Large Language Models (LLMs) have demonstrated remarkable capabilities across various applications, fundamentally reshaping the landscape of natural language processing (NLP) research. However, recent evaluation frameworks often rely on the output probabilities of LLMs for predictions, primarily due to computational constraints, diverging from real-world LLM usage scenarios. While widely employed, the efficacy of these probability-based evaluation strategies remains an open research question. This study aims to scrutinize the validity of such probability-based evaluation methods within the context of using LLMs for Multiple Choice Questions (MCQs), highlighting their inherent limitations. Our empirical investigation reveals that the prevalent probability-based evaluation method inadequately aligns with generation-based prediction. Furthermore, current evaluation frameworks typically assess LLMs through predictive tasks based on output probabilities rather than directly generating responses, owing to computational limitations. We illustrate that these probability-based approaches do not effectively correspond with generative predictions. The outcomes of our study can enhance the understanding of LLM evaluation methodologies and provide insights for future research in this domain.

Recent studies have illuminated the promising capabilities of large language models (LLMs) in handling long texts. However, their performance in machine translation (MT) of long documents remains underexplored. This paper aims to shed light on how LLMs navigate this complex task, offering a comprehensive evaluation of their capabilities and limitations in long-text MT. First, we collect and construct an instruction-based benchmark dataset, specifically designed for the finetuning and evaluation of LLMs, encompassing multilingual, multi-domain, and document-level parallel data. Second, we conduct a comprehensive comparison between MT and LLM models concerning document-level translation. Our analysis uncovers that LLMs exhibit shortcomings in long-text domains, and their performance diminishes as document size escalates. By exploiting various extrapolation strategies, we enhance the capacity of LLMs to translate longer texts. We release data, code, and models at https://github.com/longyuewangdcu/Document-MT-LLM.

Large vision-language models (LVLMs) have made significant progress in recent years. While LVLMs exhibit excellent ability in language understanding, question answering, and conversations of visual inputs, they are prone to producing hallucinations. While several methods are proposed to evaluate the hallucinations in LVLMs, most are reference-based and depend on external tools, which complicates their practical application. To assess the viability of alternative methods, it is critical to understand whether the reference-free approaches, which do not rely on any external tools, can efficiently detect hallucinations. Therefore, we initiate an exploratory study to demonstrate the effectiveness of different reference-free solutions in detecting hallucinations in LVLMs. In particular, we conduct an extensive study on three kinds of techniques: uncertainty-based, consistency-based, and supervised uncertainty quantification methods on four representative LVLMs across two different tasks. The empirical results show that the reference-free approaches are capable of effectively detecting non-factual responses in LVLMs, with the supervised uncertainty quantification method outperforming the others, achieving the best performance across different settings.

Translating literary works has perennially stood as an elusive dream in machine translation (MT), a journey steeped in intricate challenges. To foster progress in this domain, we hold a new shared task at WMT 2023, the second edition of the Discourse-Level Literary Translation. First, we (Tencent AI Lab and China Literature Ltd.) release a copyrighted and document-level Chinese-English web novel corpus. Furthermore, we put forth an industry-endorsed criteria to guide human evaluation process. This year, we totally received 10 submissions from 5 academia and industry teams. We employ both automatic and human evaluations to measure the performance of the submitted systems. The official ranking of the systems is based on the overall human judgments. In addition, our extensive analysis reveals a series of interesting findings on literary and discourse-aware MT. We release data, system outputs, and leaderboard at https://www2.statmt.org/wmt24/literary-translation-task.html.

Machine Translation (MT) has greatly advanced over the years due to the developments in deep neural networks. However, the emergence of Large Language Models (LLMs) like GPT-4 and ChatGPT is introducing a new phase in the MT domain. In this context, we believe that the future of MT is intricately tied to the capabilities of LLMs. These models not only offer vast linguistic understandings but also bring innovative methodologies, such as prompt-based techniques, that have the potential to further elevate MT. In this paper, we provide an overview of the significant enhancements in MT that are influenced by LLMs and advocate for their pivotal role in upcoming MT research and implementations. We highlight several new MT directions, emphasizing the benefits of LLMs in scenarios such as Long-Document Translation, Stylized Translation, and Interactive Translation. Additionally, we address the important concern of privacy in LLM-driven MT and suggest essential privacy-preserving strategies. By showcasing practical instances, we aim to demonstrate the advantages that LLMs offer, particularly in tasks like translating extended documents. We conclude by emphasizing the critical role of LLMs in guiding the future evolution of MT and offer a roadmap for future exploration in the sector.

Event-Level Video Question Answering (EVQA) requires complex reasoning across video events to obtain the visual information needed to provide optimal answers. However, despite significant progress in model performance, few studies have focused on using the explicit semantic connections between the question and visual information especially at the event level. There is need for using such semantic connections to facilitate complex reasoning across video frames. Therefore, we propose a semantic-aware dynamic retrospective-prospective reasoning approach for video-based question answering. Specifically, we explicitly use the Semantic Role Labeling (SRL) structure of the question in the dynamic reasoning process where we decide to move to the next frame based on which part of the SRL structure (agent, verb, patient, etc.) of the question is being focused on. We conduct experiments on a benchmark EVQA dataset - TrafficQA. Results show that our proposed approach achieves superior performance compared to previous state-of-the-art models. Our code is publicly available at https://github.com/lyuchenyang/Semantic-aware-VideoQA.

User and product information associated with a review is useful for sentiment polarity prediction. Typical approaches incorporating such information focus on modeling users and products as implicitly learned representation vectors. Most do not exploit the potential of historical reviews, or those that currently do require unnecessary modifications to model architectureor do not make full use of user/product associations. The contribution of this work is twofold: i) a method to explicitly employ historical reviews belonging to the same user/product in initializing representations, and ii) efficient incorporation of textual associations between users and products via a user-product cross-context module. Experiments on the IMDb, Yelp-2013 and Yelp-2014 English benchmarks with BERT, SpanBERT and Longformer pretrained language models show that our approach substantially outperforms previous state-of-the-art.

Machine learning (ML) systems in natural language processing (NLP) face significant challenges in generalizing to out-of-distribution (OOD) data, where the test distribution differs from the training data distribution. This poses important questions about the robustness of NLP models and their high accuracy, which may be artificially inflated due to their underlying sensitivity to systematic biases. Despite these challenges, there is a lack of comprehensive surveys on the generalization challenge from an OOD perspective in natural language understanding. Therefore, this paper aims to fill this gap by presenting the first comprehensive review of recent progress, methods, and evaluations on this topic. We further discuss the challenges involved and potential future research directions. By providing convenient access to existing work, we hope this survey will encourage future research in this area.

Large language models (LLMs) such as ChatGPT can produce coherent, cohesive, relevant, and fluent answers for various natural language processing (NLP) tasks. Taking document-level machine translation (MT) as a testbed, this paper provides an in-depth evaluation of LLMs’ ability on discourse modeling. The study focuses on three aspects: 1) Effects of Context-Aware Prompts, where we investigate the impact of different prompts on document-level translation quality and discourse phenomena; 2) Comparison of Translation Models, where we compare the translation performance of ChatGPT with commercial MT systems and advanced document-level MT methods; 3) Analysis of Discourse Modelling Abilities, where we further probe discourse knowledge encoded in LLMs and shed light on impacts of training techniques on discourse modeling. By evaluating on a number of benchmarks, we surprisingly find that LLMs have demonstrated superior performance and show potential to become a new paradigm for document-level translation: 1) leveraging their powerful long-text modeling capabilities, GPT-3.5 and GPT-4 outperform commercial MT systems in terms of human evaluation; 2) GPT-4 demonstrates a stronger ability for probing linguistic knowledge than GPT-3.5. This work highlights the challenges and opportunities of LLMs for MT, which we hope can inspire the future design and evaluation of LLMs (We release our data and annotations at https://github.com/longyuewangdcu/Document-MT-LLM).

Translating literary works has perennially stood as an elusive dream in machine translation (MT), a journey steeped in intricate challenges. To foster progress in this domain, we hold a new shared task at WMT 2023, the first edition of the Discourse-Level Literary Translation. First, we (Tencent AI Lab and China Literature Ltd.) release a copyrighted and document-level Chinese-English web novel corpus. Furthermore, we put forth an industry-endorsed criteria to guide human evaluation process. This year, we totally received 14 submissions from 7 academia and industry teams. We employ both automatic and human evaluations to measure the performance of the submitted systems. The official ranking of the systems is based on the overall human judgments. In addition, our extensive analysis reveals a series of interesting findings on literary and discourse-aware MT. We release data, system outputs, and leaderboard at http://www2.statmt.org/wmt23/literary-translation-task.html.

Evaluation of open-domain dialogue systems is highly challenging and development of better techniques is highlighted time and again as desperately needed. Despite substantial efforts to carry out reliable live evaluation of systems in recent competitions, annotations have been abandoned and reported as too unreliable to yield sensible results. This is a serious problem since automatic metrics are not known to provide a good indication of what may or may not be a high-quality conversation. Answering the distress call of competitions that have emphasized the urgent need for better evaluation techniques in dialogue, we present the successful development of human evaluation that is highly reliable while still remaining feasible and low cost. Self-replication experiments reveal almost perfectly repeatable results with a correlation of r=0.969. Furthermore, due to the lack of appropriate methods of statistical significance testing, the likelihood of potential improvements to systems occurring due to chance is rarely taken into account in dialogue evaluation, and the evaluation we propose facilitates application of standard tests. Since we have developed a highly reliable evaluation method, new insights into system performance can be revealed. We therefore include a comparison of state-of-the-art models (i) with and without personas, to measure the contribution of personas to conversation quality, as well as (ii) prescribed versus freely chosen topics. Interestingly with respect to personas, results indicate that personas do not positively contribute to conversation quality as expected.

In this paper, we describe our system for the FinNLP-2022 shared task: Evaluating the Rationales of Amateur Investors (ERAI). The ERAI shared tasks focuses on mining profitable information from financial texts by predicting the possible Maximal Potential Profit (MPP) and Maximal Loss (ML) based on the posts from amateur investors. There are two sub-tasks in ERAI: Pairwise Comparison and Unsupervised Rank, both target on the prediction of MPP and ML. To tackle the two tasks, we frame this task as a text-pair classification task where the input consists of two documents and the output is the label of whether the first document will lead to higher MPP or lower ML. Specifically, we propose to take advantage of the transferability of Sentiment Analysis data with an assumption that a more positive text will lead to higher MPP or higher ML to facilitate the prediction of MPP and ML. In experiment on the ERAI blind test set, our systems trained on Sentiment Analysis data and ERAI training data ranked 1st and 8th in ML and MPP pairwise comparison respectively. Code available in this link.

Past work that investigates out-of-domain performance of QA systems has mainly focused on general domains (e.g. news domain, wikipedia domain), underestimating the importance of subdomains defined by the internal characteristics of QA datasets. In this paper, we extend the scope of “out-of-domain” by splitting QA examples into different subdomains according to their internal characteristics including question type, text length, answer position. We then examine the performance of QA systems trained on the data from different subdomains. Experimental results show that the performance of QA systems can be significantly reduced when the train data and test data come from different subdomains. These results question the generalizability of current QA systems in multiple subdomains, suggesting the need to combat the bias introduced by the internal characteristics of QA datasets.

In this paper, we describe our DCU-Lorcan system for the FinCausal 2022 shared task: span-based cause and effect extraction from financial documents. We frame the FinCausal 2022 causality extraction task as a span extraction/sequence labeling task, our submitted systems are based on the contextualized word representations produced by pre-trained language models and linear layers predicting the label for each word, followed by post-processing heuristics. In experiments, we employ pre-trained language models including DistilBERT, BERT and SpanBERT. Our best performed system achieves F-1, Recall, Precision and Exact Match scores of 92.76, 92.77, 92.76 and 68.60 respectively. Additionally, we conduct experiments investigating the effect of data size to the performance of causality extraction model and an error analysis investigating the outputs in predictions.

In this paper, we present our system for the EvoNLP 2022 shared task Temporal Meaning Shift (TempoWiC). Different from the typically used discriminative model, we propose a generative approach based on pre-trained generation models. The basic architecture of our system is a seq2seq model where the input sequence consists of two documents followed by a question asking whether the meaning of target word changed or not, the target output sequence is a declarative sentence describing the meaning of target word changed or not. The experimental results on TempoWiC test set show that our best system (with time information) obtained an accuracy and Marco F-1 score of 68.09% and 62.59% respectively, which ranked 12th among all submitted systems. The results have shown the plausibility of using generation model for WiC tasks, meanwhile also indicate there’s still room for further improvement.

Question Generation (QG) is the task of generating a plausible question for a given <passage, answer> pair. Template-based QG uses linguistically-informed heuristics to transform declarative sentences into interrogatives, whereas supervised QG uses existing Question Answering (QA) datasets to train a system to generate a question given a passage and an answer. A disadvantage of the heuristic approach is that the generated questions are heavily tied to their declarative counterparts. A disadvantage of the supervised approach is that they are heavily tied to the domain/language of the QA dataset used as training data. In order to overcome these shortcomings, we propose a distantly-supervised QG method which uses questions generated heuristically from summaries as a source of training data for a QG system. We make use of freely available news summary data, transforming declarative summary sentences into appropriate questions using heuristics informed by dependency parsing, named entity recognition and semantic role labeling. The resulting questions are then combined with the original news articles to train an end-to-end neural QG model. We extrinsically evaluate our approach using unsupervised QA: our QG model is used to generate synthetic QA pairs for training a QA model. Experimental results show that, trained with only 20k English Wikipedia-based synthetic QA pairs, the QA model substantially outperforms previous unsupervised models on three in-domain datasets (SQuAD1.1, Natural Questions, TriviaQA) and three out-of-domain datasets (NewsQA, BioASQ, DuoRC), demonstrating the transferability of the approach.

Past work that improves document-level sentiment analysis by encoding user and product in- formation has been limited to considering only the text of the current review. We investigate incorporating additional review text available at the time of sentiment prediction that may prove meaningful for guiding prediction. Firstly, we incorporate all available historical review text belonging to the author of the review in question. Secondly, we investigate the inclusion of his- torical reviews associated with the current product (written by other users). We achieve this by explicitly storing representations of reviews written by the same user and about the same product and force the model to memorize all reviews for one particular user and product. Additionally, we drop the hierarchical architecture used in previous work to enable words in the text to directly attend to each other. Experiment results on IMDB, Yelp 2013 and Yelp 2014 datasets show improvement to state-of-the-art of more than 2 percentage points in the best case.