential applications in domains such as literary criticism, cultural heritage, marketed and branding, and intelligence analysis.

pdf bib
Can Large Language Models (or Humans) Disentangle Text?
Nicolas Pieuchon | Adel Daoud | Connor Jerzak | Moa Johansson | Richard Johansson

We investigate the potential of large language models (LLMs) to disentangle text variables—to remove the textual traces of an undesired forbidden variable in a task sometimes known as text distillation and closely related to the fairness in AI and causal inference literature. We employ a range of various LLM approaches in an attempt to disentangle text by identifying and removing information about a target variable while preserving other relevant signals. We show that in the strong test of removing sentiment, the statistical association between the processed text and sentiment is still detectable to machine learning classifiers post-LLM-disentanglement. Furthermore, we find that human annotators also struggle to disentangle sentiment while preserving other semantic content. This suggests there may be limited separability between concept variables in some text contexts, highlighting limitations of methods relying on text-level transformations and also raising questions about the robustness of disentanglement methods that achieve statistical independence in representation space.

pdf bib
Retrieval Augmented Generation of Subjective Explanations for Socioeconomic Scenarios
Razvan-Gabriel Dumitru | Maria Alexeeva | Keith Alcock | Nargiza Ludgate | Cheonkam Jeong | Zara Fatima Abdurahaman | Prateek Puri | Brian Kirchhoff | Santadarshan Sadhu | Mihai Surdeanu

We introduce a novel retrieval augmented generation approach that explicitly models causality and subjectivity. We use it to generate explanations for socioeconomic scenarios that capture beliefs of local populations. Through intrinsic and extrinsic evaluation, we show that our explanations, contextualized using causal and subjective information retrieved from local news sources, are rated higher than those produced by other large language models both in terms of mimicking the real population and the explanations quality. We also provide a discussion of the role subjectivity plays in evaluation of this natural language generation task.

pdf bib
Where on Earth Do Users Say They Are?: Geo-Entity Linking for Noisy Multilingual User Input
Tessa Masis | Brendan O’Connor

Geo-entity linking is the task of linking a location mention to the real-world geographic location. In this we explore the challenging task of geo-entity linking for noisy, multilingual social media data. There are few open-source multilingual geo-entity linking tools available and existing ones are often rule-based, which break easily in social media settings, or LLM-based, which are too expensive for large-scale datasets. We present a method which represents real-world locations as averaged embeddings from labeled user-input location names and allows for selective prediction via an interpretable confidence score. We show that our approach improves geo-entity linking on a global and multilingual social media dataset, and discuss progress and problems with evaluating at different geographic granularities.

pdf bib
News Deja Vu: Connecting Past and Present with Semantic Search
Brevin Franklin | Emily Silcock | Abhishek Arora | Tom Bryan | Melissa Dell

Social scientists and the general public often analyze contemporary events by drawing parallels with the past, a process complicated by the vast, noisy, and unstructured nature of historical texts. For example, hundreds of millions of page scans from historical newspapers have been noisily transcribed. Traditional sparse methods for searching for relevant material in these vast corpora, e.g., with keywords, can be brittle given complex vocabularies and OCR noise. This study introduces News Deja Vu, a novel semantic search tool that leverages transformer large language models and a bi-encoder approach to identify historical news articles that are most similar to modern news queries. News Deja Vu first recognizes and masks entities, in order to focus on broader parallels rather than the specific named entities being discussed. Then, a contrastively trained, lightweight bi-encoder retrieves historical articles that are most similar semantically to a modern query, illustrating how phenomena that might seem unique to the present have varied historical precedents. Aimed at social scientists, the user-friendly News Deja Vu package is designed to be accessible for those who lack extensive familiarity with deep learning. It works with large text datasets, and we show how it can be deployed to a massive scale corpus of historical, open-source news articles. While human expertise remains important for drawing deeper insights, News Deja Vu provides a powerful tool for exploring parallels in how people have perceived past and present.

pdf bib
Knowledge Distillation in Automated Annotation: Supervised Text Classification with LLM-Generated Training Labels
Nicholas Pangakis | Sam Wolken

Computational social science (CSS) practitioners often rely on human-labeled data to fine-tune supervised text classifiers. We assess the potential for researchers to augment or replace human-generated training data with surrogate training labels from generative large language models (LLMs). We introduce a recommended workflow and test this LLM application by replicating 14 classification tasks and measuring performance. We employ a novel corpus of English-language text classification data sets from recent CSS articles in high-impact journals. Because these data sets are stored in password-protected archives, our analyses are less prone to issues of contamination. For each task, we compare supervised classifiers fine-tuned using GPT-4 labels against classifiers fine-tuned with human annotations and against labels from GPT-4 and Mistral-7B with few-shot in-context learning. Our findings indicate that supervised classification models fine-tuned on LLM-generated labels perform comparably to models fine-tuned with labels from human annotators. Fine-tuning models using LLM-generated labels can be a fast, efficient and cost-effective method of building supervised text classifiers.

pdf bib
Clustering Document Parts: Detecting and Characterizing Influence Campaigns from Documents
Zhengxiang Wang | Owen Rambow

We propose a novel clustering pipeline to detect and characterize influence campaigns from documents. This approach clusters parts of document, detects clusters that likely reflect an influence campaign, and then identifies documents linked to an influence campaign via their association with the high-influence clusters. Our approach outperforms both the direct document-level classification and the direct document-level clustering approach in predicting if a document is part of an influence campaign. We propose various novel techniques to enhance our pipeline, including using an existing event factuality prediction system to obtain document parts, and aggregating multiple clustering experiments to improve the performance of both cluster and document classification. Classifying documents after clustering not only accurately extracts the parts of the documents that are relevant to influence campaigns, but also captures influence campaigns as a coordinated and holistic phenomenon. Our approach makes possible more fine-grained and interpretable characterizations of influence campaigns from documents.

pdf bib
A First Step towards Measuring Interdisciplinary Engagement in Scientific Publications: A Case Study on NLP + CSS Research
Alexandria Leto | Shamik Roy | Alexander Hoyle | Daniel Acuna | Maria Pacheco

With the rise in the prevalence of cross-disciplinary research, there is a need to develop methods to characterize its practices. Current computational methods to evaluate interdisciplinary engagement—such as affiliation diversity, keywords, and citation patterns—are insufficient to model the degree of engagement between disciplines, as well as the way in which the complementary expertise of co-authors is harnessed. In this paper, we propose an automated framework to address some of these issues on a large scale. Our framework tracks interdisciplinary citations in scientific articles and models: 1) the section and position in which they appear, and 2) the argumentative role that they play in the writing. To showcase our framework, we perform a preliminary analysis of interdisciplinary engagement in published work at the intersection of natural language processing and computational social science in the last decade.
up

pdf (full)
bib (full) Proceedings of the 18th International Workshop on Semantic Evaluation (SemEval-2024)

pdf bib
Proceedings of the 18th International Workshop on Semantic Evaluation (SemEval-2024)
Atul Kr. Ojha | A. Seza Doğruöz | Harish Tayyar Madabushi | Giovanni Da San Martino | Sara Rosenthal | Aiala Rosá

pdf bib
CUNLP at SemEval-2024 Task 8: Classify Human and AI Generated Text
Pranjal Aggarwal | Deepanshu Sachdeva

This task is a sub-part of SemEval-2024 competition which aims to classify AI vs Human Generated Text. In this paper we have experimented on an approach to automatically classify an artificially generated text and a human written text. With the advent of generative models like GPT-3.5 and GPT-4 it has become increasingly necessary to classify between the two texts due to various applications like detecting plagiarism and in tasks like fake news detection that can heavily impact real world problems, for instance stock manipulation through AI generated news articles. To achieve this, we start by using some basic models like Logistic Regression and move our way up to more complex models like transformers and GPTs for classification. This is a binary classification task where the label 1 represents AI generated text and 0 represents human generated text. The dataset was given in JSON style format which was converted to comma separated file (CSV) for better processing using the pandas library in Python as CSV files provides more readability than JSON format files. Approaches like Bagging Classifier and Voting classifier were also used.

pdf bib
OZemi at SemEval-2024 Task 1: A Simplistic Approach to Textual Relatedness Evaluation Using Transformers and Machine Translation
Hidetsune Takahashi | Xingru Lu | Sean Ishijima | Deokgyu Seo | Yongju Kim | Sehoon Park | Min Song | Kathylene Marante | Keitaro-luke Iso | Hirotaka Tokura | Emily Ohman

In this system paper for SemEval-2024 Task 1 subtask A, we present our approach to evaluating the semantic relatedness of sentence pairs in nine languages. We use a mix of statistical methods combined with fine-tuned BERT transformer models for English and use the same model and machine-translated data for the other languages. This simplistic approach shows consistently reliable scores and achieves above-average rank in all languages.

pdf bib
L3i++ at SemEval-2024 Task 8: Can Fine-tuned Large Language Model Detect Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text?
Hanh Thi Hong Tran | Tien Nam Nguyen | Antoine Doucet | Senja Pollak

This paper summarizes our participation in SemEval-2024 Task 8: Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text Detection. In this task, we aim to solve two over three Subtasks: (1) Monolingual and Multilingual Binary Human-Written vs. Machine-Generated Text Classification; and (2) Multi-Way Machine-Generated Text Classification. We conducted a comprehensive comparative study across three methodological groups: Five metric-based models (Log-Likelihood, Rank, Log-Rank, Entropy, and MFDMetric), two fine-tuned sequence-labeling language models (RoBERTA and XLM-R); and a fine-tuned large-scale language model (LS-LLaMA). Our findings suggest that our LLM outperformed both traditional sequence-labeling LM benchmarks and metric-based approaches. Furthermore, our fine-tuned classifier excelled in detecting machine-generated multilingual texts and accurately classifying machine-generated texts within a specific category, (e.g., ChatGPT, bloomz, dolly). However, they do exhibit challenges in detecting them in other categories (e.g., cohere, and davinci). This is due to potential overlap in the distribution of the metric among various LLMs. Overall, we achieved a 6th rank in both Multilingual Binary Human-Written vs. Machine-Generated Text Classification and Multi-Way Machine-Generated Text Classification on the leaderboard.

pdf bib
nicolay-r at SemEval-2024 Task 3: Using Flan-T5 for Reasoning Emotion Cause in Conversations with Chain-of-Thought on Emotion States
Nicolay Rusnachenko | Huizhi Liang

Emotion expression is one of the essential traits of conversations. It may be self-related or caused by another speaker. The variety of reasons may serve as a source of the further emotion causes: conversation history, speaker’s emotional state, etc. Inspired by the most recent advances in Chain-of-Thought, in this work, we exploit the existing three-hop reasoning approach (THOR) to perform large language model instruction-tuning for answering: emotion states (THOR-state), and emotion caused by one speaker to the other (THOR-cause). We equip THORcause with the reasoning revision (RR) for devising a reasoning path in fine-tuning. In particular, we rely on the annotated speaker emotion states to revise reasoning path. Our final submission, based on Flan-T5-base (250M) and the rule-based span correction technique, preliminary tuned with THOR-state and fine-tuned with THOR-cause-rr on competition training data, results in 3rd and 4th places (F1-proportional) and 5th place (F1-strict) among 15 participating teams. Our THOR implementation fork is publicly available: https://github.com/nicolay-r/THOR-ECAC

pdf bib
StFX-NLP at SemEval-2024 Task 9: BRAINTEASER: Three Unsupervised Riddle-Solvers
Ethan Heavey | James Hughes | Milton King

In this paper, we explore three unsupervised learning models that we applied to Task 9: BRAINTEASER of SemEval 2024. Two of these models incorporate word sense disambiguation and part-of-speech tagging, specifically leveraging SensEmBERT and the Stanford log-linear part-of-speech tagger. Our third model relies on a more traditional language modelling approach. The best performing model, a bag-of-words model leveraging word sense disambiguation and part-of-speech tagging, secured the 10th spot out of 11 places on both the sentence puzzle and word puzzle subtasks.

pdf bib
hinoki at SemEval-2024 Task 7: Numeral-Aware Headline Generation (English)
Hinoki Crum | Steven Bethard

Numerical reasoning is challenging even for large pre-trained language models. We show that while T5 models are capable of generating relevant headlines with proper numerical values, they can also make mistakes in reading comprehension and miscalculate numerical values. To overcome these issues, we propose a two-step training process: first train models to read text and generate formal representations of calculations, then train models to read calculations and generate numerical values. On the SemEval 2024 Task 7 headline fill-in-the-blank task, our two-stage Flan-T5-based approach achieved 88% accuracy. On the headline generation task, our T5-based approach achieved RougeL of 0.390, BERT F1 Score of 0.453, and MoverScore of 0.587.

pdf bib
T5-Medical at SemEval-2024 Task 2: Using T5 Medical Embedding for Natural Language Inference on Clinical Trial Data
Marco Siino

In this work, we address the challenge of identifying the inference relation between a plain language statement and Clinical Trial Reports (CTRs) by using a T5-large model embedding. The task, hosted at SemEval-2024, involves the use of the NLI4CT dataset. Each instance in the dataset has one or two CTRs, along with an annotation from domain experts, a section marker, a statement, and an entailment/contradiction label. The goal is to determine if a statement entails or contradicts the given information within a trial description. Our submission consists of a T5-large model pre-trained on the medical domain. Then the pre-trained model embedding output provides the embedding representation of the text. Eventually, after a fine-tuning phase, the provided embeddings are used to determine the CTRs’ and the statements’ cosine similarity to perform the classification. On the official test set, our submitted approach is able to reach an F1 score of 0.63, and a faithfulness and consistency score of 0.30 and 0.50 respectively.

pdf bib
CTYUN-AI at SemEval-2024 Task 7: Boosting Numerical Understanding with Limited Data Through Effective Data Alignment
Yuming Fan | Dongming Yang | Xu He

Large language models (LLMs) have demonstrated remarkable capabilities in pushing the boundaries of natural language understanding. Nevertheless, the majority of existing open-source LLMs still fall short of meeting satisfactory standards when it comes to addressing numerical problems, especially as the enhancement of their numerical capabilities heavily relies on extensive data.To bridge the gap, we aim to improve the numerical understanding of LLMs by means of efficient data alignment, utilizing only a limited amount of necessary data.Specifically, we first use a data discovery strategy to obtain the most effective portion of numerical data from large datasets. Then, self-augmentation is performed to maximize the potential of the training data. Thirdly, answers of all traning samples are aligned based on some simple rules. Finally, our method achieves the first place in the competition, offering new insights and methodologies for numerical understanding research in LLMs.

pdf bib
McRock at SemEval-2024 Task 4: Mistral 7B for Multilingual Detection of Persuasion Techniques In Memes
Marco Siino

One of the most widely used content types in internet misinformation campaigns is memes. Since they can readily reach a big number of users on social media sites, they are most successful there. Memes used in a disinformation campaign include a variety of rhetorical and psychological strategies, including smearing, name-calling, and causal oversimplification, to achieve their goal of influencing the users. The shared task’s objective is to develop models for recognizing these strategies solely in a meme’s textual content (Subtask 1) and in a multimodal context where both the textual and visual material must be analysed simultaneously (Subtasks two and three). In this paper, we discuss the application of a Mistral 7B model to address the Subtask one in English. Find the persuasive strategy that a meme employs from a hierarchy of twenty based just on its “textual content.” Only a portion of the reward is awarded if the technique’s ancestor node is chosen. This classification issue is multilabel hierarchical. Our approach based on the use of a Mistral 7B model obtains a Hierarchical F1 of 0.42 a Hierarchical Precision of 0.30 and a Hierarchical Recall of 0.71. Our selected approach is able to outperform the baseline provided for the competition.

pdf bib
Mashee at SemEval-2024 Task 8: The Impact of Samples Quality on the Performance of In-Context Learning for Machine Text Classification
Areeg Fahad Rasheed | M. Zarkoosh

Within few-shot learning, in-context learning(ICL) has become a potential method for lever-aging contextual information to improve modelperformance on small amounts of data or inresource-constrained environments where train-ing models on large datasets is prohibitive.However, the quality of the selected samplein a few shots severely limits the usefulnessof ICL. The primary goal of this paper is toenhance the performance of evaluation metricsfor in-context learning by selecting high-qualitysamples in few-shot learning scenarios. We em-ploy the chi-square test to identify high-qualitysamples and compare the results with those ob-tained using low-quality samples. Our findingsdemonstrate that utilizing high-quality samplesleads to improved performance with respect toall evaluated metrics.

pdf bib
Puer at SemEval-2024 Task 4: Fine-tuning Pre-trained Language Models for Meme Persuasion Technique Detection
Jiaxu Dao | Zhuoying Li | Youbang Su | Wensheng Gong

The paper summarizes our research on multilingual detection of persuasion techniques in memes for the SemEval-2024 Task 4. Our work focused on English-Subtask 1, implemented based on a roberta-large pre-trained model provided by the transforms tool that was fine-tuned into a corpus of social media posts. Our method significantly outperforms the officially released baseline method, and ranked 7th in English-Subtask 1 for the test set. This paper also compares the performances of different deep learning model architectures, such as BERT, ALBERT, and XLM-RoBERTa, on multilingual detection of persuasion techniques in memes. The experimental source code covered in the paper will later be sourced from Github.

pdf bib
Puer at SemEval-2024 Task 2: A BioLinkBERT Approach to Biomedical Natural Language Inference
Jiaxu Dao | Zhuoying Li | Xiuzhong Tang | Xiaoli Lan | Junde Wang

This paper delineates our investigation into the application of BioLinkBERT for enhancing clinical trials, presented at SemEval-2024 Task 2. Centering on the medical biomedical NLI task, our approach utilized the BioLinkBERT-large model, refined with a pioneering mixed loss function that amalgamates contrastive learning and cross-entropy loss. This methodology demonstrably surpassed the established benchmark, securing an impressive F1 score of 0.72 and positioning our work prominently in the field. Additionally, we conducted a comparative analysis of various deep learning architectures, including BERT, ALBERT, and XLM-RoBERTa, within the context of medical text mining. The findings not only showcase our method’s superior performance but also chart a course for future research in biomedical data processing. Our experiment source code is available on GitHub at: https://github.com/daojiaxu/semeval2024_task2.

pdf bib
NRK at SemEval-2024 Task 1: Semantic Textual Relatedness through Domain Adaptation and Ensemble Learning on BERT-based models
Nguyen Tuan Kiet | Dang Van Thin

This paper describes the system of the team NRK for Task A in the SemEval-2024 Task 1: Semantic Textual Relatedness (STR). We focus on exploring the performance of ensemble architectures based on the voting technique and different pre-trained transformer-based language models, including the multilingual and monolingual BERTology models. The experimental results show that our system has achieved competitive performance in some languages in Track A: Supervised, where our submissions rank in the Top 3 and Top 4 for Algerian Arabic and Amharic languages. Our source code is released on the GitHub site.

pdf bib
BrainLlama at SemEval-2024 Task 6: Prompting Llama to detect hallucinations and related observable overgeneration mistakes
Marco Siino

Participants in the SemEval-2024 Task 6 were tasked with executing binary classification aimed at discerning instances of fluent overgeneration hallucinations across two distinct setups: the model-aware and model-agnostic tracks. That is, participants must detect grammatically sound outputs which contain incorrect or unsupported semantic information, regardless of whether they had access to the model responsible for producing the output or not, within the model-aware and model-agnostic tracks. Two tracks were proposed for the task: a model-aware track, where organizers provided a checkpoint to a model publicly available on HuggingFace for every data point considered, and a model-agnostic track where the organizers do not. In this paper, we discuss the application of a Llama model to address both the tracks. Find the persuasive strategy that a meme employs from a hierarchy of twenty based just on its “textual content.” Only a portion of the reward is awarded if the technique’s ancestor node is chosen. This classification issue is multilabel hierarchical. Our approach reaches an accuracy of 0.62 on the agnostic track and of 0.67 on the aware track.

pdf bib
DKE-Research at SemEval-2024 Task 2: Incorporating Data Augmentation with Generative Models and Biomedical Knowledge to Enhance Inference Robustness
Yuqi Wang | Zeqiang Wang | Wei Wang | Qi Chen | Kaizhu Huang | Anh Nguyen | Suparna De

Safe and reliable natural language inference is critical for extracting insights from clinical trial reports but poses challenges due to biases in large pre-trained language models. This paper presents a novel data augmentation technique to improve model robustness for biomedical natural language inference in clinical trials. By generating synthetic examples through semantic perturbations and domain-specific vocabulary replacement and adding a new task for numerical and quantitative reasoning, we introduce greater diversity and reduce shortcut learning. Our approach, combined with multi-task learning and the DeBERTa architecture, achieved significant performance gains on the NLI4CT 2024 benchmark compared to the original language models. Ablation studies validate the contribution of each augmentation method in improving robustness. Our best-performing model ranked 12th in terms of faithfulness and 8th in terms of consistency, respectively, out of the 32 participants.

pdf bib
SATLab at SemEval-2024 Task 1: A Fully Instance-Specific Approach for Semantic Textual Relatedness Prediction
Yves Bestgen

This paper presents the SATLab participation in SemEval 2024 Task 1 on Semantic Textual Relatedness. The proposed system predicts semantic relatedness by means of the Euclidean distance between the character ngram frequencies in the two sentences to evaluate. It employs no external resources, nor information from other instances present in the material. The system performs well, coming first in five of the twelve languages. However, there is little difference between the best systems.

pdf bib
Genaios at SemEval-2024 Task 8: Detecting Machine-Generated Text by Mixing Language Model Probabilistic Features
Areg Mikael Sarvazyan | José Ángel González | Marc Franco-salvador

This paper describes the participation of the Genaios team in the monolingual track of Subtask A at SemEval-2024 Task 8. Our best system, LLMixtic, is a Transformer Encoder that mixes token-level probabilistic features extracted from four LLaMA-2 models. We obtained the best results in the official ranking (96.88% accuracy), showing a false positive ratio of 4.38% and a false negative ratio of 1.97% on the test set. We further study LLMixtic through ablation, probabilistic, and attention analyses, finding that (i) performance improves as more LLMs and probabilistic features are included, (ii) LLMixtic puts most attention on the features of the last tokens, (iii) it fails on samples where human text probabilities become consistently higher than for generated text, and (iv) LLMixtic’s false negatives exhibit a bias towards text with newlines.

pdf bib
Self-StrAE at SemEval-2024 Task 1: Making Self-Structuring AutoEncoders Learn More With Less
Mattia Opper | Siddharth Narayanaswamy

We present two simple improvements to the Self-Structuring AutoEncoder (Self-StrAE). Firstly, we show that including reconstruction to the vocabulary as an auxiliary objective improves representation quality. Secondly, we demonstrate that increasing the number of independent channels leads to significant improvements in embedding quality, while simultaneously reducing the number of parameters. Surprisingly, we demonstrate that this trend can be followed to the extreme, even to point of reducing the total number of non-embedding parameters to seven. Our system can be pre-trained from scratch with as little as 10M tokens of input data, and proves effective across English, Spanish and Afrikaans.

pdf bib
RGAT at SemEval-2024 Task 2: Biomedical Natural Language Inference using Graph Attention Network
Abir Chakraborty

In this work, we (team RGAT) describe our approaches for the SemEval 2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials (NLI4CT). The objective of this task is multi-evidence natural language inference based on different sections of clinical trial reports. We have explored various approaches, (a) dependency tree of the input query as additional features in a Graph Attention Network (GAT) along with the token and parts-of-speech features, (b) sequence-to-sequence approach using various models and synthetic data and finally, (c) in-context learning using large language models (LLMs) like GPT-4. Amongs these three approaches the best result is obtained from the LLM with 0.76 F1-score (the highest being 0.78), 0.86 in faithfulness and 0.74 in consistence.

pdf bib
BDA at SemEval-2024 Task 4: Detection of Persuasion in Memes Across Languages with Ensemble Learning and External Knowledge
Victoria Sherratt | Sedat Dogan | Ifeoluwa Wuraola | Lydia Bryan-smith | Oyinkansola Onwuchekwa | Nina Dethlefs

This paper outlines our multimodal ensemble learning system for identifying persuasion techniques in memes. We contribute an approach which utilises the novel inclusion of consistent named visual entities extracted using Google Vision’s API as an external knowledge source, joined to our multimodal ensemble via late fusion. As well as detailing our experiments in ensemble combinations, fusion methods and data augmentation, we explore the impact of including external data and summarise post-evaluation improvements to our architecture based on analysis of the task results.

pdf bib
nowhash at SemEval-2024 Task 4: Exploiting Fusion of Transformers for Detecting Persuasion Techniques in Multilingual Memes
Abu Nowhash Chowdhury | Michal Ptaszynski

Nowadays, memes are considered one of the most prominent forms of medium to disseminate information on social media. Memes are typically constructed in multilingual settings using visuals with texts. Sometimes people use memes to influence mass audiences through rhetorical and psychological techniques, such as causal oversimplification, name-calling, and smear. It is a challenging task to identify those techniques considering memes’ multimodal characteristics. To address these challenges, SemEval-2024 Task 4 introduced a shared task focusing on detecting persuasion techniques in multilingual memes. This paper presents our participation in subtasks 1 and 2(b). We use a finetuned language-agnostic BERT sentence embedding (LaBSE) model to extract effective contextual features from meme text to address the challenge of identifying persuasion techniques in subtask 1. For subtask 2(b), We finetune the vision transformer and XLM-RoBERTa to extract effective contextual information from meme image and text data. Finally, we unify those features and employ a single feed-forward linear layer on top to obtain the prediction label. Experimental results on the SemEval 2024 Task 4 benchmark dataset manifested the potency of our proposed methods for subtasks 1 and 2(b).

pdf bib
HalluSafe at SemEval-2024 Task 6: An NLI-based Approach to Make LLMs Safer by Better Detecting Hallucinations and Overgeneration Mistakes
Zahra Rahimi | Hamidreza Amirzadeh | Alireza Sohrabi | Zeinab Taghavi | Hossein Sameti

The advancement of large language models (LLMs), their ability to produce eloquent and fluent content, and their vast knowledge have resulted in their usage in various tasks and applications. Despite generating fluent content, this content can contain fabricated or false information. This problem is known as hallucination and has reduced the confidence in the output of LLMs. In this work, we have used Natural Language Inference to train classifiers for hallucination detection to tackle SemEval-2024 Task 6-SHROOM (Mickus et al., 2024) which is defined in three sub-tasks: Paraphrase Generation, Machine Translation, and Definition Modeling. We have also conducted experiments on LLMs to evaluate their ability to detect hallucinated outputs. We have achieved 75.93% and 78.33% accuracy for the modelaware and model-agnostic tracks, respectively. The shared links of our models and the codes are available on GitHub.

pdf bib
NIMZ at SemEval-2024 Task 9: Evaluating Methods in Solving Brainteasers Defying Commonsense
Zahra Rahimi | Mohammad Moein Shirzady | Zeinab Taghavi | Hossein Sameti

The goal and dream of the artificial intelligence field have long been the development of intelligent systems or agents that mimic human behavior and thinking. Creativity is an essential trait in humans that is closely related to lateral thinking. The remarkable advancements in Language Models have led to extensive research on question-answering and explicit and implicit reasoning involving vertical thinking. However, there is an increasing need to shift focus towards research and development of models that can think laterally. One must step outside the traditional frame of commonsense concepts in lateral thinking to conclude. Task 9 of SemEval-2024 is Brainteaser (Jiang et al.,2024), which requires lateral thinking to answer riddle-like multiple-choice questions. In our study, we assessed the performance of various models for the Brainteaser task. We achieved an overall accuracy of 75% for the Sentence Puzzle subtask and 66.7% for the Word Puzzle subtask. All the codes, along with the links to our saved models, are available on our GitHub.

pdf bib
Mistral at SemEval-2024 Task 5: Mistral 7B for argument reasoning in Civil Procedure
Marco Siino

At the SemEval-2024 Task 5, the organizers introduce a novel natural language processing (NLP) challenge and dataset within the realm of the United States civil procedure. Each datum within the dataset comprises a comprehensive overview of a legal case, a specific inquiry associated with it, and a potential argument in support of a solution, supplemented with an in-depth rationale elucidating the applicability of the argument within the given context. Derived from a text designed for legal education purposes, this dataset presents a multifaceted benchmarking task for contemporary legal language models. Our manuscript delineates the approach we adopted for participation in this competition. Specifically, we detail the use of a Mistral 7B model to answer the question provided. Our only and best submission reach an F1-score equal to 0.5597 and an Accuracy of 0.5714, outperforming the baseline provided for the task.

pdf bib
NCL-UoR at SemEval-2024 Task 8: Fine-tuning Large Language Models for Multigenerator, Multidomain, and Multilingual Machine-Generated Text Detection
Feng Xiong | Thanet Markchom | Ziwei Zheng | Subin Jung | Varun Ojha | Huizhi Liang

SemEval-2024 Task 8 introduces the challenge of identifying machine-generated texts from diverse Large Language Models (LLMs) in various languages and domains. The task comprises three subtasks: binary classification in monolingual and multilingual (Subtask A), multi-class classification (Subtask B), and mixed text detection (Subtask C). This paper focuses on Subtask A & B. To tackle this task, this paper proposes two methods: 1) using traditional machine learning (ML) with natural language preprocessing (NLP) for feature extraction, and 2) fine-tuning LLMs for text classification. For fine-tuning, we use the train datasets provided by the task organizers. The results show that transformer models like LoRA-RoBERTa and XLM-RoBERTa outperform traditional ML models, particularly in multilingual subtasks. However, traditional ML models performed better than transformer models for the monolingual task, demonstrating the importance of considering the specific characteristics of each subtask when selecting an appropriate approach.

pdf bib
iML at SemEval-2024 Task 2: Safe Biomedical Natural Language Interference for Clinical Trials with LLM Based Ensemble Inferencing
Abbas Akkasi | Adnan Khan | Mai A. Shaaban | Majid Komeili | Mohammad Yaqub

We engaged in the shared task 2 at SenEval-2024, employing a diverse set of solutions with a particular emphasis on leveraging a Large Language Model (LLM) based zero-shot inference approach to address the challenge.

pdf bib
CLaC at SemEval-2024 Task 4: Decoding Persuasion in Memes – An Ensemble of Language Models with Paraphrase Augmentation
Kota Shamanth Ramanath Nayak | Leila Kosseim

This paper describes our approach to SemEval-2024 Task 4 subtask 1, focusing on hierarchical multi-label detection of persuasion techniques in meme texts. Our approach was based on fine-tuning individual language models (BERT, XLM-RoBERTa, and mBERT) and leveraging a mean-based ensemble model. Additional strategies included dataset augmentation through the TC dataset and paraphrase generation as well as the fine-tuning of individual classification thresholds for each class. During testing, our system outperformed the baseline in all languages except for Arabic, where no significant improvement was reached. Analysis of the results seem to indicate that our dataset augmentation strategy and per-class threshold fine-tuning may have introduced noise and exacerbated the dataset imbalance.

pdf bib
RDproj at SemEval-2024 Task 4: An Ensemble Learning Approach for Multilingual Detection of Persuasion Techniques in Memes
Yuhang Zhu

This paper introduces our bagging-based ensemble learning approach for the SemEval-2024 Task 4 Subtask 1, focusing on multilingual persuasion detection within meme texts. This task aims to identify persuasion techniques employed within meme texts, which is a hierarchical multilabel classification task. The given text may apply multiple techniques, and persuasion techniques have a hierarchical structure. However, only a few prior persuasion detection systems have utilized the hierarchical structure of persuasion techniques. In that case, we designed a multilingual bagging-based ensemble approach, incorporating a soft voting ensemble strategy to effectively exploit persuasion techniques’ hierarchical structure. Our methodology achieved the second position in Bulgarian and North Macedonian, third in Arabic, and eleventh in English.

pdf bib
HausaNLP at SemEval-2024 Task 1: Textual Relatedness Analysis for Semantic Representation of Sentences
Saheed Abdullahi Salahudeen | Falalu Ibrahim Lawan | Yusuf Aliyu | Amina Abubakar | Lukman Aliyu | Nur Rabiu | Mahmoud Ahmad | Aliyu Rabiu Shuaibu | Alamin Musa

Semantic Text Relatedness (STR), a measure of meaning similarity between text elements, has become a key focus in the field of Natural Language Processing (NLP). We describe SemEval-2024 task 1 on Semantic Textual Relatedness featuring three tracks: supervised learning, unsupervised learning and cross-lingual learning across African and Asian languages including Afrikaans, Algerian Arabic, Amharic, Hausa, Hindi, Indonesian, Kinyarwanda, Marathi, Moroccan Arabic, Modern Standard Arabic, Punjabi, Spanish, and Telugu. Our goal is to analyse the semantic representation of sentences textual relatedness trained on mBert, all-MiniLM-L6-v2 and Bert-Based-uncased. The effectiveness of these models is evaluated using the Spearman Correlation metric, which assesses the strength of the relationship between paired data. The finding reveals the viability of transformer models in multilingual STR tasks.

pdf bib
SCaLAR NITK at SemEval-2024 Task 5: Towards Unsupervised Question Answering system with Multi-level Summarization for Legal Text
Manvith Prabhu | Haricharana Srinivasa | Anand Kumar

This paper summarizes Team SCaLAR’s work on SemEval-2024 Task 5: Legal Argument Reasoning in Civil Procedure. To address this Binary Classification task, which was daunting due to the complexity of the Legal Texts involved, we propose a simple yet novel similarity and distance-based unsupervised approach to generate labels. Further, we explore the Multi-level fusion of Legal-Bert embeddings using ensemble features, including CNN, GRU, and LSTM. To address the lengthy nature of Legal explanation in the dataset, we introduce T5-based segment-wise summarization, which successfully retained crucial information, enhancing the model’s performance. Our unsupervised system witnessed a 20-point increase in macro F1-score on the development set and a 10-point increase on the test set, which is promising given its uncomplicated architecture.

pdf bib
Abdelhak at SemEval-2024 Task 9: Decoding Brainteasers, The Efficacy of Dedicated Models Versus ChatGPT
Abdelhak Kelious | Mounir Okirim

This study introduces a dedicated model aimed at solving the BRAINTEASER task 9 , a novel challenge designed to assess models’ lateral thinking capabilities through sentence and word puzzles. Our model demonstrates remarkable efficacy, securing Rank 1 in sentence puzzle solving during the test phase with an overall score of 0.98. Additionally, we explore the comparative performance of ChatGPT, specifically analyzing how variations in temperature settings affect its ability to engage in lateral thinking and problem-solving. Our findings indicate a notable performance disparity between the dedicated model and ChatGPT, underscoring the potential of specialized approaches in enhancing creative reasoning in AI.

pdf bib
OUNLP at SemEval-2024 Task 9: Retrieval-Augmented Generation for Solving Brain Teasers with LLMs
Vineet Saravanan | Steven Wilson

The advancement of natural language processing has given rise to a variety of large language models (LLMs) with capabilities extending into the realm of complex problem-solving, including brainteasers that challenge not only linguistic fluency but also logical reasoning. This paper documents our submission to the SemEval 2024 Brainteaser task, in which we investigate the performance of state-of-the-art LLMs, such as GPT-3.5, GPT-4, and the Gemini model, on a diverse set of brainteasers using prompt engineering as a tool to enhance the models’ problem-solving abilities. We experimented with a series of structured prompts ranging from basic to those integrating task descriptions and explanations. Through a comparative analysis, we sought to determine which combinations of model and prompt yielded the highest accuracy in solving these puzzles. Our findings provide a snapshot of the current landscape of AI problem-solving and highlight the nuanced nature of LLM performance, influenced by both the complexity of the tasks and the sophistication of the prompts employed.

pdf bib
NLP-LISAC at SemEval-2024 Task 1: Transformer-based approaches for Determining Semantic Textual Relatedness
Abdessamad Benlahbib | Anass Fahfouh | Hamza Alami | Achraf Boumhidi

This paper presents our system and findings for SemEval 2024 Task 1 Track A Supervised Semantic Textual Relatedness. The main objective of this task was to detect the degree of semantic relatedness between pairs of sentences. Our submitted models (ranked 6/24 in Algerian Arabic, 7/25 in Spanish, 12/23 in Moroccan Arabic, and 13/36 in English) consist of various transformer-based models including MARBERT-V2, mDeBERTa-V3-Base, DarijaBERT, and DeBERTa-V3-Large, fine-tuned using different loss functions including Huber Loss, Mean Absolute Error, and Mean Squared Error.

pdf bib
ZXQ at SemEval-2024 Task 7: Fine-tuning GPT-3.5-Turbo for Numerical Reasoning
Zhen Qian | Xiaofei Xu | Xiuzhen Zhang

In this paper, we present our system for the SemEval-2024 Task 7, i.e., NumEval subtask 3: Numericial Reasoning. Given a news article and its headline, the numerical reasoning task involves creating a system to compute the intentionally excluded number within the news headline. We propose a fine-tuned GPT-3.5-turbo model, specifically engineered to deduce missing numerals directly from the content of news article. The model is trained with a human-engineered prompt that itegrates the news content and the masked headline, tailoring its accuracy for the designated task. It achieves an accuracy of 0.94 on the test data and secures the second position in the official leaderboard. An examination on the system’s inference results reveals its commendable accuracy in identifying correct numerals when they can be directly “copied” from the articles. However, the error rates increase when it comes to some ambiguous operations such as rounding.

pdf bib
BAMO at SemEval-2024 Task 9: BRAINTEASER: A Novel Task Defying Common Sense
Baktash Ansari | Mohammadmostafa Rostamkhani | Sauleh Eetemadi

This paper outlines our approach to SemEval 2024 Task 9, BRAINTEASER: A Novel Task Defying Common Sense. The task aims to evaluate the ability of language models to think creatively. The dataset comprises multi-choice questions that challenge models to think ‘outside of the box’. We fine-tune 2 models, BERT and RoBERTa Large. Next, we employ a Chain of Thought (CoT) zero-shot prompting approach with 6 large language models, such as GPT-3.5, Mixtral, and Llama2. Finally, we utilize ReConcile, a technique that employs a ‘round table conference’ approach with multiple agents for zero-shot learning, to generate consensus answers among 3 selected language models. Our best method achieves an overall accuracy of 85 percent on the sentence puzzles subtask.

pdf bib
yangqi at SemEval-2024 Task 9: Simulate Human Thinking by Large Language Model for Lateral Thinking Challenges
Qi Yang | Jingjie Zeng | Liang Yang | Hongfei Lin

This paper describes our system used in the SemEval-2024 Task 9 on two sub-tasks, BRAINTEASER: A Novel Task Defying Common Sense. In this work, we developed a system SHTL, which means simulate human thinking capabilities by Large Language Model (LLM). Our approach bifurcates into two main components: Common Sense Reasoning and Rationalize Defying Common Sense. To mitigate the hallucinations of LLM, we implemented a strategy that combines Retrieval-augmented Generation (RAG) with the the Self-Adaptive In-Context Learning (SAICL), thereby sufficiently leveraging the powerful language ability of LLM. The effectiveness of our method has been validated by its performance on the test set, with an average performance on two subtasks that is 30.1 higher than ChatGPT setting zero-shot and only 0.8 lower than that of humans.

pdf bib
BadRock at SemEval-2024 Task 8: DistilBERT to Detect Multigenerator, Multidomain and Multilingual Black-Box Machine-Generated Text
Marco Siino

The rise of Large Language Models (LLMs) has brought about a notable shift, rendering them increasingly ubiquitous and readily accessible. This accessibility has precipitated a surge in machine-generated content across diverse platforms encompassing news outlets, social media platforms, question-answering forums, educational platforms, and even academic domains. Recent iterations of LLMs, exemplified by entities like ChatGPT and GPT-4, exhibit a remarkable ability to produce coherent and contextually relevant responses across a broad spectrum of user inquiries. The fluidity and sophistication of these generated texts position LLMs as compelling candidates for substituting human labor in numerous applications. Nevertheless, this proliferation of machine-generated content has raised apprehensions regarding potential misuse, including the dissemination of misinformation and disruption of educational ecosystems. Given that humans marginally outperform random chance in discerning between machine-generated and human-authored text, there arises a pressing imperative to develop automated systems capable of accurately distinguishing machine-generated text. This pursuit is driven by the overarching objective of curbing the potential misuse of machine-generated content. Our manuscript delineates the approach we adopted for participation in this competition. Specifically, we detail the use of a DistilBERT model for classifying each sample in the test set provided. Our submission is able to reach an accuracy equal to 0.754 in place of the worst result obtained at the competition that is equal to 0.231.

pdf bib
WarwickNLP at SemEval-2024 Task 1: Low-Rank Cross-Encoders for Efficient Semantic Textual Relatedness
Fahad Ebrahim | Mike Joy

This work participates in SemEval 2024 Task 1 on Semantic Textural Relatedness (STR) in Track A (supervised regression) in two languages, English and Moroccan Arabic. The task consists of providing a score of how two sentences relate to each other. The system developed in this work leveraged a cross-encoder with a merged fine-tuned Low-Rank Adapter (LoRA). The system was ranked eighth in English with a Spearman coefficient of 0.842, while Moroccan Arabic was ranked seventh with a score of 0.816. Moreover, various experiments were conducted to see the impact of different models and adapters on the performance and accuracy of the system.

pdf bib
NU-RU at SemEval-2024 Task 6: Hallucination and Related Observable Overgeneration Mistake Detection Using Hypothesis-Target Similarity and SelfCheckGPT
Thanet Markchom | Subin Jung | Huizhi Liang

One of the key challenges in Natural Language Generation (NLG) is “hallucination,” in which the generated output appears fluent and grammatically sound but may contain incorrect information. To address this challenge, “SemEval-2024 Task 6 - SHROOM, a Shared-task on Hallucinations and Related Observable Overgeneration Mistakes” is introduced. This task focuses on detecting overgeneration hallucinations in texts generated from Large Language Models for various NLG tasks. To tackle this task, this paper proposes two methods: (1) hypothesis-target similarity, which measures text similarity between a generated text (hypothesis) and an intended reference text (target), and (2) a SelfCheckGPT-based method to assess hallucinations via predefined prompts designed for different NLG tasks. Experiments were conducted on the dataset provided in this task. The results show that both of the proposed methods can effectively detect hallucinations in LLM-generated texts with a possibility for improvement.

pdf bib
NCL_NLP at SemEval-2024 Task 7: CoT-NumHG: A CoT-Based SFT Training Strategy with Large Language Models for Number-Focused Headline Generation
Junzhe Zhao | Yingxi Wang | Huizhi Liang | Nicolay Rusnachenko

Headline Generation is an essential task in Natural Language Processing (NLP), where models often exhibit limited ability to accurately interpret numerals, leading to inaccuracies in generated headlines. This paper introduces CoT-NumHG, a training strategy leveraging the Chain of Thought (CoT) paradigm for Supervised Fine-Tuning (SFT) of large language models. This approach is aimed at enhancing numeral perception, interpretability, accuracy, and the generation of structured outputs. Presented in SemEval-2024 Task 7 (task 3): Numeral-Aware Headline Generation (English), this challenge is divided into two specific subtasks. The first subtask focuses on numerical reasoning, requiring models to precisely calculate and fill in the missing numbers in news headlines, while the second subtask targets the generation of complete headlines. Utilizing the same training strategy across both subtasks, this study primarily explores the first subtask as a demonstration of our training strategy. Through this competition, our CoT-NumHG-Mistral-7B model attained an accuracy rate of 94%, underscoring the effectiveness of our proposed strategy.

pdf bib
Byun at SemEval-2024 Task 6: Text Classification on Hallucinating Text with Simple Data Augmentation
Cheolyeon Byun

This paper aims to classify sentences to see if it is hallucinating, meaning the generative language model has output text that has very little to do with the user’s input, or not. This classification task is part of the Semeval 2024’s task on Hallucinations and Related Observable Over-generation Mistakes, AKA SHROOM, which aims to improve awkward-sounding texts generated by AI. This paper will first go over the first attempt at creating predictions, then show the actual scores achieved after submitting the first attempt results to Semeval, then finally go over potential improvements to be made.

pdf bib
DeepPavlov at SemEval-2024 Task 6: Detection of Hallucinations and Overgeneration Mistakes with an Ensemble of Transformer-based Models
Ivan Maksimov | Vasily Konovalov | Andrei Glinskii

The inclination of large language models (LLMs) to produce mistaken assertions, known as hallucinations, can be problematic. These hallucinations could potentially be harmful since sporadic factual inaccuracies within the generated text might be concealed by the overall coherence of the content, making it immensely challenging for users to identify them. The goal of the SHROOM shared-task is to detect grammatically sound outputs that contain incorrect or unsupported semantic information. Although there are a lot of existing hallucination detectors in generated AI content, we found out that pretrained Natural Language Inference (NLI) models yet exhibit success in detecting hallucinations. Moreover their ensemble outperforms more complicated models.

pdf bib
HIJLI_JU at SemEval-2024 Task 7: Enhancing Quantitative Question Answering Using Fine-tuned BERT Models
Partha Sengupta | Sandip Sarkar | Dipankar Das

In data and numerical analysis, Quantitative Question Answering (QQA) becomes a crucial instrument that provides deep insights for analyzing large datasets and helps make well-informed decisions in industries such as finance, healthcare, and business. This paper explores the “HIJLI_JU” team’s involvement in NumEval Task 1 within SemEval 2024, with a particular emphasis on quantitative comprehension. Specifically, our method addresses numerical complexities by fine-tuning a BERT model for sophisticated multiple-choice question answering, leveraging the Hugging Face ecosystem. The effectiveness of our QQA model is assessed using a variety of metrics, with an emphasis on the f1_score() of the scikit-learn library. Thorough analysis of the macro-F1, micro-F1, weighted-F1, average, and binary-F1 scores yields detailed insights into the model’s performance in a range of question formats.

pdf bib
NCL Team at SemEval-2024 Task 3: Fusing Multimodal Pre-training Embeddings for Emotion Cause Prediction in Conversations
Shu Li | Zicen Liao | Huizhi Liang

In this study, we introduce an MLP approach for extracting multimodal cause utterances in conversations, utilizing the multimodal conversational emotion causes from the ECF dataset. Our research focuses on evaluating a bi-modal framework that integrates video and audio embeddings to analyze emotional expressions within dialogues. The core of our methodology involves the extraction of embeddings from pre-trained models for each modality, followed by their concatenation and subsequent classification via an MLP network. We compared the accuracy performances across different modality combinations including text-audio-video, video-audio, and audio only.

pdf bib
DeBERTa at SemEval-2024 Task 9: Using DeBERTa for Defying Common Sense
Marco Siino

The widespread success of language models has spurred the natural language processing (NLP) community to tackle tasks demanding implicit and intricate reasoning, drawing upon human-like common-sense mechanisms. While endeavors in vertical thinking tasks have garnered considerable attention, there has been a relative dearth of exploration in lateral thinking puzzles. To address this gap, we introduce BRAINTEASER: a multiple-choice Question Answering task meticulously crafted to evaluate the model’s capacity for lateral thinking and its ability to challenge default common-sense associations. At the SemEval-2024 Task 9, for the first subtask (i.e., Sentence Puzzle) the organizers asked the participants to develop models able to reply to multi-answer brain-teasing questions. For this purpose, we propose the application of a DeBERTa model in a zero-shot configuration. Our proposed approach is able to reach an overall score of 0.250. Suggesting a significant room for improvements in future works.

pdf bib
TransMistral at SemEval-2024 Task 10: Using Mistral 7B for Emotion Discovery and Reasoning its Flip in Conversation
Marco Siino

The EDiReF shared task at SemEval 2024 comprises three subtasks: Emotion Recognition in Conversation (ERC) in Hindi-English code-mixed conversations, Emotion Flip Reasoning (EFR) in Hindi-English code-mixed conversations, and EFR in English conversations. The objectives for the ERC and EFR tasks are defined as follows: 1) Emotion Recognition in Conversation (ERC): In this task, participants are tasked with assigning an emotion to each utterance within a dialogue from a predefined set of possible emotions. The goal is to accurately recognize and label the emotions expressed in the conversation; 2) Emotion Flip Reasoning (EFR): This task involves identifying the trigger utterance(s) for an emotion-flip within a multi-party conversation dialogue. Participants are required to pinpoint the specific utterance(s) that serve as catalysts for a change in emotion during the conversation. In this paper we only address the first subtask (ERC) making use of an online translation strategy followed by the application of a Mistral 7B model together with a few-shot prompt strategy. Our approach obtains an F1 of 0.36, eventually exhibiting further room for improvements.

pdf bib
0x.Yuan at SemEval-2024 Task 2: Agents Debating can reach consensus and produce better outcomes in Medical NLI task
Yu-an Lu | Hung-yu Kao

In this paper, we introduce a multi-agent debating framework, experimenting on SemEval 2024 Task 2. This innovative system employs a collaborative approach involving expert agents from various medical fields to analyze Clinical Trial Reports (CTRs). Our methodology emphasizes nuanced and comprehensive analysis by leveraging the diverse expertise of agents like Biostatisticians and Medical Linguists. Results indicate that our collaborative model surpasses the performance of individual agents in terms of Macro F1-score. Additionally, our analysis suggests that while initial debates often mirror majority decisions, the debating process refines these outcomes, demonstrating the system’s capability for in-depth analysis beyond simple majority rule. This research highlights the potential of AI collaboration in specialized domains, particularly in medical text interpretation.

pdf bib
TW-NLP at SemEval-2024 Task10: Emotion Recognition and Emotion Reversal Inference in Multi-Party Dialogues.
Wei Tian | Peiyu Ji | Lei Zhang | Yue Jian

In multidimensional dialogues, emotions serve not only as crucial mediators of emotional exchanges but also carry rich information. Therefore, accurately identifying the emotions of interlocutors and understanding the triggering factors of emotional changes are paramount. This study focuses on the tasks of multilingual dialogue emotion recognition and emotion reversal reasoning based on provocateurs, aiming to enhance the accuracy and depth of emotional understanding in dialogues. To achieve this goal, we propose a novel model, MBERT-TextRCNN-PL, designed to effectively capture emotional information of interlocutors. Additionally, we introduce XGBoost-EC (Emotion Capturer) to identify emotion provocateurs, thereby delving deeper into the causal relationships behind emotional changes. By comparing with state-of-the-art models, our approach demonstrates significant improvements in recognizing dialogue emotions and provocateurs, offering new insights and methodologies for multilingual dialogue emotion understanding and emotion reversal research.

pdf bib
UWBA at SemEval-2024 Task 3: Dialogue Representation and Multimodal Fusion for Emotion Cause Analysis
Josef Baloun | Jiri Martinek | Ladislav Lenc | Pavel Kral | Matěj Zeman | Lukáš Vlček

In this paper, we present an approach for solving SemEval-2024 Task 3: The Competition of Multimodal Emotion Cause Analysis in Conversations. The task includes two subtasks that focus on emotion-cause pair extraction using text, video, and audio modalities. Our approach is composed of encoding all modalities (MFCC and Wav2Vec for audio, 3D-CNN for video, and transformer-based models for text) and combining them in an utterance-level fusion module. The model is then optimized for link and emotion prediction simultaneously. Our approach achieved 6th place in both subtasks. The full leaderboard can be found at https://codalab.lisn.upsaclay.fr/competitions/16141#results

pdf bib
GAVx at SemEval-2024 Task 10: Emotion Flip Reasoning via Stacked Instruction Finetuning of LLMs
Vy Nguyen | Xiuzhen Zhang

The Emotion Flip Reasoning task at SemEval 2024 aims at identifying the utterance(s) that trigger a speaker to shift from an emotion to another in a multi-party conversation. The spontaneous, informal, and occasionally multilingual dynamics of conversations make the task challenging. In this paper, we propose a supervised stacked instruction-based framework to finetune large language models to tackle this task. Utilising the annotated datasets provided, we curate multiple instruction sets involving chain-of-thoughts, feedback, and self-evaluation instructions, for a multi-step finetuning pipeline. We utilise the self-consistency inference strategy to enhance prediction consistency. Experimental results reveal commendable performance, achieving mean F1 scores of 0.77 and 0.76 for triggers in the Hindi-English and English-only tracks respectively. This led to us earning the second highest ranking in both tracks.

pdf bib
NLP_STR_teamS at SemEval-2024 Task1: Semantic Textual Relatedness based on MASK Prediction and BERT Model
Lianshuang Su | Xiaobing Zhou

This paper describes our participation in the SemEval-2024 Task 1, “Semantic Textual Relatedness for African and Asian Languages.” This task detects the degree of semantic relatedness between pairs of sentences. Our approach is to take out the sentence pairs of each instance to construct a new sentence as the prompt template, use MASK to predict the correlation between the two sentences, use the BERT pre-training model to process and calculate the text sequence, and use the synonym replacement method in text data augmentation to expand the size of the data set. We participate in English in track A, which uses a supervised approach, and the Spearman Correlation on the test set is 0.809.

pdf bib
Halu-NLP at SemEval-2024 Task 6: MetaCheckGPT - A Multi-task Hallucination Detection using LLM uncertainty and meta-models
Rahul Mehta | Andrew Hoblitzell | Jack O’keefe | Hyeju Jang | Vasudeva Varma

Hallucinations in large language models(LLMs) have recently become a significantproblem. A recent effort in this directionis a shared task at Semeval 2024 Task 6,SHROOM, a Shared-task on Hallucinationsand Related Observable Overgeneration Mis-takes. This paper describes our winning so-lution ranked 1st and 2nd in the 2 sub-tasksof model agnostic and model aware tracks re-spectively. We propose a meta-regressor basedensemble of LLMs based on a random forestalgorithm that achieves the highest scores onthe leader board. We also experiment with var-ious transformer based models and black boxmethods like ChatGPT, Vectara, and others. Inaddition, we perform an error analysis com-paring ChatGPT against our best model whichshows the limitations of the former

pdf bib
QFNU_CS at SemEval-2024 Task 3: A Hybrid Pre-trained Model based Approach for Multimodal Emotion-Cause Pair Extraction Task
Zining Wang | Yanchao Zhao | Guanghui Han | Yang Song

This article presents the solution of Qufu Normal University for the Multimodal Sentiment Cause Analysis competition in SemEval2024 Task 3.The competition aims to extract emotion-cause pairs from dialogues containing text, audio, and video modalities. To cope with this task, we employ a hybrid pre-train model based approach. Specifically, we first extract and fusion features from dialogues based on BERT, BiLSTM, openSMILE and C3D. Then, we adopt BiLSTM and Transformer to extract the candidate emotion-cause pairs. Finally, we design a filter to identify the correct emotion-cause pairs. The evaluation results show that, we achieve a weighted average F1 score of 0.1786 and an F1 score of 0.1882 on CodaLab.

pdf bib
NewbieML at SemEval-2024 Task 8: Ensemble Approach for Multidomain Machine-Generated Text Detection
Bao Tran | Nhi Tran

Large Language Models (LLMs) are becoming popular and easily accessible, leading to a large growth of machine-generated content over various channels. Along with this popularity, the potential misuse is also a challenge for us. In this paper, we use SemEval 2024 task A monolingual dataset with comparative study between some machine learning model with feature extraction and develop an ensemble method for our system. Our system achieved 84.31% accuracy score in the test set, ranked 36th of 137 participants. Our code is available at: https://github.com/baoivy/SemEval-Task8

pdf bib
Hidetsune at SemEval-2024 Task 3: A Simple Textual Approach to Emotion Classification and Emotion Cause Analysis in Conversations Using Machine Learning and Next Sentence Prediction
Hidetsune Takahashi

In this system paper for SemEval-2024 Task3 subtask 2, I present my simple textual approach to emotion classification and emotioncause analysis in conversations using machinelearning and next sentence prediction. I train aSpaCy model for emotion classification and usenext sentence prediction with BERT for emotion cause analysis. While speaker names andaudio-visual clips are given in addition to textof the conversations, my approach uses textualdata only to test my methodology to combinemachine learning with next sentence prediction.This paper reveals both strengths and weaknesses of my trial, suggesting a direction offuture studies to improve my introductory solution.

pdf bib
CLTeam1 at SemEval-2024 Task 10: Large Language Model based ensemble for Emotion Detection in Hinglish
Ankit Vaidya | Aditya Gokhale | Arnav Desai | Ishaan Shukla | Sheetal Sonawane

This paper outlines our approach for the ERC subtask of the SemEval 2024 EdiREF Shared Task. In this sub-task, an emotion had to be assigned to an utterance which was the part of a dialogue. The utterance had to be classified into one of the following classes- disgust, contempt, anger, neutral, joy, sadness, fear, surprise. Our proposed system makes use of an ensemble of language specific RoBERTA and BERT models to tackle the problem. A weighted F1-score of 44% was achieved by our system in this task. We conducted comprehensive ablations and suggested directions of future work. Our codebase is available publicly.

pdf bib
Hidetsune at SemEval-2024 Task 4: An Application of Machine Learning to Multilingual Propagandistic Memes Identification Using Machine Translation
Hidetsune Takahashi

In this system paper for SemEval-2024 Task4 subtask 2b, I present my approach to identifying propagandistic memes in multiple languages. I firstly establish a baseline for Englishand then implement the model into other languages (Bulgarian, North Macedonian and Arabic) by using machine translation. Data fromother subtasks (subtask 1, subtask 2a) are alsoused in addition to data for this subtask, andadditional data from Kaggle are concatenatedto these in order to enhance the model. Theresults show a high reliability of my Englishbaseline and a room for improvement of itsimplementation.

pdf bib
Hidetsune at SemEval-2024 Task 10: An English Based Approach to Emotion Recognition in Hindi-English code-mixed Conversations Using Machine Learning and Machine Translation
Hidetsune Takahashi

In this system paper for SemEval-2024 Task10 subtask 1 (ERC), I present my approach torecognizing emotions in Hindi-English codemixed conversations. I train a SpaCy modelwith English translated data and classify emotions behind Hindi-English code-mixed utterances by using the model and translating theminto English. I use machine translation to translate all the data in Hindi-English mixed language into English due to an easy access to existing data for emotion recognition in English.Some additional data in English are used to enhance my model. This English based approachdemonstrates a fundamental possibility and potential of simplifying code-mixed language intoone major language for emotion recognition.

pdf bib
All-Mpnet at SemEval-2024 Task 1: Application of Mpnet for Evaluating Semantic Textual Relatedness
Marco Siino

In this study, we tackle the task of automatically discerning the level of semantic relatedness between pairs of sentences. Specifically, Task 1 at SemEval-2024 involves predicting the Semantic Textual Relatedness (STR) of sentence pairs. Participants are tasked with ranking sentence pairs based on their proximity in meaning, quantified by their degree of semantic relatedness, across 14 different languages. Each sentence pair is assigned manually determined relatedness scores ranging from 0 (indicating complete lack of relation) to 1 (denoting maximum relatedness). In our submitted approach on the official test set, focusing on Task 1 (a supervised task in English and Spanish), we achieve a Spearman rank correlation coefficient of 0.808 for the English language and 0.611 for the Spanish language.

pdf bib
0x.Yuan at SemEval-2024 Task 5: Enhancing Legal Argument Reasoning with Structured Prompts
Yu-an Lu | Hung-yu Kao

The intersection of legal reasoning and Natural Language Processing (NLP) technologies, particularly Large Language Models (LLMs), offers groundbreaking potential for augmenting human capabilities in the legal domain. This paper presents our approach and findings from participating in SemEval-2024 Task 5, focusing on the effect of argument reasoning in civil procedures using legal reasoning prompts. We investigated the impact of structured legal reasoning methodologies, including TREACC, IRAC, IRAAC, and MIRAC, on guiding LLMs to analyze and evaluate legal arguments systematically. Our experimental setup involved crafting specific prompts based on these methodologies to instruct the LLM to dissect and scrutinize legal cases, aiming to discern the cogency of argumentative solutions within a zero-shot learning framework. The performance of our approach, as measured by F1 score and accuracy, demonstrated the efficacy of integrating structured legal reasoning into LLMs for legal analysis. The findings underscore the promise of LLMs, when equipped with legal reasoning prompts, in enhancing their ability to process and reason through complex legal texts, thus contributing to the broader application of AI in legal studies and practice.

pdf bib
Groningen team D at SemEval-2024 Task 8: Exploring data generation and a combined model for fine-tuning LLMs for Multidomain Machine-Generated Text Detection
Thijs Brekhof | Xuanyi Liu | Joris Ruitenbeek | Niels Top | Yuwen Zhou

In this system description, we describe our process and the systems that we created for the subtasks A monolingual, A multilingual, and B forthe SemEval-2024 Task 8: Multigenerator, Multidomain, and Multilingual Black-Box MachineGenerated Text Detection. This shared task aimsat detecting and differentiating between machinegenerated text and human-written text. SubtaskA is focused on detecting if a text is machinegenerated or human-written both in a monolingualand a multilingual setting. Subtask B is also focused on detecting if a text is human-written ormachine-generated, though it takes it one step further by also requiring the detection of the correct language model used for generating the text.For the monolingual aspects of this task, our approach is centered around fine-tuning a debertav3-large LM. For the multilingual setting, we created an ensemble model utilizing different monolingual models and a language identification toolto classify each text. We also experiment with thegeneration of extra training data. Our results showthat the generation of extra data aids our modelsand leads to an increase in accuracy.

pdf bib
Kathlalu at SemEval-2024 Task 8: A Comparative Analysis of Binary Classification Methods for Distinguishing Between Human and Machine-generated Text
Lujia Cao | Ece Lara Kilic | Katharina Will

This paper investigates two methods for constructing a binary classifier to distinguish between human-generated and machine-generated text. The main emphasis is on a straightforward approach based on Zipf’s law, which, despite its simplicity, achieves a moderate level of performance. Additionally, the paper briefly discusses experimentation with the utilization of unigram word counts.

pdf bib
Team Unibuc - NLP at SemEval-2024 Task 8: Transformer and Hybrid Deep Learning Based Models for Machine-Generated Text Detection
Teodor-george Marchitan | Claudiu Creanga | Liviu P. Dinu

This paper describes the approach of the UniBuc - NLP team in tackling the SemEval 2024 Task 8: Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text Detection. We explored transformer-based and hybrid deep learning architectures. For subtask B, our transformer-based model achieved a strong second-place out of 77 teams with an accuracy of 86.95%, demonstrating the architecture’s suitability for this task. However, our models showed overfitting in subtask A which could potentially be fixed with less fine-tunning and increasing maximum sequence length. For subtask C (token-level classification), our hybrid model overfit during training, hindering its ability to detect transitions between human and machine-generated text.

pdf bib
LinguisTech at SemEval-2024 Task 10: Emotion Discovery and Reasoning its Flip in Conversation
Mihaela Alexandru | Călina Ciocoiu | Ioana Măniga | Octavian Ungureanu | Daniela Gîfu | Diana Trandăbăț

The “Emotion Discovery and Reasoning Its Flip in Conversation” task at the SemEval 2024 competition focuses on the automatic recognition of emotion flips, triggered within multi-party textual conversations. This paper proposes a novel approach that draws a parallel between a mixed strategy and a comparative strategy, contrasting a Rule-Based Function with Named Entity Recognition (NER)—an approach that shows promise in understanding speaker-specific emotional dynamics. Furthermore, this method surpasses the performance of both DistilBERT and RoBERTa models, demonstrating competitive effectiveness in detecting emotion flips triggered in multi-party textual conversations, achieving a 70% F1-score. This system was ranked 6th in the SemEval 2024 competition for Subtask 3.

pdf bib
Text Mining at SemEval-2024 Task 1: Evaluating Semantic Textual Relatedness in Low-resource Languages using Various Embedding Methods and Machine Learning Regression Models
Ron Keinan

In this paper, I describe my submission to the SemEval-2024 contest. I tackled subtask 1 - “Semantic Textual Relatedness for African and Asian Languages”. To find the semantic relatedness of sentence pairs, I tackled this task by creating models for nine different languages. I then vectorized the text data using a variety of embedding techniques including doc2vec, tf-idf, Sentence-Transformers, Bert, Roberta, and more, and used 11 traditional machine learning techniques of the regression type for analysis and evaluation.

pdf bib
USMBA-NLP at SemEval-2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials using Bert
Anass Fahfouh | Abdessamad Benlahbib | Jamal Riffi | Hamid Tairi

This paper presents the application of BERT inSemEval 2024 Task 2, Safe Biomedical Natu-ral Language Inference for Clinical Trials. Themain objectives of this task were: First, to in-vestigate the consistency of BERT in its rep-resentation of semantic phenomena necessaryfor complex inference in clinical NLI settings.Second, to investigate the ability of BERT toperform faithful reasoning, i.e., make correctpredictions for the correct reasons. The submit-ted model is fine-tuned on the NLI4CT dataset,which is enhanced with a novel contrast set,using binary cross entropy loss.

pdf bib
CRCL at SemEval-2024 Task 2: Simple prompt optimizations
Clement Brutti-mairesse | Loic Verlingue

We present a baseline for the SemEval 2024 task 2 challenge, whose objective is to ascertain the inference relationship between pairs of clinical trial report sections and statements.We apply prompt optimization techniques with LLM Instruct models provided as a Language Model-as-a-Service (LMaaS).We observed, in line with recent findings, that synthetic CoT prompts significantly enhance manually crafted ones.The source code is available at this GitHub repository https://github.com/ClementBM-CLB/semeval-2024

pdf bib
SuteAlbastre at SemEval-2024 Task 4: Predicting Propaganda Techniques in Multilingual Memes using Joint Text and Vision Transformers
Ion Anghelina | Gabriel Buță | Alexandru Enache

The main goal of this year’s SemEval Task 4 isdetecting the presence of persuasion techniquesin various meme formats. While Subtask 1targets text-only posts, Subtask 2, subsectionsa and b tackle posts containing both imagesand captions. The first 2 subtasks consist ofmulti-class and multi-label classifications, inthe context of a hierarchical taxonomy of 22different persuasion techniques.This paper proposes a solution for persuasiondetection in both these scenarios and for vari-ous languages of the caption text. Our team’smain approach consists of a Multimodal Learn-ing Neural Network architecture, having Tex-tual and Vision Transformers as its backbone.The models that we have experimented with in-clude EfficientNet and ViT as visual encodersand BERT and GPT2 as textual encoders.

pdf bib
RFBES at SemEval-2024 Task 8: Investigating Syntactic and Semantic Features for Distinguishing AI-Generated and Human-Written Texts
Mohammad Heydari Rad | Farhan Farsi | Shayan Bali | Romina Etezadi | Mehrnoush Shamsfard

Nowadays, the usage of Large Language Models (LLMs) has increased, and LLMs have been used to generate texts in different languages and for different tasks. Additionally, due to the participation of remarkable companies such as Google and OpenAI, LLMs are now more accessible, and people can easily use them. However, an important issue is how we can detect AI-generated texts from human-written ones. In this article, we have investigated the problem of AI-generated text detection from two different aspects: semantics and syntax. Finally, we presented an AI model that can distinguish AI-generated texts from human-written ones with high accuracy on both multilingual and monolingual tasks using the M4 dataset. According to our results, using a semantic approach would be more helpful for detection. However, there is a lot of room for improvement in the syntactic approach, and it would be a good approach for future work.

pdf bib
BAMBAS at SemEval-2024 Task 4: How far can we get without looking at hierarchies?
Arthur Vasconcelos | Luiz Felipe De Melo | Eduardo Goncalves | Eduardo Bezerra | Aline Paes | Alexandre Plastino

This paper describes the BAMBAS team’s participation in SemEval-2024 Task 4 Subtask 1, which focused on the multilabel classification of persuasion techniques in the textual content of Internet memes. We explored a lightweight approach that does not consider the hierarchy of labels. First, we get the text embeddings leveraging the multilingual tweets-based language model, Bernice. Next, we use those embeddings to train a separate binary classifier for each label, adopting independent oversampling strategies in each model in a binary-relevance style. We tested our approach over the English dataset, exceeding the baseline by 21 percentage points, while ranking in 23th in terms of hierarchical F1 and 11st in terms of hierarchical recall.

pdf bib
Team QUST at SemEval-2024 Task 8: A Comprehensive Study of Monolingual and Multilingual Approaches for Detecting AI-generated Text
Xiaoman Xu | Xiangrun Li | Taihang Wang | Jianxiang Tian | Ye Jiang

This paper presents the participation of team QUST in Task 8 SemEval 2024. we first performed data augmentation and cleaning on the dataset to enhance model training efficiency and accuracy. In the monolingual task, we evaluated traditional deep-learning methods, multiscale positive-unlabeled framework (MPU), fine-tuning, adapters and ensemble methods. Then, we selected the top-performing models based on their accuracy from the monolingual models and evaluated them in subtasks A and B. The final model construction employed a stacking ensemble that combined fine-tuning with MPU. Our system achieved 6th (scored 6th in terms of accuracy, officially ranked 13th in order) place in the official test set in multilingual settings of subtask A. We release our system code at:https://github.com/warmth27/SemEval2024_QUST

pdf bib
YNU-HPCC at SemEval-2024 Task 9: Using Pre-trained Language Models with LoRA for Multiple-choice Answering Tasks
Jie Wang | Jin Wang | Xuejie Zhang

This study describes the model built in Task 9: brainteaser in the SemEval-2024 competition, which is a multiple-choice task. As active participants in Task 9, our system strategically employs the decoding-enhanced BERT (DeBERTa) architecture enriched with disentangled attention mechanisms. Additionally, we fine-tuned our model using low-rank adaptation (LoRA) to optimize its performance further. Moreover, we integrate focal loss into our framework to address label imbalance issues. The systematic integration of these techniques has resulted in outstanding performance metrics. Upon evaluation using the provided test dataset, our system showcases commendable results, with a remarkable accuracy score of 0.9 for subtask 1, positioning us fifth among all participants. Similarly, for subtask 2, our system exhibits a substantial accuracy rate of 0.781, securing a commendable seventh-place ranking. The code for this paper is published at: https://github.com/123yunnandaxue/Semveal-2024_task9.

pdf bib
Team jelarson at SemEval 2024 Task 8: Predicting Boundary Line Between Human and Machine Generated Text
Joseph Larson | Francis Tyers

In this paper, we handle the task of building a system that, given a document written first by a human and then finished by an LLM, the system must determine the transition word i.e. where the machine begins to write. We built a system by examining the data for textual anomalies and combining a method of heuristic approaches with a linear regression model based on the text length of each document.

pdf bib
HU at SemEval-2024 Task 8A: Can Contrastive Learning Learn Embeddings to Detect Machine-Generated Text?
Shubhashis Roy Dipta | Sadat Shahriar

This paper describes our system developed for SemEval-2024 Task 8, “Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text Detection.” Machine-generated texts have been one of the main concerns due to the use of large language models (LLM) in fake text generation, phishing, cheating in exams, or even plagiarizing copyright materials. A lot of systems have been developed to detect machine-generated text. Nonetheless, the majority of these systems rely on the text-generating model. This limitation is impractical in real-world scenarios, as it’s often impossible to know which specific model the user has used for text generation. In this work, we propose a single model based on contrastive learning, which uses ~40% of the baseline’s parameters (149M vs. 355M) but shows a comparable performance on the test dataset (21st out of 137 participants). Our key finding is that even without an ensemble of multiple models, a single base model can have comparable performance with the help of data augmentation and contrastive learning.

pdf bib
Team AT at SemEval-2024 Task 8: Machine-Generated Text Detection with Semantic Embeddings
Yuchen Wei

This study investigates the detection of machine-generated text using several semantic embedding techniques, a critical issue in the era of advanced language models. Different methodologies were examined: GloVe embeddings, N-gram embedding models, Sentence BERT, and a concatenated embedding approach, against a fine-tuned RoBERTa baseline. The research was conducted within the framework of SemEval-2024 Task 8, encompassing tasks for binary and multi-class classification of machine-generated text.

pdf bib
JN666 at SemEval-2024 Task 7: NumEval: Numeral-Aware Language Understanding and Generation
Xinyi Liu | Xintong Liu | Hengyang Lu

This paper is submitted for SemEval-2027 task 7: Enhancing the Model’s Understanding and Generation of Numerical Values. The dataset for this task is NQuAD, which requires us to select the most suitable option number from four numerical options to fill in the blank in a news article based on the context. Based on the BertForMultipleChoice model, we proposed two new models, MC BERT and SSC BERT, and improved the model’s numerical understanding ability by pre-training the model on numerical comparison tasks. Ultimately, our best-performing model achieved an accuracy rate of 79.40%, which is 9.45% higher than the accuracy rate of NEMo.

pdf bib
BERTastic at SemEval-2024 Task 4: State-of-the-Art Multilingual Propaganda Detection in Memes via Zero-Shot Learning with Vision-Language Models
Tarek Mahmoud | Preslav Nakov

Analyzing propagandistic memes in a multilingual, multimodal dataset is a challenging problem due to the inherent complexity of memes’ multimodal content, which combines images, text, and often, nuanced context. In this paper, we use a VLM in a zero-shot approach to detect propagandistic memes and achieve a state-of-the-art average macro F1 of 66.7% over all languages. Notably, we outperform other systems on North Macedonian memes, and obtain competitive results on Bulgarian and Arabic memes. We also present our early fusion approach for identifying persuasion techniques in memes in a hierarchical multilabel classification setting. This approach outperforms all other approaches in average hierarchical precision with an average score of 77.66%. The systems presented contribute to the evolving field of research on the detection of persuasion techniques in multimodal datasets by offering insights that could be of use in the development of more effective tools for combating online propaganda.

pdf bib
RKadiyala at SemEval-2024 Task 8: Black-Box Word-Level Text Boundary Detection in Partially Machine Generated Texts
Ram Mohan Rao Kadiyala

With increasing usage of generative models for text generation and widespread use of machine generated texts in various domains, being able to distinguish between human written and machine generated texts is a significant challenge. While existing models and proprietary systems focus on identifying whether given text is entirely human written or entirely machine generated, only a few systems provide insights at sentence or paragraph level at likelihood of being machine generated at a non reliable accuracy level, working well only for a set of domains and generators. This paper introduces few reliable approaches for the novel task of identifying which part of a given text is machine generated at a word level while comparing results from different approaches and methods. We present a comparison with proprietary systems , performance of our model on unseen domains’ and generators’ texts. The findings reveal significant improvements in detection accuracy along with comparison on other aspects of detection capabilities. Finally we discuss potential avenues for improvement and implications of our work. The proposed model is also well suited for detecting which parts of a text are machine generated in outputs of Instruct variants of many LLMs.

pdf bib
TLDR at SemEval-2024 Task 2: T5-generated clinical-Language summaries for DeBERTa Report Analysis
Spandan Das | Vinay Samuel | Shahriar Noroozizadeh

This paper introduces novel methodologies for the Natural Language Inference for Clinical Trials (NLI4CT) task. We present TLDR (T5-generated clinical-Language summaries for DeBERTa Report Analysis) which incorporates T5-model generated premise summaries for improved entailment and contradiction analysis in clinical NLI tasks. This approach overcomes the challenges posed by small context windows and lengthy premises, leading to a substantial improvement in Macro F1 scores: a 0.184 increase over truncated premises. Our comprehensive experimental evaluation, including detailed error analysis and ablations, confirms the superiority of TLDR in achieving consistency and faithfulness in predictions against semantically altered inputs.

pdf bib
ignore at SemEval-2024 Task 5: A Legal Classification Model with Summary Generation and Contrastive Learning
Binjie Sun | Xiaobing Zhou

This paper describes our work for SemEval-2024 Task 5: The Legal Argument Reasoning Task in Civil Procedure. After analyzing the task requirements and the training dataset, we used data augmentation, adopted the large model GPT for summary generation, and added supervised contrastive learning to the basic BERT model. Our system achieved an F1 score of 0.551, ranking 14th in the competition leaderboard. Our system achieves an F1 score improvement of 0.1241 over the official baseline model.

pdf bib
Samsung Research China-Beijing at SemEval-2024 Task 3: A multi-stage framework for Emotion-Cause Pair Extraction in Conversations
Shen Zhang | Haojie Zhang | Jing Zhang | Xudong Zhang | Yimeng Zhuang | Jinting Wu

In human-computer interaction, it is crucial for agents to respond to human by understanding their emotions. unraveling the causes of emotions is more challenging. A new task named Multimodal Emotion-Cause Pair Extraction in Conversations is responsible for recognizing emotion and identifying causal expressions. In this study, we propose a multi-stage framework to generate emotion and extract the emotion causal pairs given the target emotion. In the first stage, LLaMA2-based InstructERC is utilized to extract the emotion category of each utterance in a conversation. After emotion recognition, a two-stream attention model is employed to extract the emotion causal pairs given the target emotion for subtask 2 while MuTEC is employed to extract causal span for subtask 1. Our approach achieved first place for both of the two subtasks in the competition.

pdf bib
Werkzeug at SemEval-2024 Task 8: LLM-Generated Text Detection via Gated Mixture-of-Experts Fine-Tuning
Youlin Wu | Kaichun Wang | Kai Ma | Liang Yang | Hongfei Lin

Recent advancements in Large Language Models (LLMs) have propelled text generation to unprecedented heights, approaching human-level quality. However, it poses a new challenge to distinguish LLM-generated text from human-written text. Presently, most methods address this issue through classification, achieved by fine-tuning on small language models. Unfortunately, small language models suffer from anisotropy issue, where encoded text embeddings become difficult to differentiate in the latent space. Moreover, LLMs possess the ability to alter language styles with versatility, further complicating the classification task. To tackle these challenges, we propose Gated Mixture-of-Experts Fine-tuning (GMoEF) to detect LLM-generated text. GMoEF leverages parametric whitening to normalize text embeddings, thereby mitigating the anisotropy problem. Additionally, GMoEF employs the mixture-of-experts framework equipped with gating router to capture features of LLM-generated text from multiple perspectives. Our GMoEF achieved an impressive ranking of #8 out of 70 teams. The source code is available on https://gitlab.com/sigrs/gmoef.

pdf bib
SSN_Semeval10 at SemEval-2024 Task 10: Emotion Discovery and Reasoning its Flip in Conversations
Antony Rajesh | Supriya Abirami | Aravindan Chandrabose | Senthil Kumar

This paper presents a transformer-based model for recognizing emotions in Hindi-English code-mixed conversations, adhering to the SemEval task constraints. Leveraging BERT-based transformers, we fine-tune pre-trained models on the dataset, incorporating tokenization and attention mechanisms. Our approach achieves competitive performance (weighted F1-score of 0.4), showcasing the effectiveness of BERT in nuanced emotion analysis tasks within code-mixed conversational contexts.

pdf bib
KInIT at SemEval-2024 Task 8: Fine-tuned LLMs for Multilingual Machine-Generated Text Detection
Michal Spiegel | Dominik Macko

SemEval-2024 Task 8 is focused on multigenerator, multidomain, and multilingual black-box machine-generated text detection. Such a detection is important for preventing a potential misuse of large language models (LLMs), the newest of which are very capable in generating multilingual human-like texts. We have coped with this task in multiple ways, utilizing language identification and parameter-efficient fine-tuning of smaller LLMs for text classification. We have further used the per-language classification-threshold calibration to uniquely combine fine-tuned models predictions with statistical detection metrics to improve generalization of the system detection performance. Our submitted method achieved competitive results, ranking at the fourth place, just under 1 percentage point behind the winner.

pdf bib
Sharif-MGTD at SemEval-2024 Task 8: A Transformer-Based Approach to Detect Machine Generated Text
Seyedeh Fatemeh Ebrahimi | Karim Akhavan Azari | Amirmasoud Iravani | Arian Qazvini | Pouya Sadeghi | Zeinab Taghavi | Hossein Sameti

In this paper, we delve into the realm of detecting machine-generated text (MGT) within Natural Language Processing (NLP). Our approach involves fine-tuning a RoBERTa-base Transformer, a robust neural architecture, to tackle MGT detection as a binary classification task. Specifically focusing on Subtask A (Monolingual - English) within the SemEval-2024 competition framework, our system achieves a 78.9% accuracy on the test dataset, placing us 57th among participants. While our system demonstrates proficiency in identifying human-written texts, it faces challenges in accurately discerning MGTs.

pdf bib
IRIT-Berger-Levrault at SemEval-2024: How Sensitive Sentence Embeddings are to Hallucinations?
Nihed Bendahman | Karen Pinel-sauvagnat | Gilles Hubert | Mokhtar Billami

This article presents our participation to Task 6 of SemEval-2024, named SHROOM (a Shared-task on Hallucinations and Related Observable Overgeneration Mistakes), which aims at detecting hallucinations. We propose two types of approaches for the task: the first one is based on sentence embeddings and cosine similarity metric, and the second one uses LLMs (Large Language Model). We found that LLMs fail to improve the performance achieved by embedding generation models. The latter outperform the baseline provided by the organizers, and our best system achieves 78% accuracy.

pdf bib
CYUT at SemEval-2024 Task 7: A Numerals Augmentation and Feature Enhancement Approach to Numeral Reading Comprehension
Tsz-yeung Lau | Shih-hung Wu

This study explores Task 2 in NumEval-2024, which is SemEval-2024(Semantic Evaluation)Task 7 , focusing on the Reading Comprehension of Numerals in Text (Chinese). The datasetutilized in this study is the Numeral-related Question Answering Dataset (NQuAD), and the model employed is BERT. The data undergoes preprocessing, incorporating Numerals Augmentation and Feature Enhancement to numerical entities before model training. Additionally, fine-tuning will also be applied. The result was an accuracy rate of 77.09%, representing a 7.14% improvement compared to the initial NQuAD processing model, referred to as the Numeracy-Enhanced Model (NEMo).

pdf bib
UniBuc at SemEval-2024 Task 2: Tailored Prompting with Solar for Clinical NLI
Marius Micluta-Campeanu | Claudiu Creanga | Ana-maria Bucur | Ana Sabina Uban | Liviu P. Dinu

This paper describes the approach of the UniBuc team in tackling the SemEval 2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials. We used SOLAR Instruct, without any fine-tuning, while focusing on input manipulation and tailored prompting. By customizing prompts for individual CTR sections, in both zero-shot and few-shots settings, we managed to achieve a consistency score of 0.72, ranking 14th in the leaderboard. Our thorough error analysis revealed that our model has a tendency to take shortcuts and rely on simple heuristics, especially when dealing with semantic-preserving changes.

pdf bib
Fralak at SemEval-2024 Task 4: combining RNN-generated hierarchy paths with simple neural nets for hierarchical multilabel text classification in a multilingual zero-shot setting
Katarina Laken

This paper describes the submission of team fralak for subtask 1 of task 4 of the Semeval-2024 shared task: ‘Multilingual detection of persuasion techniques in memes’. The first subtask included only the textual content of the memes. We restructured the labels into strings that showed the full path through the hierarchy. The system includes an RNN module that is trained to generate these strings. This module was then incorporated in an ensemble model with 2 more models consisting of basic fully connected networks. Although our model did not perform particularly well on the English only setting, we found that it generalized better to other languages in a zero-shot context than most other models. Some additional experiments were performed to explain this. Findings suggest that the RNN generating the restructured labels generalized well across languages, but preprocessing did not seem to play a role. We conclude by giving suggestions for future improvements of our core idea.

pdf bib
OtterlyObsessedWithSemantics at SemEval-2024 Task 4: Developing a Hierarchical Multi-Label Classification Head for Large Language Models
Julia Wunderle | Julian Schubert | Antonella Cacciatore | Albin Zehe | Jan Pfister | Andreas Hotho

For our submission for Subtask 1, we developed a custom classification head that is designed to be applied atop of a Large Language Model. We reconstructed the hierarchy across multiple fully connected layers, allowing us to incorporate previous foundational decisions in subsequent, more fine-grained layers. To find the best hyperparameters, we conducted a grid-search and to compete in the multilingual setting, we translated all documents to English.

pdf bib
D-NLP at SemEval-2024 Task 2: Evaluating Clinical Inference Capabilities of Large Language Models
Duygu Altinok

Large language models (LLMs) have garnered significant attention and widespread usage due to their impressive performance in various tasks. However, they are not without their own set of challenges, including issues such as hallucinations, factual inconsistencies, and limitations in numerical-quantitative reasoning. Evaluating LLMs in miscellaneous reasoning tasks remains an active area of research. Prior to the breakthrough of LLMs, Transformers had already proven successful in the medical domain, effectively employed for various natural language understanding (NLU) tasks. Following this trend, LLMs have also been trained and utilized in the medical domain, raising concerns regarding factual accuracy, adherence tosafety protocols, and inherent limitations. In this paper, we focus on evaluating the natural language inference capabilities of popular open-source and closed-source LLMs using clinical trial reports as the dataset. We present the performance results of each LLM and further analyze their performance on a development set, particularly focusing on challenging instances that involve medical abbreviations and require numerical-quantitative reasoning. Gemini, our leading LLM, achieved a test set F1-score of 0.748, securing the ninth position on the task scoreboard. Our work is the first of its kind, offering a thorough examination of the inference capabilities of LLMs within the medical domain.

pdf bib
LMEME at SemEval-2024 Task 4: Teacher Student Fusion - Integrating CLIP with LLMs for Enhanced Persuasion Detection
Shiyi Li | Yike Wang | Liang Yang | Shaowu Zhang | Hongfei Lin

This paper describes our system used in the SemEval-2024 Task 4 Multilingual Detection of Persuasion Techniques in Memes. Our team proposes a detection system that employs a Teacher Student Fusion framework. Initially, a Large Language Model serves as the teacher, engaging in abductive reasoning on multimodal inputs to generate background knowledge on persuasion techniques, assisting in the training of a smaller downstream model. The student model adopts CLIP as an encoder for text and image features, and we incorporate an attention mechanism for modality alignment. Ultimately, our proposed system achieves a Macro-F1 score of 0.8103, ranking 1st out of 20 on the leaderboard of Subtask 2b in English. In Bulgarian, Macedonian and Arabic, our detection capabilities are ranked 1/15, 3/15 and 14/15.

pdf bib
Innovators at SemEval-2024 Task 10: Revolutionizing Emotion Recognition and Flip Analysis in Code-Mixed Texts
Abhay Shanbhag | Suramya Jadhav | Shashank Rathi | Siddhesh Pande | Dipali Kadam

In this paper, we introduce our system for all three tracks of the SemEval 2024 EDiReF Shared Task 10, which focuses on Emotion Recognition in Conversation (ERC) and Emotion Flip Reasoning (EFR) within the domain of conversational analysis. Task-Track 1 (ERC) aims to assign an emotion to each utterance in the Hinglish language from a predefined set of possible emotions. Tracks 2 (EFR) and 3 (EFR) aim to identify the trigger utterance(s) for an emotion flip in a multi-party conversation dialogue in Hinglish and English text, respectively. For Track 1, our study spans both traditional machine learning ensemble techniques, including Decision Trees, SVM, Logistic Regression, and Multinomial NB models, as well as advanced transformer-based models like XLM-Roberta (XLMR), DistilRoberta, and T5 from Hugging Face’s transformer library. In the EFR competition, we developed and proposed two innovative algorithms to tackle the challenges presented in Tracks 2 and 3. Specifically, our team, Innovators, developed a standout algorithm that propelled us to secure the 2nd rank in Track 2, achieving an impressive F1 score of 0.79, and the 7th rank in Track 3, with an F1 score of 0.68.

pdf bib
DUTIR938 at SemEval-2024 Task 4: Semi-Supervised Learning and Model Ensemble for Persuasion Techniques Detection in Memes
Erchen Yu | Junlong Wang | Xuening Qiao | Jiewei Qi | Zhaoqing Li | Hongfei Lin | Linlin Zong | Bo Xu

The development of social platforms has facilitated the proliferation of disinformation, with memes becoming one of the most popular types of propaganda for disseminating disinformation on the internet. Effectively detecting the persuasion techniques hidden within memes is helpful in understanding user-generated content and further promoting the detection of disinformation on the internet. This paper demonstrates the approach proposed by Team DUTIR938 in Subtask 2b of SemEval-2024 Task 4. We propose a dual-channel model based on semi-supervised learning and model ensemble. We utilize CLIP to extract image features, and employ various pretrained language models under task-adaptive pretraining for text feature extraction. To enhance the detection and generalization capabilities of the model, we implement sample data augmentation using semi-supervised pseudo-labeling methods, introduce adversarial training strategies, and design a two-stage global model ensemble strategy. Our proposed method surpasses the provided baseline method, with Macro/Micro F1 values of 0.80910/0.83667 in the English leaderboard. Our submission ranks 3rd/19 in terms of Macro F1 and 1st/19 in terms of Micro F1.

pdf bib
ISDS-NLP at SemEval-2024 Task 10: Transformer based neural networks for emotion recognition in conversations
Claudiu Creanga | Liviu P. Dinu

This paper outlines the approach of the ISDS-NLP team in the SemEval 2024 Task 10: Emotion Discovery and Reasoning its Flip in Conversation (EDiReF). For Subtask 1 we obtained a weighted F1 score of 0.43 and placed 12 in the leaderboard. We investigate two distinct approaches: Masked Language Modeling (MLM) and Causal Language Modeling (CLM). For MLM, we employ pre-trained BERT-like models in a multilingual setting, fine-tuning them with a classifier to predict emotions. Experiments with varying input lengths, classifier architectures, and fine-tuning strategies demonstrate the effectiveness of this approach. Additionally, we utilize Mistral 7B Instruct V0.2, a state-of-the-art model, applying zero-shot and few-shot prompting techniques. Our findings indicate that while Mistral shows promise, MLMs currently outperform them in sentence-level emotion classification.

pdf bib
UMUTeam at SemEval-2024 Task 4: Multimodal Identification of Persuasive Techniques in Memes through Large Language Models
Ronghao Pan | José Antonio García-díaz | Rafael Valencia-garcía

In this manuscript we describe the UMUTeam’s participation in SemEval-2024 Task 4, a shared task to identify different persuasion techniques in memes. The task is divided into three subtasks. One is a multimodal subtask of identifying whether a meme contains persuasion or not. The others are hierarchical multi-label classifications that consider textual content alone or a multimodal setting of text and visual content. This is a multilingual task, and we participated in all three subtasks but we focus only on the English dataset. Our approach is based on a fine-tuning approach with the pre-trained RoBERTa-large model. In addition, for multimodal cases with both textual and visual content, we used the LMM called LlaVa to extract image descriptions and combine them with the meme text. Our system performed well in three subtasks, achieving the tenth best result with an Hierarchical F1 of 64.774%, the fourth best in Subtask 2a with an Hierarchical F1 of 69.003%, and the eighth best in Subtask 2b with a Macro F1 of 78.660%.

pdf bib
MIPS at SemEval-2024 Task 3: Multimodal Emotion-Cause Pair Extraction in Conversations with Multimodal Language Models
Zebang Cheng | Fuqiang Niu | Yuxiang Lin | Zhi-qi Cheng | Xiaojiang Peng | Bowen Zhang

This paper presents our winning submission to Subtask 2 of SemEval 2024 Task 3 on multimodal emotion cause analysis in conversations. We propose a novel Multimodal Emotion Recognition and Multimodal Emotion Cause Extraction (MER-MCE) framework that integrates text, audio, and visual modalities using specialized emotion encoders. Our approach sets itself apart from top-performing teams by leveraging modality-specific features for enhanced emotion understanding and causality inference. Experimental evaluation demonstrates the advantages of our multimodal approach, with our submission achieving a competitive weighted F1 score of 0.3435, ranking third with a margin of only 0.0339 behind the 1st team and 0.0025 behind the 2nd team.

pdf bib
UMUTeam at SemEval-2024 Task 6: Leveraging Zero-Shot Learning for Detecting Hallucinations and Related Observable Overgeneration Mistakes
Ronghao Pan | José Antonio García-díaz | Tomás Bernal-beltrán | Rafael Valencia-garcía

In these working notes we describe the UMUTeam’s participation in SemEval-2024 shared task 6, which aims at detecting grammatically correct output of Natural Language Generation with incorrect semantic information in two different setups: model-aware and model-agnostic tracks. The task is consists of three subtasks with different model setups. Our approach is based on exploiting the zero-shot classification capability of the Large Language Models LLaMa-2, Tulu and Mistral, through prompt engineering. Our system ranked eighteenth in the model-aware setup with an accuracy of 78.4% and 29th in the model-agnostic setup with an accuracy of 76.9333%.

pdf bib
DFKI-NLP at SemEval-2024 Task 2: Towards Robust LLMs Using Data Perturbations and MinMax Training
Bhuvanesh Verma | Lisa Raithel

The NLI4CT task at SemEval-2024 emphasizes the development of robust models for Natural Language Inference on Clinical Trial Reports (CTRs) using large language models (LLMs). This edition introduces interventions specifically targeting the numerical, vocabulary, and semantic aspects of CTRs. Our proposed system harnesses the capabilities of the state-of-the-art Mistral model (Jiang et al., 2023), complemented by an auxiliary model, to focus on the intricate input space of the NLI4CT dataset. Through the incorporation of numerical and acronym-based perturbations to the data, we train a robust system capable of handling both semantic-altering and numerical contradiction interventions. Our analysis on the dataset sheds light on the challenging sections of the CTRs for reasoning.

pdf bib
UMUTeam at SemEval-2024 Task 8: Combining Transformers and Syntax Features for Machine-Generated Text Detection
Ronghao Pan | José Antonio García-díaz | Pedro José Vivancos-vicente | Rafael Valencia-garcía

These working notes describe the UMUTeam’s participation in Task 8 of SemEval-2024 entitled “Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text Detection”. This shared task aims at identifying machine-generated text in order to mitigate its potential misuse. This shared task is divided into three subtasks: Subtask A, a binary classification task to determine whether a given full-text was written by a human or generated by a machine; Subtask B, a multi-class classification problem to determine, given a full-text, who generated it. It can be written by a human or generated by a specific language model; and Subtask C, mixed human-machine text recognition. We participated in Subtask B, using an approach based on fine-tuning a pre-trained model, such as RoBERTa, combined with syntactic features of the texts. Our system placed 23rd out of a total of 77 participants, with a score of 75.350%, outperforming the baseline.

pdf bib
UMUTeam at SemEval-2024 Task 10: Discovering and Reasoning about Emotions in Conversation using Transformers
Ronghao Pan | José Antonio García-díaz | Diego Roldán | Rafael Valencia-garcía

These notes describe the participation of the UMUTeam in EDiReF, the 10th shared task of SemEval 2024. The goal is to develop systems for detecting and inferring emotional changes in the conversation. The task was divided into three related subtasks: (i) Emotion Recognition in Conversation (ERC) in Hindi-English code-mixed conversations, (ii) Emotion Flip Reasoning (EFR) in Hindi-English code-mixed conversations, and (iii) EFR in English conversations. We were involved in all three and our approach is based on a fine-tuning approach with different pre-trained models. After evaluation, we found BERT to be the best model for ERC and EFR and with this model we achieved the thirteenth best result with an F1 score of 43% in Subtask 1, the sixth best in Subtask 2 with an F1 score of 26% and the fifteenth best in Subtask 3 with an F1 score of 22%.

pdf bib
TM-TREK at SemEval-2024 Task 8: Towards LLM-Based Automatic Boundary Detection for Human-Machine Mixed Text
Xiaoyan Qu | Xiangfeng Meng

With the increasing prevalence of text gener- ated by large language models (LLMs), there is a growing concern about distinguishing be- tween LLM-generated and human-written texts in order to prevent the misuse of LLMs, such as the dissemination of misleading information and academic dishonesty. Previous research has primarily focused on classifying text as ei- ther entirely human-written or LLM-generated, neglecting the detection of mixed texts that con- tain both types of content. This paper explores LLMs’ ability to identify boundaries in human- written and machine-generated mixed texts. We approach this task by transforming it into a to- ken classification problem and regard the label turning point as the boundary. Notably, our ensemble model of LLMs achieved first place in the ‘Human-Machine Mixed Text Detection’ sub-task of the SemEval’24 Competition Task 8. Additionally, we investigate factors that in- fluence the capability of LLMs in detecting boundaries within mixed texts, including the incorporation of extra layers on top of LLMs, combination of segmentation loss, and the im- pact of pretraining. Our findings aim to provide valuable insights for future research in this area.

pdf bib
Team NP_PROBLEM at SemEval-2024 Task 7: Numerical Reasoning in Headline Generation with Preference Optimization
Pawan Rajpoot | Nut Chukamphaeng

While large language models (LLMs) exhibit impressive linguistic abilities, their numerical reasoning skills within real-world contexts re- main under-explored. This paper describes our participation in a headline-generation challenge by Numeval at Semeval 2024, which focused on numerical reasoning. Our system achieved an overall top numerical accuracy of 73.49% on the task. We explore the system’s design choices contributing to this result and analyze common error patterns. Our findings highlight the potential and ongoing challenges of integrat- ing numerical reasoning within large language model-based headline generation.

pdf bib
OPDAI at SemEval-2024 Task 6: Small LLMs can Accelerate Hallucination Detection with Weakly Supervised Data
Ze Chen | Chengcheng Wei | Songtan Fang | Jiarong He | Max Gao

This paper mainly describes a unified system for hallucination detection of LLMs, which wins the second prize in the model-agnostic track of the SemEval-2024 Task 6, and also achieves considerable results in the model-aware track. This task aims to detect hallucination with LLMs for three different text-generation tasks without labeled training data. We utilize prompt engineering and few-shot learning to verify the performance of different LLMs on the validation data. Then we select the LLMs with better performance to generate high-quality weakly supervised training data, which not only satisfies the consistency of different LLMs, but also satisfies the consistency of the optimal LLM with different sampling parameters. Furthermore, we finetune different LLMs by using the constructed training data, and finding that a relatively small LLM can achieve a competitive level of performance in hallucination detection, when compared to the large LLMs and the prompt-based approaches using GPT-4.

pdf bib
SSN_ARMM at SemEval-2024 Task 10: Emotion Detection in Multilingual Code-Mixed Conversations using LinearSVC and TF-IDF
Rohith Arumugam | Angel Deborah | Rajalakshmi Sivanaiah | Milton R S | Mirnalinee Thankanadar

Our paper explores a task involving the analysis of emotions and triggers within dialogues. We annotate each utterance with an emotion and identify triggers, focusing on binary labeling. We emphasize clear guidelines for replicability and conduct thorough analyses, including multiple system runs and experiments to highlight effective techniques. By simplifying the complexities and detailing clear methodologies, our study contributes to advancing emotion analysis and trigger identification within dialogue systems.

pdf bib
TüDuo at SemEval-2024 Task 2: Flan-T5 and Data Augmentation for Biomedical NLI
Veronika Smilga | Hazem Alabiad

This paper explores using data augmentation with smaller language models under 3 billion parameters for the SemEval-2024 Task 2 on Biomedical Natural Language Inference for Clinical Trials. We fine-tune models from the Flan-T5 family with and without using augmented data automatically generated by GPT-3.5-Turbo and find that data augmentation through techniques like synonym replacement, syntactic changes, adding random facts, and meaning reversion improves model faithfulness (ability to change predictions for semantically different inputs) and consistency (ability to give same predictions for semantic preserving changes). However, data augmentation tends to decrease performance on the original dataset distribution, as measured by F1 score. Our best system is the Flan-T5 XL model fine-tuned on the original training data combined with over 6,000 augmented examples. The system ranks in the top 10 for all three metrics.

pdf bib
FeedForward at SemEval-2024 Task 10: Trigger and sentext-height enriched emotion analysis in multi-party conversations
Zuhair Hasan Shaik | Dhivya Prasanna | Enduri Jahnavi | Rishi Thippireddy | Vamsi Madhav | Sunil Saumya | Shankar Biradar

This paper reports on an innovative approach to Emotion Recognition in Conversation and Emotion Flip Reasoning for the SemEval-2024 competition with a specific focus on analyzing Hindi-English code-mixed language. By integrating Large Language Models (LLMs) with Instruction-based Fine-tuning and Quantized Low-Rank Adaptation (QLoRA), this study introduces innovative techniques like Sentext-height and advanced prompting strategies to navigate the intricacies of emotional analysis in code-mixed conversational data. The results of the proposed work effectively demonstrate its ability to overcome label bias and the complexities of code-mixed languages. Our team achieved ranks of 5, 3, and 3 in tasks 1, 2, and 3 respectively. This study contributes valuable insights and methods for enhancing emotion recognition models, underscoring the importance of continuous research in this field.

pdf bib
YNU-HPCC at SemEval-2024 Task 5: Regularized Legal-BERT for Legal Argument Reasoning Task in Civil Procedure
Peng Shi | Jin Wang | Xuejie Zhang

This paper describes the submission of team YNU-HPCC to SemEval-2024 for Task 5: The Legal Argument Reasoning Task in Civil Procedure. The task asks candidates the topic, questions, and answers, classifying whether a given candidate’s answer is correct (True) or incorrect (False). To make a sound judgment, we propose a system. This system is based on fine-tuning the Legal-BERT model that specializes in solving legal problems. Meanwhile,Regularized Dropout (R-Drop) and focal Loss were used in the model. R-Drop is used for data augmentation, and focal loss addresses data imbalances. Our system achieved relatively good results on the competition’s official leaderboard. The code of this paper is available at https://github.com/YNU-PengShi/SemEval-2024-Task5.

pdf bib
TECHSSN at SemEval-2024 Task 10: LSTM-based Approach for Emotion Detection in Multilingual Code-Mixed Conversations
Ravindran V | Shreejith Babu G | Aashika Jetti | Rajalakshmi Sivanaiah | Angel Deborah | Mirnalinee Thankanadar | Milton R S

Emotion Recognition in Conversation (ERC) in the context of code-mixed Hindi-English interactions is a subtask addressed in SemEval-2024 as Task 10. We made our maiden attempt to solve the problem using natural language processing, machine learning and deep learning techniques, that perform well in properly assigning emotions to individual utterances from a predefined collection. The use of well-proven classifier such as Long Short Term Memory networks improve the model’s efficacy than the BERT and Glove based models. How-ever, difficulties develop in the subtle arena of emotion-flip reasoning in multi-party discussions, emphasizing the importance of specialized methodologies. Our findings shed light on the intricacies of emotion dynamics in code-mixed languages, pointing to potential areas for further research and refinement in multilingual understanding.

pdf bib
UIR-ISC at SemEval-2024 Task 3: Textual Emotion-Cause Pair Extraction in Conversations
Hongyu Guo | Xueyao Zhang | Yiyang Chen | Lin Deng | Binyang Li

The goal of Emotion Cause Pair Extraction (ECPE) is to explore the causes of emotion changes and what causes a certain emotion. This paper proposes a three-step learning approach for the task of Textual Emotion-Cause Pair Extraction in Conversations in SemEval-2024 Task 3, named ECSP. We firstly perform data preprocessing operations on the original dataset to construct negative samples. Secondly, we use a pre-trained model to construct token sequence representations with contextual information to obtain emotion prediction. Thirdly, we regard the textual emotion-cause pair extraction task as a machine reading comprehension task, and fine-tune two pre-trained models, RoBERTa and SpanBERT. Our results have achieved good results in the official rankings, ranking 3rd under the strict match with the Strict F1-score of 15.18%, which further shows that our system has a robust performance.

pdf bib
YNU-HPCC at SemEval-2024 Task10: Pre-trained Language Model for Emotion Discovery and Reasoning its Flip in Conversation
Chenyi Liang | Jin Wang | Xuejie Zhang

This paper describes the application of fine-tuning pre-trained models for SemEval-2024 Task 10: Emotion Discovery and Reasoning its Flip in Conversation (EDiReF), which requires the prediction of emotions for each utterance in a conversation and the identification of sentences where an emotional flip occurs. This model is built on the DeBERTa transformer model and enhanced for emotion detection and flip reasoning in conversations. It employs specific separators for utterance processing and utilizes specific padding to handle variable-length inputs. Methods such as R-drop, back translation, and focalloss are also employed in the training of my model. The model achieved specific results on the competition’s official leaderboard. The code of this paper is available athttps://github.com/jiaowoobjiuhao/SemEval-2024-task10.

pdf bib
YNU-HPCC at SemEval-2024 Task 2: Applying DeBERTa-v3-large to Safe Biomedical Natural Language Inference for Clinical Trials
Rengui Zhang | Jin Wang | Xuejie Zhang

This paper describes the system for the YNU-HPCC team for SemEval2024 Task 2, focusing on Safe Biomedical Natural Language Inference for Clinical Trials. The core challenge of this task lies in discerning the textual entailment relationship between Clinical Trial Reports (CTR) and statements annotated by expert annotators, including the necessity to infer the relationships in texts subjected to semantic interventions accurately. Our approach leverages a fine-tuned DeBERTa-v3-large model augmented with supervised contrastive learning and back-translation techniques. Supervised contrastive learning aims to bolster classification ac-curacy while back-translation enriches the diversity and quality of our training corpus. Our method achieves a decent F1 score. However, the results also indicate a need for further en-hancements in the system’s capacity for deep semantic comprehension, highlighting areas for future refinement. The code of this paper is available at:https://github.com/RGTnuw/RG_YNU-HPCC-at-Semeval2024-Task2.

pdf bib
YNU-HPCC at SemEval-2024 Task 1: Self-Instruction Learning with Black-box Optimization for Semantic Textual Relatedness
Weijie Li | Jin Wang | Xuejie Zhang

This paper introduces a system designed for SemEval-2024 Task 1 that focuses on assessing Semantic Textual Relatedness (STR) between sentence pairs, including its multilingual version. STR, which evaluates the coherence of sentences, is distinct from Semantic Textual Similarity (STS). However, Large Language Models (LLMs) such as ERNIE-Bot-turbo, typically trained on STS data, often struggle to differentiate between the two concepts. To address this, we developed a self-instruction method that enhances their performance distinguishing STR, particularly in cases with high STS but low STR. Beginning with a task description, the system generates new task instructions refined through human feedback. It then iteratively enhances these instructions by comparing them to the original and evaluating the differences. Utilizing the Large Language Models’ (LLMs) natural language comprehension abilities, the system aims to produce progressively optimized instructions based on the resulting scores. Through our optimized instructions, ERNIE-Bot-turbo exceeds the performance of conventional models,achieving a score enhancement of 4 to 7% on multilingual development datasets.

pdf bib
AAdaM at SemEval-2024 Task 1: Augmentation and Adaptation for Multilingual Semantic Textual Relatedness
Miaoran Zhang | Mingyang Wang | Jesujoba Alabi | Dietrich Klakow

This paper presents our system developed for the SemEval-2024 Task 1: Semantic Textual Relatedness for African and Asian Languages. The shared task aims at measuring the semantic textual relatedness between pairs of sentences, with a focus on a range of under-represented languages. In this work, we propose using machine translation for data augmentation to address the low-resource challenge of limited training data. Moreover, we apply task-adaptive pre-training on unlabeled task data to bridge the gap between pre-training and task adaptation. For model training, we investigate both full fine-tuning and adapter-based tuning, and adopt the adapter framework for effective zero-shot cross-lingual transfer. We achieve competitive results in the shared task: our system performs the best among all ranked teams in both subtask A (supervised learning) and subtask C (cross-lingual transfer).

pdf bib
BITS Pilani at SemEval-2024 Task 10: Fine-tuning BERT and Llama 2 for Emotion Recognition in Conversation
Dilip Venkatesh | Pasunti Prasanjith | Yashvardhan Sharma

Emotion Recognition in Conversation (ERC)aims to assign an emotion to a dialogue in aconversation between people. The first subtaskof EDiReF shared task aims to assign an emo-tions to a Hindi-English code mixed conversa-tion. For this, our team proposes a system toidentify the emotion based on fine-tuning largelanguage models on the MaSaC dataset. Forour study we have fine tuned 2 LLMs BERTand Llama 2 to perform sequence classificationto identify the emotion of the text.

pdf bib
BITS Pilani at SemEval-2024 Task 9: Prompt Engineering with GPT-4 for Solving Brainteasers
Dilip Venkatesh | Yashvardhan Sharma

Solving brainteasers is a task that requires complex reasoning prowess. The increase of research in natural language processing has leadto the development of massive large languagemodels with billions (or trillions) of parameters that are able to solve difficult questionsdue to their advanced reasoning capabilities.The SemEval BRAINTEASER shared tasks consists of sentence and word puzzles along withoptions containing the answer for the puzzle.Our team uses OpenAI’s GPT-4 model alongwith prompt engineering to solve these brainteasers.

pdf bib
Bridging Numerical Reasoning and Headline Generation for Enhanced Language Models
Vaishnavi R | Srimathi T | Aarthi S | Harini V

Headline generation becomes a vital tool in the dynamic world of digital media, combining creativity and scientific rigor to engage readers while maintaining accuracy. However, accuracy is currently hampered by numerical integration problems, which affect both abstractive and extractive approaches. Sentences that are extracted from the original material are typically too short to accurately represent complex information. Our research introduces an innovative two-step training technique to tackle these problems, emphasizing the significance of enhanced numerical reasoning in headline development. Promising advances are presented by utilizing text-to-text processing capabilities of the T5 model and advanced NLP approaches like BERT and RoBERTa. With the help of external contributions and our dataset, our Flan-T5 model has been improved to demonstrate how these methods may be used to overcome numerical integration issues and improve the accuracy of headline production.

pdf bib
TueSents at SemEval-2024 Task 8: Predicting the Shift from Human Authorship to Machine-generated Output in a Mixed Text
Valentin Pickard | Hoa Do

This paper describes our approach and resultsfor the SemEval 2024 task of identifying thetoken index in a mixed text where a switchfrom human authorship to machine-generatedtext occurs. We explore two BiLSTMs, oneover sentence feature vectors to predict theindex of the sentence containing such a changeand another over character embeddings of thetext. As sentence features, we compute tokencount, mean token length, standard deviationof token length, counts for punctuation andspace characters, various readability scores,word frequency class and word part-of-speechclass counts for each sentence. class counts.The evaluation is performed on mean absoluteerror (MAE) between predicted and actualboundary token index. While our competitionresults were notably below the baseline, theremay still be useful aspects to our approach.

pdf bib
TECHSSN1 at SemEval-2024 Task 10: Emotion Classification in Hindi-English Code-Mixed Dialogue using Transformer-based Models
Venkatasai Ojus Yenumulapalli | Pooja Premnath | Parthiban Mohankumar | Rajalakshmi Sivanaiah | Angel Deborah

The increase in the popularity of code mixed languages has resulted in the need to engineer language models for the same . Unlike pure languages, code-mixed languages lack clear grammatical structures, leading to ambiguous sentence constructions. This ambiguity presents significant challenges for natural language processing tasks, including syntactic parsing, word sense disambiguation, and language identification. This paper focuses on emotion recognition of conversations in Hinglish, a mix of Hindi and English, as part of Task 10 of SemEval 2024. The proposed approach explores the usage of standard machine learning models like SVM, MNB and RF, and also BERT-based models for Hindi-English code-mixed data- namely, HingBERT, Hing mBERT and HingRoBERTa for subtask A.

pdf bib
SHROOM-INDElab at SemEval-2024 Task 6: Zero- and Few-Shot LLM-Based Classification for Hallucination Detection
Bradley Allen | Fina Polat | Paul Groth

We describe the University of Amsterdam Intelligent Data Engineering Lab team’s entry for the SemEval-2024 Task 6 competition. The SHROOM-INDElab system builds on previous work on using prompt programming and in-context learning with large language models (LLMs) to build classifiers for hallucination detection, and extends that work through the incorporation of context-specific definition of task, role, and target concept, and automated generation of examples for use in a few-shot prompting approach. The resulting system achieved fourth-best and sixth-best performance in the model-agnostic track and model-aware tracks for Task 6, respectively, and evaluation using the validation sets showed that the system’s classification decisions were consistent with those of the crowdsourced human labelers. We further found that a zero-shot approach provided better accuracy than a few-shot approach using automatically generated examples. Code for the system described in this paper is available on Github.

pdf bib
I2C-Huelva at SemEval-2024 Task 8: Boosting AI-Generated Text Detection with Multimodal Models and Optimized Ensembles
Alberto Rodero Peña | Jacinto Mata Vazquez | Victoria Pachón Álvarez

With the rise of AI-based text generators, the need for effective detection mechanisms has become paramount. This paper presents new techniques for building adaptable models and optimizing training aspects for identifying synthetically produced texts across multiple generators and domains. The study, divided into binary and multilabel classification tasks, avoids overfitting through strategic training data limitation. A key innovation is the incorporation of multimodal models that blend numerical text features with conventional NLP approaches. The work also delves into optimizing ensemble model combinations via various voting methods, focusing on accuracy as the official metric. The optimized ensemble strategy demonstrates significant efficacy in both subtasks, highlighting the potential of multimodal and ensemble methods in enhancing the robustness of detection systems against emerging text generators.

pdf bib
Snarci at SemEval-2024 Task 4: Themis Model for Binary Classification of Memes
Luca Zedda | Alessandra Perniciano | Andrea Loddo | Cecilia Di Ruberto | Manuela Sanguinetti | Maurizio Atzori

This paper introduces an approach developed for multimodal meme analysis, specifically targeting the identification of persuasion techniques embedded within memes. Our methodology integrates Large Language Models (LLMs) and contrastive learning image encoders to discern the presence of persuasive elements in memes across diverse platforms. By capitalizing on the contextual understanding facilitated by LLMs and the discriminative power of contrastive learning for image encoding, our framework provides a robust solution for detecting and classifying memes with persuasion techniques. The system was used in Task 4 of Semeval 2024, precisely for Substask 2b (binary classification of presence of persuasion techniques). It showed promising results overall, achieving a Macro-F1=0.7986 on the English test data (i.e., the language the system was trained on) and Macro-F1=0.66777/0.47917/0.5554, respectively, on the other three “surprise” languages proposed by the task organizers, i.e., Bulgarian, North Macedonian and Arabic. The paper provides an overview of the system, along with a discussion of the results obtained and its main limitations.

pdf bib
Fired_from_NLP at SemEval-2024 Task 1: Towards Developing Semantic Textual Relatedness Predictor - A Transformer-based Approach
Anik Shanto | Md. Sajid Alam Chowdhury | Mostak Chowdhury | Udoy Das | Hasan Murad

Predicting semantic textual relatedness (STR) is one of the most challenging tasks in the field of natural language processing. Semantic relatedness prediction has real-life practical applications while developing search engines and modern text generation systems. A shared task on semantic textual relatedness has been organized by SemEval 2024, where the organizer has proposed a dataset on semantic textual relatedness in the English language under Shared Task 1 (Track A3). In this work, we have developed models to predict semantic textual relatedness between pairs of English sentences by training and evaluating various transformer-based model architectures, deep learning, and machine learning methods using the shared dataset. Moreover, we have utilized existing semantic textual relatedness datasets such as the stsb multilingual benchmark dataset, the SemEval 2014 Task 1 dataset, and the SemEval 2015 Task 2 dataset. Our findings show that in the SemEval 2024 Shared Task 1 (Track A3), the fine-tuned-STS-BERT model performed the best, scoring 0.8103 on the test set and placing 25th out of all participants.

pdf bib
BITS Pilani at SemEval-2024 Task 1: Using text-embedding-3-large and LaBSE embeddings for Semantic Textual Relatedness
Dilip Venkatesh | Sundaresan Raman

Semantic Relatedness of a pair of text (sentences or words) is the degree to which theirmeanings are close. The Track A of the Semantic Textual Relatedness shared task aimsto find the semantic relatedness for the English language along with multiple other lowresource languages with the use of pretrainedlanguage models. We proposes a system tofind the Spearman coefficient of a textual pairusing pretrained embedding models like textembedding-3-large and LaBSE.

pdf bib
SmurfCat at SemEval-2024 Task 6: Leveraging Synthetic Data for Hallucination Detection
Elisei Rykov | Yana Shishkina | Ksenia Petrushina | Ksenia Titova | Sergey Petrakov | Alexander Panchenko

In this paper, we present our novel systems developed for the SemEval-2024 hallucination detection task. Our investigation spans a range of strategies to compare model predictions with reference standards, encompassing diverse baselines, the refinement of pre-trained encoders through supervised learning, and an ensemble approaches utilizing several high-performing models. Through these explorations, we introduce three distinct methods that exhibit strong performance metrics. To amplify our training data, we generate additional training samples from unlabelled training subset. Furthermore, we provide a detailed comparative analysis of our approaches. Notably, our premier method achieved a commendable 9th place in the competition’s model-agnostic track and 20th place in model-aware track, highlighting its effectiveness and potential.

pdf bib
USTCCTSU at SemEval-2024 Task 1: Reducing Anisotropy for Cross-lingual Semantic Textual Relatedness Task
Jianjian Li | Shengwei Liang | Yong Liao | Hongping Deng | Haiyang Yu

Cross-lingual semantic textual relatedness task is an important research task that addresses challenges in cross-lingual communication and text understanding. It helps establish semantic connections between different languages, crucial for downstream tasks like machine translation, multilingual information retrieval, and cross-lingual text understanding.Based on extensive comparative experiments, we choose the XLM-R-base as our base model and use pre-trained sentence representations based on whitening to reduce anisotropy.Additionally, for the given training data, we design a delicate data filtering method to alleviate the curse of multilingualism. With our approach, we achieve a 2nd score in Spanish, a 3rd in Indonesian, and multiple entries in the top ten results in the competition’s track C. We further do a comprehensive analysis to inspire future research aimed at improving performance on cross-lingual tasks.

pdf bib
GreyBox at SemEval-2024 Task 4: Progressive Fine-tuning (for Multilingual Detection of Propaganda Techniques)
Nathan Roll | Calbert Graham

We introduce a novel fine-tuning approach that effectively primes transformer-based language models to detect rhetorical and psychological techniques within internet memes. Our end-to-end system retains multilingual and task-general capacities from pretraining stages while adapting to domain intricacies using an increasingly targeted set of examples– achieving competitive rankings across English, Bulgarian, and North Macedonian. We find that our monolingual post-training regimen is sufficient to improve task performance in 17 language varieties beyond equivalent zero-shot capabilities despite English-only data. To promote further research, we release our code publicly on GitHub.

pdf bib
NLU-STR at SemEval-2024 Task 1: Generative-based Augmentation and Encoder-based Scoring for Semantic Textual Relatedness
Sanad Malaysha | Mustafa Jarrar | Mohammed Khalilia

Semantic textual relatedness is a broader concept of semantic similarity. It measures the extent to which two chunks of text convey similar meaning or topics, or share related concepts or contexts. This notion of relatedness can be applied in various applications, such as document clustering and summarizing. SemRel-2024, a shared task in SemEval-2024, aims at reducing the gap in the semantic relatedness task by providing datasets for fourteen languages and dialects including Arabic. This paper reports on our participation in Track A (Algerian and Moroccan dialects) and Track B (Modern Standard Arabic). A BERT-based model is augmented and fine-tuned for regression scoring in supervised track (A), while BERT-based cosine similarity is employed for unsupervised track (B). Our system ranked 1st in SemRel-2024 for MSA with a Spearman correlation score of 0.49. We ranked 5th for Moroccan and 12th for Algerian with scores of 0.83 and 0.53, respectively.

pdf bib
scaLAR SemEval-2024 Task 1: Semantic Textual Relatednes for English
Anand Kumar | Hemanth Kumar

This study investigates Semantic TextualRelated- ness (STR) within Natural LanguageProcessing (NLP) through experiments conducted on a dataset from the SemEval-2024STR task. The dataset comprises train instances with three features (PairID, Text, andScore) and test instances with two features(PairID and Text), where sentence pairs areseparated by '/n’ in the Text column. UsingBERT(sentence transformers pipeline), we explore two approaches: one with fine-tuning(Track A: Supervised) and another without finetuning (Track B: UnSupervised). Fine-tuningthe BERT pipeline yielded a Spearman correlation coefficient of 0.803, while without finetuning, a coefficient of 0.693 was attained usingcosine similarity. The study concludes by emphasizing the significance of STR in NLP tasks,highlighting the role of pre-trained languagemodels like BERT and Sentence Transformersin enhancing semantic relatedness assessments.

pdf bib
TECHSSN at SemEval-2024 Task 1: Multilingual Analysis for Semantic Textual Relatedness using Boosted Transformer Models
Shreejith Babu G | Ravindran V | Aashika Jetti | Rajalakshmi Sivanaiah | Angel Deborah

This paper presents our approach to SemEval- 2024 Task 1: Semantic Textual Relatedness (STR). Out of the 14 languages provided, we specifically focused on English and Telugu. Our proposal employs advanced natural language processing techniques and leverages the Sentence Transformers library for sentence embeddings. For English, a Gradient Boosting Regressor trained on DistilBERT embeddingsachieves competitive results, while for Telugu, a multilingual model coupled with hyperparameter tuning yields enhanced performance. The paper discusses the significance of semantic relatedness in various languages, highlighting the challenges and nuances encountered. Our findings contribute to the understanding of semantic textual relatedness across diverse linguistic landscapes, providing valuable insights for future research in multilingual natural language processing.

pdf bib
Noot Noot at SemEval-2024 Task 7: Numerical Reasoning and Headline Generation
Sankalp Bahad | Yash Bhaskar | Parameswari Krishnamurthy

Natural language processing (NLP) modelshave achieved remarkable progress in recentyears, particularly in tasks related to semanticanalysis. However, many existing benchmarksprimarily focus on lexical and syntactic un-derstanding, often overlooking the importanceof numerical reasoning abilities. In this pa-per, we argue for the necessity of incorporatingnumeral-awareness into NLP evaluations andpropose two distinct tasks to assess this capabil-ity: Numerical Reasoning and Headline Gener-ation. We present datasets curated for each taskand evaluate various approaches using both au-tomatic and human evaluation metrics. Ourresults demonstrate the diverse strategies em-ployed by participating teams and highlight thepromising performance of emerging modelslike Mixtral 8x7b instruct. We discuss the im-plications of our findings and suggest avenuesfor future research in advancing numeral-awarelanguage understanding and generation.

pdf bib
Fine-tuning Language Models for AI vs Human Generated Text detection
Sankalp Bahad | Yash Bhaskar | Parameswari Krishnamurthy

In this paper, we introduce a machine-generated text detection system designed totackle the challenges posed by the prolifera-tion of large language models (LLMs). Withthe rise of LLMs such as ChatGPT and GPT-4,there is a growing concern regarding the po-tential misuse of machine-generated content,including misinformation dissemination. Oursystem addresses this issue by automating theidentification of machine-generated text acrossmultiple subtasks: binary human-written vs.machine-generated text classification, multi-way machine-generated text classification, andhuman-machine mixed text detection. We em-ploy the RoBERTa Base model and fine-tuneit on a diverse dataset encompassing variousdomains, languages, and sources. Throughrigorous evaluation, we demonstrate the effec-tiveness of our system in accurately detectingmachine-generated text, contributing to effortsaimed at mitigating its potential misuse.

pdf bib
eagerlearners at SemEval2024 Task 5: The Legal Argument Reasoning Task in Civil Procedure
Hoorieh Sabzevari | Mohammadmostafa Rostamkhani | Sauleh Eetemadi

This study investigates the performance of the zero-shot method in classifying data using three large language models, alongside two models with large input token sizes and the two pre-trained models on legal data. Our main dataset comes from the domain of U.S. civil procedure. It includes summaries of legal cases, specific questions, potential answers, and detailed explanations for why each solution is relevant, all sourced from a book aimed at law students. By comparing different methods, we aimed to understand how effectively they handle the complexities found in legal datasets. Our findings show how well the zero-shot method of large language models can understand complicated data. We achieved our highest F1 score of 64% in these experiments.

pdf bib
TrustAI at SemEval-2024 Task 8: A Comprehensive Analysis of Multi-domain Machine Generated Text Detection Techniques
Ashok Urlana | Aditya Saibewar | Bala Mallikarjunarao Garlapati | Charaka Vinayak Kumar | Ajeet Singh | Srinivasa Rao Chalamala

The Large Language Models (LLMs) exhibit remarkable ability to generate fluent content across a wide spectrum of user queries. However, this capability has raised concerns regarding misinformation and personal information leakage. In this paper, we present our methods for the SemEval2024 Task8, aiming to detect machine-generated text across various domains in both mono-lingual and multi-lingual contexts. Our study comprehensively analyzes various methods to detect machine-generated text, including statistical, neural, and pre-trained model approaches. We also detail our experimental setup and perform a in-depth error analysis to evaluate the effectiveness of these methods. Our methods obtain an accuracy of 86.9% on the test set of subtask-A mono and 83.7% for subtask-B. Furthermore, we also highlight the challenges and essential factors for consideration in future studies.

pdf bib
Pinealai at SemEval-2024 Task 1: Exploring Semantic Relatedness Prediction using Syntactic, TF-IDF, and Distance-Based Features.
Alex Eponon | Luis Ramos Perez

The central aim of this experiment is to establish a system proficient in predicting semantic relatedness between pairs of English texts. Additionally, the study seeks to delve into diverse features capable of enhancing the ability of models to identify semantic relatedness within given sentences. Several strategies have been used that combine TF-IDF, syntactic features, and similarity measures to train machine learning to predict semantic relatedness between pairs of sentences. The results obtained were above the baseline with an approximate Spearman score of 0.84.

pdf bib
Infrrd.ai at SemEval-2024 Task 7: RAG-based end-to-end training to generate headlines and numbers
Jianglong He | Saiteja Tallam | Srirama Nakshathri | Navaneeth Amarnath | Pratiba Kr | Deepak Kumar

We propose a training algorithm based on retrieval-augmented generation (RAG) to obtain the most similar training samples. The training samples obtained are used as a reference to perform contextual learning-based fine-tuning of large language models (LLMs). We use the proposed method to generate headlines and extract numerical values from unstructured text. Models are made aware of the presence of numbers in the unstructured text with extended markup language (XML) tags specifically designed to capture the numbers. The headlines of unstructured text are preprocessed to wrap the number and then presented to the model. A number of mathematical operations are also passed as references to cover the chain-of-thought (COT) approach. Therefore, the model can calculate the final value passed to a mathematical operation. We perform the validation of numbers as a post-processing step to verify whether the numerical value calculated by the model is correct or not. The automatic validation of numbers in the generated headline helped the model achieve the best results in human evaluation among the methods involved.

pdf bib
AlphaIntellect at SemEval-2024 Task 6: Detection of Hallucinations in Generated Text
Sohan Choudhury | Priyam Saha | Subharthi Ray | Shankha Das | Dipankar Das

One major issue in natural language generation (NLG) models is detecting hallucinations (semantically inaccurate outputs). This study investigates a hallucination detection system designed for three distinct NLG tasks: definition modeling, paraphrase generation, and machine translation. The system uses feedforward neural networks for classification and SentenceTransformer models for similarity scores and sentence embeddings. Even though the SemEval-2024 benchmark shows good results, there is still room for improvement. Promising paths toward improving performance include considering multi-task learning methods, including strategies for handling out-of-domain data minimizing bias, and investigating sophisticated architectures.

pdf bib
YSP at SemEval-2024 Task 1: Enhancing Sentence Relatedness Assessment using Siamese Networks
Yasamin Aali | Sardar Hamidian | Parsa Farinneya

In this paper we present the system for Track A in the SemEval-2024 Task 1: Semantic Textual Relatedness for African and Asian Languages (STR). The proposed system integrates a Siamese Network architecture with pre-trained language models, including BERT, RoBERTa, and the Universal Sentence Encoder (USE). Through rigorous experimentation and analysis, we evaluate the performance of these models across multiple languages. Our findings reveal that the Universal Sentence Encoder excels in capturing semantic similarities, outperforming BERT and RoBERTa in most scenarios. Particularly notable is the USE’s exceptional performance in English and Marathi. These results emphasize the importance of selecting appropriate pre-trained models based on linguistic considerations and task requirements.

pdf bib
NootNoot At SemEval-2024 Task 6: Hallucinations and Related Observable Overgeneration Mistakes Detection
Sankalp Bahad | Yash Bhaskar | Parameswari Krishnamurthy

Semantic hallucinations in neural language gen-eration systems pose a significant challenge tothe reliability and accuracy of natural languageprocessing applications. Current neural mod-els often produce fluent but incorrect outputs,undermining the usefulness of generated text.In this study, we address the task of detectingsemantic hallucinations through the SHROOM(Semantic Hallucinations Real Or Mistakes)dataset, encompassing data from diverse NLGtasks such as definition modeling, machinetranslation, and paraphrase generation. We in-vestigate three methodologies: fine-tuning onlabelled training data, fine-tuning on labelledvalidation data, and a zero-shot approach usingthe Mixtral 8x7b instruct model. Our resultsdemonstrate the effectiveness of these method-ologies in identifying semantic hallucinations,with the zero-shot approach showing compet-itive performance without additional training.Our findings highlight the importance of robustdetection mechanisms for ensuring the accu-racy and reliability of neural language genera-tion systems.

pdf bib
Transformers at SemEval-2024 Task 5: Legal Argument Reasoning Task in Civil Procedure using RoBERTa
Kriti Singhal | Jatin Bedi

Legal argument reasoning task in civil procedure is a new NLP task utilizing a dataset from the domain of the U.S. civil procedure. The task aims at identifying whether the solution to a question in the legal domain is correct or not. This paper describes the team “Transformers” submission to the Legal Argument Reasoning Task in Civil Procedure shared task at SemEval-2024 Task 5. We use a BERT-based architecture for the shared task. The highest F1-score score and accuracy achieved was 0.6172 and 0.6531 respectively. We secured the 13th rank in the Legal Argument Reasoning Task in Civil Procedure shared task.

pdf bib
YNU-HPCC at SemEval-2024 Task 7: Instruction Fine-tuning Models for Numerical Understanding and Generation
Kaiyuan Chen | Jin Wang | Xuejie Zhang

This paper presents our systems for Task 7, Numeral-Aware Language Understanding and Generation of SemEval 2024. As participants of Task 7, we engage in all subtasks and implement corresponding systems for each subtask. All subtasks cover three aspects: Quantitative understanding (English), Reading Comprehension of the Numbers in the text (Chinese), and Numeral-Aware Headline Generation (English). Our approach explores employing instruction-tuned models (Flan-T5) or text-to-text models (T5) to accomplish the respective subtasks. We implement the instruction fine-tuning with or without demonstrations and employ similarity-based retrieval or manual methods to construct demonstrations for each example in instruction fine-tuning. Moreover, we reformulate the model’s output into a chain-of-thought format with calculation expressions to enhance its reasoning performance for reasoning subtasks. The competitive results in all subtasks demonstrate the effectiveness of our systems.

pdf bib
CAILMD-23 at SemEval-2024 Task 1: Multilingual Evaluation of Semantic Textual Relatedness
Srushti Sonavane | Sharvi Endait | Ridhima Sinare | Pritika Rohera | Advait Naik | Dipali Kadam

The explosive growth of online content demands robust Natural Language Processing (NLP) techniques that can capture nuanced meanings and cultural context across diverse languages. Semantic Textual Relatedness (STR) goes beyond superficial word overlap, considering linguistic elements and non-linguistic factors like topic, sentiment, and perspective. Despite its pivotal role, prior NLP research has predominantly focused on English, limiting its applicability across languages. Addressing this gap, our paper dives into capturing deeper connections between sentences beyond simple word overlap. Going beyond English-centric NLP research, we explore STR in Marathi, Hindi, Spanish, and English, unlocking the potential for information retrieval, machine translation, and more. Leveraging the SemEval-2024 shared task, we explore various language models across three learning paradigms: supervised, unsupervised, and cross-lingual. Our comprehensive methodology gains promising results, demonstrating the effectiveness of our approach. This work aims to not only showcase our achievements but also inspire further research in multilingual STR, particularly for low-resourced languages.

pdf bib
SEME at SemEval-2024 Task 2: Comparing Masked and Generative Language Models on Natural Language Inference for Clinical Trials
Mathilde Aguiar | Pierre Zweigenbaum | Nona Naderi

This paper describes our submission to Task 2 of SemEval-2024: Safe Biomedical Natural Language Inference for Clinical Trials. The Multi-evidence Natural Language Inference for Clinical Trial Data (NLI4CT) consists of a Textual Entailment (TE) task focused on the evaluation of the consistency and faithfulness of Natural Language Inference (NLI) models applied to Clinical Trial Reports (CTR). We test 2 distinct approaches, one based on finetuning and ensembling Masked Language Models and the other based on prompting Large Language Models using templates, in particular, using Chain-Of-Thought and Contrastive Chain-Of-Thought. Prompting Flan-T5-large in a 2-shot setting leads to our best system that achieves 0.57 F1 score, 0.64 Faithfulness, and 0.56 Consistency.

pdf bib
MAINDZ at SemEval-2024 Task 5: CLUEDO - Choosing Legal oUtcome by Explaining Decision through Oversight
Irene Benedetto | Alkis Koudounas | Lorenzo Vaiani | Eliana Pastor | Luca Cagliero | Francesco Tarasconi

Large language models (LLMs) have recently obtained strong performance on complex reasoning tasks. However, their capabilities in specialized domains like law remain relatively unexplored. We present CLUEDO, a system to tackle a novel legal reasoning task that involves determining if a provided answer correctly addresses a legal question derived from U.S. civil procedure cases. CLUEDO utilizes multiple collaborator models that are trained using multiple-choice prompting to choose the right label and generate explanations. These collaborators are overseen by a final “detective” model that identifies the most accurate answer in a zero-shot manner. Our approach achieves an F1 macro score of 0.74 on the development set and 0.76 on the test set, outperforming individual models. Unlike the powerful GPT-4, CLUEDO provides more stable predictions thanks to the ensemble approach. Our results showcase the promise of tailored frameworks to enhance legal reasoning capabilities in LLMs.

pdf bib
Groningen Group E at SemEval-2024 Task 8: Detecting machine-generated texts through pre-trained language models augmented with explicit linguistic-stylistic features
Patrick Darwinkel | Sijbren Van Vaals | Marieke Van Der Holt | Jarno Van Houten

Our approach to detecting machine-generated text for the SemEval-2024 Task 8 combines a wide range of linguistic-stylistic features with pre-trained language models (PLM). Experiments using random forests and PLMs resulted in an augmented DistilBERT system for subtask A and B and an augmented Longformer for subtask C. These systems achieved accuracies of 0.63 and 0.77 for the mono- and multilingual tracks of subtask A, 0.64 for subtask B and a MAE of 26.07 for subtask C. Although lower than the task organizer’s baselines, we demonstrate that linguistic-stylistic features are predictors for whether a text was authored by a model (and if so, which one).

pdf bib
Magnum JUCSE at SemEval-2024 Task 4: Multilingual Detection of Persuasion Techniques in Memes
Adnan Khurshid | Dipankar Das

This paper focuses on the task of detecting persuasion techniques organised in a hierarchy within meme text in multiple languages like English, North Macedonian, Arabic and Bulgarian, exploring the ways in which textual elements contribute to the dissemination of persuasive messages.The main strategy of the system is to train a binary classifier for each node in the hierarchy and predict labels in a top down fashion by seeing the confidence value of the prediction at any node. For each unique label in the hierarchy, a dataset is created from the original dataset which is then used to train the binary classifier for that label

pdf bib
Tübingen-CL at SemEval-2024 Task 1: Ensemble Learning for Semantic Relatedness Estimation
Leixin Zhang | Çağrı Çöltekin

The paper introduces our system for SemEval-2024 Task 1, which aims to predict the relatedness of sentence pairs. Operating under the hypothesis that semantic relatedness is a broader concept that extends beyond mere similarity of sentences, our approach seeks to identify useful features for relatedness estimation. We employ an ensemble approach integrating various systems, including statistical textual features and outputs of deep learning models to predict relatedness scores. The findings suggest that semantic relatedness can be inferred from various sources and ensemble models outperform many individual systems in estimating semantic relatedness.

pdf bib
SemEval Task 8: A Comparison of Traditional and Neural Models for Detecting Machine Authored Text
Srikar Kashyap Pulipaka | Shrirang Mhalgi | Joseph Larson | Sandra Kübler

Since Large Language Models have reached a stage where it is becoming more and more difficult to distinguish between human and machine written text, there is an increasing need for automated systems to distinguish between them. As part of SemEval Task 8, Subtask A: Binary Human-Written vs. Machine-Generated Text Classification, we explore a variety of machine learning classifiers, from traditional statistical methods, such as Naïve Bayes and Decision Trees, to fine-tuned transformer models, suchas RoBERTa and ALBERT. Our findings show that using a fine-tuned RoBERTa model with optimizedhyperparameters yields the best accuracy. However, the improvement does not translate to the test set because of the differences in distribution in the development and test sets.

pdf bib
RACAI at SemEval-2024 Task 10: Combining algorithms for code-mixed Emotion Recognition in Conversation
Sara Niță | Vasile Păiș

Code-mixed emotion recognition constitutes a challenge for NLP research due to the text’s deviation from the traditional grammatical structure of the original languages. This paper describes the system submitted by the RACAI Team for the SemEval 2024 Task 10 - EDiReF subtasks 1: Emotion Recognition in Conversation (ERC) in Hindi-English code-mixed conversations. We propose a system that combines a transformer-based model with two simple neural networks.

pdf bib
ROSHA at SemEval-2024 Task 9: BRAINTEASER A Novel Task Defying Common Sense
Mohammadmostafa Rostamkhani | Shayan Mousavinia | Sauleh Eetemadi

In our exploration of SemEval 2024 Task 9, specifically the challenging BRAINTEASER: A Novel Task Defying Common Sense, we employed various strategies for the BRAINTEASER QA task, which encompasses both sentence and word puzzles. In the initial approach, we applied the XLM-RoBERTa model both to the original training dataset and concurrently to the original dataset alongside the BiRdQA dataset and the original dataset alongside RiddleSense for comprehensive model training.Another strategy involved expanding each word within our BiRdQA dataset into a full sentence. This unique perspective aimed to enhance the semantic impact of individual words in our training regimen for word puzzle (WP) riddles. Utilizing ChatGPT-3.5, we extended each word into an extensive sentence, applying this process to all options within each riddle.Furthermore, we explored the implementation of RECONCILE (Round-table conference) using three prominent large language models—ChatGPT, Gemini, and the Mixtral-8x7B Large Language Model (LLM). As a final approach, we leveraged GPT-4 results. Remarkably, our most successful experiment yielded noteworthy results, achieving a score of 0.900 for sentence puzzles (S_ori) and 0.906 for word puzzles (W_ori).

pdf bib
Sharif-STR at SemEval-2024 Task 1: Transformer as a Regression Model for Fine-Grained Scoring of Textual Semantic Relations
Seyedeh Fatemeh Ebrahimi | Karim Akhavan Azari | Amirmasoud Iravani | Hadi Alizadeh | Zeinab Taghavi | Hossein Sameti

This paper explores semantic textual relatedness (STR) using fine-tuning techniques on the RoBERTa transformer model, focusing on sentence-level STR within Track A (Supervised). The study evaluates the effectiveness of this approach across different languages, with promising results in English and Spanish but encountering challenges in Arabic.

pdf bib
DUTh at SemEval 2024 Task 5: A multi-task learning approach for the Legal Argument Reasoning Task in Civil Procedure
Ioannis Maslaris | Avi Arampatzis

Text-generative models have proven to be good reasoners. Although reasoning abilities are mostly observed in larger language models, a number of strategies try to transfer this skill to smaller language models. This paper presents our approach to SemEval 2024 Task-5: The Legal Argument Reasoning Task in Civil Procedure. This shared task aims to develop a system that efficiently handles a multiple-choice question-answering task in the context of the US civil procedure domain. The dataset provides a human-generated rationale for each answer. Given the complexity of legal issues, this task certainly challenges the reasoning abilities of LLMs and AI systems in general. Our work explores fine-tuning an LLM as a correct/incorrect answer classifier. In this context, we are making use of multi-task learning toincorporate the rationales into the fine-tuning process.

pdf bib
MAMET at SemEval-2024 Task 7: Supervised Enhanced Reasoning Agent Model
Mahmood Kalantari | Mehdi Feghhi | Taha Khany Alamooti

In the intersection of language understanding and numerical reasoning, a formidable challenge arises in natural language processing (NLP). Our study delves into the realm of NumEval, focusing on numeral-aware language understanding and generation using the QP, QQA and QNLI datasets. We harness the potential of the Orca2 model, Fine-tuning it in both normal and Chain-of-Thought modes with prompt tuning to enhance accuracy. Despite initial conjectures, our findings reveal intriguing disparities in model performance. While standard training methodologies yield commendable accuracy rates. The core contribution of this work lies in its elucidation of the intricate interplay between dataset sequencing and model performance. We expected to achieve a general model with the Fine Tuning model on the QP and QNLI datasets respectively, which has good accuracy in all three datasets. However, this goal was not achieved, and in order to achieve this goal, we introduce our structure.

pdf bib
DUTh at SemEval-2024 Task 6: Comparing Pre-trained Models on Sentence Similarity Evaluation for Detecting of Hallucinations and Related Observable Overgeneration Mistakes
Ioanna Iordanidou | Ioannis Maslaris | Avi Arampatzis

In this paper, we present our approach toSemEval-2024 Task 6: SHROOM, a Sharedtask on Hallucinations and Related ObservableOvergeneration Mistakes, which aims to determine weather AI generated text is semanticallycorrect or incorrect. This work is a comparative study of Large Language Models (LLMs)in the context of the task, shedding light ontheir effectiveness and nuances. We present asystem that leverages pre-trained LLMs, suchas LaBSE, T5, and DistilUSE, for binary classification of given sentences into ‘Hallucination’or ‘Not Hallucination’ classes by evaluatingthe model’s output against the reference correct text. Moreover, beyond utilizing labeleddatasets, our methodology integrates syntheticlabel creation in unlabeled datasets, followedby the prediction of test labels.

pdf bib
MBZUAI-UNAM at SemEval-2024 Task 1: Sentence-CROBI, a Simple Cross-Bi-Encoder-Based Neural Network Architecture for Semantic Textual Relatedness
Jesus German Ortiz Barajas | Gemma Bel-enguix | Helena Goméz-adorno

The Semantic Textual Relatedness (STR) shared task aims at detecting the degree of semantic relatedness between pairs of sentences on low-resource languages from Afroasiatic, Indoeuropean, Austronesian, Dravidian, and Nigercongo families. We use the Sentence-CROBI architecture to tackle this problem. The model is adapted from its original purpose of paraphrase detection to explore its capacities in a related task with limited resources and in multilingual and monolingual settings. Our approach combines the vector representation of cross-encoders and bi-encoders and possesses high adaptable capacity by combining several pre-trained models. Our system obtained good results on the low-resource languages of the dataset using a multilingual fine-tuning approach.

pdf bib
DUTh at SemEval 2024 Task 8: Comparing classic Machine Learning Algorithms and LLM based methods for Multigenerator, Multidomain and Multilingual Machine-Generated Text Detection
Theodora Kyriakou | Ioannis Maslaris | Avi Arampatzis

Text-generative models evolve rapidly nowadays. Although, they are very useful tools for a lot of people, they have also raised concerns for different reasons. This paper presents our work for SemEval2024 Task-8 on 2 out of the 3 subtasks. This shared task aims at finding automatic models for making AI vs. human written text classification easier. Our team, after trying different preprocessing, several Machine Learning algorithms, and some LLMs, ended up with mBERT, XLM-RoBERTa, and BERT for the tasks we submitted. We present both positive and negative methods, so that future researchers are informed about what works and what doesn’t.

pdf bib
 Sina Alinejad at SemEval-2024 Task 7: Numeral Prediction using gpt3.5
Sina Alinejad | Erfan Moosavi Monazzah

pdf bib
IUSTNLPLAB at SemEval-2024 Task 4: Multilingual Detection of Persuasion Techniques in Memes
Mohammad Osoolian | Erfan Moosavi Monazzah | Sauleh Eetemadi

This paper outlines our approach to SemEval-2024 Task 4: Multilingual Detection of Persuasion Techniques in Memes, specifically addressing subtask 1. The study focuses on model fine-tuning using language models, including BERT, GPT-2, and RoBERTa, with the experiment results demonstrating optimal performance with GPT-2. Our system submission achieved a competitive ranking of 17th out of 33 teams in subtask 1, showcasing the effectiveness of the employed methodology in the context of persuasive technique identification within meme texts.

pdf bib
PWEITINLP at SemEval-2024 Task 3: Two Step Emotion Cause Analysis
Sofiia Levchenko | Rafał Wolert | Piotr Andruszkiewicz

ECPE (emotion cause pair extraction) task was introduced to solve the shortcomings of ECE (emotion cause extraction). Models with sequential data processing abilities or complex architecture can be utilized to solve this task. Our contribution to solving Subtask 1: Textual Emotion-Cause Pair Extraction in Conversations defined in the SemEval-2024 Task 3: The Competition of Multimodal Emotion Cause Analysis in Conversations is to create a two-step solution to the ECPE task utilizing GPT-3 for emotion classification and SpanBERT for extracting the cause utterances.

pdf bib
IUST-NLPLAB at SemEval-2024 Task 9: BRAINTEASER By MPNet (Sentence Puzzle)
Mohammad Hossein Abbaspour | Erfan Moosavi Monazzah | Sauleh Eetemadi

This study addresses a task encompassing two distinct subtasks: Sentence-puzzle and Word-puzzle. Our primary focus lies within the Sentence-puzzle subtask, which involves discerning the correct answer from a set of three options for a given riddle constructed from sentence fragments. We propose four distinct methodologies tailored to address this subtask effectively. Firstly, we introduce a zero-shot approach leveraging the capabilities of the GPT-3.5 model. Additionally, we present three fine-tuning methodologies utilizing MPNet as the underlying architecture, each employing a different loss function. We conduct comprehensive evaluations of these methodologies on the designated task dataset and meticulously document the obtained results. Furthermore, we conduct an in-depth analysis to ascertain the respective strengths and weaknesses of each method. Through this analysis, we aim to provide valuable insights into the challenges inherent to this task domain.

pdf bib
iimasNLP at SemEval-2024 Task 8: Unveiling structure-aware language models for automatic generated text identification
Andric Valdez | Fernando Márquez | Jorge Pantaleón | Helena Gómez | Gemma Bel-enguix

Large language models (LLMs) are artificial intelligence systems that can generate text, translate languages, and answer questions in a human-like way. While these advances are impressive, there is concern that LLMs could also be used to generate fake or misleading content. In this work, as a part of our participation in SemEval-2024 Task-8, we investigate the ability of LLMs to identify whether a given text was written by a human or by a specific AI. We believe that human and machine writing style patterns are different from each other, so integrating features at different language levels can help in this classification task. For this reason, we evaluate several LLMs that aim to extract valuable multilevel information (such as lexical, semantic, and syntactic) from the text in their training processing. Our best scores on Sub- taskA (monolingual) and SubtaskB were 71.5% and 38.2% in accuracy, respectively (both using the ConvBERT LLM); for both subtasks, the baseline (RoBERTa) achieved an accuracy of 74%.

pdf bib
INGEOTEC at SemEval-2024 Task 10: Bag of Words Classifiers
Daniela Moctezuma | Eric Tellez | Jose Ortiz Bejar | Mireya Paredes

The Emotion Recognition in Conversation subtask aims to predict the emotions of the utterance of a conversation. In its most basic form, one can treat each utterance separately without considering that it is part of a conversation. Using this simplification, one can use any text classification algorithm to tackle this problem. This contribution follows this approach by solving the problem with different text classifiers based on Bag of Words. Nonetheless, the best approach takes advantage of the dynamics of the conversation; however, this algorithm is not statistically different than a Bag of Words with a Linear Support Vector Machine.

pdf bib
IIMAS at SemEval-2024 Task 9: A Comparative Approach for Brainteaser Solutions
Cecilia Reyes | Orlando Ramos-flores | Diego Martínez-maqueda

In this document, we detail our participation experience in SemEval-2024 Task 9: BRAINTEASER-A Novel Task Defying Common Sense. We tackled this challenge by applying fine-tuning techniques with pre-trained models (BERT and RoBERTa Winogrande), while also augmenting the dataset with the LLMs ChatGPT and Gemini. We achieved an accuracy of 0.93 with our best model, along with an F1 score of 0.87 for the Entailment class, 0.94 for the Contradiction class, and 0.96 for the Neutral class

pdf bib
PetKaz at SemEval-2024 Task 3: Advancing Emotion Classification with an LLM for Emotion-Cause Pair Extraction in Conversations
Roman Kazakov | Kseniia Petukhova | Ekaterina Kochmar

In this paper, we present our submission to the SemEval-2023 Task 3 “The Competition of Multimodal Emotion Cause Analysis in Conversations”, focusing on extracting emotion-cause pairs from dialogs. Specifically, our approach relies on combining fine-tuned GPT-3.5 for emotion classification and using a BiLSTM-based neural network to detect causes. We score 2nd in the ranking for Subtask 1, demonstrating the effectiveness of our approach through one of the highest weighted-average proportional F1 scores recorded at 0.264.

pdf bib
SCaLAR at SemEval-2024 Task 8: Unmasking the machine : Exploring the power of RoBERTa Ensemble for Detecting Machine Generated Text
Anand Kumar | Abhin B | Sidhaarth Murali

SemEval SubtaskB, a shared task that is concerned with the detection of text generated by one out of the 5 different models - davinci, bloomz, chatGPT, cohere and dolly. This is an important task considering the boom of generative models in the current day scenario and their ability to draft mails, formal documents, write and qualify exams and many more which keep evolving every passing day. The purpose of classifying text as generated by which pre-trained model helps in analyzing how each of the training data has affected the ability of the model in performing a certain given task. In the proposed approach, data augmentation was done in order to handle lengthier sentences and also labelling them with the same parent label. Upon the augmented data three RoBERTa models were trained on different segments of data which were then ensembled using a voting classifier based on their R2 score to achieve a higher accuracy than the individual models itself. The proposed model achieved an overall validation accuracy of 97.05% and testing accuracy of 76.25%. and our standing was 18th position on the leaderboard.

pdf bib
PetKaz at SemEval-2024 Task 8: Can Linguistics Capture the Specifics of LLM-generated Text?
Kseniia Petukhova | Roman Kazakov | Ekaterina Kochmar

In this paper, we present our submission to the SemEval-2024 Task 8 “Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text Detection”, focusing on the detection of machine-generated texts (MGTs) in English. Specifically, our approach relies on combining embeddings from the RoBERTa-base with diversity features and uses a resampled training set. We score 16th from 139 in the ranking for Subtask A, and our results show that our approach is generalizable across unseen models and domains, achieving an accuracy of 0.91.

pdf bib
SLPL SHROOM at SemEval2024 Task 06 : A comprehensive study on models ability to detect hallucination
Pouya Fallah | Soroush Gooran | Mohammad Jafarinasab | Pouya Sadeghi | Reza Farnia | Amirreza Tarabkhah | Zeinab Sadat Taghavi | Hossein Sameti

Language models, particularly generative models, are susceptible to hallucinations, generating outputs that contradict factual knowledgeor the source text. This study explores methodsfor detecting hallucinations in three SemEval2024 Task 6 tasks: Machine Translation, Definition Modeling, and Paraphrase Generation.We evaluate two methods: semantic similaritybetween the generated text and factual references, and an ensemble of language modelsthat judge each other’s outputs. Our resultsshow that semantic similarity achieves moderate accuracy and correlation scores in trial data,while the ensemble method offers insights intothe complexities of hallucination detection butfalls short of expectations. This work highlights the challenges of hallucination detectionand underscores the need for further researchin this critical area.

pdf bib
INGEOTEC at SemEval-2024 Task 1: Bag of Words and Transformers
Daniela Moctezuma | Eric Tellez | Mario Graff

Understanding the meaning of a written message is crucial in solving problems related to Natural Language Processing; the relatedness of two or more messages is a semantic problem tackled with supervised and unsupervised learning. This paper outlines our submissions to the Semantic Textual Relatedness (STR) challenge at SemEval 2024, which is devoted to evaluating the degree of semantic similarity and relatedness between two sentences across multiple languages. We use two main strategies in our submissions. The first approach is based on the Bag-of-Word scheme, while the second one uses pre-trained Transformers for text representation. We found some attractive results, especially in cases where different models adjust better to certain languages over others.

pdf bib
OctavianB at SemEval-2024 Task 6: An exploration of humanlike qualities of hallucinated LLM texts
Octavian Brodoceanu

The tested method for detection involves utilizing models, trained for differentiating machine-generated text, in order to distinguish between regular and hallucinated sequences. The hypothesis under investigation is that the patterns learned in pretraining will be transferable to the task at hand. The rationale is as follows: the training data of the model is human-written text, therefore deviations from the training set could be detected in this manner.A second method has been added post competition as a further exploration of the dataset involving using the loss of the generation as determined by a pretrained LLM.

pdf bib
FI Group at SemEval-2024 Task 8: A Syntactically Motivated Architecture for Multilingual Machine-Generated Text Detection
Maha Ben-fares | Urchade Zaratiana | Simon Hernandez | Pierre Holat

In this paper, we present the description of our proposed system for Subtask A - multilingual track at SemEval-2024 Task 8, which aims to classify if text has been generated by an AI or Human. Our approach treats binary text classification as token-level prediction, with the final classification being the average of token-level predictions. Through the use of rich representations of pre-trained transformers, our model is trained to selectively aggregate information from across different layers to score individual tokens, given that each layer may contain distinct information. Notably, our model demonstrates competitive performance on the test dataset, achieving an accuracy score of 95.8%. Furthermore, it secures the 2nd position in the multilingual track of Subtask A, with a mere 0.1% behind the leading system.

pdf bib
Team Innovative at SemEval-2024 Task 8: Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text Detection
Surbhi Sharma | Irfan Mansuri

With the widespread adoption of large language models (LLMs), such as ChatGPT and GPT-4, in various domains, concerns regarding their potential misuse, including spreading misinformation and disrupting education, have escalated. The need to discern between human-generated and machine-generated text has become increasingly crucial. This paper addresses the challenge of automatic text classification with a focus on distinguishing between human-written and machine-generated text. Leveraging the robust capabilities of the RoBERTa model, we propose an approach for text classification, termed as RoBERTa hybrid, which involves fine-tuning the pre-trained Roberta model coupled with additional dense layers and softmax activation for authorship attribution. In this paper, we present an approach that leverages Stylometric features, hybrid features, and the output probabilities of a fine-tuned RoBERTa model. Our method achieves a test accuracy of 73% and a validation accuracy of 89%, demonstrating promising advancements in the field of machine-generated text detection. These results mark significant progress in the domain of machine-generated text detection, as evidenced by our 74th position on the leaderboard for Subtask-A of SemEval-2024 Task 8.

pdf bib
EURECOM at SemEval-2024 Task 4: Hierarchical Loss and Model Ensembling in Detecting Persuasion Techniques
Youri Peskine | Raphael Troncy | Paolo Papotti

This paper describes the submission of team EURECOM at SemEval-2024 Task 4: Multilingual Detection of Persuasion Techniques in Memes. We only tackled the first sub-task, consisting of detecting 20 named persuasion techniques in the textual content of memes. We trained multiple BERT-based models (BERT, RoBERTa, BERT pre-trained on harmful detection) using different losses (Cross Entropy, Binary Cross Entropy, Focal Loss and a custom-made hierarchical loss). The best results were obtained by leveraging the hierarchical nature of the data, by outputting ancestor classes and with a hierarchical loss. Our final submission consist of an ensembling of our top-3 best models for each persuasion techniques. We obtain hierarchical F1 scores of 0.655 (English), 0.345 (Bulgarian), 0.442 (North Macedonian) and 0.178 (Arabic) on the test set.

pdf bib
TU Wien at SemEval-2024 Task 6: Unifying Model-Agnostic and Model-Aware Techniques for Hallucination Detection
Varvara Arzt | Mohammad Mahdi Azarbeik | Ilya Lasy | Tilman Kerl | Gábor Recski

This paper discusses challenges in Natural Language Generation (NLG), specifically addressing neural networks producing output that is fluent but incorrect, leading to “hallucinations”. The SHROOM shared task involves Large Language Models in various tasks, and our methodology employs both model-agnostic and model-aware approaches for hallucination detection. The limited availability of labeled training data is addressed through automatic label generation strategies. Model-agnostic methods include word alignment and fine-tuning a BERT-based pretrained model, while model-aware methods leverage separate classifiers trained on LLMs’ internal data (layer activations and attention values). Ensemble methods combine outputs through various techniques such as regression metamodels, voting, and probability fusion. Our best performing systems achieved an accuracy of 80.6% on the model-aware track and 81.7% on the model-agnostic track, ranking 3rd and 8th among all systems, respectively.

pdf bib
silp_nlp at SemEval-2024 Task 1: Cross-lingual Knowledge Transfer for Mono-lingual Learning
Sumit Singh | Pankaj Goyal | Uma Tiwary

Our team, silp_nlp, participated in all three tracks of SemEval2024 Task 1: Semantic Textual Relatedness (STR). We created systems for a total of 29 subtasks across all tracks: nine subtasks for track A, 10 subtasks for track B, and ten subtasks for track C. To make the most of our knowledge across all subtasks, we used transformer-based pre-trained models, which are known for their strong cross-lingual transferability. For track A, we trained our model in two stages. In the first stage, we focused on multi-lingual learning from all tracks. In the second stage, we fine-tuned the model for individual tracks. For track B, we used a unigram and bigram representation with suport vector regression (SVR) and eXtreme Gradient Boosting (XGBoost) regression. For track C, we again utilized cross-lingual transferability without the use of targeted subtask data. Our work highlights the fact that knowledge gained from all subtasks can be transferred to an individual subtask if the base language model has strong cross-lingual characteristics. Our system ranked first in the Indonesian subtask of Track B (C7) and in the top three for four other subtasks.

pdf bib
LastResort at SemEval-2024 Task 3: Exploring Multimodal Emotion Cause Pair Extraction as Sequence Labelling Task
Suyash Vardhan Mathur | Akshett Jindal | Hardik Mittal | Manish Shrivastava

Conversation is the most natural form of human communication, where each utterance can range over a variety of possible emotions. While significant work has been done towards the detection of emotions in text, relatively little work has been done towards finding the cause of the said emotions, especially in multimodal settings. SemEval 2024 introduces the task of Multimodal Emotion Cause Analysis in Conversations, which aims to extract emotions reflected in individual utterances in a conversation involving multiple modalities (textual, audio, and visual modalities) along with the corresponding utterances that were the cause for the emotion. In this paper, we propose models that tackle this task as an utterance labeling and a sequence labeling problem and perform a comparative study of these models, involving baselines using different encoders, using BiLSTM for adding contextual information of the conversation, and finally adding a CRF layer to try to model the inter-dependencies between adjacent utterances more effectively. In the official leaderboard for the task, our architecture was ranked 8th, achieving an F1-score of 0.1759 on the leaderboard.

pdf bib
DaVinci at SemEval-2024 Task 9: Few-shot prompting GPT-3.5 for Unconventional Reasoning
Suyash Vardhan Mathur | Akshett Jindal | Manish Shrivastava

While significant work has been done in the field of NLP on vertical thinking, which involves primarily logical thinking, little work has been done towards lateral thinking, which involves looking at problems from an unconventional perspective defying existing conceptions and notions. Towards this direction, SemEval 2024 introduces the task of BRAINTEASER, which involves two types of questions – Sentence Puzzle and Word Puzzle that defy conventional common-sense reasoning and constraints. In this paper, we tackle both the questions using few-shot prompting on GPT-3.5 and gain insights regarding the difference in the nature of the two types of questions. Our prompting strategy placed us 26th on the leaderboard for the Sentence Puzzle and 15th on the Word Puzzle task.

pdf bib
MorphingMinds at SemEval-2024 Task 10: Emotion Recognition in Conversation in Hindi-English Code-Mixed Conversations
Monika Vyas

The research focuses on emotion detection in multilingual conversations, particularly in Romanized Hindi and English, with applications in sentiment analysis and mental health assessments. The study employs Machine learning, deep learning techniques, including Transformer-based models like XLM-RoBERTa, for feature extraction and emotion classification. Various experiments are conducted to evaluate model performance, including fine-tuning, data augmentation, and addressing dataset imbalances. The findings highlight challenges and opportunities in emotion detection across languages and emphasize culturally sensitive approaches. The study contributes to advancing emotion analysis in multilingual contexts and provides practical guidance for developing more accurate emotion detection systems.

pdf bib
SemanticCUETSync at SemEval-2024 Task 1: Finetuning Sentence Transformer to Find Semantic Textual Relatedness
Md. Sajjad Hossain | Ashraful Islam Paran | Symom Hossain Shohan | Jawad Hossain | Mohammed Moshiul Hoque

Semantic textual relatedness is crucial to Natural Language Processing (NLP). Methodologies often exhibit superior performance in high-resource languages such as English compared to low-resource ones like Marathi, Telugu, and Spanish. This study leverages various machine learning (ML) approaches, including Support Vector Regression (SVR) and Random Forest, deep learning (DL) techniques such as Siamese Neural Networks, and transformer-based models such as MiniLM-L6-v2, Marathi-sbert, Telugu-sentence-bert-nli, and Roberta-bne-sentiment-analysis-es, to assess semantic relatedness across English, Marathi, Telugu, and Spanish. The developed transformer-based methods notably outperformed other models in determining semantic textual relatedness across these languages, achieving a Spearman correlation coefficient of 0.822 (for English), 0.870 (for Marathi), 0.820 (for Telugu), and 0.677 (for Spanish). These results led to our work attaining rankings of 22th (for English), 11th (for Marathi), 11th (for Telegu) and 14th (for Spanish), respectively.

pdf bib
IASBS at SemEval-2024 Task 10: Delving into Emotion Discovery and Reasoning in Code-Mixed Conversations
Mehrzad Tareh | Aydin Mohandesi | Ebrahim Ansari

In this paper, we detail the IASBS team’s approach and findings from participating in SemEval-2024 Task 10, “Emotion Discovery and Reasoning in Hindi-English Code-mixed Conversations (EDiReF).” This task encompasses three critical subtasks: Emotion Recognition in Conversation (ERC), and Emotion Flip Reasoning (EFR) in both Hindi-English code-mixed and English dialogues. Our methodology integrates advanced NLP and machine learning techniques, focusing on the unique challenges of code-mixing, such as linguistic diversity and shifts in emotional context. By implementing a robust framework that includes data preprocessing, and feature engineering using models like GPT-4 and DistilBERT, we extend our analysis beyond mere emotion identification to explore the triggers behind emotion flips. This endeavor not only achieved third place on the leaderboard, demonstrating a high proficiency in emotion and flip detection with an F1-Score of 0.70 but also significantly contributed to the advancement of emotional AI. Our findings offer valuable insights into the complex interplay of emotions in communication, showcasing the potential for enhancing applications across various domains, from social media analytics to healthcare, and underscore the importance of understanding emotional dynamics in code-mixed conversations for future research and practical applications.

pdf bib
Deja Vu at SemEval 2024 Task 9: A Comparative Study of Advanced Language Models for Commonsense Reasoning
Trina Chakraborty | Marufur Rahman | Md Omar Faruqe

This research systematically forms an impression of the capabilities of advanced language models in addressing the BRAINTEASER task introduced at SemEval 2024, which is specifically designed to explore the models’ proficiency in lateral commonsense reasoning. The task sets forth an array of Sentence and Word Puzzles, carefully crafted to challenge the models with scenarios requiring unconventional thought processes. Our methodology encompasses a holistic approach, incorporating pre-processing of data, fine-tuning of transformer-based language models, and strategic data augmentation to explore the depth and flexibility of each model’s understanding. The preliminary results of our analysis are encouraging, highlighting significant potential for advancements in the models’ ability to engage in lateral reasoning. Further insights gained from post-competition evaluations suggest scopes for notable enhancements in model performance, emphasizing the continuous evolution of the models in mastering complex reasoning tasks.

pdf bib
FtG-CoT at SemEval-2024 Task 9: Solving Sentence Puzzles Using Fine-Tuned Language Models and Zero-Shot CoT Prompting
Micah Zhang | Shafiuddin Rehan Ahmed | James H. Martin

Recent large language models (LLMs) can solve puzzles that require creativity and lateral thinking. To advance this front of research, we tackle SemEval-2024 Task 9: BRAINTEASER: A Novel Task Defying Common Sense. We approach this task by introducing a technique that we call Fine-tuned Generated Chain-of-Thought (FtG-CoT). It is a novel few-shot prompting method that combines a fine-tuned BERT classifier encoder with zero-shot chain-of-thought generation and a fine-tuned LLM. The fine-tuned BERT classifier provides a context-rich encoding of each example question and choice list. Zero-shot chain-of-thought generation leverages the benefits of chain-of-thought prompting without requiring manual creation of the reasoning chains. We fine-tune the LLM on the generated chains-of-thought and include a set of generated reasoning chains in the final few-shot LLM prompt to maximize the relevance and correctness of the final generated response. In this paper, we show that FtG-CoT outperforms the zero-shot prompting baseline presented in the task paper and is highly effective at solving challenging sentence puzzles achieving a perfect score on the practice set and a 0.9 score on the evaluation set.

pdf bib
LyS at SemEval-2024 Task 3: An Early Prototype for End-to-End Multimodal Emotion Linking as Graph-Based Parsing
Ana Ezquerro | David Vilares

This paper describes our participation in SemEval 2024 Task 3, which focused on Multimodal Emotion Cause Analysis in Conversations. We developed an early prototype for an end-to-end system that uses graph-based methods from dependency parsing to identify causal emotion relations in multi-party conversations. Our model comprises a neural transformer-based encoder for contextualizing multimodal conversation data and a graph-based decoder for generating the adjacency matrix scores of the causal graph. We ranked 7th out of 15 valid and official submissions for Subtask 1, using textual inputs only. We also discuss our participation in Subtask 2 during post-evaluation using multi-modal inputs.

pdf bib
NumDecoders at SemEval-2024 Task 7: FlanT5 and GPT enhanced with CoT for Numerical Reasoning
Andres Gonzalez | Md Zobaer Hossain | Jahedul Alam Junaed

In this paper we present a Chain-of-Thought enhanced solution for large language models, including flanT5 and GPT 3.5 Turbo, aimed at solving mathematical problems to fill in blanks from news headlines. Our approach builds on adata augmentation strategy that incorporates additional mathematical reasoning observations into the original dataset sourced from another mathematical corpus. Both automatic and manual annotations are applied to explicitly describe the reasoning steps required for models to reach the target answer. We employ an ensemble majority voting method to generate finalpredictions across our best-performing models. Our analysis reveals that while larger models trained with our enhanced dataset achieve significant gains (91% accuracy, ranking 5th on the NumEval Task 3 leaderboard), smaller models do not experience improvements and may even see a decrease in overall accuracy. We conclude that improving our automatic an-notations via crowdsourcing methods can be a worthwhile endeavor to train larger models than the ones from this study to see the most accurate results.

pdf bib
FZI-WIM at SemEval-2024 Task 2: Self-Consistent CoT for Complex NLI in Biomedical Domain
Jin Liu | Steffen Thoma

This paper describes the inference system of FZI-WIM at the SemEval-2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials. Our system utilizes the chain of thought (CoT) paradigm to tackle this complex reasoning problem and further improve the CoT performance with self-consistency. Instead of greedy decoding, we sample multiple reasoning chains with the same prompt and make thefinal verification with majority voting. The self-consistent CoT system achieves a baseline F1 score of 0.80 (1st), faithfulness score of 0.90 (3rd), and consistency score of 0.73 (12th). We release the code and data publicly.

pdf bib
Lisbon Computational Linguists at SemEval-2024 Task 2: Using a Mistral-7B Model and Data Augmentation
Artur Guimarães | Bruno Martins | João Magalhães

ABSTRACT: This paper describes our approach to the SemEval-2024 safe biomedical Natural Language Inference for Clinical Trials (NLI4CT) task, which concerns classifying statements about Clinical Trial Reports (CTRs). We explored the capabilities of Mistral-7B, a generalistic open-source Large Language Model (LLM). We developed a prompt for the NLI4CT task, and fine-tuning a quantized version of the model using a slightly augmented version of the training dataset. The experimental results show that this approach can produce notable results in terms of the macro F1-score, while having limitations in terms of faithfulness and consistency. All the developed code is publicly available on a GitHub repository.

pdf bib
GIL-IIMAS UNAM at SemEval-2024 Task 1: SAND: An In Depth Analysis of Semantic Relatedness Using Regression and Similarity Characteristics
Francisco Lopez-ponce | Ángel Cadena | Karla Salas-jimenez | Gemma Bel-enguix | David Preciado-márquez

The STR shared task aims at detecting the degree of semantic relatedness between sentence pairs in multiple languages. Semantic relatedness relies on elements such as topic similarity, point of view agreement, entailment, and even human intuition, making it a broader field than sentence similarity. The GIL-IIMAS UNAM team proposes a model based in the SAND characteristics composition (Sentence Transformers, AnglE Embeddings, N-grams, Sentence Length Difference coefficient) and classical regression algorithms. This model achieves a 0.83 Spearman Correlation score in the English test, and a 0.73 in the Spanish counterpart, finishing just above the SemEval baseline in English, and second place in Spanish.

pdf bib
Team UTSA-NLP at SemEval 2024 Task 5: Prompt Ensembling for Argument Reasoning in Civil Procedures with GPT4
Dan Schumacher | Anthony Rios

In this paper, we present our system for the SemEval Task 5, The Legal Argument Reasoning Task in Civil Procedure Challenge. Legal argument reasoning is an essential skill that all law students must master. Moreover, it is important to develop natural language processing solutions that can reason about a question given terse domain-specific contextual information. Our system explores a prompt-based solution using GPT4 to reason over legal arguments. We also evaluate an ensemble of prompting strategies, including chain-of-thought reasoning and in-context learning. Overall, our system results in a Macro F1 of .8095 on the validation dataset and .7315 (5th out of 21 teams) on the final test set. Code for this project is available at https://github.com/danschumac1/CivilPromptReasoningGPT4.

pdf bib
BD-NLP at SemEval-2024 Task 2: Investigating Generative and Discriminative Models for Clinical Inference with Knowledge Augmentation
Shantanu Nath | Ahnaf Mozib Samin

Healthcare professionals rely on evidence from clinical trial records (CTRs) to devise treatment plans. However, the increasing quantity of CTRs poses challenges in efficiently assimilating the latest evidence to provide personalized evidence-based care. In this paper, we present our solution to the SemEval- 2024 Task 2 titled “Safe Biomedical Natural Language Inference for Clinical Trials”. Given a statement and one/two CTRs as inputs, the task is to determine whether or not the statement entails or contradicts the CTRs. We explore both generative and discriminative large language models (LLM) to investigate their performance for clinical inference. Moreover, we contrast the general-purpose LLMs with the ones specifically tailored for the clinical domain to study the potential advantage in mitigating distributional shifts. Furthermore, the benefit of augmenting additional knowledge within the prompt/statement is examined in this work. Our empirical study suggests that DeBERTa-lg, a discriminative general-purpose natural language inference model, obtains the highest F1 score of 0.77 on the test set, securing the fourth rank on the leaderboard. Intriguingly, the augmentation of knowledge yields subpar results across most cases.

pdf bib
NLP at UC Santa Cruz at SemEval-2024 Task 5: Legal Answer Validation using Few-Shot Multi-Choice QA
Anish Pahilajani | Samyak Jain | Devasha Trivedi

This paper presents our submission to the SemEval 2024 Task 5: The Legal Argument Reasoning Task in Civil Procedure. We present two approaches to solving the task of legal answer validation, given an introduction to the case, a question and an answer candidate. Firstly, we fine-tuned pre-trained BERT-based models and found that models trained on domain knowledge perform better. Secondly, we performed few-shot prompting on GPT models and found that reformulating the answer validation task to be a multiple-choice QA task remarkably improves the performance of the model. Our best submission is a BERT-based model that achieved the 7th place out of 20.

pdf bib
CoT-based Data Augmentation Strategy for Persuasion Techniques Detection
Dailin Li | Chuhan Wang | Xin Zou | Junlong Wang | Peng Chen | Jian Wang | Liang Yang | Hongfei Lin

Detecting persuasive communication is an important topic in Natural Language Processing (NLP), as it can be useful in identifying fake information on social media. We have developed a system to identify applied persuasion techniques in text fragments across four languages: English, Bulgarian, North Macedonian, and Arabic. Our system uses data augmentation methods and employs an ensemble strategy that combines the strengths of both RoBERTa and DeBERTa models. Due to limited resources, we concentrated solely on task 1, and our solution achieved the top ranking in the English track during the official assessments. We also analyse the impact of architectural decisions, data constructionand training strategies.

pdf bib
HaRMoNEE at SemEval-2024 Task 6: Tuning-based Approaches to Hallucination Recognition
Timothy Obiso | Jingxuan Tu | James Pustejovsky

This paper presents the Hallucination Recognition Model for New Experiment Evaluation (HaRMoNEE) team’s winning (#1) and #10 submissions for SemEval-2024 Task 6: Shared- task on Hallucinations and Related Observable Overgeneration Mistakes (SHROOM)’s two subtasks. This task challenged its participants to design systems to detect hallucinations in Large Language Model (LLM) outputs. Team HaRMoNEE proposes two architectures: (1) fine-tuning an off-the-shelf transformer-based model and (2) prompt tuning large-scale Large Language Models (LLMs). One submission from the fine-tuning approach outperformed all other submissions for the model-aware subtask; one submission from the prompt-tuning approach is the 10th-best submission on the leaderboard for the model-agnostic subtask. Our systems also include pre-processing, system-specific tuning, post-processing, and evaluation.

pdf bib
VerbaNexAI Lab at SemEval-2024 Task 10: Emotion recognition and reasoning in mixed-coded conversations based on an NRC VAD approach
Santiago Garcia | Elizabeth Martinez | Juan Cuadrado | Juan Martinez-santos | Edwin Puertas

This study introduces an innovative approach to emotion recognition and reasoning about emotional shifts in code-mixed conversations, leveraging the NRC VAD Lexicon and computational models such as Transformer and GRU. Our methodology systematically identifies and categorizes emotional triggers, employing Emotion Flip Reasoning (EFR) and Emotion Recognition in Conversation (ERC). Through experiments with the MELD and MaSaC datasets, we demonstrate the model’s precision in accurately identifying emotional shift triggers and classifying emotions, evidenced by a significant improvement in accuracy as shown by an increase in the F1 score when including VAD analysis. These results underscore the importance of incorporating complex emotional dimensions into conversation analysis, paving new pathways for understanding emotional dynamics in code-mixed texts.

pdf bib
VerbaNexAI Lab at SemEval-2024 Task 3: Deciphering emotional causality in conversations using multimodal analysis approach
Victor Pacheco | Elizabeth Martinez | Juan Cuadrado | Juan Carlos Martinez Santos | Edwin Puertas

This study delineates our participation in the SemEval-2024 Task 3: Multimodal Emotion Cause Analysis in Conversations, focusing on developing and applying an innovative methodology for emotion detection and cause analysis in conversational contexts. Leveraging logistic regression, we analyzed conversational utterances to identify emotions per utterance. Subsequently, we employed a dependency analysis pipeline, utilizing SpaCy to extract significant chunk features, including object, subject, adjectival modifiers, and adverbial clause modifiers. These features were analyzed within a graph-like framework, conceptualizing the dependency relationships as edges connecting emotional causes (tails) to their corresponding emotions (heads). Despite the novelty of our approach, the preliminary results were unexpectedly humbling, with a consistent score of 0.0 across all evaluated metrics. This paper presents our methodology, the challenges encountered, and an analysis of the potential factors contributing to these outcomes, offering insights into the complexities of emotion-cause analysis in multimodal conversational data.

pdf bib
VerbaNexAI Lab at SemEval-2024 Task 1: A Multilayer Artificial Intelligence Model for Semantic Relationship Detection
Anderson Morillo | Daniel Peña | Juan Carlos Martinez Santos | Edwin Puertas

This paper presents an artificial intelligence model designed to detect semantic relationships in natural language, addressing the challenges of SemEval 2024 Task 1. Our goal is to advance machine understanding of the subtleties of human language through semantic analysis. Using a novel combination of convolutional neural networks (CNNs), long short-term memory (LSTM) networks, and an attention mechanism, our model is trained on the STR-2022 dataset. This approach enhances its ability to detect semantic nuances in different texts. The model achieved an 81.92% effectiveness rate and ranked 24th in SemEval 2024 Task 1. These results demonstrate its robustness and adaptability in detecting semantic relationships and validate its performance in diverse linguistic contexts. Our work contributes to natural language processing by providing insights into semantic textual relatedness. It sets a benchmark for future research and promises to inspire innovations that could transform digital language processing and interaction.

pdf bib
UMBCLU at SemEval-2024 Task 1: Semantic Textual Relatedness with and without machine translation
Shubhashis Roy Dipta | Sai Vallurupalli

The aim of SemEval-2024 Task 1, “Semantic Textual Relatedness for African and Asian Languages” is to develop models for identifying semantic textual relatedness (STR) between two sentences using multiple languages (14 African and Asian languages) and settings (supervised, unsupervised, and cross-lingual). Large language models (LLMs) have shown impressive performance on several natural language understanding tasks such as multilingual machine translation (MMT), semantic similarity (STS), and encoding sentence embeddings. Using a combination of LLMs that perform well on these tasks, we developed two STR models, TranSem and FineSem, for the supervised and cross-lingual settings. We explore the effectiveness of several training methods and the usefulness of machine translation. We find that direct fine-tuning on the task is comparable to using sentence embeddings and translating to English leads to better performance for some languages. In the supervised setting, our model performance is better than the official baseline for 3 languages with the remaining 4 performing on par. In the cross-lingual setting, our model performance is better than the baseline for 3 languages (leading to 1st place for Africaans and 2nd place for Indonesian), is on par for 2 languages and performs poorly on the remaining 7 languages.

pdf bib
MasonTigers at SemEval-2024 Task 9: Solving Puzzles with an Ensemble of Chain-of-Thought Prompts
Nishat Raihan | Dhiman Goswami | Al Nahian Bin Emran | Sadiya Sayara Chowdhury Puspo | Amrita Ganguly | Marcos Zampieri

Our paper presents team MasonTigers submission to the SemEval-2024 Task 9 - which provides a dataset of puzzles for testing natural language understanding. We employ large language models (LLMs) to solve this task through several prompting techniques. Zero-shot and few-shot prompting generate reasonably good results when tested with proprietary LLMs, compared to the open-source models. We obtain further improved results with chain-of-thought prompting, an iterative prompting method that breaks down the reasoning process step-by-step. We obtain our best results by utilizing an ensemble of chain-of-thought prompts, placing 2nd in the word puzzle subtask and 13th in the sentence puzzle subtask. The strong performance of prompted LLMs demonstrates their capability for complex reasoning when provided with a decomposition of the thought process. Our work sheds light on how step-wise explanatory prompts can unlock more of the knowledge encoded in the parameters of large models.

pdf bib
MasonTigers at SemEval-2024 Task 8: Performance Analysis of Transformer-based Models on Machine-Generated Text Detection
Sadiya Sayara Chowdhury Puspo | Nishat Raihan | Dhiman Goswami | Al Nahian Bin Emran | Amrita Ganguly | Özlem Uzuner

This paper presents the MasonTigers entryto the SemEval-2024 Task 8 - Multigenerator, Multidomain, and Multilingual BlackBox Machine-Generated Text Detection. Thetask encompasses Binary Human-Written vs.Machine-Generated Text Classification (TrackA), Multi-Way Machine-Generated Text Classification (Track B), and Human-Machine MixedText Detection (Track C). Our best performing approaches utilize mainly the ensemble ofdiscriminator transformer models along withsentence transformer and statistical machinelearning approaches in specific cases. Moreover, Zero shot prompting and fine-tuning ofFLAN-T5 are used for Track A and B.

pdf bib
UIC NLP GRADS at SemEval-2024 Task 3: Two-Step Disjoint Modeling for Emotion-Cause Pair Extraction
Sharad Chandakacherla | Vaibhav Bhargava | Natalie Parde

Disentangling underlying factors contributing to the expression of emotion in multimodal data is challenging but may accelerate progress toward many real-world applications. In this paper we describe our approach for solving SemEval-2024 Task #3, Sub-Task #1, focused on identifying utterance-level emotions and their causes using the text available from the multimodal F.R.I.E.N.D.S. television series dataset. We propose to disjointly model emotion detection and causal span detection, borrowing a paradigm popular in question answering (QA) to train our model. Through our experiments we find that (a) contextual utterances before and after the target utterance play a crucial role in emotion classification; and (b) once the emotion is established, detecting the causal spans resulting in that emotion using our QA-based technique yields promising results.

pdf bib
MasonTigers at SemEval-2024 Task 1: An Ensemble Approach for Semantic Textual Relatedness
Dhiman Goswami | Sadiya Sayara Chowdhury Puspo | Nishat Raihan | Al Nahian Bin Emran | Amrita Ganguly | Marcos Zampieri

This paper presents the MasonTigers’ entry to the SemEval-2024 Task 1 - Semantic Textual Relatedness. The task encompasses supervised (Track A), unsupervised (Track B), and cross-lingual (Track C) approaches to semantic textual relatedness across 14 languages. MasonTigers stands out as one of the two teams who participated in all languages across the three tracks. Our approaches achieved rankings ranging from 11th to 21st in Track A, from 1st to 8th in Track B, and from 5th to 12th in Track C. Adhering to the task-specific constraints, our best performing approaches utilize an ensemble of statistical machine learning approaches combined with language-specific BERT based models and sentence transformers.

pdf bib
RiddleMasters at SemEval-2024 Task 9: Comparing Instruction Fine-tuning with Zero-Shot Approaches
Kejsi Take | Chau Tran

This paper describes our contribution to SemEval 2023 Task 8: Brainteaser. We compared multiple zero-shot approaches using GPT-4, the state of the art model with Mistral-7B, a much smaller open-source LLM. While GPT-4 remains a clear winner in all the zero-shot approaches, we show that finetuning Mistral-7B can achieve comparable, even though marginally lower results.

pdf bib
IITK at SemEval-2024 Task 2: Exploring the Capabilities of LLMs for Safe Biomedical Natural Language Inference for Clinical Trials
Shreyasi Mandal | Ashutosh Modi

Large Language models (LLMs) have demonstrated state-of-the-art performance in various natural language processing (NLP) tasks across multiple domains, yet they are prone to shortcut learning and factual inconsistencies. This research investigates LLMs’ robustness, consistency, and faithful reasoning when performing Natural Language Inference (NLI) on breast cancer Clinical Trial Reports (CTRs) in the context of SemEval 2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials. We examine the reasoning capabilities of LLMs and their adeptness at logical problem-solving. A comparative analysis is conducted on pre-trained language models (PLMs), GPT-3.5, and Gemini Pro under zero-shot settings using Retrieval-Augmented Generation (RAG) framework, integrating various reasoning chains. The evaluation yields an F1 score of 0.69, consistency of 0.71, and a faithfulness score of 0.90 on the test dataset.

pdf bib
PEAR at SemEval-2024 Task 1: Pair Encoding with Augmented Re-sampling for Semantic Textual Relatedness
Tollef Jørgensen

This paper describes a system submitted to the supervised track (Track A) at SemEval-24: Semantic Textual Relatedness for African and Asian Languages. Challenged with datasets of varying sizes, some as small as 800 samples, we observe that the PEAR system, using smaller pre-trained masked language models to process sentence pairs (Pair Encoding), results in models that efficiently adapt to the task.In addition to the simplistic modeling approach, we experiment with hyperparameter optimization and data expansion from the provided training sets using multilingual bi-encoders, sampling a dynamic number of nearest neighbors (Augmented Re-sampling). The final models are lightweight, allowing fast experimentation and integration of new languages.

pdf bib
BCAmirs at SemEval-2024 Task 4: Beyond Words: A Multimodal and Multilingual Exploration of Persuasion in Memes
Amirhossein Abaskohi | Amirhossein Dabiriaghdam | Lele Wang | Giuseppe Carenini

Memes, combining text and images, frequently use metaphors to convey persuasive messages, shaping public opinion. Motivated by this, our team engaged in SemEval-2024 Task 4, a hierarchical multi-label classification task designed to identify rhetorical and psychological persuasion techniques embedded within memes. To tackle this problem, we introduced a caption generation step to assess the modality gap and the impact of additional semantic information from images, which improved our result. Our best model utilizes GPT-4 generated captions alongside meme text to fine-tune RoBERTa as the text encoder and CLIP as the image encoder. It outperforms the baseline by a large margin in all 12 subtasks. In particular, it ranked in top-3 across all languages in Subtask 2a, and top-4 in Subtask 2b, demonstrating quantitatively strong performance. The improvement achieved by the introduced intermediate step is likely attributable to the metaphorical essence of images that challenges visual encoders. This highlights the potential for improving abstract visual semantics encoding.

pdf bib
Pauk at SemEval-2024 Task 4: A Neuro-Symbolic Method for Consistent Classification of Propaganda Techniques in Memes
Matt Pauk | Maria Leonor Pacheco

Memes play a key role in most modern informa-tion campaigns, particularly propaganda cam-paigns. Identifying the persuasive techniquespresent in memes is an important step in de-veloping systems to recognize and curtail pro-paganda. This work presents a framework toidentify the persuasive techniques present inmemes for the SemEval 2024 Task 4, accordingto a hierarchical taxonomy of propaganda tech-niques. The framework involves a knowledgedistillation method, where the base model is acombination of DeBERTa and ResNET usedto classify the text and image, and the teachermodel consists of a group of weakly enforcedlogic rules that promote the hierarchy of per-suasion techniques. The addition of the logicrule layer for knowledge distillation shows im-provement in respecting the hierarchy of thetaxonomy with a slight boost in performance.

pdf bib
Saama Technologies at SemEval-2024 Task 2: Three-module System for NLI4CT Enhanced by LLM-generated Intermediate Labels
Hwanmun Kim | Kamal Raj Kanakarajan | Malaikannan Sankarasubbu

Participating in SemEval 2024 Task 2, we built a three-module system to predict entailment labels for NLI4CT, which consists of a sequence of the query generation module, the query answering module, and the aggregation module. We fine-tuned or prompted each module with the intermediate labels we generated with LLMs, and we optimized the combinations of different modules through experiments. Our system is ranked 19th ~ 24th in the SemEval 2024 Task 2 leaderboard in different metrics. We made several interesting observations regarding the correlation between different metrics and the sensitivity of our system on the aggregation module. We performed the error analysis on our system which can potentially help to improve our system further.

pdf bib
AmazUtah_NLP at SemEval-2024 Task 9: A MultiChoice Question Answering System for Commonsense Defying Reasoning
Soumya Mishra | Mina Ghashami

The SemEval 2024 BRAINTEASER task represents a pioneering venture in Natural Language Processing (NLP) by focusing on lateral thinking, a dimension of cognitive reasoning that is often overlooked in traditional linguistic analyses. This challenge comprises of Sentence Puzzle and Word Puzzle sub-tasks and aims to test language models’ capacity for divergent thinking. In this paper, we present our approach to the BRAINTEASER task. We employ a holistic strategy by leveraging cutting-edge pre-trained models in multiple choice architecture, and diversify the training data with Sentence and Word Puzzle datasets. To gain further improvement, we fine-tuned the model with synthetic humor/jokes dataset and the RiddleSense dataset which helped augmenting the model’s lateral thinking abilities. Empirical results show that our approach achieve 92.5% accuracy in Sentence Puzzle subtask and 80.2% accuracy in Word Puzzle subtask.

pdf bib
IITK at SemEval-2024 Task 1: Contrastive Learning and Autoencoders for Semantic Textual Relatedness in Multilingual Texts
Udvas Basak | Rajarshi Dutta | Shivam Pandey | Ashutosh Modi

This paper describes our system developed for the SemEval-2024 Task 1: Semantic Textual Relatedness. The challenge is focused on automatically detecting the degree of relatedness between pairs of sentences for 14 languages including both high and low-resource Asian and African languages. Our team participated in two subtasks consisting of Track A: supervised and Track B: unsupervised. This paper focuses on a BERT-based contrastive learning and similarity metric based approach primarily for the supervised track while exploring autoencoders for the unsupervised track. It also aims on the creation of a bigram relatedness corpus using negative sampling strategy, thereby producing refined word embeddings.

pdf bib
Compos Mentis at SemEval2024 Task6: A Multi-Faceted Role-based Large Language Model Ensemble to Detect Hallucination
Souvik Das | Rohini Srihari

Hallucinations in large language models (LLMs), where they generate fluent but factually incorrect outputs, pose challenges for applications requiring strict truthfulness. This work proposes a multi-faceted approach to detect such hallucinations across various language tasks. We leverage automatic data annotation using a proprietary LLM, fine-tuning of the Mistral-7B-instruct-v0.2 model on annotated and benchmark data, role-based and rationale-based prompting strategies, and an ensemble method combining different model outputs through majority voting. This comprehensive framework aims to improve the robustness and reliability of hallucination detection for LLM generations.

pdf bib
NYCU-NLP at SemEval-2024 Task 2: Aggregating Large Language Models in Biomedical Natural Language Inference for Clinical Trials
Lung-hao Lee | Chen-ya Chiou | Tzu-mi Lin

This study describes the model design of the NYCU-NLP system for the SemEval-2024 Task 2 that focuses on natural language inference for clinical trials. We aggregate several large language models to determine the inference relation (i.e., entailment or contradiction) between clinical trial reports and statements that may be manipulated with designed interventions to investigate the faithfulness and consistency of the developed models. First, we use ChatGPT v3.5 to augment original statements in training data and then fine-tune the SOLAR model with all augmented data. During the testing inference phase, we fine-tune the OpenChat model to reduce the influence of interventions and fed a cleaned statement into the fine-tuned SOLAR model for label prediction. Our submission produced a faithfulness score of 0.9236, ranking second of 32 participating teams, and ranked first for consistency with a score of 0.8092.

pdf bib
Team MLab at SemEval-2024 Task 8: Analyzing Encoder Embeddings for Detecting LLM-generated Text
Kevin Li | Kenan Hasanaliyev | Sally Zhu | George Altshuler | Alden Eberts | Eric Chen | Kate Wang | Emily Xia | Eli Browne | Ian Chen

This paper explores solutions to the challenges posed by the widespread use of LLMs, particularly in the context of identifying human-written versus machine-generated text. Focusing on Subtask B of SemEval 2024 Task 8, we compare the performance of RoBERTa and DeBERTa models. Subtask B involved identifying not only human or machine text but also the specific LLM responsible for generating text, where our DeBERTa model outperformed the RoBERTa baseline by over 10% in leaderboard accuracy. The results highlight the rapidly growing capabilities of LLMs and importance of keeping up with the latest advancements. Additionally, our paper presents visualizations using PCA and t-SNE that showcase the DeBERTa model’s ability to cluster different LLM outputs effectively. These findings contribute to understanding and improving AI methods for detecting machine-generated text, allowing us to build more robust and traceable AI systems in the language ecosystem.

pdf bib
Calc-CMU at SemEval-2024 Task 7: Pre-Calc - Learning to Use the Calculator Improves Numeracy in Language Models
Vishruth Veerendranath | Vishwa Shah | Kshitish Ghate

Quantitative and numerical comprehension in language is an important task in many fields like education and finance, but still remains a challenging task for language models. While tool and calculator usage has shown to be helpful to improve mathematical reasoning in large pretrained decoder-only language models, this remains unexplored for smaller language models with encoders. In this paper, we propose Pre-Calc, a simple pre-finetuning objective of learning to use the calculator for both encoder-only and encoder-decoder architectures, formulated as a discriminative and generative task respectively. We pre-train BERT and RoBERTa for discriminative calculator use and Flan-T5 for generative calculator use on the MAWPS, SVAMP, and AsDiv-A datasets, which improves performance on downstream tasks that require numerical understanding. Our code and data are available at https://github.com/calc-cmu/pre-calc.

pdf bib
AISPACE at SemEval-2024 task 8: A Class-balanced Soft-voting System for Detecting Multi-generator Machine-generated Text
Renhua Gu | Xiangfeng Meng

SemEval-2024 Task 8 provides a challenge to detect human-written and machine-generated text. There are 3 subtasks for different detection scenarios. This paper proposes a system that mainly deals with Subtask B. It aims to detect if given full text is written by human or is generated by a specific Large Language Model (LLM), which is actually a multi-class text classification task. Our team AISPACE conducted a systematic study of fine-tuning transformer-based models, including encoder-only, decoder-only and encoder-decoder models. We compared their performance on this task and identified that encoder-only models performed exceptionally well. We also applied a weighted Cross Entropy loss function to address the issue of data imbalance of different class samples. Additionally, we employed soft-voting strategy over multi-models ensemble to enhance the reliability of our predictions. Our system ranked top 1 in Subtask B, which sets a state-of-the-art benchmark for this new challenge.

pdf bib
SemEval-2024 Task 7: Numeral-Aware Language Understanding and Generation
Chung-chi Chen | Jian-tao Huang | Hen-hsen Huang | Hiroya Takamura | Hsin-hsi Chen

Numbers are frequently utilized in both our daily narratives and professional documents, such as clinical notes, scientific papers, financial documents, and legal court orders. The ability to understand and generate numbers is thus one of the essential aspects of evaluating large language models. In this vein, we propose a collection of datasets in SemEval-2024 Task 7 - NumEval. This collection encompasses several tasks focused on numeral-aware instances, including number prediction, natural language inference, question answering, reading comprehension, reasoning, and headline generation. This paper offers an overview of the dataset and presents the results of all subtasks in NumEval. Additionally, we contribute by summarizing participants’ methods and conducting an error analysis. To the best of our knowledge, NumEval represents one of the early tasks that perform peer evaluation in SemEval’s history. We will further share observations from this aspect and provide suggestions for future SemEval tasks.

pdf bib
UCSC NLP at SemEval-2024 Task 10: Emotion Discovery and Reasoning its Flip in Conversation (EDiReF)
Neng Wan | Steven Au | Esha Ubale | Decker Krogh

We describe SemEval-2024 Task 10: EDiReF consisting of three sub-tasks involving emotion in conversation across Hinglish code-mixed and English datasets. Subtasks include classification of speaker emotion in multiparty conversations (Emotion Recognition in Conversation) and reasoning around shifts in speaker emotion state (Emotion Flip Reasoning). We deployed a BERT model for emotion recognition and two GRU-based models for emotion flip. Our model achieved F1 scores of 0.45, 0.79, and 0.68 for subtasks 1, 2, and 3, respectively.

pdf bib
CLULab-UofA at SemEval-2024 Task 8: Detecting Machine-Generated Text Using Triplet-Loss-Trained Text Similarity and Text Classification
Mohammadhossein Rezaei | Yeaeun Kwon | Reza Sanayei | Abhyuday Singh | Steven Bethard

Detecting machine-generated text is a critical task in the era of large language models. In this paper, we present our systems for SemEval-2024 Task 8, which focuses on multi-class classification to discern between human-written and maching-generated texts by five state-of-the-art large language models. We propose three different systems: unsupervised text similarity, triplet-loss-trained text similarity, and text classification. We show that the triplet-loss trained text similarity system outperforms the other systems, achieving 80% accuracy on the test set and surpassing the baseline model for this subtask. Additionally, our text classification system, which takes into account sentence paraphrases generated by the candidate models, also outperforms the unsupervised text similarity system, achieving 74% accuracy.

pdf bib
SINAI at SemEval-2024 Task 8: Fine-tuning on Words and Perplexity as Features for Detecting Machine Written Text
Alberto Gutiérrez Megías | L. Alfonso Ureña-lópez | Eugenio Martínez Cámara

This work presents the proposed systems of the SINAI team for the subtask A of the Task 8 in SemEval 2024. We present the evaluation of two disparate systems, and our final submitted system. We claim that the perplexity value of a text may be used as classification signal. Accordingly, we conduct a study on the utility of perplexity for discerning text authorship, and we perform a comparative analysis of the results obtained on the datasets of the task. This comparative evaluation includes results derived from the systems evaluated, such as fine-tuning using an XLM-RoBERTa-Large transformer or using perplexity as a classification criterion. In addition, we discuss the results reached on the test set, where we show that there is large differences among the language probability distribution of the training and test sets. These analysis allows us to open new research lines to improve the detection of machine-generated text.

pdf bib
USTC-BUPT at SemEval-2024 Task 8: Enhancing Machine-Generated Text Detection via Domain Adversarial Neural Networks and LLM Embeddings
Zikang Guo | Kaijie Jiao | Xingyu Yao | Yuning Wan | Haoran Li | Benfeng Xu | Licheng Zhang | Quan Wang | Yongdong Zhang | Zhendong Mao

This paper introduces the system developed by USTC-BUPT for SemEval-2024 Task 8. The shared task comprises three subtasks across four tracks, aiming to develop automatic systems to distinguish between human-written and machine-generated text across various domains, languages and generators. Our system comprises four components: DATeD, LLAM, TLE, and AuDM, which empower us to effectively tackle all subtasks posed by the challenge. In the monolingual track, DATeD improves machine-generated text detection by incorporating a gradient reversal layer and integrating additional domain labels through Domain Adversarial Neural Networks, enhancing adaptation to diverse text domains. In the multilingual track, LLAM employs different strategies based on language characteristics. For English text, the LLM Embeddings approach utilizes embeddings from a proxy LLM followed by a two-stage CNN for classification, leveraging the broad linguistic knowledge captured during pre-training to enhance performance. For text in other languages, the LLM Sentinel approach transforms the classification task into a next-token prediction task, which facilitates easier adaptation to texts in various languages, especially low-resource languages. TLE utilizes the LLM Embeddings method with a minor modification in the classification strategy for subtask B. AuDM employs data augmentation and fine-tunes the DeBERTa model specifically for subtask C. Our system wins the multilingual track and ranks second in the monolingual track. Additionally, it achieves third place in both subtask B and C.

pdf bib
ALF at SemEval-2024 Task 9: Exploring Lateral Thinking Capabilities of LMs through Multi-task Fine-tuning
Seyed Ali Farokh | Hossein Zeinali

Recent advancements in natural language processing (NLP) have prompted the development of sophisticated reasoning benchmarks. This paper presents our system for the SemEval 2024 Task 9 competition and also investigates the efficacy of fine-tuning language models (LMs) on BrainTeaser—a benchmark designed to evaluate NLP models’ lateral thinking and creative reasoning abilities. Our experiments focus on two prominent families of pre-trained models, BERT and T5. Additionally, we explore the potential benefits of multi-task fine-tuning on commonsense reasoning datasets to enhance performance. Our top-performing model, DeBERTa-v3-large, achieves an impressive overall accuracy of 93.33%, surpassing human performance.

pdf bib
Pollice Verso at SemEval-2024 Task 6: The Roman Empire Strikes Back
Konstantin Kobs | Jan Pfister | Andreas Hotho

We present an intuitive approach for hallucination detection in LLM outputs that is modeled after how humans would go about this task. We engage several LLM “experts” to independently assess whether a response is hallucinated. For this we select recent and popular LLMs smaller than 7B parameters. By analyzing the log probabilities for tokens that signal a positive or negative judgment, we can determine the likelihood of hallucination. Additionally, we enhance the performance of our “experts” by automatically refining their prompts using the recently introduced OPRO framework. Furthermore, we ensemble the replies of the different experts in a uniform or weighted manner, which builds a quorum from the expert replies. Overall this leads to accuracy improvements of up to 10.6 p.p. compared to the challenge baseline. We show that a Zephyr 3B model is well suited for the task. Our approach can be applied in the model-agnostic and model-aware subtasks without modification and is flexible and easily extendable to related tasks.

pdf bib
whatdoyoumeme at SemEval-2024 Task 4: Hierarchical-Label-Aware Persuasion Detection using Translated Texts
Nishan Chatterjee | Marko Pranjic | Boshko Koloski | Lidia Pivovarova | Senja Pollak

In this paper, we detail the methodology of team whatdoyoumeme for the SemEval 2024 Task on Multilingual Persuasion Detection in Memes. We integrate hierarchical label information to refine detection capabilities, and employ a cross-lingual approach, utilizing translation to adapt the model to Macedonian, Arabic, and Bulgarian. Our methodology encompasses both the analysis of meme content and extending labels to include hierarchical structure. The effectiveness of the approach is demonstrated through improved model performance in multilingual contexts, highlighting the utility of translation-based methods and hierarchy-aware learning, over traditional baselines.

pdf bib
LomonosovMSU at SemEval-2024 Task 4: Comparing LLMs and embedder models to identifying propaganda techniques in the content of memes in English for subtasks No1, No2a, and No2b
Gleb Skiba | Mikhail Pukemo | Dmitry Melikhov | Konstantin Vorontsov

This paper presents the solution of the LomonosovMSU team for the SemEval-2024 Task 4 “Multilingual Detection of Persuasion Techniques in Memes” competition for the English language task. During the task solving process, generative and BERT-like (training classifiers on top of embedder models) approaches were tested for subtask No1, as well as an BERT-like approach on top of multimodal embedder models for subtasks No2a/No2b. The models were trained using datasets provided by the competition organizers, enriched with filtered datasets from previous SemEval competitions. The following results were achieved: 18th place for subtask No1, 9th place for subtask No2a, and 11th place for subtask No2b.

pdf bib
AILS-NTUA at SemEval-2024 Task 6: Efficient model tuning for hallucination detection and analysis
Natalia Grigoriadou | Maria Lymperaiou | George Filandrianos | Giorgos Stamou

In this paper, we present our team’s submissions for SemEval-2024 Task-6 - SHROOM, a Shared-task on Hallucinations and Related Observable Overgeneration Mistakes. The participants were asked to perform binary classification to identify cases of fluent overgeneration hallucinations. Our experimentation included fine-tuning a pre-trained model on hallucination detection and a Natural Language Inference (NLI) model. The most successful strategy involved creating an ensemble of these models, resulting in accuracy rates of 77.8% and 79.9% on model-agnostic and model-aware datasets respectively, outperforming the organizers’ baseline and achieving notable results when contrasted with the top-performing results in the competition, which reported accuracies of 84.7% and 81.3% correspondingly.

pdf bib
JMI at SemEval 2024 Task 3: Two-step approach for multimodal ECAC using in-context learning with GPT and instruction-tuned Llama models
Arefa . | Mohammed Abbas Ansari | Chandni Saxena | Tanvir Ahmad

This paper presents our system development for SemEval-2024 Task 3: “The Competition of Multimodal Emotion Cause Analysis in Conversations”. Effectively capturing emotions in human conversations requires integrating multiple modalities such as text, audio, and video. However, the complexities of these diverse modalities pose challenges for developing an efficient multimodal emotion cause analysis (ECA) system. Our proposed approach addresses these challenges by a two-step framework. We adopt two different approaches in our implementation. In Approach 1, we employ instruction-tuning with two separate Llama 2 models for emotion and cause prediction. In Approach 2, we use GPT-4V for conversation-level video description and employ in-context learning with annotated conversation using GPT 3.5. Our system wins rank 4, and system ablation experiments demonstrate that our proposed solutions achieve significant performance gains.

pdf bib
LMU-BioNLP at SemEval-2024 Task 2: Large Diverse Ensembles for Robust Clinical NLI
Zihang Sun | Danqi Yan | Anyi Wang | Tanalp Agustoslu | Qi Feng | Chengzhi Hu | Longfei Zuo | Shijia Zhou | Hermine Kleiner | Pingjun Hong

In this paper, we describe our submission for the NLI4CT 2024 shared task on robust Natural Language Inference over clinical trial reports. Our system is an ensemble of nine diverse models which we aggregate via majority voting. The models use a large spectrum of different approaches ranging from a straightforward Convolutional Neural Network over fine-tuned Large Language Models to few-shot-prompted language models using chain-of-thought reasoning.Surprisingly, we find that some individual ensemble members are not only more accurate than the final ensemble model but also more robust.

pdf bib
MARiA at SemEval 2024 Task-6: Hallucination Detection Through LLMs, MNLI, and Cosine similarity
Reza Sanayei | Abhyuday Singh | Mohammadhossein Rezaei | Steven Bethard

The advent of large language models (LLMs) has revolutionized Natural Language Generation (NLG), offering unmatched text generation capabilities. However, this progress introduces significant challenges, notably hallucinations—semantically incorrect yet fluent outputs. This phenomenon undermines content reliability, as traditional detection systems focus more on fluency than accuracy, posing a risk of misinformation spread.Our study addresses these issues by proposing a unified strategy for detecting hallucinations in neural model-generated text, focusing on the SHROOM task in SemEval 2024. We employ diverse methodologies to identify output divergence from the source content. We utilized Sentence Transformers to measure cosine similarity between source-hypothesis and source-target embeddings, experimented with omitting source content in the cosine similarity computations, and Leveragied LLMs’ In-Context Learning with detailed task prompts as our methodologies. The varying performance of our different approaches across the subtasks underscores the complexity of Natural Language Understanding tasks, highlighting the importance of addressing the nuances of semantic correctness in the era of advanced language models.

pdf bib
NUS-Emo at SemEval-2024 Task 3: Instruction-Tuning LLM for Multimodal Emotion-Cause Analysis in Conversations
Meng Luo | Han Zhang | Shengqiong Wu | Bobo Li | Hong Han | Hao Fei

This paper describes the architecture of our system developed for participation in Task 3 of SemEval-2024: Multimodal Emotion-Cause Analysis in Conversations. Our project targets the challenges of subtask 2, dedicated to Multimodal Emotion-Cause Pair Extraction with Emotion Category (MECPE-Cat), and constructs a dual-component system tailored to the unique challenges of this task. We divide the task into two subtasks: emotion recognition in conversation (ERC) and emotion-cause pair extraction (ECPE). To address these subtasks, we capitalize on the abilities of Large Language Models (LLMs), which have consistently demonstrated state-of-the-art performance across various natural language processing tasks and domains. Most importantly, we design an approach of emotion-cause-aware instruction-tuning for LLMs, to enhance the perception of the emotions with their corresponding causal rationales. Our method enables us to adeptly navigate the complexities of MECPE-Cat, achieving an average 34.71% F1 score of the task, and securing the 2nd rank on the leaderboard. The code and metadata to reproduce our experiments are all made publicly available.

pdf bib
TueCICL at SemEval-2024 Task 8: Resource-efficient approaches for machine-generated text detection
Daniel Stuhlinger | Aron Winkler

Recent developments in the field of NLP have brought large language models (LLMs) to the forefront of both public and research attention. As the use of language generation technologies becomes more widespread, the problem arises of determining whether a given text is machine generated or not. Task 8 at SemEval 2024 consists of a shared task with this exact objective. Our approach aims at developing models and strategies that strike a good balance between performance and model size. We show that it is possible to compete with large transformer-based solutions with smaller systems.

pdf bib
GeminiPro at SemEval-2024 Task 9: BrainTeaser on Gemini
Kyu Hyun Choi | Seung-hoon Na

It is known that human thought can be distinguished into lateral and vertical thinking. The development of language models has thus far been focused on evaluating and advancing vertical thinking, while lateral thinking has been somewhat neglected. To foster progress in this area, SemEval has created and distributed a brainteaser dataset based on lateral thinking consist of sentence puzzles and word puzzle QA. In this paper, we test and discuss the performance of the currently known best model, Gemini, on this dataset.

pdf bib
Archimedes-AUEB at SemEval-2024 Task 5: LLM explains Civil Procedure
Odysseas Chlapanis | Ion Androutsopoulos | Dimitrios Galanis

The SemEval task on Argument Reasoning in Civil Procedure is challenging in that it requires understanding legal concepts and inferring complex arguments. Currently, most Large Language Models (LLM) excelling in the legal realm are principally purposed for classification tasks, hence their reasoning rationale is subject to contention. The approach we advocate involves using a powerful teacher-LLM (ChatGPT) to extend the training dataset with explanations and generate synthetic data. The resulting data are then leveraged to fine-tune a small student-LLM. Contrary to previous work, our explanations are not directly derived from the teacher’s internal knowledge. Instead they are grounded in authentic human analyses, therefore delivering a superior reasoning signal. Additionally, a new ‘mutation’ method generates artificial data instances inspired from existing ones. We are publicly releasing the explanations as an extension to the original dataset, along with the synthetic dataset and the prompts that were used to generate both. Our system ranked 15th in the SemEval competition. It outperforms its own teacher and can produce explanations aligned with the original human analyses, as verified by legal experts.

pdf bib
Weighted Layer Averaging RoBERTa for Black-Box Machine-Generated Text Detection
Ayan Datta | Aryan Chandramania | Radhika Mamidi

We propose a novel approach for machine-generated text detection using a RoBERTa model with weighted layer averaging and AdaLoRA for parameter-efficient fine-tuning. Our method incorporates information from all model layers, capturing diverse linguistic cues beyond those accessible from the final layer alone. To mitigate potential overfitting and improve generalizability, we leverage AdaLoRA, which injects trainable low-rank matrices into each Transformer layer, significantly reducing the number of trainable parameters. Furthermore, we employ data mixing to ensure our model encounters text from various domains and generators during training, enhancing its ability to generalize to unseen data. This work highlights the potential of combining layer-wise information with parameter-efficient fine-tuning and data mixing for effective machine-generated text detection.

pdf bib
Mast Kalandar at SemEval-2024 Task 8: On the Trail of Textual Origins: RoBERTa-BiLSTM Approach to Detect AI-Generated Text
Jainit Bafna | Hardik Mittal | Suyash Sethia | Manish Shrivastava | Radhika Mamidi

Large Language Models (LLMs) have showcased impressive abilities in generating fluent responses to diverse user queries. However, concerns regarding the potential misuse ofsuch texts in journalism, educational, and academic contexts have surfaced. SemEval 2024introduces the task of Multigenerator, Multidomain, and Multilingual Black-Box MachineGenerated Text Detection, aiming to developautomated systems for identifying machinegenerated text and detecting potential misuse. In this paper, we i) propose a RoBERTaBiLSTM based classifier designed to classifytext into two categories: AI-generated or human ii) conduct a comparative study of ourmodel with baseline approaches to evaluate itseffectiveness. This paper contributes to the advancement of automatic text detection systemsin addressing the challenges posed by machinegenerated text misuse. Our architecture ranked46th on the official leaderboard with an accuracy of 80.83 among 125.

pdf bib
HW-TSC 2024 Submission for the SemEval-2024 Task 1: Semantic Textual Relatedness (STR)
Mengyao Piao | Su Chang | Yuang Li | Xiaosong Qiao | Xiaofeng Zhao | Yinglu Li | Min Zhang | Hao Yang

The degree of semantic relatedness of two units of language has long been considered fundamental to understanding meaning. In this paper, we present the system of Huawei Translation Services Center (HW-TSC) for Task 1 of SemEval 2024, which aims to automatically measure the semantic relatedness of sentence pairs in African and Asian languages. The task dataset for this task covers about 14 different languages, These languages originate from five distinct language families and are predominantly spoken in Africa and Asia. For this shared task, we describe our proposed solutions, including ideas and the implementation steps of the task, as well as the outcomes of each experiment on the development dataset. To enhance the performance, we leverage these experimental outcomes and construct an ensemble one. Our results demonstrate that our system achieves impressive performance on test datasets in unsupervised track B and ranked first place for the Punjabi language pair.

pdf bib
KnowComp at SemEval-2024 Task 9: Conceptualization-Augmented Prompting with Large Language Models for Lateral Reasoning
Weiqi Wang | Baixuan Xu | Haochen Shi | Jiaxin Bai | Qi Hu | Yangqiu Song

Lateral thinking is essential in breaking away from conventional thought patterns and finding innovative solutions to problems. Despite this, language models often struggle with reasoning tasks that require lateral thinking. In this paper, we present our system for SemEval-2024 Task 9’s BrainTeaser challenge, which requires language models to answer brain teaser questions that typically involve lateral reasoning scenarios. Our framework is based on large language models and incorporates a zero-shot prompting method that integrates conceptualizations of automatically detected instances in the question. We also transform the task of question answering into a declarative format to enhance the discriminatory ability of large language models. Our zero-shot evaluation results with ChatGPT indicate that our approach outperforms baselines, including zero-shot and few-shot prompting and chain-of-thought reasoning. Additionally, our system ranks ninth on the official leaderboard, demonstrating its strong performance.

pdf bib
HW-TSC at SemEval-2024 Task 9: Exploring Prompt Engineering Strategies for Brain Teaser Puzzles Through LLMs
Yinglu Li | Zhao Yanqing | Min Zhang | Yadong Deng | Aiju Geng | Xiaoqin Liu | Mengxin Ren | Yuang Li | Su Chang | Xiaofeng Zhao

Large Language Models (LLMs) have demonstrated impressive performance on many Natural Language Processing (NLP) tasks. However, their ability to solve more creative, lateral thinking puzzles remains relatively unexplored. In this work, we develop methods to enhance the lateral thinking and puzzle-solving capabilities of LLMs. We curate a dataset of word-type and sentence-type brain teasers requiring creative problem-solving abilities beyond commonsense reasoning. We first evaluate the zero-shot performance of models like GPT-3.5 and GPT-4 on this dataset. To improve their puzzle-solving skills, we employ prompting techniques like providing reasoning clues and chaining multiple examples to demonstrate the desired thinking process. We also fine-tune the state-of-the-art Mixtral 7x8b LLM on ourdataset. Our methods enable the models to achieve strong results, securing 2nd and 3rd places in the brain teaser task. Our work highlights the potential of LLMs in acquiring complex reasoning abilities with the appropriate training. The efficacy of our approaches opens up new research avenues into advancing lateral thinking and creative problem-solving with AI systems.

pdf bib
SU-FMI at SemEval-2024 Task 5: From BERT Fine-Tuning to LLM Prompt Engineering - Approaches in Legal Argument Reasoning
Kristiyan Krumov | Svetla Boytcheva | Ivan Koytchev

This paper presents our approach and findings for SemEval-2024 Task 5, focusing on legal argument reasoning. We explored the effectiveness of fine-tuning pre-trained BERT models and the innovative application of large language models (LLMs) through prompt engineering in the context of legal texts. Our methodology involved a combination of techniques to address the challenges posed by legal language processing, including handling long texts and optimizing natural language understanding (NLU) capabilities for the legal domain. Our contributions were validated by achieving a third-place ranking on the SemEval 2024 Task 5 Leaderboard. The results underscore the potential of LLMs and prompt engineering in enhancing legal reasoning tasks, offering insights into the evolving landscape of NLU technologies within the legal field.

pdf bib
Challenges at SemEval 2024 Task 7: Contrastive Learning Approach on Numeral-Aware Language Generation
Ali Zhunis | Hao-yun Chuang

Although Large Language Model (LLM) excels on generating headline on ROUGE evaluation, it still fails to reason number and generate news article headline with accurate number. Attending SemEval-2024 Task 7 subtask 3, our team aims on using contrastive loss to increase the understanding of the number from their different expression, and knows to identify between different number and its respective expression. This system description paper uses T5 and BART as the baseline model, comparing its result with and without the constrative loss. The result shows that BART with contrastive loss have excelled all the models, and its performance on the number accuracy has the highest performance among all.

pdf bib
Team Bolaca at SemEval-2024 Task 6: Sentence-transformers are all you need
Béla Rösener | Hong-bo Wei | Ilinca Vandici

Our team tackled the SemEval-2024 Task 6, focusing on identifying fluent over-generation hallucinations in NLP outputs. We proposed a pragmatic solution using a logistic regression classifier and a feed-forward ANN, harnessing SBERT embeddings for feature extraction.

pdf bib
AIpom at SemEval-2024 Task 8: Detecting AI-produced Outputs in M4
Alexander Shirnin | Nikita Andreev | Vladislav Mikhailov | Ekaterina Artemova

This paper describes AIpom, a system designed to detect a boundary between human-written and machine-generated text (SemEval-2024 Task 8, Subtask C: Human-Machine Mixed Text Detection). We propose a two-stage pipeline combining predictions from an instruction-tuned decoder-only model and encoder-only sequence taggers. AIpom is ranked second on the leaderboard while achieving a Mean Absolute Error of 15.94. Ablation studies confirm the benefits of pipelining encoder and decoder models, particularly in terms of improved performance.

pdf bib
CLaC at SemEval-2024 Task 2: Faithful Clinical Trial Inference
Jennifer Marks | Mohammadreza Davari | Leila Kosseim

This paper presents the methodology used for our participation in SemEval 2024 Task 2 (Jullien et al., 2024) – Safe Biomedical Natural Language Inference for Clinical Trials. The task involved Natural Language Inference (NLI) on clinical trial data, where statements were provided regarding information within Clinical Trial Reports (CTRs). These statements could pertain to a single CTR or compare two CTRs, requiring the identification of the inference relation (entailment vs contradiction) between CTR-statement pairs. Evaluation was based on F1, Faithfulness, and Consistency metrics, with priority given to the latter two by the organizers. Our approach aims to maximize Faithfulness and Consistency, guided by intuitive definitions provided by the organizers, without detailed metric calculations. Experimentally, our approach yielded models achieving maximal Faithfulness (top rank) and average Consistency (mid rank) at the expense of F1 (low rank). Future work will focus on refining our approach to achieve a balance among all three metrics.

pdf bib
MALTO at SemEval-2024 Task 6: Leveraging Synthetic Data for LLM Hallucination Detection
Federico Borra | Claudio Savelli | Giacomo Rosso | Alkis Koudounas | Flavio Giobergia

In Natural Language Generation (NLG), contemporary Large Language Models (LLMs) face several challenges, such as generating fluent yet inaccurate outputs and reliance on fluency-centric metrics. This often leads to neural networks exhibiting “hallucinations.” The SHROOM challenge focuses on automatically identifying these hallucinations in the generated text. To tackle these issues, we introduce two key components, a data augmentation pipeline incorporating LLM-assisted pseudo-labelling and sentence rephrasing, and a voting ensemble from three models pre-trained on Natural Language Inference (NLI) tasks and fine-tuned on diverse datasets.

pdf bib
Maha Bhaashya at SemEval-2024 Task 6: Zero-Shot Multi-task Hallucination Detection
Patanjali Bhamidipati | Advaith Malladi | Manish Shrivastava | Radhika Mamidi

In recent studies, the extensive utilization oflarge language models has underscored the importance of robust evaluation methodologiesfor assessing text generation quality and relevance to specific tasks. This has revealeda prevalent issue known as hallucination, anemergent condition in the model where generated text lacks faithfulness to the source anddeviates from the evaluation criteria. In thisstudy, we formally define hallucination and propose a framework for its quantitative detectionin a zero-shot setting, leveraging our definitionand the assumption that model outputs entailtask and sample specific inputs. In detectinghallucinations, our solution achieves an accuracy of 0.78 in a model-aware setting and 0.61in a model-agnostic setting. Notably, our solution maintains computational efficiency, requiring far less computational resources than other SOTA approaches, aligning with the trendtowards lightweight and compressed models.

pdf bib
Team art-nat-HHU at SemEval-2024 Task 8: Stylistically Informed Fusion Model for MGT-Detection
Vittorio Ciccarelli | Cornelia Genz | Nele Mastracchio | Wiebke Petersen | Anna Stein | Hanxin Xia

This paper presents our solution for subtask A of shared task 8 of SemEval 2024 for classifying human- and machine-written texts in English across multiple domains. We propose a fusion model consisting of RoBERTa based pre-classifier and two MLPs that have been trained to correct the pre-classifier using linguistic features. Our model achieved an accuracy of 85%.

pdf bib
AIMA at SemEval-2024 Task 3: Simple Yet Powerful Emotion Cause Pair Analysis
Alireza Ghahramani Kure | Mahshid Dehghani | Mohammad Mahdi Abootorabi | Nona Ghazizadeh | Seyed Arshan Dalili | Ehsaneddin Asgari

The SemEval-2024 Task 3 presents two subtasks focusing on emotion-cause pair extraction within conversational contexts. Subtask 1 revolves around the extraction of textual emotion-cause pairs, where causes are defined and annotated as textual spans within the conversation. Conversely, Subtask 2 extends the analysis to encompass multimodal cues, including language, audio, and vision, acknowledging instances where causes may not be exclusively represented in the textual data. Our proposed model for emotion-cause analysis is meticulously structured into three core segments: (i) embedding extraction, (ii) cause-pair extraction & emotion classification, and (iii) cause extraction using QA after finding pairs. Leveraging state-of-the-art techniques and fine-tuning on task-specific datasets, our model effectively unravels the intricate web of conversational dynamics and extracts subtle cues signifying causality in emotional expressions. Our team, AIMA, demonstrated strong performance in the SemEval-2024 Task 3 competition. We ranked as the 10th in subtask 1 and the 6th in subtask 2 out of 23 teams.

pdf bib
AIMA at SemEval-2024 Task 10: History-Based Emotion Recognition in Hindi-English Code-Mixed Conversations
Mohammad Mahdi Abootorabi | Nona Ghazizadeh | Seyed Arshan Dalili | Alireza Ghahramani Kure | Mahshid Dehghani | Ehsaneddin Asgari

In this study, we introduce a solution to the SemEval 2024 Task 10 on subtask 1, dedicated to Emotion Recognition in Conversation (ERC) in code-mixed Hindi-English conversations. ERC in code-mixed conversations presents unique challenges, as existing models are typically trained on monolingual datasets and may not perform well on code-mixed data. To address this, we propose a series of models that incorporate both the previous and future context of the current utterance, as well as the sequential information of the conversation. To facilitate the processing of code-mixed data, we developed a Hinglish-to-English translation pipeline to translate the code-mixed conversations into English. We designed four different base models, each utilizing powerful pre-trained encoders to extract features from the input but with varying architectures. By ensembling all of these models, we developed a final model that outperforms all other baselines.

pdf bib
Team MGTD4ADL at SemEval-2024 Task 8: Leveraging (Sentence) Transformer Models with Contrastive Learning for Identifying Machine-Generated Text
Huixin Chen | Jan Büssing | David Rügamer | Ercong Nie

This paper outlines our approach to SemEval-2024 Task 8 (Subtask B), which focuses on discerning machine-generated text from human-written content, while also identifying the text sources, i.e., from which Large Language Model (LLM) the target text is generated. Our detection system is built upon Transformer-based techniques, leveraging various pre-trained language models (PLMs), including sentence transformer models. Additionally, we incorporate Contrastive Learning (CL) into the classifier to improve the detecting capabilities and employ Data Augmentation methods. Ultimately, our system achieves a peak accuracy of 76.96% on the test set of the competition, configured using a sentence transformer model integrated with CL methodology.

pdf bib
ClusterCore at SemEval-2024 Task 7: Few Shot Prompting With Large Language Models for Numeral-Aware Headline Generation
Monika Singh | Sujit Kumar | Tanveen . | Sanasam Ranbir Singh

The generation of headlines, a crucial aspect of abstractive summarization, aims to compress an entire article into a concise, single line of text despite the effectiveness of modern encoder-decoder models for text generation and summarization tasks. The encoder-decoder model commonly faces challenges in accurately generating numerical content within headlines. This study empirically explored LLMs for numeral-aware headline generation and proposed few-shot prompting with LLMs for numeral-aware headline generations. Experiments conducted on the NumHG dataset and NumEval-2024 test set suggest that fine-tuning LLMs on NumHG dataset enhances the performance of LLMs for numeral aware headline generation. Furthermore, few-shot prompting with LLMs surpassed the performance of fine-tuned LLMs for numeral-aware headline generation.

pdf bib
HierarchyEverywhere at SemEval-2024 Task 4: Detection of Persuasion Techniques in Memes Using Hierarchical Text Classifier
Omid Ghahroodi | Ehsaneddin Asgari

Text classification is an important task in natural language processing. Hierarchical Text Classification (HTC) is a subset of text classification task-type. HTC tackles multi-label classification challenges by leveraging tree structures that delineate relationships between classes, thereby striving to enhance classification accuracy through the utilization of inter-class relationships. Memes, as prevalent vehicles of modern communication within social networks, hold immense potential as instruments for propagandistic dissemination due to their profound impact on users. In SemEval-2024 Task 4, the identification of propaganda and its various forms in memes is explored through two sub-tasks: (i) utilizing only the textual component of memes, and (ii) incorporating both textual and pictorial elements. In this study, we address the proposed problem through the lens of HTC, using state-of-the-art hierarchical text classification methodologies to detect propaganda in memes. Our system achieved first place in English Sub-task 2a, underscoring its efficacy in tackling the complexities inherent in propaganda detection within the meme landscape.

pdf bib
AILS-NTUA at SemEval-2024 Task 9: Cracking Brain Teasers: Transformer Models for Lateral Thinking Puzzles
Ioannis Panagiotopoulos | George Filandrianos | Maria Lymperaiou | Giorgos Stamou

In this paper, we outline our submission for the SemEval-2024 Task 9 competition: ‘BRAINTEASER: A Novel Task Defying Common Sense’. We engage in both sub-tasks: Sub-task A-Sentence Puzzle and Sub-task B-Word Puzzle. We evaluate a plethora of pre-trained transformer-based language models of different sizes through fine-tuning. Subsequently, we undertake an analysis of their scores and responses to aid future researchers in understanding and utilizing these models effectively. Our top-performing approaches secured competitive positions on the competition leaderboard across both sub-tasks. In the evaluation phase, our best submission attained an average accuracy score of 81.7% in the Sentence Puzzle, and 85.4% in the Word Puzzle, significantly outperforming the best neural baseline (ChatGPT) by more than 20% and 30% respectively.

pdf bib
DeepPavlov at SemEval-2024 Task 3: Multimodal Large Language Models in Emotion Reasoning
Julia Belikova | Dmitrii Kosenko

This paper presents the solution of the DeepPavlov team for the Multimodal Sentiment Cause Analysis competition in SemEval-2024 Task 3, Subtask 2 (Wang et al., 2024). In the evaluation leaderboard, our approach ranks 7th with an F1-score of 0.2132. Large Language Models (LLMs) are transformative in their ability to comprehend and generate human-like text. With recent advancements, Multimodal Large Language Models (MLLMs) have expanded LLM capabilities, integrating different modalities such as audio, vision, and language. Our work delves into the state-of-the-art MLLM Video-LLaMA, its associated modalities, and its application to the emotion reasoning downstream task, Multimodal Emotion Cause Analysis in Conversations (MECAC). We investigate the model’s performance in several modes: zero-shot, few-shot, individual embeddings, and fine-tuned, providing insights into their limits and potential enhancements for emotion understanding.

pdf bib
iREL at SemEval-2024 Task 9: Improving Conventional Prompting Methods for Brain Teasers
Harshit Gupta | Manav Chaudhary | Shivansh Subramanian | Tathagata Raha | Vasudeva Varma

This paper describes our approach for SemEval-2024 Task 9: BRAINTEASER: A Novel Task Defying Common Sense. The BRAINTEASER task comprises multiple-choice Question Answering designed to evaluate the models’ lateral thinking capabilities. It consists of Sentence Puzzle and Word Puzzle subtasks that require models to defy default commonsense associations and exhibit unconventional thinking. We propose a unique strategy to improve the performance of pre-trained language models, notably the Gemini 1.0 Pro Model, in both subtasks. We employ static and dynamic few-shot prompting techniques and introduce a model-generated reasoning strategy that utilizes the LLM’s reasoning capabilities to improve performance. Our approach demonstrated significant improvements, showing that it performed better than the baseline models by a considerable margin but fell short of performing as well as the human annotators, thus highlighting the efficacy of the proposed strategies.

pdf bib
uTeBC-NLP at SemEval-2024 Task 9: Can LLMs be Lateral Thinkers?
Pouya Sadeghi | Amirhossein Abaskohi | Yadollah Yaghoobzadeh

Inspired by human cognition, Jiang et al. 2023 create a benchmark for assessing LLMs’ lateral thinking—thinking outside the box. Building upon this benchmark, we investigate how different prompting methods enhance LLMs’ performance on this task to reveal their inherent power for outside-the-box thinking ability. Through participating in SemEval-2024, task 9, Sentence Puzzle sub-task, we explore prompt engineering methods: chain of thoughts (CoT) and direct prompting, enhancing with informative descriptions, and employing contextualizing prompts using a retrieval augmented generation (RAG) pipeline. Our experiments involve three LLMs including GPT-3.5, GPT-4, and Zephyr-7B-beta. We generate a dataset of thinking paths between riddles and options using GPT-4, validated by humans for quality. Findings indicate that compressed informative prompts enhance performance. Dynamic in-context learning enhances model performance significantly. Furthermore, fine-tuning Zephyr on our dataset enhances performance across other commonsense datasets, underscoring the value of innovative thinking.

pdf bib
IITK at SemEval-2024 Task 4: Hierarchical Embeddings for Detection of Persuasion Techniques in Memes
Shreenaga Chikoti | Shrey Mehta | Ashutosh Modi

Memes are one of the most popular types of content used in an online disinformation campaign. They are primarily effective on social media platforms since they can easily reach many users. Memes in a disinformation campaign achieve their goal of influencing the users through several rhetorical and psychological techniques, such as causal oversimplification, name-calling, and smear. The SemEval 2024 Task 4 Multilingual Detection of Persuasion Technique in Memes on identifying such techniques in the memes is divided across three sub-tasks: (1) Hierarchical multi-label classification using only textual content of the meme, (2) Hierarchical multi-label classification using both, textual and visual content of the meme and (3) Binary classification of whether the meme contains a persuasion technique or not using it’s textual and visual content. This paper proposes an ensemble of Class Definition Prediction (CDP) and hyperbolic embeddings-based approaches for this task. We enhance meme classification accuracy and comprehensiveness by integrating HypEmo’s hierarchical label embeddings (Chen et al., 2023) and a multi-task learning framework for emotion prediction. We achieve a hierarchical F1-score of 0.60, 0.67, and 0.48 on the respective sub-tasks.

pdf bib
HIT-MI&T Lab at SemEval-2024 Task 6: DeBERTa-based Entailment Model is a Reliable Hallucination Detector
Wei Liu | Wanyao Shi | Zijian Zhang | Hui Huang

This paper describes our submission for SemEval-2024 Task 6: SHROOM, a Shared-task on Hallucinations and Related Observable Overgeneration Mistakes. We propose four groups of methods for hallucination detection: 1) Entailment Recognition; 2) Similarity Search; 3) Factuality Verification; 4) Confidence Estimation. The four methods rely on either the semantic relationship between the hypothesis and its source (target) or on the model-aware features during decoding. We participated in both the model-agnostic and model-aware tracks. Our method’s effectiveness is validated by our high rankings 3rd in the model-agnostic track and 5th in the model-aware track. We have released our code on GitHub.

pdf bib
UAlberta at SemEval-2024 Task 1: A Potpourri of Methods for Quantifying Multilingual Semantic Textual Relatedness and Similarity
Ning Shi | Senyu Li | Guoqing Luo | Amirreza Mirzaei | Ali Rafiei | Jai Riley | Hadi Sheikhi | Mahvash Siavashpour | Mohammad Tavakoli | Bradley Hauer

We describe our systems for SemEval-2024 Task 1: Semantic Textual Relatedness. We investigate the correlation between semantic relatedness and semantic similarity. Specifically, we test two hypotheses: (1) similarity is a special case of relatedness, and (2) semantic relatedness is preserved under translation. We experiment with a variety of approaches which are based on explicit semantics, downstream applications, contextual embeddings, large language models (LLMs), as well as ensembles of methods. We find empirical support for our theoretical insights. In addition, our best ensemble system yields highly competitive results in a number of diverse categories. Our code and data are available on GitHub.

pdf bib
HW-TSC at SemEval-2024 Task 5: Self-Eval? A Confident LLM System for Auto Prediction and Evaluation for the Legal Argument Reasoning Task
Xiaofeng Zhao | Xiaosong Qiao | Kaiwen Ou | Min Zhang | Su Chang | Mengyao Piao | Yuang Li | Yinglu Li | Ming Zhu | Yilun Liu

In this article, we present an effective system for semeval-2024 task 5. The task involves assessing the feasibility of a given solution in civil litigation cases based on relevant legal provisions, issues, solutions, and analysis. This task demands a high level of proficiency in U.S. law and natural language reasoning. In this task, we designed a self-eval LLM system that simultaneously performs reasoning and self-assessment tasks. We created a confidence interval and a prompt instructing the LLM to output the answer to a question along with its confidence level. We designed a series of experiments to prove the effectiveness of the self-eval mechanism. In order to avoid the randomness of the results, the final result is obtained by voting on three results generated by the GPT-4. Our submission was conducted under zero-resource setting, and we achieved first place in the task with an F1-score of 0.8231 and an accuracy of 0.8673.

pdf bib
IITK at SemEval-2024 Task 10: Who is the speaker? Improving Emotion Recognition and Flip Reasoning in Conversations via Speaker Embeddings
Shubham Patel | Divyaksh Shukla | Ashutosh Modi

This paper presents our approach for the SemEval-2024 Task 10: Emotion Discovery and Reasoning its Flip in Conversations. We propose a transformer-based speaker-centric model for the Emotion Flip Reasoning (EFR) task and a masked-memory network along with a speaker participation vector for the Emotion Recognition in Conversations (ERC) task. We propose a Probable Trigger Zone, which is more likely to contain the utterances causing the emotion of a speaker to flip. In EFR, sub-task 3, the proposed approach archives a 5.9 (F1 score) improvement over the provided task baseline. The ablation study results highlight the significance of various design choices in the proposed method.

pdf bib
DeepPavlov at SemEval-2024 Task 8: Leveraging Transfer Learning for Detecting Boundaries of Machine-Generated Texts
Anastasia Voznyuk | Vasily Konovalov

The Multigenerator, Multidomain, and Multilingual Black-Box Machine-Generated Text Detection shared task in the SemEval-2024 competition aims to tackle the problem of misusing collaborative human-AI writing. Although there are a lot of existing detectors of AI content, they are often designed to give a binary answer and thus may not be suitable for more nuanced problem of finding the boundaries between human-written and machine-generated texts, while hybrid human-AI writing becomes more and more popular. In this paper, we address the boundary detection problem. Particularly, we present a pipeline for augmenting data for supervised fine-tuning of DeBERTaV3. We receive new best MAE score, according to the leaderboard of the competition, with this pipeline.

pdf bib
Bit_numeval at SemEval-2024 Task 7: Enhance Numerical Sensitivity and Reasoning Completeness for Quantitative Understanding
Xinyue Liang | Jiawei Li | Yizhe Yang | Yang Gao

In this paper, we describe the methods used for Quantitative Natural Language Inference (QNLI), and Quantitative Question Answering (QQA) in task1 of Semeval2024 NumEval. The challenge’s focus is to enhance the model’s quantitative understanding consequently improving its performance on certain tasks. We accomplish this task from two perspectives: (1) By integrating real-world numerical comparison data during the supervised fine-tuning (SFT) phase, we enhanced the model’s numerical sensitivity. (2) We develop an innovative reward model scoring mechanism, leveraging reinforcement learning from human feedback (RLHF) techniques to improve the model’s reasoning completeness.

pdf bib
MaiNLP at SemEval-2024 Task 1: Analyzing Source Language Selection in Cross-Lingual Textual Relatedness
Shijia Zhou | Huangyan Shan | Barbara Plank | Robert Litschko

This paper presents our system developed for the SemEval-2024 Task 1: Semantic Textual Relatedness (STR), on Track C: Cross-lingual. The task aims to detect semantic relatedness of two sentences from the same languages. For cross-lingual approach we developed a set of linguistics-inspired models trained with several task-specific strategies. We 1) utilize language vectors for selection of donor languages; 2) investigate the multi-source approach for training; 3) use transliteration of non-latin script to study impact of “script gap”; 4) opt machine translation for data augmentation. We additionally compare the performance of XLM-RoBERTa and Furina with the same training strategy. Our submission achieved the first place in the C8 (Kinyarwanda) test.

pdf bib
NLP_Team1@SSN at SemEval-2024 Task 1: Impact of language models in Sentence-BERT for Semantic Textual Relatedness in Low-resource Languages
Senthil Kumar | Aravindan Chandrabose | Gokulakrishnan B | Karthikraja Tp

Semantic Textual Relatedness (STR) will provide insight into the limitations of existing models and support ongoing work on semantic representations. Track A in Shared Task-1, provides pairs of sentences with semantic relatedness scores for 9 languages out of which 7 are low-resources. These languages are from four different language families. We developed models for 8 languages (except for Amharic) in Track A, using Sentence Transformers (SBERT) architecture, and fine-tuned them with multilingual and monolingual pre-trained language models (PLM). Our models for English (eng), Algerian Arabic (arq), andKinyarwanda (kin) languages were ranked 12, 5, and 8 respectively. Our submissions are ranked 5th among 40 submissions in Track A with an average Spearman correlation score of 0.74. However, we observed that the usage of monolingual PLMs did not guarantee better than multilingual PLMs in Marathi (mar), and Telugu (tel) languages in our case.

pdf bib
ShefCDTeam at SemEval-2024 Task 4: A Text-to-Text Model for Multi-Label Classification
Meredith Gibbons | Maggie Mi | Xingyi Song | Aline Villavicencio

This paper presents our findings for SemEval2024 Task 4. We submit only to subtask 1, applying the text-to-text framework using a FLAN-T5 model with a combination of parameter efficient fine-tuning methods - low-rankadaptation and prompt tuning. Overall, we find that the system performs well in English, but performance is limited in Bulgarian, North Macedonian and Arabic. Our analysis raises interesting questions about the effects of labelorder and label names when applying the text-to-text framework.

pdf bib
NLPNCHU at SemEval-2024 Task 4: A Comparison of MDHC Strategy and In-domain Pre-training for Multilingual Detection of Persuasion Techniques in Memes
Shih-wei Guo | Yao-chung Fan

This study presents a systematic method for identifying 22 persuasive techniques used in multilingual memes. We explored various fine-tuning techniques and classification strategies, such as data augmentation, problem transformation, and hierarchical multi-label classification strategies. Identifying persuasive techniques in memes involves a multimodal task. We fine-tuned the XLM-RoBERTA-large-twitter language model, focusing on domain-specific language modeling, and integrated it with the CLIP visual model’s embedding to consider image and text features simultaneously. In our experiments, we evaluated the effectiveness of our approach by using official validation data in English. Our system in the competition, achieving competitive rankings in Subtask1 and Subtask2b across four languages: English, Bulgarian, North Macedonian, and Arabic. Significantly, we achieved 2nd place ranking for Arabic language in Subtask 1.

pdf bib
Mothman at SemEval-2024 Task 9: An Iterative System for Chain-of-Thought Prompt Optimization
Alvin Chen | Ray Groshan | Sean Von Bayern

Extensive research exists on the performance of large language models on logic-based tasks, whereas relatively little has been done on their ability to generate creative solutions on lateral thinking tasks. The BrainTeaser shared task tests lateral thinking and uses adversarial datasets to prevent memorization, resulting in poor performance for out-of-the-box models. We propose a system for iterative, chain-of-thought prompt engineering which optimizes prompts using human evaluation. Using this shared task, we demonstrate our system’s ability to significantly improve model performance by optimizing prompts and evaluate the input dataset.

pdf bib
Zero Shot is All You Need at SemEval-2024 Task 9: A study of State of the Art LLMs on Lateral Thinking Puzzles
Erfan Moosavi Monazzah | Mahdi Feghhi

The successful deployment of large language models in numerous NLP tasks has spurred the demand for tackling more complex tasks, which were previously unattainable. SemEval-2024 Task 9 introduces the brainteaser dataset that necessitates intricate, human-like reasoning to solve puzzles that challenge common sense. At first glance, the riddles in the dataset may appear trivial for humans to solve. However, these riddles demand lateral thinking, which deviates from vertical thinking that is the dominant form when it comes to current reasoning tasks. In this paper, we examine the ability of current state-of-the-art LLMs to solve this task. Our study is diversified by selecting both open and closed source LLMs with varying numbers of parameters. Additionally, we extend the task dataset with synthetic explanations derived from the LLMs’ reasoning processes during task resolution. These could serve as a valuable resource for further expanding the task dataset and developing more robust methods for tasks that require complex reasoning. All the codes and datasets are available in paper’s GitHub repository.

pdf bib
Edinburgh Clinical NLP at SemEval-2024 Task 2: Fine-tune your model unless you have access to GPT-4
Aryo Gema | Giwon Hong | Pasquale Minervini | Luke Daines | Beatrice Alex

The NLI4CT task assesses Natural Language Inference systems in predicting whether hypotheses entail or contradict evidence from Clinical Trial Reports. In this study, we evaluate various Large Language Models (LLMs) with multiple strategies, including Chain-of-Thought, In-Context Learning, and Parameter-Efficient Fine-Tuning (PEFT). We propose a PEFT method to improve the consistency of LLMs by merging adapters that were fine-tuned separately using triplet and language modelling objectives. We found that merging the two PEFT adapters improves the F1 score (+0.0346) and consistency (+0.152) of the LLMs. However, our novel methods did not produce more accurate results than GPT-4 in terms of faithfulness and consistency. Averaging the three metrics, GPT-4 ranks joint-first in the competition with 0.8328. Finally, our contamination analysis with GPT-4 indicates that there was no test data leakage. Our code is available at https://github.com/EdinburghClinicalNLP/semeval_nli4ct.

pdf bib
CaresAI at SemEval-2024 Task 2: Improving Natural Language Inference in Clinical Trial Data using Model Ensemble and Data Explanation
Reem Abdel-salam | Mary Adewunmi | Mercy Akinwale

Large language models (LLMs) have demonstrated state-of-the-art performance across multiple domains in various natural language tasks. Entailment tasks, however, are more difficult to achieve with a high-performance model. The task is to use safe natural language models to conclude biomedical clinical trial reports (CTRs). The Natural Language Inference for Clinical Trial Data (NLI4CT) task aims to define a given entailment and hypothesis based on CTRs. This paper aims to address the challenges of medical abbreviations and numerical data that can be logically inferred from one another due to acronyms, using different data pre-processing techniques to explain such data. This paper presents a model for NLI4CT SemEval 2024 task 2 that trains the data with DeBERTa, BioLink, BERT, GPT2, BioGPT, and Clinical BERT using the best training approaches, such as fine-tuning, prompt tuning, and contrastive learning. Furthermore, to validate these models, different experiments have been carried out. Our best system is built on an ensemble of different models with different training settings, which achieves an F1 score of 0.77, a faithfulness score of 0.76, and a consistency score of 0.75 and secures the sixth rank in the official leaderboard. In conclusion, this paper has addressed challenges in medical text analysis by exploring various NLP techniques, evaluating multiple advanced natural languagemodels(NLM) models and achieving good results with the ensemble model. Additionally, this project has contributed to the advancement of safe and effective NLMs for analysing complex medical data in CTRs.

pdf bib
CVcoders on Semeval-2024 Task 4
Fatemezahra Bakhshande | Mahdieh Naderi

In this paper, we present our methodology for addressing the SemEval 2024 Task 4 on “Multilingual Detection of Persuasion Techniques in Memes.” Our method focuses on identifying persuasion techniques within textual and multimodal meme content using a combination of preprocessing techniques and established models. By integrating advanced preprocessing methods, such as the OpenAI API for text processing, and utilizing a multimodal architecture combining VGG for image feature extraction and GPT-2 for text feature extraction, we achieve improved model performance. To handle class imbalance, we employ Focal Loss as the loss function and AdamW as the optimizer. Experimental results demonstrate the effectiveness of our approach, achieving competitive performance in the task. Notably, our system attains an F1 macro score of 0.67 and an F1 micro score of 0.74 on the test dataset, ranking third among all participants in the competition. Our findings highlight the importance of robust preprocessing techniques and model selection in effectively analyzing memes for persuasion techniques, contributing to efforts to combat misinformation on social media platforms.

pdf bib
Groningen Team F at SemEval-2024 Task 8: Detecting Machine-Generated Text using Feature-Based Machine Learning Models
Rina Donker | Björn Overbeek | Dennis Thulden | Oscar Zwagers

Large language models (LLMs) have shown remarkable capability of creating fluent responses to a wide variety of user queries. However, this also comes with concerns regarding the spread of misinformation and potential misuse within educational context. In this paper we describe our contribution to SemEval-2024 Task 8 (Wang et al., 2024), a shared task created around detecting machine-generated text. We aim to create several feature-based models that can detect whether a text is machine-generated or human-written. In the end, we obtained an accuracy of 0.74 on the binary human-written vs. machine-generated text classification task (Subtask A monolingual) and an accuracy of 0.61 on the multi-way machine-generated text-classification task (Subtask B). For future work, more features and models could be implemented.

pdf bib
Groningen Team A at SemEval-2024 Task 8: Human/Machine Authorship Attribution Using a Combination of Probabilistic and Linguistic Features
Huseyin Alecakir | Puja Chakraborty | Pontus Henningsson | Matthijs Van Hofslot | Alon Scheuer

Our approach primarily centers on feature-based systems, where a diverse array of features pertinent to the text’s linguistic attributes is extracted. Alongside those, we incorporate token-level probabilistic features which are fed into a Bidirectional Long Short-Term Memory (BiLSTM) model. Both resulting feature arrays are concatenated and fed into our final prediction model. Our method under-performed compared to the baseline, despite the fact that previous attempts by others have successfully used linguistic features for the purpose of discerning machine-generated text. We conclude that our examined subset of linguistically motivated features alongside probabilistic features was not able to contribute almost any performance at all to a hybrid classifier of human and machine texts.

pdf bib
SemEval 2024 - Task 10: Emotion Discovery and Reasoning its Flip in Conversation (EDiReF)
Shivani Kumar | Md. Shad Akhtar | Erik Cambria | Tanmoy Chakraborty

We present SemEval-2024 Task 10, a shared task centred on identifying emotions and finding the rationale behind their flips within monolingual English and Hindi-English code-mixed dialogues. This task comprises three distinct subtasks – emotion recognition in conversation for code-mixed dialogues, emotion flip reasoning for code-mixed dialogues, and emotion flip reasoning for English dialogues. Participating systems were tasked to automatically execute one or more of these subtasks. The datasets for these tasks comprise manually annotated conversations focusing on emotions and triggers for emotion shifts.1 A total of 84 participants engaged in this task, with the most adept systems attaining F1-scores of 0.70, 0.79, and 0.76 for the respective subtasks. This paper summarises the results and findings from 24 teams alongside their system descriptions.

pdf bib
SemEval-2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials
Mael Jullien | Marco Valentino | André Freitas

Large Language Models (LLMs) are at the forefront of NLP achievements but fall short in dealing with shortcut learning, factual inconsistency, and vulnerability to adversarial inputs. These shortcomings are especially critical in medical contexts, where they can misrepresent actual model capabilities. Addressing this, we present SemEval-2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials. Our contributions include the refined NLI4CT-P dataset (i.e. Natural Language Inference for Clinical Trials - Perturbed), designed to challenge LLMs with interventional and causal reasoning tasks, along with a comprehensive evaluation of methods and results for participant submissions. A total of 106 participants registered for the task contributing to over 1200 individual submissions and 25 system overview papers. This initiative aims to advance the robustness and applicability of NLI models in healthcare, ensuring safer and more dependable AI assistance in clinical decision-making. We anticipate that the dataset, models, and outcomes of this task can support future research in the field of biomedical NLI. The dataset, competition leaderboard, and website are publicly available.

pdf bib
SemEval Task 1: Semantic Textual Relatedness for African and Asian Languages
Nedjma Ousidhoum | Shamsuddeen Hassan Muhammad | Mohamed Abdalla | Idris Abdulmumin | Ibrahim Said Ahmad | Sanchit Ahuja | Alham Fikri Aji | Vladimir Araujo | Meriem Beloucif | Christine De Kock

We present the first shared task on Semantic Textual Relatedness (STR). While earlier shared tasks primarily focused on semantic similarity, we instead investigate the broader phenomenon of semantic relatedness across 14 languages: Afrikaans, Algerian Arabic, Amharic, English, Hausa, Hindi, Indonesian, Kinyarwanda, Marathi, Moroccan Arabic, Modern Standard Arabic, Punjabi, Spanish, and Telugu. These languages originate from five distinct language families and are predominantly spoken in Africa and Asia – regions characterised by the relatively limited availability of NLP resources. Each instance in the datasets is a sentence pair associated with a score that represents the degree of semantic textual relatedness between the two sentences. Participating systems were asked to rank sentence pairs by their closeness in meaning (i.e., their degree of semantic relatedness) in the 14 languages in three main tracks: (a) supervised, (b) unsupervised, and (c) crosslingual. The task attracted 163 participants. We received 70 submissions in total (across all tasks) from 51 different teams, and 38 system description papers. We report on the best-performing systems as well as the most common and the most effective approaches for the three different tracks.

pdf bib
SemEval-2024 Task 6: SHROOM, a Shared-task on Hallucinations and Related Observable Overgeneration Mistakes
Timothee Mickus | Elaine Zosa | Raul Vazquez | Teemu Vahtola | Jörg Tiedemann | Vincent Segonne | Alessandro Raganato | Marianna Apidianaki

This paper presents the results of the SHROOM, a shared task focused on detecting hallucinations: outputs from natural language generation (NLG) systems that are fluent, yet inaccurate. Such cases of overgeneration put in jeopardy many NLG applications, where correctness is often mission-critical. The shared task was conducted with a newly constructed dataset of 4000 model outputs labeled by 5 annotators each, spanning 3 NLP tasks: machine translation, paraphrase generation and definition modeling.The shared task was tackled by a total of 58 different users grouped in 42 teams, out of which 26 elected to write a system description paper; collectively, they submitted over 300 prediction sets on both tracks of the shared task. We observe a number of key trends in how this approach was tackled—many participants rely on a handful of model, and often rely either on synthetic data for fine-tuning or zero-shot prompting strategies. While a majority of the teams did outperform our proposed baseline system, the performances of top-scoring systems are still consistent with a random handling of the more challenging items.

pdf bib
SemEval-2024 Task 9: BRAINTEASER: A Novel Task Defying Common Sense
Yifan Jiang | Filip Ilievski | Kaixin Ma

While vertical thinking relies on logical and commonsense reasoning, lateral thinking requires systems to defy commonsense associations and overwrite them through unconventional thinking. Lateral thinking has been shown to be challenging for current models but has received little attention. A recent benchmark, BRAINTEASER, aims to evaluate current models’ lateral thinking ability in a zero-shot setting. In this paper, we split the original benchmark to also support fine-tuning setting and present SemEval Task 9, BRAINTEASER(S), the first task at this competition designed to test the system’s reasoning and lateral thinking ability. As a popular task, BRAINTEASER(S)’s two subtasks receive 483 team submissions from 182 participants during the competition. This paper provides a fine-grained system analysis of the competition results, together with a reflection on what this means for the ability of the systems to reason laterally.We hope that the BRAINTEASER(S) subtasks and findings in this paper can stimulate future work on lateral thinking and robust reasoning by computational models

pdf bib
SemEval-2024 Task 4: Multilingual Detection of Persuasion Techniques in Memes
Dimitar Dimitrov | Firoj Alam | Maram Hasanain | Abul Hasnat | Fabrizio Silvestri | Preslav Nakov | Giovanni Da San Martino

The automatic identification of misleading and persuasive content has emerged as a significant issue among various stakeholders, including social media platforms, policymakers, and the broader society. To tackle this issue within the context of memes, we organized a shared task at SemEval-2024, focusing on the multilingual detection of persuasion techniques. This paper outlines the dataset, the organization of the task, the evaluation framework, the outcomes, and the systems that participated. The task targets memes in four languages, with the inclusion of three surprise test datasets in Bulgarian, North Macedonian, and Arabic. It encompasses three subtasks: (i) identifying whether a meme utilizes a persuasion technique; (ii) identifying persuasion techniques within the meme’s ”textual content”; and (iii) identifying persuasion techniques across both the textual and visual components of the meme (a multimodal task). Furthermore, due to the complex nature of persuasion techniques, we present a hierarchy that groups the 22 persuasion techniques into several levels of categories. This became one of the attractive shared tasks in SemEval 2024, with 153 teams registered, 48 teams submitting results, and finally, 32 system description papers submitted.

pdf bib
SemEval-2024 Task 5: Argument Reasoning in Civil Procedure
Lena Held | Ivan Habernal

This paper describes the results of SemEval-2024 Task 5: Argument Reasoning in Civil Procedure, consisting of a single task on judging and reasoning about the answers to questions in U.S. civil procedure. The dataset for this task contains question, answer and explanation pairs taken from The Glannon Guide To Civil Procedure (Glannon, 2018). The task was to classify in a binary manner if the answer is a correct choice for the question or not. Twenty participants submitted their solutions, with the best results achieving a remarkable 82.31% F1-score. We summarize and analyze the results from all participating systems and provide an overview over the systems of 14 participants.

pdf bib
SemEval-2024 Task 3: Multimodal Emotion Cause Analysis in Conversations
Fanfan Wang | Heqing Ma | Rui Xia | Jianfei Yu | Erik Cambria

The ability to understand emotions is an essential component of human-like artificial intelligence, as emotions greatly influence human cognition, decision making, and social interactions. In addition to emotion recognition in conversations, the task of identifying the potential causes behind an individual’s emotional state in conversations, is of great importance in many application scenarios. We organize SemEval-2024 Task 3, named Multimodal Emotion Cause Analysis in Conversations, which aims at extracting all pairs of emotions and their corresponding causes from conversations. Under different modality settings, it consists of two subtasks: Textual Emotion-Cause Pair Extraction in Conversations (TECPE) and Multimodal Emotion-Cause Pair Extraction in Conversations (MECPE). The shared task has attracted 143 registrations and 216 successful submissions.In this paper, we introduce the task, dataset and evaluation settings, summarize the systems of the top teams, and discuss the findings of the participants.

pdf bib
SheffieldVeraAI at SemEval-2024 Task 4: Prompting and fine-tuning a Large Vision-Language Model for Binary Classification of Persuasion Techniques in Memes
Charlie Grimshaw | Kalina Bontcheva | Xingyi Song

This paper describes our approach for SemEval-2024 Task 4: Multilingual Detection of Persuasion Techniques in Memes. Specifically, we concentrate on Subtask 2b, a binary classification challenge that entails categorizing memes as either “propagandistic” or “non-propagandistic”. To address this task, we utilized the large multimodal pretrained model, LLaVa. We explored various prompting strategies and fine-tuning methods, and observed that the model, when not fine-tuned but provided with a few-shot learning examples, achieved the best performance. Additionally, we enhanced the model’s multilingual capabilities by integrating a machine translation model. Our system secured the 2nd place in the Arabic language category.

pdf bib
SemEval-2024 Task 8: Multidomain, Multimodel and Multilingual Machine-Generated Text Detection
Yuxia Wang | Jonibek Mansurov | Petar Ivanov | Jinyan Su | Artem Shelmanov | Akim Tsvigun | Osama Mohammed Afzal | Tarek Mahmoud | Giovanni Puccetti | Thomas Arnold

We present the results and the main findings of SemEval-2024 Task 8: Multigenerator, Multidomain, and Multilingual Machine-Generated Text Detection. The task featured three subtasks. Subtask A is a binary classification task determining whether a text is written by a human or generated by a machine. This subtask has two tracks: a monolingual track focused solely on English texts and a multilingual track. Subtask B is to detect the exact source of a text, discerning whether it is written by a human or generated by a specific LLM. Subtask C aims to identify the changing point within a text, at which the authorship transitions from human to machine. The task attracted a large number of participants: subtask A monolingual (126), subtask A multilingual (59), subtask B (70), and subtask C (30). In this paper, we present the task, analyze the results, and discuss the system submissions and the methods they used. For all subtasks, the best systems used LLMs.
up

pdf (full)
bib (full) Proceedings of the 21st SIGMORPHON workshop on Computational Research in Phonetics, Phonology, and Morphology

pdf bib
Proceedings of the 21st SIGMORPHON workshop on Computational Research in Phonetics, Phonology, and Morphology
Garrett Nicolai | Eleanor Chodroff | Frederic Mailhot | Çağrı Çöltekin

pdf bib
VeLePa: a Verbal Lexicon of Pame
Borja Herce

This paper presents VeLePa, an inflected verbal lexicon of Central Pame (pbs, cent2154), an Otomanguean language from Mexico. This resource contains 12528 words in phonological form representing the complete inflectional paradigms of 216 verbs, supplemented with use frequencies. Computer-operable (CLDF) inflected lexicons of non-WEIRD underresourced languages are urgently needed to expand digital capacities in this languages (e.g. in NLP). VeLePa contributes to this, and does so with data from a language which is morphologically extraordinary, with unusually high levels of irregularity and multiple conjugations at various loci within the word: prefixes, stems, tone, and suffixes constitute different albeit interrelated subsystems of inflection.

pdf bib
J-UniMorph: Japanese Morphological Annotation through the Universal Feature Schema
Kosuke Matsuzaki | Masaya Taniguchi | Kentaro Inui | Keisuke Sakaguchi

We introduce a Japanese Morphology dataset, J-UniMorph, developed based on the UniMorph feature schema. This dataset addresses the unique and rich verb forms characteristic of the language’s agglutinative nature. J-UniMorph distinguishes itself from the existing Japanese subset of UniMorph, which is automatically extracted from Wiktionary. On average, the Wiktionary Edition features around 12 inflected forms for each word and is primarily dominated by denominal verbs (i.e., [noun] + suru (do-PRS)). Morphologically, this inflection pattern is same as the verb suru (do). In contrast, J-UniMorph explores a much broader and more frequently used range of verb forms, offering 118 inflected forms for each word on average. It includes honorifics, a range of politeness levels, and other linguistic nuances, emphasizing the distinctive characteristics of the Japanese language. This paper presents detailed statistics and characteristics of J-UniMorph, comparing it with the Wiktionary Edition. We will release J-UniMorph and its interactive visualizer publicly available, aiming to support cross-linguistic research and various applications.

pdf bib
More than Just Statistical Recurrence: Human and Machine Unsupervised Learning of Māori Word Segmentation across Morphological Processes
Ashvini Varatharaj | Simon Todd

Non-Māori-speaking New Zealanders (NMS) are able to segment Māori words in a highly similar way to fluent speakers (Panther et al., 2024). This ability is assumed to derive through the identification and extraction of statistically recurrent forms. We examine this assumption by asking how NMS segmentations compare to those produced by Morfessor, an unsupervised machine learning model that operates based on statistical recurrence, across words formed by a variety of morphological processes. Both NMS and Morfessor succeed in segmenting words formed by concatenative processes (compounding and affixation without allomorphy), but NMS also succeed for words that invoke templates (reduplication and allomorphy) and other cues to morphological structure, implying that their learning process is sensitive to more than just statistical recurrence.

pdf bib
Different Tokenization Schemes Lead to Comparable Performance in Spanish Number Agreement
Catherine Arnett | Tyler Chang | Sean Trott

The relationship between language model tokenization and performance is an open area of research. Here, we investigate how different tokenization schemes impact number agreement in Spanish plurals. We find that morphologically-aligned tokenization performs similarly to other tokenization schemes, even when induced artificially for words that would not be tokenized that way during training. We then present exploratory analyses demonstrating that language model embeddings for different plural tokenizations have similar distributions along the embedding space axis that maximally distinguishes singular and plural nouns. Our results suggest that morphologically-aligned tokenization is a viable tokenization approach, and existing models already generalize some morphological patterns to new items. However, our results indicate that morphological tokenization is not strictly required for performance.

pdf bib
Ye Olde French: Effect of Old and Middle French on SIGMORPHON-UniMorph Shared Task Data
William Kezerian | Kristine Yu

We offer one explanation for the historically low performance of French in the SIGMORPHON-UniMorph shared tasks. We conducted experiments replicating the 2023 task on French with the non-neural and neural baselines, first using the original task splits, and then using splits that excluded Old and Middle French lemmas. We applied a taxonomy on our errors using a framework based on Kyle Gorman’s “Weird Inflects but OK” 2019 annotation scheme, finding that a high portion of the French errors produced with the original splits were due to the inclusion of Old French forms, which was resolved with cleaned data.

pdf bib
The Effect of Model Capacity and Script Diversity on Subword Tokenization for Sorani Kurdish
Ali Salehi | Cassandra L. Jacobs

Tokenization and morphological segmentation continue to pose challenges for text processing and studies of human language. Here, we focus on written Soranî Kurdish, which uses a modified script based on Persian and Arabic, and its transliterations into the Kurdish Latin script. Importantly, Perso-Arabic and Latin-based writing systems demonstrate different statistical and structural properties, which may have significant effects on subword vocabulary learning. This has major consequences for frequency- or probability-based models of morphological induction. We explore the possibility that jointly training subword vocabularies using a source script along with its transliteration would improve morphological segmentation, subword tokenization, and whether gains are observed for one system over others. We find that joint training has a similar effect to increasing vocabulary size, while keeping subwords shorter in length, which produces higher-quality subwords that map onto morphemes.

pdf bib
Decomposing Fusional Morphemes with Vector Embeddings
Michael Ginn | Alexis Palmer

Distributional approaches have proven effective in modeling semantics and phonology through vector embeddings. We explore whether distributional representations can also effectively model morphological information. We train static vector embeddings over morphological sequences. Then, we explore morpheme categories for fusional morphemes, which encode multiple linguistic dimensions, and often have close relationships to other morphemes. We study whether the learned vector embeddings align with these linguistic dimensions, finding strong evidence that this is the case. Our work uses two low-resource languages, Uspanteko and Tsez, demonstrating that distributional morphological representations are effective even with limited data.

pdf bib
Acoustic barycenters as exemplar production targets
Frederic Mailhot | Cassandra L. Jacobs

We present a solution to the problem of exemplar-based language production from variable-duration tokens, leveraging algorithms from the domain of time-series clustering and classification. Our model stores and outputs tokens of phonetically rich and temporally variable representations of recorded speech. We show qualitatively and quantitatively that model outputs retain essential acoustic/phonetic characteristics despite the noise introduced by averaging, and also demonstrate the effects of similarity and indexical information as constraints on exemplar cloud selection.

pdf bib
Japanese Rule-based Grapheme-to-phoneme Conversion System and Multilingual Named Entity Dataset with International Phonetic Alphabet
Yuhi Matogawa | Yusuke Sakai | Taro Watanabe | Chihiro Taguchi

In Japanese, loanwords are primarily written in Katakana, a syllabic writing system, based on their pronunciation. However, the transliterated loanwords often exhibit spelling variations, such as the word “Hepburn” being written as “ヘボン (hebon)”, “ヘプバーン (hepubaan)”, “ヘップバーン (heppubaan)”. These orthographical variants pose a bottleneck in multilingual Named Entity Recognition (NER), because named entities (NEs) do not have one-to-one matches. In this study, we introduce a rule-based grapheme-to-phoneme (G2P) system for Japanese based on literature in linguistics and a large-scale multilingual NE dataset with annotations of the International Phonetic Alphabet (IPA), focusing on IPA to address the Katakana spelling variations in loanwords. These rules and dataset are expected to be beneficial for tasks such as NE aggregation, G2P system, construction of cross-lingual language models, and entity linking. We hope our work advances research on Japanese NER with multilingual loanwords by solving the spelling ambiguities.
up

pdf (full)
bib (full) Proceedings of the 13th Joint Conference on Lexical and Computational Semantics (*SEM 2024)

pdf bib
Proceedings of the 13th Joint Conference on Lexical and Computational Semantics (*SEM 2024)
Danushka Bollegala | Vered Shwartz

pdf bib
MASSIVE Multilingual Abstract Meaning Representation: A Dataset and Baselines for Hallucination Detection
Michael Regan | Shira Wein | George Baker | Emilio Monti

Abstract Meaning Representation (AMR) is a semantic formalism that captures the core meaning of an utterance. There has been substantial work developing AMR corpora in English and more recently across languages, though the limited size of existing datasets and the cost of collecting more annotations are prohibitive. With both engineering and scientific questions in mind, we introduce MASSIVE-AMR, a dataset with more than 84,000 text-to-graph annotations, currently the largest and most diverse of its kind: AMR graphs for 1,685 information-seeking utterances mapped to 50+ typologically diverse languages. We describe how we built our resource and its unique features before reporting on experiments using large language models for multilingual AMR and SPARQL parsing as well as applying AMRs for hallucination detection in the context of knowledge base question answering, with results shedding light on persistent issues using LLMs for structured parsing.

pdf bib
How Does Stereotype Content Differ across Data Sources?
Kathleen Fraser | Svetlana Kiritchenko | Isar Nejadgholi

For decades, psychologists have been studying stereotypes using specially-designed rating scales to capture people’s beliefs and opinions about different social groups. Now, using NLP tools on extensive collections of text, we have the opportunity to study stereotypes “in the wild” and on a large scale. However, are we truly capturing the same information? In this paper we compare measurements along six psychologically-motivated, stereotype-relevant dimensions (Sociability, Morality, Ability, Assertiveness, Beliefs, and Status) for 10 groups, defined by occupation. We compute these measurements on stereotypical English sentences written by crowd-workers, stereotypical sentences generated by ChatGPT, and more general data collected from social media, and contrast the findings with traditional, survey-based results, as well as a spontaneous word-list generation task. We find that while the correlation with the traditional scales varies across dimensions, the free-text data can be used to specify the particular traits associated with each group, and provide context for numerical survey data.

pdf bib
Polysemy through the lens of psycholinguistic variables: a dataset and an evaluation of static and contextualized language models
Andrea Bruera | Farbod Zamani | Massimo Poesio

Polysemes are words that can have different senses depending on the context of utterance: for instance, ‘newspaper’ can refer to an organization (as in ‘manage the newspaper’) or to an object (as in ‘open the newspaper’). Contrary to a large body of evidence coming from psycholinguistics, polysemy has been traditionally modelled in NLP by assuming that each sense should be given a separate representation in a lexicon (e.g. WordNet). This led to the current situation, where datasets used to evaluate the ability of computational models of semantics miss crucial details about the representation of polysemes, thus limiting the amount of evidence that can be gained from their use. In this paper we propose a framework to approach polysemy as a continuous variation in psycholinguistic properties of a word in context. This approach accommodates different sense interpretations, without postulating clear-cut jumps between senses. First we describe a publicly available English dataset that we collected, where polysemes in context (verb-noun phrases) are annotated for their concreteness and body sensory strength. Then, we evaluate static and contextualized language models in their ability to predict the ratings of each polyseme in context, as well as in their ability to capture the distinction among senses, revealing and characterizing in an interpretable way the models’ flaws.

pdf bib
Post-Hoc Answer Attribution for Grounded and Trustworthy Long Document Comprehension: Task, Insights, and Challenges
Abhilasha Sancheti | Koustava Goswami | Balaji Srinivasan

Attributing answer text to its source document for information-seeking questions is crucial for building trustworthy, reliable, and accountable systems. We formulate a new task of post-hoc answer attribution for long document comprehension (LDC). Owing to the lack of long-form abstractive and information-seeking LDC datasets, we refactor existing datasets to assess the strengths and weaknesses of existing retrieval-based and proposed answer decomposition and textual entailment-based optimal selection attribution systems for this task. We throw light on the limitations of existing datasets and the need for datasets to assess the actual performance of systems on this task.

pdf bib
A Benchmark Suite of Japanese Natural Questions
Takuya Uematsu | Hao Wang | Daisuke Kawahara | Tomohide Shibata

To develop high-performance and robust natural language processing (NLP) models, it is important to have various question answering (QA) datasets to train, evaluate, and analyze them. Although there are various QA datasets available in English, there are only a few QA datasets in other languages. We focus on Japanese, a language with only a few basic QA datasets, and aim to build a Japanese version of Natural Questions (NQ) consisting of questions that naturally arise from human information needs. We collect natural questions from query logs of a Japanese search engine and build the dataset using crowdsourcing. We construct Japanese Natural Questions (JNQ) and a Japanese version of BoolQ (JBoolQ), which is derived from NQ and consists of yes/no questions. JNQ consists of 16,871 questions, and JBoolQ consists of 6,467 questions. We also define two tasks from JNQ and one from JBoolQ and establish baselines using competitive methods drawn from related literature. We hope that these datasets will facilitate research on QA and NLP models in Japanese. We are planning to release JNQ and JBoolQ.

pdf bib
ROUGE-K: Do Your Summaries Have Keywords?
Sotaro Takeshita | Simone Ponzetto | Kai Eckert

Keywords, that is, content-relevant words in summaries play an important role in efficient information conveyance, making it critical to assess if system-generated summaries contain such informative words during evaluation. However, existing evaluation metrics for extreme summarization models do not pay explicit attention to keywords in summaries, leaving developers ignorant of their presence. To address this issue, we present a keyword-oriented evaluation metric, dubbed ROUGE-K, which provides a quantitative answer to the question of – How well do summaries include keywords? Through the lens of this keyword-aware metric, we surprisingly find that a current strong baseline model often misses essential information in their summaries. Our analysis reveals that human annotators indeed find the summaries with more keywords to be more relevant to the source documents. This is an important yet previously overlooked aspect in evaluating summarization systems. Finally, to enhance keyword inclusion, we propose four approaches for incorporating word importance into a transformer-based model and experimentally show that it enables guiding models to include more keywords while keeping the overall quality.

pdf bib
Investigating Aspect Features in Contextualized Embeddings with Semantic Scales and Distributional Similarity
Yuxi Li | Emmanuele Chersoni | Yu-Yin Hsu

Aspect, a linguistic category describing how actions and events unfold over time, is traditionally characterized by three semantic properties: stativity, durativity and telicity. In this study, we investigate whether and to what extent these properties are encoded in the verb token embeddings of the contextualized spaces of two English language models – BERT and GPT-2. First, we propose an experiment using semantic projections to examine whether the values of the vector dimensions of annotated verbs for stativity, durativity and telicity reflect human linguistic distinctions. Second, we use distributional similarity to replicate the notorious Imperfective Paradox described by Dowty (1977), and assess whether the embedding models are sensitive to capture contextual nuances of the verb telicity. Our results show that both models encode the semantic distinctions for the aspect properties of stativity and telicity in most of their layers, while durativity is the most challenging feature. As for the Imperfective Paradox, only the embedding similarities computed with the vectors from the early layers of the BERT model align with the expected pattern.

pdf bib
WikiScenes with Descriptions: Aligning Paragraphs and Sentences with Images in Wikipedia Articles
Özge Alaçam | Ronja Utescher | Hannes Grönner | Judith Sieker | Sina Zarrieß

Research in Language & Vision rarely uses naturally occurring multimodal documents as Wikipedia articles, since they feature complex image-text relations and implicit image-text alignments. In this paper, we provide one of the first datasets that provides ground-truth annotations of image-text alignments in multi-paragraph multi-image articles. The dataset can be used to study phenomena of visual language grounding in longer documents and assess retrieval capabilities of language models trained on, e.g., captioning data. Our analyses show that there are systematic linguistic differences between the image captions and descriptive sentences from the article’s text and that intra-document retrieval is a challenging task for state-of-the-art models in L&V (CLIP, VILT, MCSE).

pdf bib
Relevance, Diversity, and Exclusivity: Designing Keyword-augmentation Strategy for Zero-shot Classifiers
Taro Yano | Kunihiro Takeoka | Masafumi Oyamada

Zero-shot text classification involves categorizing text into classes without labeled data, typically using a pre-trained language model to compute the correlation between text and class names. This makes it essential for class names to contain sufficient information. Existing methods incorporate semantically similar keywords related to class names, but the properties of effective keywords remain unclear. We demonstrate that effective keywords should possess three properties: 1) keyword relevance to the task objective, 2) inter-class exclusivity, and 3) intra-class diversity. We also propose an automatic method for acquiring keywords that satisfy these properties without additional knowledge bases or data. Experiments on nine real-world datasets show our method outperforms existing approaches in fully zero-shot and generalized zero-shot settings. Ablation studies further confirm the importance of all three properties for superior performance.

pdf bib
Lexical Substitution as Causal Language Modeling
Ning Shi | Bradley Hauer | Grzegorz Kondrak

Causal language models such as the GPT series have achieved significant success across various domains. However, their application to the lexical substitution task (LST) remains largely unexplored due to inherent limitations in autoregressive decoding. Our work is motivated by our observation that existing LST approaches tend to suffer from a misalignment between the pre-training objectives of the language models that they employ, and their subsequent fine-tuning and application for substitute generation. We introduce PromptSub, the first system to use causal language modeling (CLM) for LST. Through prompt-aware fine-tuning, PromptSub not only enriches the given context with additional knowledge, but also leverages the unidirectional nature of autoregressive decoding. PromptSub consistently outperforms GeneSis, the best previously published supervised LST method. Further analysis demonstrates the potential of PromptSub to further benefit from increased model capacity, expanded data resources, and retrieval of external knowledge. By framing LST within the paradigm of CLM, our approach indicates the versatility of general CLM-based systems, such as ChatGPT, in catering to specialized tasks, including LST.

pdf bib
Paraphrase Identification via Textual Inference
Ning Shi | Bradley Hauer | Jai Riley | Grzegorz Kondrak

Paraphrase identification (PI) and natural language inference (NLI) are two important tasks in natural language processing. Despite their distinct objectives, an underlying connection exists, which has been notably under-explored in empirical investigations. We formalize the relationship between these semantic tasks and introduce a method for solving PI using an NLI system, including the adaptation of PI datasets for fine-tuning NLI models. Through extensive evaluations on six PI benchmarks, across both zero-shot and fine-tuned settings, we showcase the efficacy of NLI models for PI through our proposed reduction. Remarkably, our fine-tuning procedure enables NLI models to outperform dedicated PI models on PI datasets. In addition, our findings provide insights into the limitations of current PI benchmarks.

pdf bib
Identifying Emotional and Polar Concepts via Synset Translation
Logan Woudstra | Moyo Dawodu | Frances Igwe | Senyu Li | Ning Shi | Bradley Hauer | Grzegorz Kondrak

Emotion identification and polarity classification seek to determine the sentiment expressed by a writer. Sentiment lexicons that provide classifications at the word level fail to distinguish between different senses of polysemous words. To address this problem, we propose a translation-based method for labeling each individual lexical concept and word sense. Specifically, we translate synsets into 20 different languages and verify the sentiment of these translations in multilingual sentiment lexicons. By applying our method to all WordNet synsets, we produce SentiSynset, a synset-level sentiment resource containing 12,429 emotional synsets and 15,567 polar synsets, which is significantly larger than previous resources. Experimental evaluation shows that our method outperforms prior automated methods that classify word senses, in addition to outperforming ChatGPT. We make the resulting resource publicly available on GitHub.

pdf bib
A Closer Look at Claim Decomposition
Miriam Wanner | Seth Ebner | Zhengping Jiang | Mark Dredze | Benjamin Van Durme

As generated text becomes more commonplace, it is increasingly important to evaluate how well-supported such text is by external knowledge sources. Many approaches for evaluating textual support rely on some method for decomposing text into its individual subclaims which are scored against a trusted reference. We investigate how various methods of claim decomposition—especially LLM-based methods—affect the result of an evaluation approach such as the recently proposed FActScore, finding that it is sensitive to the decomposition method used. This sensitivity arises because such metrics attribute overall textual support to the model that generated the text even though error can also come from the metric’s decomposition step. To measure decomposition quality, we introduce an adaptation of FActScore, which we call DecompScore. We then propose an LLM-based approach to generating decompositions inspired by Bertrand Russell’s theory of logical atomism and neo-Davidsonian semantics and demonstrate its improved decomposition quality over previous methods.

pdf bib
Speedy Gonzales: A Collection of Fast Task-Specific Models for Spanish
José Cañete | Felipe Bravo-Marquez

Large language models (LLM) are now a very common and successful path to approach language and retrieval tasks. While these LLM achieve surprisingly good results it is a challenge to use them on more constrained resources. Techniques to compress these LLM into smaller and faster models have emerged for English or Multilingual settings, but it is still a challenge for other languages. In fact, Spanish is the second language with most native speakers but lacks of these kind of resources. In this work, we evaluate all the models publicly available for Spanish on a set of 6 tasks and then, by leveraging on Knowledge Distillation, we present Speedy Gonzales, a collection of inference-efficient task-specific language models based on the ALBERT architecture. All of our models (fine-tuned and distilled) are publicly available on: https://huggingface.co/dccuchile.

pdf bib
Exploring Factual Entailment with NLI: A News Media Study
Guy Mor-Lan | Effi Levi

We explore the relationship between factuality and Natural Language Inference (NLI) by introducing FactRel – a novel annotation scheme that models factual rather than textual entailment, and use it to annotate a dataset of naturally occurring sentences from news articles. Our analysis shows that 84% of factually supporting pairs and 63% of factually undermining pairs do not amount to NLI entailment or contradiction, respectively, suggesting that factual relationships are more apt for analyzing media discourse. We experiment with models for pairwise classification on the new dataset, and find that in some cases, generating synthetic data with GPT-4 on the basis of the annotated dataset can improve performance. Surprisingly, few-shot learning with GPT-4 yields strong results on par with medium LMs (DeBERTa) trained on the labelled dataset. We hypothesize that these results indicate the fundamental dependence of this task on both world knowledge and advanced reasoning abilities.

pdf bib
The Emergence of High-Level Semantics in a Signaling Game
Timothée Bernard | Timothee Mickus | Hiroya Takamura

The symbol grounding problem—how to connect a symbolic system to the outer world—is a longstanding question in AI that has recently gained prominence with the progress made in NLP in general and surrounding large language models in particular. In this article, we study the emergence of semantic categories in the communication protocol developed by neural agents involved in a well-established type of signaling game. In its basic form, the game requires one agent to retrieve an image based on a message produced by a second agent. We first show that the agents are able to, and do, learn to communicate high-level semantic concepts rather than low-level features of the images even from very indirect training signal to that end. Second, we demonstrate that the introduction of an adversarial agent in the game fosters the emergence of semantics by producing an appropriate training signal when no other method is available.

pdf bib
PDDLEGO: Iterative Planning in Textual Environments
Li Zhang | Peter Jansen | Tianyi Zhang | Peter Clark | Chris Callison-Burch | Niket Tandon

Planning in textual environments have been shown to be a long-standing challenge even for current models. A recent, promising line of work uses LLMs to generate a formal representation of the environment that can be solved by a symbolic planner. However, existing methods rely on a fully-observed environment where all entity states are initially known, so a one-off representation can be constructed, leading to a complete plan. In contrast, we tackle partially-observed environments where there is initially no sufficient information to plan for the end-goal. We propose PDDLEGO that iteratively construct a planning representation that can lead to a partial plan for a given sub-goal. By accomplishing the sub-goal, more information is acquired to augment the representation, eventually achieving the end-goal. We show that plans produced by few-shot PDDLEGO are 43% more efficient than generating plans end-to-end on the Coin Collector simulation, with strong performance (98%) on the more complex Cooking World simulation where end-to-end LLMs fail to generate coherent plans (4%).

pdf bib
VOLIMET: A Parallel Corpus of Literal and Metaphorical Verb-Object Pairs for English–German and English–French
Prisca Piccirilli | Alexander Fraser | Sabine Schulte im Walde

The interplay of cultural and linguistic elements that characterizes metaphorical language poses a substantial challenge for both human comprehension and machine processing. This challenge goes beyond monolingual settings and becomes particularly complex in translation, even more so in automatic translation. We present VOLIMET, a corpus of 2,916 parallel sentences containing gold standard alignments of metaphorical verb-object pairs and their literal paraphrases, e.g., tackle/address question, from English to German and French. On the one hand, the parallel nature of our corpus enables us to explore monolingual patterns for metaphorical vs. literal uses in English. On the other hand, we investigate different aspects of cross-lingual translations into German and French and the extent to which metaphoricity and literalness in the source language are transferred to the target languages. Monolingually, our findings reveal clear preferences in using metaphorical or literal uses of verb-object pairs. Cross-lingually, we observe a rich variability in translations as well as different behaviors for our two target languages.

pdf bib
Improving Word Sense Induction through Adversarial Forgetting of Morphosyntactic Information
Deniz Yava | Timothée Bernard | Laura Kallmeyer | Benoît Crabbé

This paper addresses the problem of word sense induction (WSI) via clustering of word embeddings. It starts from the hypothesis that contextualized word representations obtained from pre-trained language models (LMs), while being a valuable source for WSI, encode more information than what is necessary for the identification of word senses and some of this information affect the performance negatively in unsupervised settings. We investigate whether using contextualized representations that are invariant to these ‘nuisance features’ can increase WSI performance. For this purpose, we propose an adaptation of the adversarial training framework proposed by Jaiswal et al. (2020) to erase specific information from the representations of LMs, thereby creating feature-invariant representations. We experiment with erasing (i) morphological and (ii) syntactic features. The results of subsequent clustering for WSI show that these features indeed act like noise: Using feature-invariant representations, compared to using the original representations, increases clustering-based WSI performance. Furthermore, we provide an in-depth analysis of how the information about the syntactic and morphological features of words relate to and affect WSI performance.

pdf bib
What’s wrong with your model? A Quantitative Analysis of Relation Classification
Elisa Bassignana | Rob van der Goot | Barbara Plank

With the aim of improving the state-of-the-art (SOTA) on a target task, a standard strategy in Natural Language Processing (NLP) research is to design a new model, or modify the existing SOTA, and then benchmark its performance on the target task. We argue in favor of enriching this chain of actions by a preliminary error-guided analysis: First, explore weaknesses by analyzing the hard cases where the existing model fails, and then target the improvement based on those. Interpretable evaluation has received little attention for structured prediction tasks. Therefore we propose the first in-depth analysis suite for Relation Classification (RC), and show its effectiveness through a case study. We propose a set of potentially influential attributes to focus on (e.g., entity distance, sentence length). Then, we bucket our datasets based on these attributes, and weight the importance of them through correlations. This allows us to identify highly challenging scenarios for the RC model. By exploiting the findings of our analysis, with a carefully targeted adjustment to our architecture, we effectively improve the performance over the baseline by >3 Micro-F1.

pdf bib
Disambiguating Emotional Connotations of Words Using Contextualized Word Representations
Akram Sadat Hosseini | Steffen Staab

Understanding emotional nuances in written content is crucial for effective communication; however, the context-dependent nature of language poses challenges in precisely discerning emotions in text. This study contributes to the understanding of how the emotional connotations of a word are influenced by the sentence context in which it appears. Leveraging the contextual understanding embedded in contextualized word representations, we conduct an empirical investigation to (i) evaluate the varying abilities of these representations in distinguishing the diverse emotional connotations evoked by the same word across different contexts, (ii) explore potential biases in these representations toward specific emotions of a word, and (iii) assess the capability of these representations in estimating the number of emotional connotations evoked by a word in diverse contexts. Our experiments, utilizing four popular models—BERT, RoBERTa, XLNet, and GPT-2—and drawing on the GoEmotions and SemEval 2018 datasets, demonstrate that these models effectively discern emotional connotations of words. RoBERTa, in particular, shows superior performance and greater resilience against biases. Our further analysis reveals that disambiguating the emotional connotations of words significantly enhances emotion identification at the sentence level.

pdf bib
Length-Aware Multi-Kernel Transformer for Long Document Classification
Guangzeng Han | Jack Tsao | Xiaolei Huang

Lengthy documents pose a unique challenge to neural language models due to substantial memory consumption. While existing state-of-the-art (SOTA) models segment long texts into equal-length snippets (e.g., 128 tokens per snippet) or deploy sparse attention networks, these methods have new challenges of context fragmentation and generalizability due to sentence boundaries and varying text lengths. For example, our empirical analysis has shown that SOTA models consistently overfit one set of lengthy documents (e.g., 2000 tokens) while performing worse on texts with other lengths (e.g., 1000 or 4000). In this study, we propose a Length-Aware Multi-Kernel Transformer (LAMKIT) to address the new challenges for the long document classification. LAMKIT encodes lengthy documents by diverse transformer-based kernels for bridging context boundaries and vectorizes text length by the kernels to promote model robustness over varying document lengths. Experiments on five standard benchmarks from health and law domains show LAMKIT outperforms SOTA models up to an absolute 10.9% improvement. We conduct extensive ablation analyses to examine model robustness and effectiveness over varying document lengths.

pdf bib
Investigating Wit, Creativity, and Detectability of Large Language Models in Domain-Specific Writing Style Adaptation of Reddit’s Showerthoughts
Tolga Buz | Benjamin Frost | Nikola Genchev | Moritz Schneider | Lucie-Aimée Kaffee | Gerard de Melo

Recent Large Language Models (LLMs) have shown the ability to generate content that is difficult or impossible to distinguish from human writing. We investigate the ability of differently-sized LLMs to replicate human writing style in short, creative texts in the domain of Showerthoughts, thoughts that may occur during mundane activities. We compare GPT-2 and GPT-Neo fine-tuned on Reddit data as well as GPT-3.5 invoked in a zero-shot manner, against human-authored texts. We measure human preference on the texts across the specific dimensions that account for the quality of creative, witty texts. Additionally, we compare the ability of humans versus fine-tuned RoBERTa-based classifiers to detect AI-generated texts. We conclude that human evaluators rate the generated texts slightly worse on average regarding their creative quality, but they are unable to reliably distinguish between human-written and AI-generated texts. We further provide the dataset for creative, witty text generation based on Reddit Showerthoughts posts.

pdf bib
Multilingual and Code-Switched Sentence Ordering
Alexandre Salle | Shervin Malmasi

Sentence Ordering (SO) is a linguistic task which requires re-ordering of shuffled sentences into a coherent paragraph. SO has downstream applications, but also serves as a semantic probe for computational models as this capability is essential for understanding narrative structures, causal and temporal relations within texts. Despite its importance, prior research has been limited to predictable English language structures and has not thoroughly addressed the complexities of multilingual and varied narrative contexts. To fill this gap, we introduce a novel and comprehensive Multilingual Sentence Ordering task that extends SO to diverse narratives across 12 languages, including challenging code-switched texts. We have developed MultiSO, a new benchmark dataset that represents these challenges. Our findings reveal that both specialized sentence ordering models and advanced Large Language Models like GPT-4 face significant challenges with this task.

pdf bib
HANS, are you clever? Clever Hans Effect Analysis of Neural Systems
Leonardo Ranaldi | Fabio Zanzotto

Large Language Models (LLMs) have been exhibiting outstanding abilities to reason around cognitive states, intentions, and reactions of all people involved, letting humans guide and comprehend day-to-day social interactions effectively. In fact, several multiple-choice questions (MCQ) benchmarks have been proposed to construct solid assessments of the models’ abilities. However, earlier works demonstrate the presence of inherent “order bias” in LLMs, posing challenges to the appropriate evaluation. In this paper, we investigate LLMs’ resilience abilities through a series of probing tests using four MCQ benchmarks. Introducing adversarial examples, we show a significant performance gap, mainly when varying the order of the choices, which reveals a selection bias and brings into discussion reasoning abilities. Following a correlation between first positions and model choices due to positional bias, we hypothesized the presence of structural heuristics in the decision-making process of the LLMs, strengthened by including significant examples in few-shot scenarios. Finally, by using the Chain-of-Thought (CoT) technique, we elicit the model to reason and mitigate the bias by obtaining more robust models.

pdf bib
Exploring Semantics in Pretrained Language Model Attention
Frédéric Charpentier | Jairo Cugliari | Adrien Guille

Abstract Meaning Representations (AMRs) encode the semantics of sentences in the form of graphs. Vertices represent instances of concepts, and labeled edges represent semantic relations between those instances. Language models (LMs) operate by computing weights of edges of per layer complete graphs whose vertices are words in a sentence or a whole paragraph. In this work, we investigate the ability of the attention heads of two LMs, RoBERTa and GPT2, to detect the semantic relations encoded in an AMR. This is an attempt to show semantic capabilities of those models without finetuning. To do so, we apply both unsupervised and supervised learning techniques.

pdf bib
Enhancing Self-Attention via Knowledge Fusion: Deriving Sentiment Lexical Attention from Semantic-Polarity Scores
Dongjun Jang | Jinwoong Kim | Hyopil Shin

In recent years, pre-trained language models have demonstrated exceptional performance across various natural language processing (NLP) tasks. One fundamental component of these models is the self-attention mechanism, which has played a vital role in capturing meaningful relationships between tokens. However, a question still remains as to whether injecting lexical features into the self-attention mechanism can further enhance the understanding and performance of language models. This paper presents a novel approach for injecting semantic-polarity knowledge, referred to as Sentiment Lexical Attention, directly into the self-attention mechanism of Transformer-based models. The primary goal is to improve performance on sentiment classification task. Our approach involves consistently injecting Sentiment Lexical Attention derived from the lexicon corpus into the attention scores throughout the training process. We empirically evaluate our method on the NSMC sentiment classification benchmark, showcasing significant performance improvements and achieving state-of-the-art results. Furthermore, our approach demonstrates robustness and effectiveness in out-of-domain tasks, indicating its potential for broad applicability. Additionally, we analyze the impact of Sentiment Lexical Attention on the view of the CLS token’s attention distribution. Our method offers a fresh perspective on synergizing lexical features and attention scores, thereby encouraging further investigations in the realm of knowledge injection utilizing the lexical features.

pdf bib
Handling Ontology Gaps in Semantic Parsing
Andrea Bacciu | Marco Damonte | Marco Basaldella | Emilio Monti

The majority of Neural Semantic Parsing (NSP) models are developed with the assumption that there are no concepts outside the ones such models can represent with their target symbols (closed-world assumption). This assumption leads to generate hallucinated outputs rather than admitting their lack of knowledge. Hallucinations can lead to wrong or potentially offensive responses to users. Hence, a mechanism to prevent this behavior is crucial to build trusted NSP-based Question Answering agents. To that end, we propose the Hallucination Simulation Framework (HSF), a general setting for stimulating and analyzing NSP model hallucinations. The framework can be applied to any NSP task with a closed-ontology. Using the proposed framework and KQA Pro as the benchmark dataset, we assess state-of-the-art techniques for hallucination detection. We then present a novel hallucination detection strategy that exploits the computational graph of the NSP model to detect the NSP hallucinations in the presence of ontology gaps, out-of-domain utterances, and to recognize NSP errors, improving the F1-Score respectively by ~21%, ~24% and ~1%. This is the first work in closed-ontology NSP that addresses the problem of recognizing ontology gaps. We release our code and checkpoints at https://github.com/amazon-science/handling-ontology-gaps-in-semantic-parsing.

pdf bib
PipeNet: Question Answering with Semantic Pruning over Knowledge Graphs
Ying Su | Jipeng Zhang | Yangqiu Song | Tong Zhang

It is well acknowledged that incorporating explicit knowledge graphs (KGs) can benefit question answering. Existing approaches typically follow a grounding-reasoning pipeline in which entity nodes are first grounded for the query (question and candidate answers), and then a reasoning module reasons over the matched multi-hop subgraph for answer prediction. Although the pipeline largely alleviates the issue of extracting essential information from giant KGs, efficiency is still an open challenge when scaling up hops in grounding the subgraphs. In this paper, we target at finding semantically related entity nodes in the subgraph to improve the efficiency of graph reasoning with KG. We propose a grounding-pruning-reasoning pipeline to prune noisy nodes, remarkably reducing the computation cost and memory usage while also obtaining decent subgraph representation. In detail, the pruning module first scores concept nodes based on the dependency distance between matched spans and then prunes the nodes according to score ranks. To facilitate the evaluation of pruned subgraphs, we also propose a graph attention network (GAT) based module to reason with the subgraph data. Experimental results on CommonsenseQA and OpenBookQA demonstrate the effectiveness of our method.

pdf bib
A Trip Towards Fairness: Bias and De-Biasing in Large Language Models
Leonardo Ranaldi | Elena Ruzzetti | Davide Venditti | Dario Onorati | Fabio Zanzotto

Cheap-to-Build Very Large-Language Models (CtB-LLMs) with affordable training are emerging as the next big revolution in natural language processing and understanding. These CtB-LLMs are democratizing access to trainable Very Large-Language Models (VLLMs) and, thus, may represent the building blocks of many NLP systems solving downstream tasks. Hence, a little or a large bias in CtB-LLMs may cause huge harm. In this paper, we performed a large investigation of the bias of three families of CtB-LLMs, and we showed that debiasing techniques are effective and usable. Indeed, according to current tests, the LLaMA and the OPT families have an important bias in gender, race, religion, and profession. In contrast to the analysis for other LMMs, we discovered that bias depends not on the number of parameters but on the perplexity. Finally, the debiasing of OPT using LORA reduces bias up to 4.12 points in the normalized stereotype score.

pdf bib
Compositional Structured Explanation Generation with Dynamic Modularized Reasoning
Xiyan Fu | Anette Frank

In this work, we propose a new task, compositional structured explanation generation (CSEG), to facilitate research on compositional generalization in reasoning. Despite the success of language models in solving reasoning tasks, their compositional generalization capabilities are under-researched. Our new CSEG task tests a model’s ability to generalize from generating entailment trees with a limited number of inference steps – to more steps, focusing on the length and shapes of entailment trees. CSEG is challenging in requiring both reasoning and compositional generalization abilities, and by being framed as a generation task. Besides the CSEG task, we propose a new dynamic modularized reasoning model, MORSE, that factorizes the inference process into modules, where each module represents a functional unit. We adopt modularized self-attention to dynamically select and route inputs to dedicated heads, which specializes them to specific functions. Using CSEG, we compare MORSE to models from prior work. Our analyses show that the task is challenging, but that the dynamic reasoning modules of MORSE are effective, showing competitive compositional generalization abilities in a generation setting.

pdf bib
Inspecting Soundness of AMR Similarity Metrics in terms of Equivalence and Inequivalence
Kyung Seo Ki | Bugeun Kim | Gahgene Gweon

In this study, we investigate soundness of current Abstract Meaning Representation (AMR) similarity metrics in terms of equivalence and inequivalence. Specifically, AMR guidelines provide several equivalence and inequivalence conditions to reflect the meaning aspect of the semantics. Thus, it is important to examine an AMR metric’s soundness, i.e., whether the metric correctly reflects the guidelines. However, the existing metrics have less investigated their soundness. In this work, we propose a new experimental method using simulated data and a series of statistical tests to verify the metric’s soundness. Our experimental result revealed that all existing metrics such as Smatch, SemBLEU, S2match, Smatch++, WWLK-theta, WWLK-k3e2n, and SEMA did not fully meet the AMR guidelines in terms of equivalence and inequivalence aspects. Also, to alleviate this soundness problem, we suggest a revised metric called Smatch#, which adopts simple graph standardization technique that can improve the soundness of an existing metric.

pdf bib
Sõnajaht: Definition Embeddings and Semantic Search for Reverse Dictionary Creation
Aleksei Dorkin | Kairit Sirts

We present an information retrieval based reverse dictionary system using modern pre-trained language models and approximate nearest neighbors search algorithms. The proposed approach is applied to an existing Estonian language lexicon resource, Sõnaveeb (word web), with the purpose of enhancing and enriching it by introducing cross-lingual reverse dictionary functionality powered by semantic search. The performance of the system is evaluated using both an existing labeled English dataset of words and definitions that is extended to contain also Estonian and Russian translations, and a novel unlabeled evaluation approach that extracts the evaluation data from the lexicon resource itself using synonymy relations. Evaluation results indicate that the information retrieval based semantic search approach without any model training is feasible, producing median rank of 1 in the monolingual setting and median rank of 2 in the cross-lingual setting using the unlabeled evaluation approach, with models trained for cross-lingual retrieval and including Estonian in their training data showing superior performance in our particular task.

pdf bib
Do large language models and humans have similar behaviours in causal inference with script knowledge?
Xudong Hong | Margarita Ryzhova | Daniel Biondi | Vera Demberg

Recently, large pre-trained language models (LLMs) have demonstrated superior language understanding abilities, including zero-shot causal reasoning. However, it is unclear to what extent their capabilities are similar to human ones. We here study the processing of an event B in a script-based story, which causally depends on a previous event A. In our manipulation, event A is stated, negated, or omitted in an earlier section of the text. We first conducted a self-paced reading experiment, which showed that humans exhibit significantly longer reading times when causal conflicts exist (¬ A → B) than under logical conditions (A → B). However, reading times remain similar when cause A is not explicitly mentioned, indicating that humans can easily infer event B from their script knowledge. We then tested a variety of LLMs on the same data to check to what extent the models replicate human behavior. Our experiments show that 1) only recent LLMs, like GPT-3 or Vicuna, correlate with human behavior in the ¬ A → B condition. 2) Despite this correlation, all models still fail to predict that nil → B is less surprising than ¬ A → B, indicating that LLMs still have difficulties integrating script knowledge.

pdf bib
EDM3: Event Detection as Multi-task Text Generation
Ujjwala Anantheswaran | Himanshu Gupta | Mihir Parmar | Kuntal Pal | Chitta Baral

We present EDM3, a novel approach for Event Detection (ED) based on decomposing and reformulating ED, and fine-tuning over its atomic subtasks. EDM3 enhances knowledge transfer while mitigating prediction error propagation inherent in pipelined approaches. EDM3 infers dataset-specific knowledge required for the complex primary task from its atomic tasks, making it adaptable to any set of event types. We evaluate EDM3 on multiple ED datasets, achieving state-of-the-art results on RAMS (71.3% vs 65.1% F1), and competitive performance on WikiEvents, MAVEN (∆ = 0.2%), and MLEE (∆ = 1.8%). We present an ablation study over rare event types (<15 instances in training data) in MAVEN, where EDM3 achieves ~90% F1. To the best of the authors’ knowledge, we are the first to analyze ED performance over non-standard event configurations (i.e., multi-word and multi-class triggers). Experimental results show that EDM3 achieves ~90% exact match accuracy on multi-word triggers and ~61% prediction accuracy on multi-class triggers. This work establishes the effectiveness of EDM3 in enhancing performance on a complex information extraction task.
up

pdf (full)
bib (full) Proceedings of the 4th Workshop on Trustworthy Natural Language Processing (TrustNLP 2024)

pdf bib
Proceedings of the 4th Workshop on Trustworthy Natural Language Processing (TrustNLP 2024)
Kai-Wei Chang | Anaelia Ovalle | Jieyu Zhao | Yang Trista Cao | Ninareh Mehrabi | Aram Galstyan | Jwala Dhamala | Anoop Kumar | Rahul Gupta

pdf bib
Beyond Turing: A Comparative Analysis of Approaches for Detecting Machine-Generated Text
Muhammad Adilazuarda

Significant progress has been made on text generation by pre-trained language models (PLMs), yet distinguishing between human and machine-generated text poses an escalating challenge. This paper offers an in-depth evaluation of three distinct methods used to address this task: traditional shallow learning, Language Model (LM) fine-tuning, and Multilingual Model fine-tuning. These approaches are rigorously tested on a wide range of machine-generated texts, providing a benchmark of their competence in distinguishing between human-authored and machine-authored linguistic constructs. The results reveal considerable differences in performance across methods, thus emphasizing the continued need for advancement in this crucial area of NLP. This study offers valuable insights and paves the way for future research aimed at creating robust and highly discriminative models.

pdf bib
Automated Adversarial Discovery for Safety Classifiers
Yash Kumar Lal | Preethi Lahoti | Aradhana Sinha | Yao Qin | Ananth Balashankar

Safety classifiers are critical in mitigating toxicity on online forums such as social media and in chatbots. Still, they continue to be vulnerable to emergent, and often innumerable, adversarial attacks.Traditional automated adversarial data generation methods, however, tend to produce attacks that are not diverse, but variations of previously observed harm types.We formalize the task of automated adversarial discovery for safety classifiers - to find new attacks along previously unseen harm dimensions that expose new weaknesses in the classifier.We measure progress on this task along two key axes (1) adversarial success: does the attack fool the classifier? and (2) dimensional diversity: does the attack represent a previously unseen harm type?Our evaluation of existing attack generation methods on the CivilComments toxicity task reveals their limitations: Word perturbation attacks fail to fool classifiers, while prompt-based LLM attacks have more adversarial success, but lack dimensional diversity.Even our best-performing prompt-based method finds new successful attacks on unseen harm dimensions of attacks only 5% of the time.Automatically finding new harmful dimensions of attack is crucial and there is substantial headroom for future research on our new task.

pdf bib
FairBelief - Assessing Harmful Beliefs in Language Models
Mattia Setzu | Marta Marchiori Manerba | Pasquale Minervini | Debora Nozza

Language Models (LMs) have been shown to inherit undesired biases that might hurt minorities and underrepresented groups if such systems were integrated into real-world applications without careful fairness auditing.This paper proposes FairBelief, an analytical approach to capture and assess beliefs, i.e., propositions that an LM may embed with different degrees of confidence and that covertly influence its predictions. With FairBelief, we leverage prompting to study the behavior of several state-of-the-art LMs across different previously neglected axes, such as model scale and likelihood, assessing predictions on a fairness dataset specifically designed to quantify LMs’ outputs’ hurtfulness.Finally, we conclude with an in-depth qualitative assessment of the beliefs emitted by the models.We apply FairBelief to English LMs, revealing that, although these architectures enable high performances on diverse natural language processing tasks, they show hurtful beliefs about specific genders. Interestingly, training procedure and dataset, model scale, and architecture induce beliefs of different degrees of hurtfulness.

pdf bib
The Trade-off between Performance, Efficiency, and Fairness in Adapter Modules for Text Classification
Minh Duc Bui | Katharina Von Der Wense

Current natural language processing (NLP) research tends to focus on only one or, less frequently, two dimensions – e.g., performance, interpretability, or efficiency – at a time, which may lead to suboptimal conclusions. Work on adapter modulesfocuses on improving performance and efficiency, with no investigation of unintended consequences on other aspects such as fairness. To address this gap, we conduct experiments on three text classification datasets by either (1) finetuning all parameters or (2) using adapter modules. Regarding performance and efficiency, we confirm prior findings that the accuracy of adapter-enhanced models is roughly on par with that of fully finetuned models, while training time is substantially reduced. Regarding fairness, we show that adapter modules result in mixed fairness across sensitive groups. Further investigation reveals that, when the standard finetuned model exhibits limited biases, adapter modules typically do not introduce extra bias. On the other hand, when the finetuned model exhibits increased bias, the use of adapter modules poses the potential danger of amplifying these biases to a significant extent. Our findings highlight the need for a case-by-case evaluation rather than a one-size-fits-all judgment.

pdf bib
When XGBoost Outperforms GPT-4 on Text Classification: A Case Study
Matyas Bohacek | Michal Bravansky

Large language models (LLMs) are increasingly used for applications beyond text generation, ranging from text summarization to instruction following. One popular example of exploiting LLMs’ zero- and few-shot capabilities is the task of text classification. This short paper compares two popular LLM-based classification pipelines (GPT-4 and LLAMA 2) to a popular pre-LLM-era classification pipeline on the task of news trustworthiness classification, focusing on performance, training, and deployment requirements. We find that, in this case, the pre-LLM-era ensemble pipeline outperforms the two popular LLM pipelines while being orders of magnitude smaller in parameter size.

pdf bib
Towards Healthy AI: Large Language Models Need Therapists Too
Baihan Lin | Djallel Bouneffouf | Guillermo Cecchi | Kush Varshney

Recent advances in large language models (LLMs) have led to the development of powerful chatbots capable of engaging in fluent human-like conversations. However, these chatbots may be harmful, exhibiting manipulation, gaslighting, narcissism, and other toxicity. To work toward safer and more well-adjusted models, we propose a framework that uses psychotherapy to identify and mitigate harmful chatbot behaviors. The framework involves four different artificial intelligence (AI) agents: the Chatbot whose behavior is to be adjusted, a User, a Therapist, and a Critic that can be paired with reinforcement learning-based LLM tuning. We illustrate the framework with a working example of a social conversation involving four instances of ChatGPT, showing that the framework may mitigate the toxicity in conversations between LLM-driven chatbots and people. Although there are still several challenges and directions to be addressed in the future, the proposed framework is a promising approach to improving the alignment between LLMs and human values.

pdf bib
Exploring Causal Mechanisms for Machine Text Detection Methods
Kiyoon Yoo | Wonhyuk Ahn | Yeji Song | Nojun Kwak

The immense attraction towards text generation garnered by ChatGPT has spurred the need for discriminating machine-text from human text. In this work, we provide preliminary evidence that the scores computed by existing zero-shot and supervised machine-generated text detection methods are not solely determined by the generated texts, but are affected by prompts and real texts as well. Using techniques from causal inference, we show the existence of backdoor paths that confounds the relationships between text and its detection score and how the confounding bias can be partially mitigated. We open up new research directions in identifying other factors that may be interwoven in the detection of machine text. Our study calls for a deeper investigation into which kinds of prompts make the detection of machine text more difficult or easier

pdf bib
FactAlign: Fact-Level Hallucination Detection and Classification Through Knowledge Graph Alignment
Mohamed Rashad | Ahmed Zahran | Abanoub Amin | Amr Abdelaal | Mohamed Altantawy

This paper proposes a novel black-box approach for fact-level hallucination detection and classification by transforming the problem into a knowledge graph alignment task. This approach allows us to classify detected hallucinations as either intrinsic or extrinsic. The paper starts by discussing the field of hallucination detection and introducing several approaches to related work. Then, we introduce the proposed FactAlign approach for hallucination detection and discuss how we can use it to classify hallucinations as either intrinsic or extrinsic. Experiments are carried out to evaluate the proposed method against state-of-the-art methods on the hallucination detection task using the WikiBio GPT-3 hallucination dataset, and on the hallucination type classification task using the XSum hallucination annotations dataset. The experimental results show that our method achieves a 0.889 F1 score for the hallucination detection and 0.825 F1 for the hallucination type classification, without any further training, fine-tuning, or producing multiple samples of the LLM response.

pdf bib
Cross-Task Defense: Instruction-Tuning LLMs for Content Safety
Yu Fu | Wen Xiao | Jia Chen | Jiachen Li | Evangelos Papalexakis | Aichi Chien | Yue Dong

Recent studies reveal that Large Language Models (LLMs) face challenges in balancing safety with utility, particularly when processing long texts for NLP tasks like summarization and translation. Despite defenses against malicious short questions, the ability of LLMs to safely handle dangerous long content, such as manuals teaching illicit activities, remains unclear. Our work aims to develop robust defenses for LLMs in processing malicious documents alongside benign NLP task queries. We introduce a defense dataset comprised of safety-related examples and propose single-task and mixed-task losses for instruction tuning. Our empirical results demonstrate that LLMs can significantly enhance their capacity to safely manage dangerous content with appropriate instruction tuning. Additionally, strengthening the defenses of tasks most susceptible to misuse is effective in protecting LLMs against processing harmful information. We also observe that trade-offs between utility and safety exist in defense strategies, where Llama2, utilizing our proposed approach, displays a significantly better balance compared to Llama1.

pdf bib
On the Interplay between Fairness and Explainability
Stephanie Brandl | Emanuele Bugliarello | Ilias Chalkidis

In order to build reliable and trustworthy NLP applications, models need to be both fair across different demographics and explainable. Usually these two objectives, fairness and explainability, are optimized and/or examined independently of each other. Instead, we argue that forthcoming, trustworthy NLP systems should consider both.In this work, we perform a first study to understand how they influence each other: do fair(er) models rely on more plausible explanations? and vice versa. To this end, we conduct experiments on two English multi-class text classification datasets, BIOS and ECtHR, that provide information on gender and nationality, respectively, as well as human-annotated rationales. We fine-tune pre-trained language models with several methods for (i) bias mitigation, which aims to improve fairness; (ii) rationale extraction, which aims to produce plausible explanations.We find that bias mitigation algorithms do not always lead to fairer models. Moreover, in our analysis, we see that empirical fairness and explainability are orthogonal.

pdf bib
Holistic Evaluation of Large Language Models: Assessing Robustness, Accuracy, and Toxicity for Real-World Applications
David Cecchini | Arshaan Nazir | Kalyan Chakravarthy | Veysel Kocaman

Large Language Models (LLMs) have been widely used in real-world applications. However, as LLMs evolve and new datasets are released, it becomes crucial to build processes to evaluate and control the models’ performance. In this paper, we describe how to add Robustness, Accuracy, and Toxicity scores to model comparison tables, or leaderboards. We discuss the evaluation metrics, the approaches considered, and present the results of the first evaluation round for model Robustness, Accuracy, and Toxicity scores. Our results show that GPT 4 achieves top performance on robustness and accuracy test, while Llama 2 achieves top performance on the toxicity test. We note that newer open-source models such as open chat 3.5 and neural chat 7B can perform well on these three test categories. Finally, domain-specific tests and models are also planned to be added to the leaderboard to allow for a more detailed evaluation of models in specific areas such as healthcare, legal, and finance.

pdf bib
HGOT: Hierarchical Graph of Thoughts for Retrieval-Augmented In-Context Learning in Factuality Evaluation
Yihao Fang | Stephen Thomas | Xiaodan Zhu

With the widespread adoption of large language models (LLMs) in numerous applications, the challenge of factuality and the propensity for hallucinations has emerged as a significant concern. To address this issue, particularly in retrieval-augmented in-context learning, we introduce the hierarchical graph of thoughts (HGOT), a structured, multi-layered graph approach designed to enhance the retrieval of pertinent passages during in-context learning. The framework utilizes the emergent planning capabilities of LLMs, employing the divide-and-conquer strategy to break down complex queries into manageable sub-queries. It refines self-consistency majority voting for answer selection, which incorporates the recently proposed citation recall and precision metrics to assess the quality of thoughts, linking an answer’s credibility intrinsically to the thought’s quality. This methodology introduces a weighted system in majority voting, prioritizing answers based on the citation quality of their thoughts. Additionally, we propose a scoring mechanism for evaluating retrieved passages, considering factors such as citation frequency and quality, self-consistency confidence, and the retrieval module’s ranking. Experiments indicate that HGOT excels as a versatile approach, outperforming competing models in FEVER by up to 7% and matching leading models such as Retrieve-then-Read in Open-SQuAD, and DSP in HotPotQA, demonstrating its efficacy in enhancing LLMs’ factuality.

pdf bib
Overconfidence is Key: Verbalized Uncertainty Evaluation in Large Language and Vision-Language Models
Tobias Groot | Matias Valdenegro - Toro

Language and Vision-Language Models (LLMs/VLMs) have revolutionized the field of AI by their ability to generate human-like text and understand images, but ensuring their reliability is crucial. This paper aims to evaluate the ability of LLMs (GPT4, GPT-3.5, LLaMA2, and PaLM 2) and VLMs (GPT4V and Gemini Pro Vision) to estimate their verbalized uncertainty via prompting. We propose the new Japanese Uncertain Scenes (JUS) dataset, aimed at testing VLM capabilities via difficult queries and object counting, and the Net Calibration Error (NCE) to measure direction of miscalibration.Results show that both LLMs and VLMs have a high calibration error and are overconfident most of the time, indicating a poor capability for uncertainty estimation. Additionally we develop prompts for regression tasks, and we show that VLMs have poor calibration when producing mean/standard deviation and 95% confidence intervals.

pdf bib
Tweak to Trust: Assessing the Reliability of Summarization Metrics in Contact Centers via Perturbed Summaries
Kevin Patel | Suraj Agrawal | Ayush Kumar

In the dynamic realm of call center communications, the potential of abstractive summarization to transform information condensation is evident. However, evaluating the performance of abstractive summarization systems within contact center domain poses a significant challenge. Traditional evaluation metrics prove inadequate in capturing the multifaceted nature of call center conversations, characterized by diverse topics, emotional nuances, and dynamic contexts. This paper uses domain-specific perturbed summaries to scrutinize the robustness of summarization metrics in the call center domain. Through extensive experiments on call center data, we illustrate how perturbed summaries uncover limitations in existing metrics. We additionally utilize perturbation as data augmentation strategy to train domain-specific metrics. Our findings underscore the potential of perturbed summaries to complement current evaluation techniques, advancing reliable and adaptable summarization solutions in the call center domain.

pdf bib
Flatness-Aware Gradient Descent for Safe Conversational AI
Leila Khalatbari | Saeid Hosseini | Hossein Sameti | Pascale Fung

As generative dialog models become ubiquitous in real-world applications, it is paramount to ensure a harmless generation. There are two major challenges when enforcing safety to open-domain chatbots. Firstly, it is impractical to provide training data reflecting the desired response to all emerging forms of toxicity (generalisation challenge). Secondly, implementing safety features may compromise the quality of the conversation (trade-off challenge). To tackle the challenges, this paper introduces a regularized fine-tuning approach called FlatGD. By employing a safety-tailored loss, we translate better optimization to more safety. To ensure better optimization, FlatGD penalizes sharp trajectories of loss curve, encouraging flatness of the converged local minima. Experimental results on datasets of “BAD” and “prosocial dialog” demonstrate that our model outperforms the current baselines in reducing toxicity while preserving the conversation quality. Moreover, compared to other baselines, FlatGD can better generalize to unseen toxic data.

pdf bib
Introducing GenCeption for Multimodal LLM Benchmarking: You May Bypass Annotations
Lele Cao | Valentin Buchner | Zineb Senane | Fangkai Yang

Multimodal Large Language Models (MLLMs) are commonly evaluated using costly annotated multimodal benchmarks. However, these benchmarks often struggle to keep pace with the rapidly advancing requirements of MLLM evaluation. We propose GenCeption, a novel and annotation-free MLLM evaluation framework that merely requires unimodal data to assess inter-modality semantic coherence and inversely reflects the models’ inclination to hallucinate. Analogous to the popular DrawCeption game, GenCeption initiates with a non-textual sample and undergoes a series of iterative description and generation steps. Semantic drift across iterations is quantified using the GC@T metric. Our empirical findings validate GenCeption’s efficacy, showing strong correlations with popular MLLM benchmarking results. GenCeption may be extended to mitigate training data contamination by utilizing ubiquitous, previously unseen unimodal data.

pdf bib
Semantic-Preserving Adversarial Example Attack against BERT
Chongyang Gao | Kang Gu | Soroush Vosoughi | Shagufta Mehnaz

Adversarial example attacks against textual data have been drawing increasing attention in both the natural language processing (NLP) and security domains. However, most of the existing attacks overlook the importance of semantic similarity and yield easily recognizable adversarial samples. As a result, the defense methods developed in response to these attacks remain vulnerable and could be evaded by advanced adversarial examples that maintain high semantic similarity with the original, non-adversarial text. Hence, this paper aims to investigate the extent of textual adversarial examples in maintaining such high semantic similarity. We propose Reinforce attack, a reinforcement learning-based framework to generate adversarial text that preserves high semantic similarity with the original text. In particular, the attack process is controlled by a reward function rather than heuristics, as in previous methods, to encourage higher semantic similarity and lower query costs. Through automatic and human evaluations, we show that our generated adversarial texts preserve significantly higher semantic similarity than state-of-the-art attacks while achieving similar attack success rates (outperforming at times), thus uncovering novel challenges for effective defenses.

pdf bib
Sandwich attack: Multi-language Mixture Adaptive Attack on LLMs
Bibek Upadhayay | Vahid Behzadan

A significant challenge in reliable deployment of Large Language Models (LLMs) is malicious manipulation via adversarial prompting techniques such as jailbreaks. Employing mechanisms such as safety training have proven useful in addressing this challenge. However, in multilingual LLMs, adversaries can exploit the imbalanced representation of low-resource languages in datasets used for pretraining and safety training. In this paper, we introduce a new black-box attack vector called the Sandwich Attack: a multi-language mixture attack, which manipulates state-of-the-art LLMs into generating harmful and misaligned responses. Our experiments with five different models, namely Bard, Gemini Pro, LLaMA-2-70-B-Chat, GPT-3.5-Turbo, GPT-4, and Claude-3-OPUS, show that this attack vector can be used by adversaries to elicit harmful responses from these models. By detailing both the mechanism and impact of the Sandwich attack, this paper aims to guide future research and development towards more secure and resilient LLMs, ensuring they serve the public good while minimizing potential for misuse. Content Warning: This paper contains examples of harmful language.

pdf bib
Masking Latent Gender Knowledge for Debiasing Image Captioning
Fan Yang | Shalini Ghosh | Emre Barut | Kechen Qin | Prashan Wanigasekara | Chengwei Su | Weitong Ruan | Rahul Gupta

Large language models incorporate world knowledge and present breakthrough performances on zero-shot learning. However, these models capture societal bias (e.g., gender or racial bias) due to bias during the training process which raises ethical concerns or can even be potentially harmful. The issue is more pronounced in multi-modal settings, such as image captioning, as images can also add onto biases (e.g., due to historical non-equal representation of genders in different occupations). In this study, we investigate the removal of potentially problematic knowledge from multi-modal models used for image captioning. We relax the gender bias issue in captioning models by degenderizing generated captions through the use of a simple linear mask, trained via adversarial training. Our proposal makes no assumption on the architecture of the model and freezes the model weights during the procedure, which also enables the mask to be turned off. We conduct experiments on COCO caption datasets using our masking solution. The results suggest that the proposed mechanism can effectively mask the targeted biased knowledge, by replacing more than 99% gender words with neutral ones, and maintain a comparable captioning quality performance with minimal (e.g., -1.4 on BLEU4 and ROUGE) impact to accuracy metrics.

pdf bib
BELIEVE: Belief-Enhanced Instruction Generation and Augmentation for Zero-Shot Bias Mitigation
Lisa Bauer | Ninareh Mehrabi | Palash Goyal | Kai-Wei Chang | Aram Galstyan | Rahul Gupta

Language models, pre-trained on large amounts of unmoderated content, have been shown to contain societal biases. Mitigating such biases typically requires access to model parameters and training schemas. In this work, we address bias mitigation at inference time, such that it can be applied to any black-box model. To this end, we propose a belief generation and augmentation framework, BELIEVE, that demonstrates effective bias mitigation for natural language generation by augmenting input prompts with automatically generated instruction-based beliefs. Our framework eases the bottleneck required for manually crafting these instruction-based beliefs, by extending a recently proposed iterative in-context learning framework to automatically generate beliefs via a language model. We assess the impact of this system on fairness, and demonstrate effective bias mitigation on pretrained and instruction-tuned models for both sentiment and regard with respect to multiple protected classes including race, gender, and political ideology.

pdf bib
Tell Me Why: Explainable Public Health Fact-Checking with Large Language Models
Majid Zarharan | Pascal Wullschleger | Babak Behkam Kia | Mohammad Taher Pilehvar | Jennifer Foster

This paper presents a comprehensive analysis of explainable fact-checking through a series of experiments, focusing on the ability of large language models to verify public health claims and provide explanations or justifications for their veracity assessments. We examine the effectiveness of zero/few-shot prompting and parameter-efficient fine-tuning across various open and closed-source models, examining their performance in both isolated and joint tasks of veracity prediction and explanation generation. Importantly, we employ a dual evaluation approach comprising previously established automatic metrics and a novel set of criteria through human evaluation. Our automatic evaluation indicates that, within the zero-shot scenario, GPT-4 emerges as the standout performer, but in few-shot and parameter-efficient fine-tuning contexts, open-source models demonstrate their capacity to not only bridge the performance gap but, in some instances, surpass GPT-4. Human evaluation reveals yet more nuance as well as indicating potential problems with the gold explanations.
up

pdf (full)
bib (full) Proceedings of the Eleventh Workshop on NLP for Similar Languages, Varieties, and Dialects (VarDial 2024)

pdf bib
Proceedings of the Eleventh Workshop on NLP for Similar Languages, Varieties, and Dialects (VarDial 2024)
Yves Scherrer | Tommi Jauhiainen | Nikola Ljubešić | Marcos Zampieri | Preslav Nakov | Jörg Tiedemann

pdf bib
VarDial Evaluation Campaign 2024: Commonsense Reasoning in Dialects and Multi-Label Similar Language Identification
Adrian-Gabriel Chifu | Goran Glavaš | Radu Tudor Ionescu | Nikola Ljubešić | Aleksandra Miletić | Filip Miletić | Yves Scherrer | Ivan Vulić

This report presents the results of the shared tasks organized as part of the VarDial Evaluation Campaign 2024. The campaign is part of the eleventh workshop on Natural Language Processing (NLP) for Similar Languages, Varieties and Dialects (VarDial), co-located with NAACL 2024. Two shared tasks were included this year: dialectal causal commonsense reasoning (DIALECT-COPA), and Multi-label classification of similar languages (DSL-ML). Both tasks were organized for the first time this year, but DSL-ML partially overlaps with the DSL-TL task organized in 2023.

pdf bib
What Drives Performance in Multilingual Language Models?
Sina Bagheri Nezhad | Ameeta Agrawal

This study investigates the factors influencing the performance of multilingual large language models (MLLMs) across diverse languages. We study 6 MLLMs, including masked language models, autoregressive models, and instruction-tuned LLMs, on the SIB-200 dataset, a topic classification dataset encompassing 204 languages. Our analysis considers three scenarios: ALL languages, SEEN languages (present in the model’s pretraining data), and UNSEEN languages (not present or documented in the model’s pretraining data in any meaningful way). We examine the impact of factors such as pretraining data size, general resource availability, language family, and script type on model performance. Decision tree analysis reveals that pretraining data size is the most influential factor for SEEN languages. However, interestingly, script type and language family become more crucial for UNSEEN languages, highlighting the importance of cross-lingual transfer learning. Notably, model size and architecture do not significantly alter the most important features identified. Our findings provide valuable insights into the strengths and limitations of current MLLMs and hope to guide the development of more effective and equitable multilingual NLP systems.

pdf bib
Does Whisper Understand Swiss German? An Automatic, Qualitative, and Human Evaluation
Eyal Dolev | Clemens Lutz | Noëmi Aepli

Whisper is a state-of-the-art automatic speech recognition (ASR) model (Radford et al., 2022). Although Swiss German dialects are allegedly not part of Whisper’s training data, preliminary experiments showed Whisper can transcribe Swiss German quite well, with the output being a speech translation into Standard German. To gain a better understanding of Whisper’s performance on Swiss German, we systematically evaluate it using automatic, qualitative, and human evaluation. We test its performance on three existing test sets: SwissDial (Dogan-Schönberger et al., 2021), STT4SG-350 (Plüss et al., 2023), and Swiss Parliaments Corpus (Plüss et al., 2021). In addition, we create a new test set for this study based on short mock clinical interviews. To automatically evaluate performance, we used word error rate (WER) and BLEU. We also conducted a qualitative analysis of Whisper’s performance, discussing its strengths and weaknesses. Finally, 28 people participated in a survey evaluating Whisper’s performance. All of our evaluations showed that Whisper is a viable ASR system for Swiss German, so long as the Standard German output is desired.

pdf bib
How Well Do Tweets Represent Sub-Dialects of Egyptian Arabic?
Mai Mohamed Eida | Mayar Nassar | Jonathan Dunn

How well does naturally-occurring digital text, such as Tweets, represent sub-dialects of Egyptian Arabic (EA)? This paper focuses on two EA sub-dialects: Cairene Egyptian Arabic (CEA) and Sa’idi Egyptian Arabic (SEA). We use morphological markers from ground-truth dialect surveys as a distance measure across four geo-referenced datasets. Results show that CEA markers are prevalent as expected in CEA geo-referenced tweets, while SEA markers are limited across SEA geo-referenced tweets. SEA tweets instead show a prevalence of CEA markers and higher usage of Modern Standard Arabic. We conclude that corpora intended to represent sub-dialects of EA do not accurately represent sub-dialects outside of the Cairene variety. This finding calls into question the validity of relying on tweets alone to represent dialectal differences.

pdf bib
When Elote, Choclo and Mazorca are not the Same. Isomorphism-Based Perspective to the Spanish Varieties Divergences
Cristina España-Bonet | Ankur Bhatt | Koel Dutta Chowdhury | Alberto Barrón-Cedeño

Spanish is an official language in 20 countries; in 19 of them, it arrived by means of overseas colonisation. Its close contact with several coexistent languages and the rich regional and cultural diversity has produced varieties which divert from each other. We study these divergences in a data-based approach and according to their qualitative and quantitative effects in word embeddings. We generate embeddings for Spanish in 24 countries and examine the topology of the spaces. Due to the similarities between varieties —in contrast to what happens to different languages in bilingual topological studies— we first scrutinise the behaviour of three isomorphism measures in (quasi-)isomorphic settings: relational similarity, Eigenvalue similarity and Gromov-Hausdorff distance. We then use the most trustworthy measure to quantify the divergences among varieties. Finally, we use the departures from isomorphism to build relational trees for the Spanish varieties by hierarchical clustering.

pdf bib
Modeling Orthographic Variation in Occitan’s Dialects
Zachary Hopton | Noëmi Aepli

Effectively normalizing spellings in textual data poses a considerable challenge, especially for low-resource languages lacking standardized writing systems. In this study, we fine-tuned a multilingual model with data from several Occitan dialects and conducted a series of experiments to assess the model’s representations of these dialects. For evaluation purposes, we compiled a parallel lexicon encompassing four Occitan dialects.Intrinsic evaluations of the model’s embeddings revealed that surface similarity between the dialects strengthened representations. When the model was further fine-tuned for part-of-speech tagging, its performance was robust to dialectical variation, even when trained solely on part-of-speech data from a single dialect. Our findings suggest that large multilingual models minimize the need for spelling normalization during pre-processing.

pdf bib
DIALECT-COPA: Extending the Standard Translations of the COPA Causal Commonsense Reasoning Dataset to South Slavic Dialects
Nikola Ljubešić | Nada Galant | Sonja Benčina | Jaka Čibej | Stefan Milosavljević | Peter Rupnik | Taja Kuzman

The paper presents new causal commonsense reasoning datasets for South Slavic dialects, based on the Choice of Plausible Alternatives (COPA) dataset. The dialectal datasets are built by translating by native dialect speakers from the English original and the corresponding standard translation. Three dialects are covered – the Cerkno dialect of Slovenian, the Chakavian dialect of Croatian and the Torlak dialect of Serbian. The datasets are the first resource for evaluation of large language models on South Slavic dialects, as well as among the first commonsense reasoning datasets on dialects overall. The paper describes specific challenges met during the translation process. A comparison of the dialectal datasets with their standard language counterparts shows a varying level of character-level, word-level and lexicon-level deviation of dialectal text from the standard datasets. The observed differences are well reproduced in initial zero-shot and 10-shot experiments, where the Slovenian Cerkno dialect and the Croatian Chakavian dialect show significantly lower results than the Torlak dialect. These results show also for the dialectal datasets to be significantly more challenging than the standard datasets. Finally, in-context learning on just 10 examples shows to improve the results dramatically, especially for the dialects with the lowest results.

pdf bib
The Role of Adverbs in Language Variety Identification: The Case of Portuguese Multi-Word Adverbs
Izabela Müller | Nuno Mamede | Jorge Baptista

This paper aims to assess the role of multiword compound adverbs in distinguishing Brazilian Portuguese (PT-BR) from European Portuguese (PT-PT). Two key factors underpin this focus: Firstly, multiword expressions often provide less ambiguity compared to single words, even when their meaning is idiomatic (non-compositional). Secondly, despite constituting a significant portion of lexicons in many languages, they are frequently overlooked in Natural Language Processing, possibly due to their heterogeneous nature and lexical range.For this study, a large lexicon of Portuguese multiword adverbs (3,665) annotated with diatopic information regarding language variety was utilized. The paper investigates the distribution of this category in a corpus consisting in excerpts from journalistic texts sourced from the DSL (Dialect and Similar Language) corpus, representing Brazilian (PT-BR) and European Portuguese (PT-PT), respectively, each partition containing 18,000 sentences.Results indicate a substantial similarity between the two varieties, with a considerable overlap in the lexicon of multiword adverbs. Additionally, specific adverbs unique to each language variety were identified. Lexical entries recognized in the corpus represent 18.2% (PT-BR) to 19.5% (PT-PT) of the lexicon, and approximately 5,700 matches in each partition. While many of the matches are spurious due to ambiguity with otherwise non-idiomatic, free strings, occurrences of adverbs marked as exclusive to one variety in texts from the other variety are rare.

pdf bib
NoMusic - The Norwegian Multi-Dialectal Slot and Intent Detection Corpus
Petter Mæhlum | Yves Scherrer

This paper presents a new textual resource for Norwegian and its dialects. The NoMusic corpus contains Norwegian translations of the xSID dataset, an evaluation dataset for spoken language understanding (slot and intent detection). The translations cover Norwegian Bokmål, as well as eight dialects from three of the four major Norwegian dialect areas. To our knowledge, this is the first multi-parallel resource for written Norwegian dialects, and the first evaluation dataset for slot and intent detection focusing on non-standard Norwegian varieties. In this paper, we describe the annotation process and provide some analyses on the types of linguistic variation that can be found in the dataset.

pdf bib
Understanding Position Bias Effects on Fairness in Social Multi-Document Summarization
Olubusayo Olabisi | Ameeta Agrawal

Text summarization models have typically focused on optimizing aspects of quality such as fluency, relevance, and coherence, particularly in the context of news articles. However, summarization models are increasingly being used to summarize diverse sources of text, such as social media data, that encompass a wide demographic user base. It is thus crucial to assess not only the quality of the generated summaries, but also the extent to which they can fairly represent the opinions of diverse social groups. Position bias, a long-known issue in news summarization, has received limited attention in the context of social multi-document summarization. We deeply investigate this phenomenon by analyzing the effect of group ordering in input documents when summarizing tweets from three distinct linguistic communities: African-American English, Hispanic-aligned Language, and White-aligned Language. Our empirical analysis shows that although the textual quality of the summaries remains consistent regardless of the input document order, in terms of fairness, the results vary significantly depending on how the dialect groups are presented in the input data. Our results suggest that position bias manifests differently in social multi-document summarization, severely impacting the fairness of summarization models.

pdf bib
Can LLMs Handle Low-Resource Dialects? A Case Study on Translation and Common Sense Reasoning in Šariš
Viktória Ondrejová | Marek Šuppa

While Large Language Models (LLMs) have demonstrated considerable potential in advancing natural language processing in dialect-specific contexts, their effectiveness in these settings has yet to be thoroughly assessed. This study introduces a case study on Šariš, a dialect of Slovak, which is itself a language with fewer resources, focusing on Machine Translation and Common Sense Reasoning tasks. We employ LLMs in a zero-shot configuration and for data augmentation to refine Slovak-Šariš and Šariš-Slovak translation models. The accuracy of these models is then manually verified by native speakers. Additionally, we introduce ŠarišCOPA, a new dataset for causal common sense reasoning, which, alongside SlovakCOPA, serves to evaluate LLM’s performance in a zero-shot framework. Our findings highlight LLM’s capabilities in processing low-resource dialects and suggest a viable approach for initiating dialect-specific translation models in such contexts.

pdf bib
Experiments in Multi-Variant Natural Language Processing for Nahuatl
Robert Pugh | Francis Tyers

Linguistic variation is a complicating factor for digital language technologies. This is particularly true for languages that lack an official “standard” variety, including many regional and minoritized languages. In this paper, we describe a set of experiments focused on multivariant natural language processing for the Nahuatl, an indigenous Mexican language with a high level of linguistic variation and no single recognized standard variant. Using small (10k tokens), recently-published annotated datasets for two Nahuatl variants, we compare the performance of single-variant, cross-variant, and joint training, and explore how different models perform on a third Nahuatl variant, unseen in training. These results and the subsequent discussion contribute to efforts of developing low-resource NLP that is robust to diatopic variation. We share all code used to process the data and run the experiments.

pdf bib
Highly Granular Dialect Normalization and Phonological Dialect Translation for Limburgish
Andreas Simons | Stefano De Pascale | Karlien Franco

We study highly granular dialect normalization and phonological dialect translation on Limburgish, a non-standardized low-resource language with a wide variation in spelling conventions and phonology. We find improvements to the traditional transformer by embedding the geographic coordinates of dialects in dialect normalization tasks and use these geographically-embedded transformers to translate words between the phonologies of different dialects. These results are found to be consistent with notions in traditional Limburgish dialectology.

pdf bib
Multilingual Identification of English Code-Switching
Igor Sterner

This work addresses the task of identifying English code-switching in multilingual text. We train two token-level classifiers on data of high-resource language pairs. The first distinguishes between English, not English, morphologically mixed, and other words. The second is a binary classifier that identifies named entities. Results indicate that our system is on-par with SoTA for high-resource language pairs. Meanwhile we show that on low-resource language pairs not in the training data our system outperforms SoTA by between 2.31 and 4.59% F1. We also analyse the correlation between typological similarity of the languages and difficulty in recognizing code-switching. Our system is a new strong baseline system for code-switching research between any language and English.

pdf bib
Studying Language Variation Considering the Re-Usability of Modern Theories, Tools and Resources for Annotating Explicit and Implicit Events in Centuries Old Text
Stella Verkijk | Pia Sommerauer | Piek Vossen

This paper discusses the re-usibility of existing approaches, tools and automatic techniques for the annotation and detection of events in a challenging variant of centuries old Dutch written in the archives of the Dutch East India Company. We describe our annotation process and provide a thorough analysis of different versions of manually annotated data and the first automatic results from two fine-tuned Language Models. Through the analysis of this complete process, the paper studies two things: to what extent we can use NLP theories and tasks formulated for modern English to formulate an annotation task for Early Modern Dutch and to what extent we can use NLP models and tools built for modern Dutch (and other languages) on Early Modern Dutch. We believe these analyses give us insight into how to deal with the large variation language showcases in describing events, and how this variation may differ accross domains. We release the annotation guidelines, annotated data, and code.

pdf bib
Language Identification of Philippine Creole Spanish: Discriminating Chavacano From Related Languages
Aileen Joan Vicente | Charibeth Cheng

Chavacano is a Spanish Creole widely spoken in the southern regions of the Philippines. It is one of the many Philippine languages yet to be studied computationally. This paper presents the development of a language identification model of Chavacano to distinguish it from languages that influence its creolization using character convolutional networks. Unlike studies that discriminated similar languages based on geographical proximity, this paper reports a similarity focused on the creolization of a language. We established the similarity of Chavacano and its related languages, Spanish, Portuguese, Cebuano, and Hiligaynon, from the number of common words in the corpus for all languages. We report an accuracy of 93% for the model generated using ten filters with a filter width of 5. The training experiments reveal that increasing the filter width, number of filters, or training epochs is unnecessary even if the accuracy increases because the generated models present irregular learning behavior or may have already been overfitted. This study also demonstrates that the character features extracted from convolutional neural networks, similar to n-grams, are sufficient in identifying Chavacano. Future work on the language identification of Chavacano includes improving classification accuracy for short or code-switched texts for practical applications such as social media sensors for disaster response and management.

pdf bib
Data-Augmentation-Based Dialectal Adaptation for LLMs
Fahim Faisal | Antonios Anastasopoulos

This report presents gmnlp’s participation to the Dialect-Copa shared task at VarDial 2024 (Chifu et al., 2024), which focuses on evaluating the commonsense reasoning capabilities of large language models (LLMs) on South Slavic micro-dialects. The task aims to assess how well LLMs can handle non-standard dialectal varieties, as their performance on standard languages is already well-established. We propose an approach that combines the strengths of different types of language models and leverages data augmentation techniques to improve task performance on three South Slavic dialects: Chakavian, Cherkano, and Torlak. We conduct experiments using a language-family-focused encoder-based model (BERTić) and a domain-agnostic multilingual model (AYA-101). Our results demonstrate that the proposed data augmentation techniques lead to substantial performance gains across all three test datasets in the open-source model category. This work highlights the practical utility of data augmentation and the potential of LLMs in handling non-standard dialectal varieties, contributing to the broader goal of advancing natural language understanding in low-resource and dialectal settings.

pdf bib
JSI and WüNLP at the DIALECT-COPA Shared Task: In-Context Learning From Just a Few Dialectal Examples Gets You Quite Far
Nikola Ljubešić | Taja Kuzman | Peter Rupnik | Ivan Vulić | Fabian Schmidt | Goran Glavaš

The paper presents the JSI and WüNLP systems submitted to the DIALECT-COPA shared task on causal commonsense reasoning in dialectal texts. Jointly, we compare LLM-based zero-shot and few-shot in-context inference (JSI team), and task-specific few-shot fine-tuning, in English and respective standard language, with zero-shot cross-lingual transfer (ZS-XLT) to the test dialects (WüNLP team). Given the very strong zero-shot and especially few-shot in-context learning (ICL) performance, we further investigate whether task semantics, or language/dialect semantics explain the strong performance, showing that a significant part of the improvement indeed stems from learning the language or dialect semantics from the in-context examples, with only a minor contribution from understanding the nature of the task. The higher importance of the dialect semantics to the task semantics is further shown by the finding that the in-context learning with only a few dialectal instances achieves comparable results to the supervised fine-tuning approach on hundreds of instances in standard language.

pdf bib
Incorporating Dialect Understanding Into LLM Using RAG and Prompt Engineering Techniques for Causal Commonsense Reasoning
Benedikt Perak | Slobodan Beliga | Ana Meštrović

The choice of plausible alternatives (COPA) task requires selecting the most plausible outcome from two choices based on understanding the causal relationships presented in a given text.This paper outlines several approaches and model adaptation strategies to the VarDial 2024 DIALECT-COPA shared task, focusing on causal commonsense reasoning in South-Slavic dialects. We utilize and evaluate the GPT-4 model in combination with various prompts engineering and the Retrieval-Augmented Generation (RAG) technique. Initially, we test and compare the performance of GPT-4 with simple and advanced prompts on the COPA task across three dialects: Cerkno, Chakavian and Torlak. Next, we enhance prompts using the RAG technique specifically for the Chakavian and Cerkno dialect. This involves creating an extended Chakavian-English and Cerkno-Slovene lexical dictionary and integrating it into the prompts. Our findings indicate that the most complex approach, which combines an advanced prompt with an injected dictionary, yields the highest performance on the DIALECT-COPA task.

pdf bib
One-Shot Prompt for Language Variety Identification
Nat Gillin

We present a one-shot prompting approach to multi-class classification for similar language identification with off-the-shelf pre-trained large language model that is not particularly trained or tuned for the language identification task. Without post-training or fine-tuning the model, we simply include one example per class when prompting the model and surprisingly the model to generate the language andlocale labels accordingly.

pdf bib
Improving Multi-Label Classification of Similar Languages by Semantics-Aware Word Embeddings
The Ngo | Thi Anh Nguyen | My Ha | Thi Minh Nguyen | Phuong Le-Hong

The VLP team participated in the DSL-ML shared task of the VarDial 2024 workshop which aims to distinguish texts in similar languages. This paper presents our approach to solving the problem and discusses our experimental and official results. We propose to integrate semantics-aware word embeddings which are learned from ConceptNet into a bidirectional long short-term memory network. This approach achieves good performance – our sys- tem is ranked in the top two or three of the best performing teams for the task.

pdf bib
Brandeis at VarDial 2024 DSL-ML Shared Task: Multilingual Models, Simple Baselines and Data Augmentation
Jonne Sälevä | Chester Palen-Michel

This paper describes the Brandeis University submission to VarDial 2024 DSL-ML Shared Task on multilabel classification for discriminating between similar languages. Our submission consists of three entries per language to the closed track, where no additional data was permitted. Our approach involves a set of simple non-neural baselines using logistic regression, random forests and support vector machines. We follow this by experimenting with finetuning multilingual BERT, either on a single language or all the languages concatenated together.In addition to benchmarking the model architectures against one another on the development set, we perform extensive hyperparameter tuning, which is afforded by the small size of the training data.Our experiments on the development set suggest that finetuned mBERT systems significantly benefit most languages compared to the baseline.However, on the test set, our results indicate that simple models based on scikit-learn can perform surprisingly well and even outperform pretrained language models, as we see with BCMS.Our submissions achieve the best performance on all languages as reported by the organizers. Except for Spanish and French, our non-neural baseline also ranks in the top 3 for all other languages.
up

pdf (full)
bib (full) Proceedings of the 8th Workshop on Online Abuse and Harms (WOAH 2024)

pdf bib
Proceedings of the 8th Workshop on Online Abuse and Harms (WOAH 2024)
Yi-Ling Chung | Zeerak Talat | Debora Nozza | Flor Miriam Plaza-del-Arco | Paul Röttger | Aida Mostafazadeh Davani | Agostina Calabrese

pdf bib
Investigating radicalisation indicators in online extremist communities
Christine De Kock | Eduard Hovy

We identify and analyse three sociolinguistic indicators of radicalisation within online extremist forums: hostility, longevity and social connectivity. We develop models to predict the maximum degree of each indicator measured over an individual’s lifetime, based on a minimal number of initial interactions. Drawing on data from two diverse extremist communities, our results demonstrate that NLP methods are effective at prioritising at-risk users. This work offers practical insights for intervention strategies and policy development, and highlights an important but under-studied research direction.

pdf bib
Detection of Conspiracy Theories Beyond Keyword Bias in German-Language Telegram Using Large Language Models
Milena Pustet | Elisabeth Steffen | Helena Mihaljevic

To protect users from massive hateful content, existing works studied automated hate speech detection. Despite the existing efforts, one question remains: do automated hate speech detectors conform to social media content policies? A platform’s content policies are a checklist of content moderated by the social media platform. Because content moderation rules are often uniquely defined, existing hate speech datasets cannot directly answer this question. This work seeks to answer this question by creating HateModerate, a dataset for testing the behaviors of automated content moderators against content policies. First, we engage 28 annotators and GPT in a six-step annotation process, resulting in a list of hateful and non-hateful test suites matching each of Facebook’s 41 hate speech policies. Second, we test the performance of state-of-the-art hate speech detectors against HateModerate, revealing substantial failures these models have in their conformity to the policies. Third, using HateModerate, we augment the training data of a top-downloaded hate detector on HuggingFace. We observe significant improvement in the models’ conformity to content policies while having comparable scores on the original test data. Our dataset and code can be found in the attachment.

pdf bib
EkoHate: Abusive Language and Hate Speech Detection for Code-switched Political Discussions on Nigerian Twitter
Comfort Ilevbare | Jesujoba Alabi | David Ifeoluwa Adelani | Firdous Bakare | Oluwatoyin Abiola | Oluwaseyi Adeyemo

Nigerians have a notable online presence and actively discuss political and topical matters. This was particularly evident throughout the 2023 general election, where Twitter was used for campaigning, fact-checking and verification, and even positive and negative discourse. However, little or none has been done in the detection of abusive language and hate speech in Nigeria. In this paper, we curated code-switched Twitter data directed at three musketeers of the governorship election on the most populous and economically vibrant state in Nigeria; Lagos state, with the view to detect offensive speech in political discussions. We developed EkoHate—an abusive language and hate speech dataset for political discussions between the three candidates and their followers using a binary (normal vs offensive) and fine-grained four-label annotation scheme. We analysed our dataset and provided an empirical evaluation of state-of-the-art methods across both supervised and cross-lingual transfer learning settings. In the supervised setting, our evaluation results in both binary and four-label annotation schemes show that we can achieve 95.1 and 70.3 F1 points respectively. Furthermore, we show that our dataset adequately transfers very well to three publicly available offensive datasets (OLID, HateUS2020, and FountaHate), generalizing to political discussions in other regions like the US.

pdf bib
A Study of the Class Imbalance Problem in Abusive Language Detection
Yaqi Zhang | Viktor Hangya | Alexander Fraser

Abusive language detection has drawn increasing interest in recent years. However, a less systematically explored obstacle is label imbalance, i.e., the amount of abusive data is much lower than non-abusive data, leading to performance issues. The aim of this work is to conduct a comprehensive comparative study of popular methods for addressing the class imbalance issue. We explore 10 well-known approaches on 8 datasets with distinct characteristics: binary or multi-class, moderately or largely imbalanced, focusing on various types of abuse, etc. Additionally, we pro-pose two novel methods specialized for abuse detection: AbusiveLexiconAug and ExternalDataAug, which enrich the training data using abusive lexicons and external abusive datasets, respectively. We conclude that: 1) our AbusiveLexiconAug approach, random oversampling, and focal loss are the most versatile methods on various datasets; 2) focal loss tends to yield peak model performance; 3) oversampling and focal loss provide promising results for binary datasets and small multi-class sets, while undersampling and weighted cross-entropy are more suitable for large multi-class sets; 4) most methods are sensitive to hyperparameters, yet our suggested choice of hyperparameters provides a good starting point.

pdf bib
HausaHate: An Expert Annotated Corpus for Hausa Hate Speech Detection
Francielle Vargas | Samuel Guimarães | Shamsuddeen Hassan Muhammad | Diego Alves | Ibrahim Said Ahmad | Idris Abdulmumin | Diallo Mohamed | Thiago Pardo | Fabrício Benevenuto

We introduce the first expert annotated corpus of Facebook comments for Hausa hate speech detection. The corpus titled HausaHate comprises 2,000 comments extracted from Western African Facebook pages and manually annotated by three Hausa native speakers, who are also NLP experts. Our corpus was annotated using two different layers. We first labeled each comment according to a binary classification: offensive versus non-offensive. Then, offensive comments were also labeled according to hate speech targets: race, gender and none. Lastly, a baseline model using fine-tuned LLM for Hausa hate speech detection is presented, highlighting the challenges of hate speech detection tasks for indigenous languages in Africa, as well as future advances.

pdf bib
VIDA: The Visual Incel Data Archive. A Theory-oriented Annotated Dataset To Enhance Hate Detection Through Visual Culture
Selenia Anastasi | Florian Schneider | Chris Biemann | Tim Fischer

Images increasingly constitute a larger portion of internet content, encoding even more complex meanings. Recent studies have highlight the pivotal role of visual communication in the spread of extremist content, particularly that associated with right-wing political ideologies. However, the capability of machine learning systems to recognize such meanings, sometimes implicit, remains limited. To enable future research in this area, we introduce and release VIDA, the Visual Incel Data Archive, a multimodal dataset comprising visual material and internet memes collected from two main Incel communities (Italian and Anglophone) known for their extremist misogynistic content. Following the analytical framework of Shifman (2014), we propose a new taxonomy for annotation across three main levels of analysis: content, form, and stance (hate). This allows for the association of images with fine-grained contextual information that help to identify the presence of offensiveness and a broader set of cultural references, enhancing the understanding of more nuanced aspects in visual communication. In this work we present a statistical analysis of the annotated dataset as well as discuss annotation examples and future line of research.

pdf bib
Towards a Unified Framework for Adaptable Problematic Content Detection via Continual Learning
Ali Omrani | Alireza Salkhordeh Ziabari | Preni Golazizian | Jeffrey Sorensen | Morteza Dehghani

Detecting problematic content, such as hate speech, is a multifaceted and ever-changing task, influenced by social dynamics, user populations, diversity of sources, and evolving language. There has been significant efforts, both in academia and in industry, to develop annotated resources that capture various aspects of problematic content. Due to researchers’ diverse objectives, these annotations are often inconsistent and hence, reports of progress on the detection of problematic content are fragmented. This pattern is expected to persist unless we pool these resources, taking into account the dynamic nature of this issue. In this paper, we propose integrating the available resources, leveraging their dynamic nature to break this pattern, and introduce a continual learning framework and benchmark for problematic content detection. Our benchmark, comprising 84 related tasks, creates a novel measure of progress: prioritizing the adaptability of classifiers to evolving tasks over excelling in specific tasks. To ensure continuous relevance, our benchmark is designed for seamless integration of new tasks. Our results demonstrate that continual learning methods outperform static approaches by up to 17% and 4% AUC in capturing the evolving content and adapting to novel forms of problematic content

pdf bib
From Linguistics to Practice: a Case Study of Offensive Language Taxonomy in Hebrew
Chaya Liebeskind | Marina Litvak | Natalia Vanetik

The perception of offensive language varies based on cultural, social, and individual perspectives. With the spread of social media, there has been an increase in offensive content online, necessitating advanced solutions for its identification and moderation. This paper addresses the practical application of an offensive language taxonomy, specifically targeting Hebrew social media texts. By introducing a newly annotated dataset, modeled after the taxonomy of explicit offensive language of (Lewandowska-Tomaszczyk et al., 2023)„ we provide a comprehensive examination of various degrees and aspects of offensive language. Our findings indicate the complexities involved in the classification of such content. We also outline the implications of relying on fixed taxonomies for Hebrew.

pdf bib
Estimating the Emotion of Disgust in Greek Parliament Records
Vanessa Lislevand | John Pavlopoulos | Panos Louridas | Konstantina Dritsa

We present an analysis of the sentiment in Greek political speech, by focusing on the most frequently occurring emotion in electoral data, the emotion of “disgust”. We show that emotion classification is generally tough, but high accuracy can be achieved for that particular emotion. Using our best-performing model to classify political records of the Greek Parliament Corpus from 1989 to 2020, we studied the points in time when this emotion was frequently occurring and we ranked the Greek political parties based on their estimated score. We then devised an algorithm to investigate the emotional context shift of words that describe specific conditions and that can be used to stigmatise. Given that early detection of such word usage is essential for policy-making, we report two words we found being increasingly used in a negative emotional context, and one that is likely to be carrying stigma, in the studied parliamentary records.

pdf bib
Simple LLM based Approach to Counter Algospeak
Jan Fillies | Adrian Paschke

With the use of algorithmic moderation on online communication platforms, an increase in adaptive language aiming to evade the automatic detection of problematic content has been observed. One form of this adapted language is known as “Algospeak” and is most commonly associated with large social media platforms, e.g., TikTok. It builds upon Leetspeak or online slang with its explicit intention to avoid machine readability. The machine-learning algorithms employed to automate the process of content moderation mostly rely on human-annotated datasets and supervised learning, often not adjusted for a wide variety of languages and changes in language. This work uses linguistic examples identified in research literature to introduce a taxonomy for Algospeak and shows that with the use of an LLM (GPT-4), 79.4% of the established terms can be corrected to their true form, or if needed, their underlying associated concepts. With an example sentence, 98.5% of terms are correctly identified. This research demonstrates that LLMs are the future in solving the current problem of moderation avoidance by Algospeak.

pdf bib
Harnessing Personalization Methods to Identify and Predict Unreliable Information Spreader Behavior
Shaina Ashraf | Fabio Gruschka | Lucie Flek | Charles Welch

Studies on detecting and understanding the spread of unreliable news on social media have identified key characteristic differences between reliable and unreliable posts. These differences in language use also vary in expression across individuals, making it important to consider personal factors in unreliable news detection. The application of personalization methods for this has been made possible by recent publication of datasets with user histories, though this area is still largely unexplored. In this paper we present approaches to represent social media users in order to improve performance on three tasks: (1) classification of unreliable news posts, (2) classification of unreliable news spreaders, and, (3) prediction of the spread of unreliable news. We compare the User2Vec method from previous work to two other approaches; a learnable user embedding layer trained with the downstream task, and a representation derived from an authorship attribution classifier. We demonstrate that the implemented strategies substantially improve classification performance over state-of-the-art and provide initial results on the task of unreliable news prediction.

pdf bib
Robust Safety Classifier Against Jailbreaking Attacks: Adversarial Prompt Shield
Jinhwa Kim | Ali Derakhshan | Ian Harris

Large Language Models’ safety remains a critical concern due to their vulnerability to jailbreaking attacks, which can prompt these systems to produce harmful and malicious responses. Safety classifiers, computational models trained to discern and mitigate potentially harmful, offensive, or unethical outputs, offer a practical solution to address this issue. However, despite their potential, existing safety classifiers often fail when exposed to adversarial attacks such as gradient-optimized suffix attacks. In response, our study introduces Adversarial Prompt Shield (APS), a lightweight safety classifier model that excels in detection accuracy and demonstrates resilience against unseen jailbreaking prompts. We also introduce efficiently generated adversarial training datasets, named Bot Adversarial Noisy Dialogue (BAND), which are designed to fortify the classifier’s robustness. Through extensive testing on various safety tasks and unseen jailbreaking attacks, we demonstrate the effectiveness and resilience of our models. Evaluations show that our classifier has the potential to significantly reduce the Attack Success Rate by up to 44.9%. This advance paves the way for the next generation of more reliable and resilient Large Language Models.

pdf bib
Improving aggressiveness detection using a data augmentation technique based on a Diffusion Language Model
Antonio Reyes-Ramírez | Mario Aragón | Fernando Sánchez-Vega | Adrian López-Monroy

Cyberbullying has grown in recent years, largely attributed to the proliferation of social media users. This phenomenon manifests in various forms, such as hate speech and offensive language, increasing the necessity of effective detection models to tackle this problem. Most approaches focus on supervised algorithms, which have an important drawback—they heavily depend on the availability of ample training data. This paper attempts to tackle this insufficient data problem using data augmentation (DA) techniques. Concretely, we propose a novel data augmentation technique based on a Diffusion Language Model (DLA). We compare our proposed method against well-known DA techniques, such as contextual augmentation and Easy Data Augmentation (EDA). Our findings reveal a slight but promising improvement, leading to more robust results with very low variance. Additionally, we provide a comprehensive qualitative analysis using classification errors, and complementary analysis, shedding light on the nuances of our approach.

pdf bib
The Mexican Gayze: A Computational Analysis of the Attitudes towards the LGBT+ Population in Mexico on Social Media Across a Decade
Scott Andersen | Segio-Luis Ojeda-Trueba | Juan Vásquez | Gemma Bel-Enguix

Thanks to the popularity of social media, data generated by online communities provides an abundant source of diverse language information. This abundance of data allows NLP practitioners and computational linguists to analyze sociolinguistic phenomena occurring in digital communication. In this paper, we analyze the Twitter discourse around the Mexican Spanish-speaking LGBT+ community. For this, we evaluate how the polarity of some nouns related to the LGBT+ community has evolved in conversational settings using a corpus of tweets that cover a time span of ten years. We hypothesize that social media’s fast-moving, turbulent linguistic environment encourages language evolution faster than ever before. Our results indicate that most of the inspected terms have undergone some shift in denotation or connotation. No other generalizations can be observed in the data, given the difficulty that current NLP methods have to account for polysemy, and the wide differences between the various subgroups that make up the LGBT+ community. A fine-grained analysis of a series of LGBT+-related lexical terms is also included in this work.

pdf bib
X-posing Free Speech: Examining the Impact of Moderation Relaxation on Online Social Networks
Arvindh Arun | Saurav Chhatani | Jisun An | Ponnurangam Kumaraguru

We investigate the impact of free speech and the relaxation of moderation on online social media platforms using Elon Musk’s takeover of Twitter as a case study. By curating a dataset of over 10 million tweets, our study employs a novel framework combining content and network analysis. Our findings reveal a significant increase in the distribution of certain forms of hate content, particularly targeting the LGBTQ+ community and liberals. Network analysis reveals the formation of cohesive hate communities facilitated by influential bridge users, with substantial growth in interactions hinting at increased hate production and diffusion. By tracking the temporal evolution of PageRank, we identify key influencers, primarily self-identified far-right supporters disseminating hate against liberals and woke culture. Ironically, embracing free speech principles appears to have enabled hate speech against the very concept of freedom of expression and free speech itself. Our findings underscore the delicate balance platforms must strike between open expression and robust moderation to curb the proliferation of hate online.

pdf bib
The Uli Dataset: An Exercise in Experience Led Annotation of oGBV
Arnav Arora | Maha Jinadoss | Cheshta Arora | Denny George | Brindaalakshmi | Haseena Khan | Kirti Rawat | Div | Ritash | Seema Mathur

Online gender-based violence has grown concomitantly with the adoption of the internet and social media. Its effects are worse in the Global majority where many users use social media in languages other than English. The scale and volume of conversations on the internet have necessitated the need for automated detection of hate speech and, more specifically, gendered abuse. There is, however, a lack of language-specific and contextual data to build such automated tools. In this paper, we present a dataset on gendered abuse in three languages- Hindi, Tamil and Indian English. The dataset comprises of tweets annotated along three questions pertaining to the experience of gender abuse, by experts who identify as women or a member of the LGBTQIA+ community in South Asia. Through this dataset, we demonstrate a participatory approach to creating datasets that drive AI systems.

pdf bib
Towards Interpretable Hate Speech Detection using Large Language Model-extracted Rationales
Ayushi Nirmal | Amrita Bhattacharjee | Paras Sheth | Huan Liu

Although social media platforms are a prominent arena for users to engage in interpersonal discussions and express opinions, the facade and anonymity offered by social media may allow users to spew hate speech and offensive content. Given the massive scale of such platforms, there arises a need to automatically identify and flag instances of hate speech. Although several hate speech detection methods exist, most of these black-box methods are not interpretable or explainable by design. To address the lack of interpretability, in this paper, we propose to use state-of-the-art Large Language Models (LLMs) to extract features in the form of rationales from the input text, to train a base hate speech classifier, thereby enabling faithful interpretability by design. Our framework effectively combines the textual understanding capabilities of LLMs and the discriminative power of state-of-the-art hate speech classifiers to make these classifiers faithfully interpretable. Our comprehensive evaluation on a variety of social media hate speech datasets demonstrate: (1) the goodness of the LLM-extracted rationales, and (2) the surprising retention of detector performance even after training to ensure interpretability. All code and data will be made available at https://github.com/AmritaBh/shield.

pdf bib
A Bayesian Quantification of Aporophobia and the Aggravating Effect of Low–Wealth Contexts on Stigmatization
Ryan Brate | Marieke Van Erp | Antal Van Den Bosch

Aporophobia, a negative social bias against poverty and the poor, has been highlighted asan overlooked phenomenon in toxicity detec-tion in texts. Aporophobia is potentially im-portant both as a standalone form of toxicity,but also given its potential as an aggravatingfactor in the wider stigmatization of groups. Asyet, there has been limited quantification of thisphenomenon. In this paper, we first quantifythe extent of aporophobia, as observable in Red-dit data: contrasting estimates of stigmatisingtopic propensity between low–wealth contextsand high–wealth contexts via Bayesian estima-tion. Next, we consider aporophobia as a causalfactor in the prejudicial association of groupswith stigmatising topics, by introducing peoplegroup as a variable, specifically Black people.This group is selected given its history of be-ing the subject of toxicity. We evaluate theaggravating effect on the observed n–grams in-dicative of stigmatised topics observed in com-ments which refer to Black people, due to thepresence of low–wealth contexts. We performthis evaluation via a Structural Causal Mod-elling approach, performing interventions onsimulations via Bayesian models, for three hy-pothesised causal mechanisms.

pdf bib
Toxicity Classification in Ukrainian
Daryna Dementieva | Valeriia Khylenko | Nikolay Babakov | Georg Groh

The task of toxicity detection is still a relevant task, especially in the context of safe and fair LMs development. Nevertheless, labeled binary toxicity classification corpora are not available for all languages, which is understandable given the resource-intensive nature of the annotation process. Ukrainian, in particular, is among the languages lacking such resources. To our knowledge, there has been no existing toxicity classification corpus in Ukrainian. In this study, we aim to fill this gap by investigating cross-lingual knowledge transfer techniques and creating labeled corpora by: (i)~translating from an English corpus, (ii)~filtering toxic samples using keywords, and (iii)~annotating with crowdsourcing. We compare LLMs prompting and other cross-lingual transfer approaches with and without fine-tuning offering insights into the most robust and efficient baselines.

pdf bib
A Strategy Labelled Dataset of Counterspeech
Aashima Poudhar | Ioannis Konstas | Gavin Abercrombie

Increasing hateful conduct online demands effective counterspeech strategies to mitigate its impact. We introduce a novel dataset annotated with such strategies, aimed at facilitating the generation of targeted responses to hateful language. We labelled 1000 hate speech/counterspeech pairs from an existing dataset with strategies established in the social sciences. We find that a one-shot prompted classification model achieves promising accuracy in classifying the strategies according to the manual labels, demonstrating the potential of generative Large Language Models (LLMs) to distinguish between counterspeech strategies.

pdf bib
Improving Covert Toxicity Detection by Retrieving and Generating References
Dong-Ho Lee | Hyundong Cho | Woojeong Jin | Jihyung Moon | Sungjoon Park | Paul Röttger | Jay Pujara | Roy Ka-wei Lee

Models for detecting toxic content play an important role in keeping people safe online. There has been much progress in detecting overt toxicity. Covert toxicity, however, remains a challenge because its detection requires an understanding of implicit meaning and subtle connotations. In this paper, we explore the potential of leveraging references, such as external knowledge and textual interpretations, to enhance the detection of covert toxicity. We run experiments on two covert toxicity datasets with two types of references: 1) information retrieved from a search API, and 2) interpretations generated by large language models. We find that both types of references improve detection, with the latter being more useful than the former. We also find that generating interpretations grounded on properties of covert toxicity, such as humor and irony, lead to the largest improvements

pdf bib
Subjective Isms? On the Danger of Conflating Hate and Offence in Abusive Language Detection
Amanda Cercas Curry | Gavin Abercrombie | Zeerak Talat

Natural language processing research has begun to embrace the notion of annotator subjectivity, motivated by variations in labelling. This approach understands each annotator’s view as valid, which can be highly suitable for tasks that embed subjectivity, e.g., sentiment analysis. However, this construction may be inappropriate for tasks such as hate speech detection, as it affords equal validity to all positions on e.g., sexism or racism. We argue that the conflation of hate and offence can invalidate findings on hate speech, and call for future work to be situated in theory, disentangling hate from its orthogonal concept, offence.

pdf bib
From Languages to Geographies: Towards Evaluating Cultural Bias in Hate Speech Datasets
Manuel Tonneau | Diyi Liu | Samuel Fraiberger | Ralph Schroeder | Scott Hale | Paul Röttger

Perceptions of hate can vary greatly across cultural contexts. Hate speech (HS) datasets, however, have traditionally been developed by language. This hides potential cultural biases, as one language may be spoken in different countries home to different cultures. In this work, we evaluate cultural bias in HS datasets by leveraging two interrelated cultural proxies: language and geography. We conduct a systematic survey of HS datasets in eight languages and confirm past findings on their English-language bias, but also show that this bias has been steadily decreasing in the past few years. For three geographically-widespread languages—English, Arabic and Spanish—we then leverage geographical metadata from tweets to approximate geo-cultural contexts by pairing language and country information. We find that HS datasets for these languages exhibit a strong geo-cultural bias, largely overrepresenting a handful of countries (e.g., US and UK for English) relative to their prominence in both the broader social media population and the general population speaking these languages. Based on these findings, we formulate recommendations for the creation of future HS datasets.

pdf bib
SGHateCheck: Functional Tests for Detecting Hate Speech in Low-Resource Languages of Singapore
Ri Chi Ng | Nirmalendu Prakash | Ming Shan Hee | Kenny Tsu Wei Choo | Roy Ka-wei Lee

To address the limitations of current hate speech detection models, we introduce SGHateCheck, a novel framework designed for the linguistic and cultural context of Singapore and Southeast Asia. It extends the functional testing approach of HateCheck and MHC, employing large language models for translation and paraphrasing into Singapore’s main languages, and refining these with native annotators. SGHateCheck reveals critical flaws in state-of-the-art models, highlighting their inadequacy in sensitive content moderation. This work aims to foster the development of more effective hate speech detection tools for diverse linguistic environments, particularly for Singapore and Southeast Asia contexts.