The emergence of Large Language Models (LLMs), has opened exciting possibilities for constructing computational simulations designed to replicate human behavior accurately. Current research suggests that LLM-based agents become increasingly human-like in their performance, sparking interest in using these AI agents as substitutes for human participants in behavioral studies. However, LLMs are complex statistical learners without straightforward deductive rules, making them prone to unexpected behaviors. Hence, it is crucial to study and pinpoint the key behavioral distinctions between humans and LLM-based agents. In this study, we highlight the limitations of LLMs in simulating human interactions, particularly focusing on LLMs’ ability to simulate political debates on topics that are important aspects of people’s day-to-day lives and decision-making processes. Our findings indicate a tendency for LLM agents to conform to the model’s inherent social biases despite being directed to debate from certain political perspectives. This tendency results in behavioral patterns that seem to deviate from well-established social dynamics among humans. We reinforce these observations using an automatic self-fine-tuning method, which enables us to manipulate the biases within the LLM and demonstrate that agents subsequently align with the altered biases. These results underscore the need for further research to develop methods that help agents overcome these biases, a critical step toward creating more realistic simulations.

When large language models are aligned via supervised fine-tuning, they may encounter new factual information that was not acquired through pre-training. It is often conjectured that this can teach the model the behavior of hallucinating factually incorrect responses, as the model is trained to generate facts that are not grounded in its pre-existing knowledge. In this work, we study the impact of such exposure to new knowledge on the capability of the fine-tuned model to utilize its pre-existing knowledge. To this end, we design a controlled setup, focused on closed-book QA, where we vary the proportion of the fine-tuning examples that introduce new knowledge. We demonstrate that large language models struggle to acquire new factual knowledge through fine-tuning, as fine-tuning examples that introduce new knowledge are learned significantly slower than those consistent with the model’s knowledge. However, we also find that as the examples with new knowledge are eventually learned, they linearly increase the model’s tendency to hallucinate. Taken together, our results highlight the risk in introducing new factual knowledge through fine-tuning, and support the view that large language models mostly acquire factual knowledge through pre-training, whereas fine-tuning teaches them to use it more efficiently.

Existing research on Domain Robustness (DR) suffers from disparate setups, limited task variety, and scarce research on recent capabilities such as in-context learning. Furthermore, the common practice of measuring DR might not be fully accurate. Current research focuses on challenge sets and relies solely on the Source Drop (SD): Using the source in-domain performance as a reference point for degradation. However, we argue that the Target Drop (TD), which measures degradation from the target in-domain performance, should be used as a complementary point of view. To address these issues, we first curated a DR benchmark comprised of 7 diverse NLP tasks, which enabled us to measure both the SD and the TD. We then conducted a comprehensive large-scale DR study involving over 14,000 domain shifts across 21 fine-tuned models and few-shot LLMs. We found that both model types suffer from drops upon domain shifts. While fine-tuned models excel in-domain, few-shot LLMs often surpass them cross-domain, showing better robustness. In addition, we found that a large SD can often be explained by shifting to a harder domain rather than by a genuine DR challenge, and this highlights the importance of TD as a complementary metric. We hope our study will shed light on the current DR state of NLP models and promote improved evaluation practices toward more robust models.

We address the challenge of building task-agnostic classifiers using only text descriptions, demonstrating a unified approach to image classification, 3D point cloud classification, and action recognition from scenes. Unlike approaches that learn a fixed representation of the output classes, we generate at inference time a model tailored to a query classification task. To generate task-based zero-shot classifiers, we train a hypernetwork that receives class descriptions and outputs a multi-class model. The hypernetwork is designed to be equivariant with respect to the set of descriptions and the classification layer, thus obeying the symmetries of the problem and improving generalization. Our approach generates non-linear classifiers, handles rich textual descriptions, and may be adapted to produce lightweight models efficient enough for on-device applications. We evaluate this approach in a series of zero-shot classification tasks, for image, point-cloud, and action recognition, using a range of text descriptions: From single words to rich descriptions. Our results demonstrate strong improvements over previous approaches, showing that zero-shot learning can be applied with little training data. Furthermore, we conduct an analysis with foundational vision and language models, demonstrating that they struggle to generalize when describing what attributes the class lacks.

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

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.

Modern Natural Language Generation (NLG) models come with massive computational and storage requirements. In this work, we study the potential of compressing them, which is crucial for real-world applications serving millions of users. We focus on Knowledge Distillation (KD) techniques, in which a small student model learns to imitate a large teacher model, allowing to transfer knowledge from the teacher to the student. In contrast to much of the previous work, our goal is to optimize the model for a specific NLG task and a specific dataset. Typically in real-world applications, in addition to labeled data there is abundant unlabeled task-specific data, which is crucial for attaining high compression rates via KD. In this work, we conduct a systematic study of task-specific KD techniques for various NLG tasks under realistic assumptions. We discuss the special characteristics of NLG distillation and particularly the exposure bias problem. Following, we derive a family of Pseudo-Target (PT) augmentation methods, substantially extending prior work on sequence-level KD. We propose the Joint-Teaching method, which applies word-level KD to multiple PTs generated by both the teacher and the student. Finally, we validate our findings in an extreme setup with no labeled examples using GPT-4 as the teacher. Our study provides practical model design observations and demonstrates the effectiveness of PT training for task-specific KD in NLG.

As Natural Language Processing (NLP) algorithms continually achieve new milestones, out-of-distribution generalization remains a significant challenge. This paper addresses the issue of multi-source adaptation for unfamiliar domains: We leverage labeled data from multiple source domains to generalize to unknown target domains at training. Our innovative framework employs example-based Hypernetwork adaptation: a T5 encoder-decoder initially generates a unique signature from an input example, embedding it within the source domains’ semantic space. This signature is subsequently utilized by a Hypernetwork to generate the task classifier’s weights. In an advanced version, the signature also enriches the input example’s representation. We evaluated our method across two tasks—sentiment classification and natural language inference—in 29 adaptation scenarios, where it outpaced established algorithms. We also compare our finetuned architecture to few-shot GPT-3, demonstrating its effectiveness in essential use cases. To the best of our knowledge, this marks the first application of Hypernetworks to the adaptation for unknown domains.

This paper investigates the problem-solving capabilities of Large Language Models (LLMs) by evaluating their performance on stumpers, unique single-step intuition problems that pose challenges for human solvers but are easily verifiable. We compare the performance of four state-of-the-art LLMs (Davinci-2, Davinci-3, GPT-3.5-Turbo, GPT-4) to human participants. Our findings reveal that the new-generation LLMs excel in solving stumpers and surpass human performance. However, humans exhibit superior skills in verifying solutions to the same problems. This research enhances our understanding of LLMs’ cognitive abilities and provides insights for enhancing their problem-solving potential across various domains.

Most work on modeling the conversation history in Conversational Question Answering (CQA) reports a single main result on a common CQA benchmark. While existing models show impressive results on CQA leaderboards, it remains unclear whether they are robust to shifts in setting (sometimes to more realistic ones), training data size (e.g., from large to small sets) and domain. In this work, we design and conduct the first large-scale robustness study of history modeling approaches for CQA. We find that high benchmark scores do not necessarily translate to strong robustness, and that various methods can perform extremely differently under different settings. Equipped with the insights from our study, we design a novel prompt-based history modeling approach and demonstrate its strong robustness across various settings. Our approach is inspired by existing methods that highlight historic answers in the passage. However, instead of highlighting by modifying the passage token embeddings, we add textual prompts directly in the passage text. Our approach is simple, easy to plug into practically any model, and highly effective, thus we recommend it as a starting point for future model developers. We also hope that our study and insights will raise awareness to the importance of robustness-focused evaluation, in addition to obtaining high leaderboard scores, leading to better CQA systems.1

Natural language processing (NLP) algorithms have become very successful, but they still struggle when applied to out-of-distribution examples. In this paper we propose a controllable generation approach in order to deal with this domain adaptation (DA) challenge. Given an input text example, our DoCoGen algorithm generates a domain-counterfactual textual example (D-con) - that is similar to the original in all aspects, including the task label, but its domain is changed to a desired one. Importantly, DoCoGen is trained using only unlabeled examples from multiple domains - no NLP task labels or parallel pairs of textual examples and their domain-counterfactuals are required. We show that DoCoGen can generate coherent counterfactuals consisting of multiple sentences. We use the D-cons generated by DoCoGen to augment a sentiment classifier and a multi-label intent classifier in 20 and 78 DA setups, respectively, where source-domain labeled data is scarce. Our model outperforms strong baselines and improves the accuracy of a state-of-the-art unsupervised DA algorithm.

Persuasion games are fundamental in economics and AI research and serve as the basis for important applications. However, work on this setup assumes communication with stylized messages that do not consist of rich human language. In this paper we consider a repeated sender (expert) – receiver (decision maker) game, where the sender is fully informed about the state of the world and aims to persuade the receiver to accept a deal by sending one of several possible natural language reviews. We design an automatic expert that plays this repeated game, aiming to achieve the maximal payoff. Our expert is implemented within the Monte Carlo Tree Search (MCTS) algorithm, with deep learning models that exploit behavioral and linguistic signals in order to predict the next action of the decision maker, and the future payoff of the expert given the state of the game and a candidate review. We demonstrate the superiority of our expert over strong baselines and its adaptability to different decision makers and potential proposed deals.1

Natural Language Processing algorithms have made incredible progress, but they still struggle when applied to out-of-distribution examples. We address a challenging and underexplored version of this domain adaptation problem, where an algorithm is trained on several source domains, and then applied to examples from unseen domains that are unknown at training time. Particularly, no examples, labeled or unlabeled, or any other knowledge about the target domain are available to the algorithm at training time. We present PADA: An example-based autoregressive Prompt learning algorithm for on-the-fly Any-Domain Adaptation, based on the T5 language model. Given a test example, PADA first generates a unique prompt for it and then, conditioned on this prompt, labels the example with respect to the NLP prediction task. PADA is trained to generate a prompt that is a token sequence of unrestricted length, consisting of Domain Related Features (DRFs) that characterize each of the source domains. Intuitively, the generated prompt is a unique signature that maps the test example to a semantic space spanned by the source domains. In experiments with 3 tasks (text classification and sequence tagging), for a total of 14 multi-source adaptation scenarios, PADA substantially outperforms strong baselines.1

A fundamental goal of scientific research is to learn about causal relationships. However, despite its critical role in the life and social sciences, causality has not had the same importance in Natural Language Processing (NLP), which has traditionally placed more emphasis on predictive tasks. This distinction is beginning to fade, with an emerging area of interdisciplinary research at the convergence of causal inference and language processing. Still, research on causality in NLP remains scattered across domains without unified definitions, benchmark datasets and clear articulations of the challenges and opportunities in the application of causal inference to the textual domain, with its unique properties. In this survey, we consolidate research across academic areas and situate it in the broader NLP landscape. We introduce the statistical challenge of estimating causal effects with text, encompassing settings where text is used as an outcome, treatment, or to address confounding. In addition, we explore potential uses of causal inference to improve the robustness, fairness, and interpretability of NLP models. We thus provide a unified overview of causal inference for the NLP community.1

Multi-task learning, in which several tasks are jointly learned by a single model, allows NLP models to share information from multiple annotations and may facilitate better predictions when the tasks are inter-related. This technique, however, requires annotating the same text with multiple annotation schemes, which may be costly and laborious. Active learning (AL) has been demonstrated to optimize annotation processes by iteratively selecting unlabeled examples whose annotation is most valuable for the NLP model. Yet, multi-task active learning (MT-AL) has not been applied to state-of-the-art pre-trained Transformer-based NLP models. This paper aims to close this gap. We explore various multi-task selection criteria in three realistic multi-task scenarios, reflecting different relations between the participating tasks, and demonstrate the effectiveness of multi-task compared to single-task selection. Our results suggest that MT-AL can be effectively used in order to minimize annotation efforts for multi-task NLP models.1

Few-shot crosslingual transfer has been shown to outperform its zero-shot counterpart with pretrained encoders like multilingual BERT. Despite its growing popularity, little to no attention has been paid to standardizing and analyzing the design of few-shot experiments. In this work, we highlight a fundamental risk posed by this shortcoming, illustrating that the model exhibits a high degree of sensitivity to the selection of few shots. We conduct a large-scale experimental study on 40 sets of sampled few shots for six diverse NLP tasks across up to 40 languages. We provide an analysis of success and failure cases of few-shot transfer, which highlights the role of lexical features. Additionally, we show that a straightforward full model finetuning approach is quite effective for few-shot transfer, outperforming several state-of-the-art few-shot approaches. As a step towards standardizing few-shot crosslingual experimental designs, we make our sampled few shots publicly available.

Deep learning algorithms have shown promising results in visual question answering (VQA) tasks, but a more careful look reveals that they often do not understand the rich signal they are being fed with. To understand and better measure the generalization capabilities of VQA systems, we look at their robustness to counterfactually augmented data. Our proposed augmentations are designed to make a focused intervention on a specific property of the question such that the answer changes. Using these augmentations, we propose a new robustness measure, Robustness to Augmented Data (RAD), which measures the consistency of model predictions between original and augmented examples. Through extensive experimentation, we show that RAD, unlike classical accuracy measures, can quantify when state-of-the-art systems are not robust to counterfactuals. We find substantial failure cases which reveal that current VQA systems are still brittle. Finally, we connect between robustness and generalization, demonstrating the predictive power of RAD for performance on unseen augmentations.

Most combinations of NLP tasks and language varieties lack in-domain examples for supervised training because of the paucity of annotated data. How can neural models make sample-efficient generalizations from task–language combinations with available data to low-resource ones? In this work, we propose a Bayesian generative model for the space of neural parameters. We assume that this space can be factorized into latent variables for each language and each task. We infer the posteriors over such latent variables based on data from seen task–language combinations through variational inference. This enables zero-shot classification on unseen combinations at prediction time. For instance, given training data for named entity recognition (NER) in Vietnamese and for part-of-speech (POS) tagging in Wolof, our model can perform accurate predictions for NER in Wolof. In particular, we experiment with a typologically diverse sample of 33 languages from 4 continents and 11 families, and show that our model yields comparable or better results than state-of-the-art, zero-shot cross-lingual transfer methods. Our code is available at github.com/cambridgeltl/parameter-factorization.

Recent improvements in the predictive quality of natural language processing systems are often dependent on a substantial increase in the number of model parameters. This has led to various attempts of compressing such models, but existing methods have not considered the differences in the predictive power of various model components or in the generalizability of the compressed models. To understand the connection between model compression and out-of-distribution generalization, we define the task of compressing language representation models such that they perform best in a domain adaptation setting. We choose to address this problem from a causal perspective, attempting to estimate the average treatment effect (ATE) of a model component, such as a single layer, on the model’s predictions. Our proposed ATE-guided Model Compression scheme (AMoC), generates many model candidates, differing by the model components that were removed. Then, we select the best candidate through a stepwise regression model that utilizes the ATE to predict the expected performance on the target domain. AMoC outperforms strong baselines on dozens of domain pairs across three text classification and sequence tagging tasks.1

Understanding predictions made by deep neural networks is notoriously difficult, but also crucial to their dissemination. As all machine learning–based methods, they are as good as their training data, and can also capture unwanted biases. While there are tools that can help understand whether such biases exist, they do not distinguish between correlation and causation, and might be ill-suited for text-based models and for reasoning about high-level language concepts. A key problem of estimating the causal effect of a concept of interest on a given model is that this estimation requires the generation of counterfactual examples, which is challenging with existing generation technology. To bridge that gap, we propose CausaLM, a framework for producing causal model explanations using counterfactual language representation models. Our approach is based on fine-tuning of deep contextualized embedding models with auxiliary adversarial tasks derived from the causal graph of the problem. Concretely, we show that by carefully choosing auxiliary adversarial pre-training tasks, language representation models such as BERT can effectively learn a counterfactual representation for a given concept of interest, and be used to estimate its true causal effect on model performance. A byproduct of our method is a language representation model that is unaffected by the tested concept, which can be useful in mitigating unwanted bias ingrained in the data.1

Semi-supervised learning through deep generative models and multi-lingual pretraining techniques have orchestrated tremendous success across different areas of NLP. Nonetheless, their development has happened in isolation, while the combination of both could potentially be effective for tackling task-specific labelled data shortage. To bridge this gap, we combine semi-supervised deep generative models and multi-lingual pretraining to form a pipeline for document classification task. Compared to strong supervised learning baselines, our semi-supervised classification framework is highly competitive and outperforms the state-of-the-art counterparts in low-resource settings across several languages.

The rise of pre-trained language models has yielded substantial progress in the vast majority of Natural Language Processing (NLP) tasks. However, a generic approach towards the pre-training procedure can naturally be sub-optimal in some cases. Particularly, fine-tuning a pre-trained language model on a source domain and then applying it to a different target domain, results in a sharp performance decline of the eventual classifier for many source-target domain pairs. Moreover, in some NLP tasks, the output categories substantially differ between domains, making adaptation even more challenging. This, for example, happens in the task of aspect extraction, where the aspects of interest of reviews of, e.g., restaurants or electronic devices may be very different. This paper presents a new fine-tuning scheme for BERT, which aims to address the above challenges. We name this scheme DILBERT: Domain Invariant Learning with BERT, and customize it for aspect extraction in the unsupervised domain adaptation setting. DILBERT harnesses the categorical information of both the source and the target domains to guide the pre-training process towards a more domain and category invariant representation, thus closing the gap between the domains. We show that DILBERT yields substantial improvements over state-of-the-art baselines while using a fraction of the unlabeled data, particularly in more challenging domain adaptation setups.

We present a neural framework for learning associations between interrelated groups of words such as the ones found in Subject-Verb-Object (SVO) structures. Our model induces a joint function-specific word vector space, where vectors of e.g. plausible SVO compositions lie close together. The model retains information about word group membership even in the joint space, and can thereby effectively be applied to a number of tasks reasoning over the SVO structure. We show the robustness and versatility of the proposed framework by reporting state-of-the-art results on the tasks of estimating selectional preference and event similarity. The results indicate that the combinations of representations learned with our task-independent model outperform task-specific architectures from prior work, while reducing the number of parameters by up to 95%.

Sports competitions are widely researched in computer and social science, with the goal of understanding how players act under uncertainty. Although there is an abundance of computational work on player metrics prediction based on past performance, very few attempts to incorporate out-of-game signals have been made. Specifically, it was previously unclear whether linguistic signals gathered from players’ interviews can add information that does not appear in performance metrics. To bridge that gap, we define text classification tasks of predicting deviations from mean in NBA players’ in-game actions, which are associated with strategic choices, player behavior, and risk, using their choice of language prior to the game. We collected a data set of transcripts from key NBA players’ pre-game interviews and their in-game performance metrics, totalling 5,226 interview-metric pairs. We design neural models for players’ action prediction based on increasingly more complex aspects of the language signals in their open-ended interviews. Our models can make their predictions based on the textual signal alone, or on a combination of that signal with signals from past-performance metrics. Our text-based models outperform strong baselines trained on performance metrics only, demonstrating the importance of language usage for action prediction. Moreover, the models that utilize both textual input and past-performance metrics produced the best results. Finally, as neural networks are notoriously difficult to interpret, we propose a method for gaining further insight into what our models have learned. Particularly, we present a latent Dirichlet allocation–based analysis, where we interpret model predictions in terms of correlated topics. We find that our best performing textual model is most associated with topics that are intuitively related to each prediction task and that better models yield higher correlation with more informative topics.1

We introduce Multi-SimLex, a large-scale lexical resource and evaluation benchmark covering data sets for 12 typologically diverse languages, including major languages (e.g., Mandarin Chinese, Spanish, Russian) as well as less-resourced ones (e.g., Welsh, Kiswahili). Each language data set is annotated for the lexical relation of semantic similarity and contains 1,888 semantically aligned concept pairs, providing a representative coverage of word classes (nouns, verbs, adjectives, adverbs), frequency ranks, similarity intervals, lexical fields, and concreteness levels. Additionally, owing to the alignment of concepts across languages, we provide a suite of 66 crosslingual semantic similarity data sets. Because of its extensive size and language coverage, Multi-SimLex provides entirely novel opportunities for experimental evaluation and analysis. On its monolingual and crosslingual benchmarks, we evaluate and analyze a wide array of recent state-of-the-art monolingual and crosslingual representation models, including static and contextualized word embeddings (such as fastText, monolingual and multilingual BERT, XLM), externally informed lexical representations, as well as fully unsupervised and (weakly) supervised crosslingual word embeddings. We also present a step-by-step data set creation protocol for creating consistent, Multi-Simlex–style resources for additional languages. We make these contributions—the public release of Multi-SimLex data sets, their creation protocol, strong baseline results, and in-depth analyses which can be helpful in guiding future developments in multilingual lexical semantics and representation learning—available via a Web site that will encourage community effort in further expansion of Multi-Simlex to many more languages. Such a large-scale semantic resource could inspire significant further advances in NLP across languages.

Pivot-based neural representation models have led to significant progress in domain adaptation for NLP. However, previous research following this approach utilize only labeled data from the source domain and unlabeled data from the source and target domains, but neglect to incorporate massive unlabeled corpora that are not necessarily drawn from these domains. To alleviate this, we propose PERL: A representation learning model that extends contextualized word embedding models such as BERT (Devlin et al., 2019) with pivot-based fine-tuning. PERL outperforms strong baselines across 22 sentiment classification domain adaptation setups, improves in-domain model performance, yields effective reduced-size models, and increases model stability.1

Sentence fusion is the task of joining related sentences into coherent text. Current training and evaluation schemes for this task are based on single reference ground-truths and do not account for valid fusion variants. We show that this hinders models from robustly capturing the semantic relationship between input sentences. To alleviate this, we present an approach in which ground-truth solutions are automatically expanded into multiple references via curated equivalence classes of connective phrases. We apply this method to a large-scale dataset and use the augmented dataset for both model training and evaluation. To improve the learning of semantic representation using multiple references, we enrich the model with auxiliary discourse classification tasks under a multi-tasking framework. Our experiments highlight the improvements of our approach over state-of-the-art models.

Performance in cross-lingual NLP tasks is impacted by the (dis)similarity of languages at hand: e.g., previous work has suggested there is a connection between the expected success of bilingual lexicon induction (BLI) and the assumption of (approximate) isomorphism between monolingual embedding spaces. In this work we present a large-scale study focused on the correlations between monolingual embedding space similarity and task performance, covering thousands of language pairs and four different tasks: BLI, parsing, POS tagging and MT. We hypothesize that statistics of the spectrum of each monolingual embedding space indicate how well they can be aligned. We then introduce several isomorphism measures between two embedding spaces, based on the relevant statistics of their individual spectra. We empirically show that (1) language similarity scores derived from such spectral isomorphism measures are strongly associated with performance observed in different cross-lingual tasks, and (2) our spectral-based measures consistently outperform previous standard isomorphism measures, while being computationally more tractable and easier to interpret. Finally, our measures capture complementary information to typologically driven language distance measures, and the combination of measures from the two families yields even higher task performance correlations.

Comparing between Deep Neural Network (DNN) models based on their performance on unseen data is crucial for the progress of the NLP field. However, these models have a large number of hyper-parameters and, being non-convex, their convergence point depends on the random values chosen at initialization and during training. Proper DNN comparison hence requires a comparison between their empirical score distributions on unseen data, rather than between single evaluation scores as is standard for more simple, convex models. In this paper, we propose to adapt to this problem a recently proposed test for the Almost Stochastic Dominance relation between two distributions. We define the criteria for a high quality comparison method between DNNs, and show, both theoretically and through analysis of extensive experimental results with leading DNN models for sequence tagging tasks, that the proposed test meets all criteria while previously proposed methods fail to do so. We hope the test we propose here will set a new working practice in the NLP community.

We consider a zero-shot semantic parsing task: parsing instructions into compositional logical forms, in domains that were not seen during training. We present a new dataset with 1,390 examples from 7 application domains (e.g. a calendar or a file manager), each example consisting of a triplet: (a) the application’s initial state, (b) an instruction, to be carried out in the context of that state, and (c) the state of the application after carrying out the instruction. We introduce a new training algorithm that aims to train a semantic parser on examples from a set of source domains, so that it can effectively parse instructions from an unknown target domain. We integrate our algorithm into the floating parser of Pasupat and Liang (2015), and further augment the parser with features and a logical form candidate filtering logic, to support zero-shot adaptation. Our experiments with various zero-shot adaptation setups demonstrate substantial performance gains over a non-adapted parser.

Pivot Based Language Modeling (PBLM) (Ziser and Reichart, 2018a), combining LSTMs with pivot-based methods, has yielded significant progress in unsupervised domain adaptation. However, this approach is still challenged by the large pivot detection problem that should be solved, and by the inherent instability of LSTMs. In this paper we propose a Task Refinement Learning (TRL) approach, in order to solve these problems. Our algorithms iteratively train the PBLM model, gradually increasing the information exposed about each pivot. TRL-PBLM achieves stateof- the-art accuracy in six domain adaptation setups for sentiment classification. Moreover, it is much more stable than plain PBLM across model configurations, making the model much better fitted for practical use.

The best solution of structured prediction models in NLP is often inaccurate because of limited expressive power of the model or to non-exact parameter estimation. One way to mitigate this problem is sampling candidate solutions from the model’s solution space, reasoning that effective exploration of this space should yield high-quality solutions. Unfortunately, sampling is often computationally hard and many works hence back-off to sub-optimal strategies, such as extraction of the best scoring solutions of the model, which are not as diverse as sampled solutions. In this paper we propose a perturbation-based approach where sampling from a probabilistic model is computationally efficient. We present a learning algorithm for the variance of the perturbations, and empirically demonstrate its importance. Moreover, while finding the argmax in our model is intractable, we propose an efficient and effective approximation. We apply our framework to cross-lingual dependency parsing across 72 corpora from 42 languages and to lightly supervised dependency parsing across 13 corpora from 12 languages, and demonstrate strong results in terms of both the quality of the entire solution list and of the final solution.1

Neural dependency parsing has proven very effective, achieving state-of-the-art results on numerous domains and languages. Unfortunately, it requires large amounts of labeled data, which is costly and laborious to create. In this paper we propose a self-training algorithm that alleviates this annotation bottleneck by training a parser on its own output. Our Deep Contextualized Self-training (DCST) algorithm utilizes representation models trained on sequence labeling tasks that are derived from the parser’s output when applied to unlabeled data, and integrates these models with the base parser through a gating mechanism. We conduct experiments across multiple languages, both in low resource in-domain and in cross-domain setups, and demonstrate that DCST substantially outperforms traditional self-training as well as recent semi-supervised training methods.1

Recent work has validated the importance of subword information for word representation learning. Since subwords increase parameter sharing ability in neural models, their value should be even more pronounced in low-data regimes. In this work, we therefore provide a comprehensive analysis focused on the usefulness of subwords for word representation learning in truly low-resource scenarios and for three representative morphological tasks: fine-grained entity typing, morphological tagging, and named entity recognition. We conduct a systematic study that spans several dimensions of comparison: 1) type of data scarcity which can stem from the lack of task-specific training data, or even from the lack of unannotated data required to train word embeddings, or both; 2) language type by working with a sample of 16 typologically diverse languages including some truly low-resource ones (e.g. Rusyn, Buryat, and Zulu); 3) the choice of the subword-informed word representation method. Our main results show that subword-informed models are universally useful across all language types, with large gains over subword-agnostic embeddings. They also suggest that the effective use of subwords largely depends on the language (type) and the task at hand, as well as on the amount of available data for training the embeddings and task-based models, where having sufficient in-task data is a more critical requirement.

While neural dependency parsers provide state-of-the-art accuracy for several languages, they still rely on large amounts of costly labeled training data. We demonstrate that in the small data regime, where uncertainty around parameter estimation and model prediction matters the most, Bayesian neural modeling is very effective. In order to overcome the computational and statistical costs of the approximate inference step in this framework, we utilize an efficient sampling procedure via stochastic gradient Langevin dynamics to generate samples from the approximated posterior. Moreover, we show that our Bayesian neural parser can be further improved when integrated into a multi-task parsing and POS tagging framework, designed to minimize task interference via an adversarial procedure. When trained and tested on 6 languages with less than 5k training instances, our parser consistently outperforms the strong bilstm baseline (Kiperwasser and Goldberg, 2016). Compared with the biaffine parser (Dozat et al., 2017) our model achieves an improvement of up to 3% for Vietnames and Irish, while our multi-task model achieves an improvement of up to 9% across five languages: Farsi, Russian, Turkish, Vietnamese, and Irish.

Semantic specialization integrates structured linguistic knowledge from external resources (such as lexical relations in WordNet) into pretrained distributional vectors in the form of constraints. However, this technique cannot be leveraged in many languages, because their structured external resources are typically incomplete or non-existent. To bridge this gap, we propose a novel method that transfers specialization from a resource-rich source language (English) to virtually any target language. Our specialization transfer comprises two crucial steps: 1) Inducing noisy constraints in the target language through automatic word translation; and 2) Filtering the noisy constraints via a state-of-the-art relation prediction model trained on the source language constraints. This allows us to specialize any set of distributional vectors in the target language with the refined constraints. We prove the effectiveness of our method through intrinsic word similarity evaluation in 8 languages, and with 3 downstream tasks in 5 languages: lexical simplification, dialog state tracking, and semantic textual similarity. The gains over the previous state-of-art specialization methods are substantial and consistent across languages. Our results also suggest that the transfer method is effective even for lexically distant source-target language pairs. Finally, as a by-product, our method produces lists of WordNet-style lexical relations in resource-poor languages.

Can we construct a neural language model which is inductively biased towards learning human language? Motivated by this question, we aim at constructing an informative prior for held-out languages on the task of character-level, open-vocabulary language modelling. We obtain this prior as the posterior over network weights conditioned on the data from a sample of training languages, which is approximated through Laplace’s method. Based on a large and diverse sample of languages, the use of our prior outperforms baseline models with an uninformative prior in both zero-shot and few-shot settings, showing that the prior is imbued with universal linguistic knowledge. Moreover, we harness broad language-specific information available for most languages of the world, i.e., features from typological databases, as distant supervision for held-out languages. We explore several language modelling conditioning techniques, including concatenation and meta-networks for parameter generation. They appear beneficial in the few-shot setting, but ineffective in the zero-shot setting. Since the paucity of even plain digital text affects the majority of the world’s languages, we hope that these insights will broaden the scope of applications for language technology.

Recent efforts in cross-lingual word embedding (CLWE) learning have predominantly focused on fully unsupervised approaches that project monolingual embeddings into a shared cross-lingual space without any cross-lingual signal. The lack of any supervision makes such approaches conceptually attractive. Yet, their only core difference from (weakly) supervised projection-based CLWE methods is in the way they obtain a seed dictionary used to initialize an iterative self-learning procedure. The fully unsupervised methods have arguably become more robust, and their primary use case is CLWE induction for pairs of resource-poor and distant languages. In this paper, we question the ability of even the most robust unsupervised CLWE approaches to induce meaningful CLWEs in these more challenging settings. A series of bilingual lexicon induction (BLI) experiments with 15 diverse languages (210 language pairs) show that fully unsupervised CLWE methods still fail for a large number of language pairs (e.g., they yield zero BLI performance for 87/210 pairs). Even when they succeed, they never surpass the performance of weakly supervised methods (seeded with 500-1,000 translation pairs) using the same self-learning procedure in any BLI setup, and the gaps are often substantial. These findings call for revisiting the main motivations behind fully unsupervised CLWE methods.

Linguistic typology aims to capture structural and semantic variation across the world’s languages. A large-scale typology could provide excellent guidance for multilingual Natural Language Processing (NLP), particularly for languages that suffer from the lack of human labeled resources. We present an extensive literature survey on the use of typological information in the development of NLP techniques. Our survey demonstrates that to date, the use of information in existing typological databases has resulted in consistent but modest improvements in system performance. We show that this is due to both intrinsic limitations of databases (in terms of coverage and feature granularity) and under-utilization of the typological features included in them. We advocate for a new approach that adapts the broad and discrete nature of typological categories to the contextual and continuous nature of machine learning algorithms used in contemporary NLP. In particular, we suggest that such an approach could be facilitated by recent developments in data-driven induction of typological knowledge.

Representation learning with pivot-based methods and with Neural Networks (NNs) have lead to significant progress in domain adaptation for Natural Language Processing. However, most previous work that follows these approaches does not explicitly exploit the structure of the input text, and its output is most often a single representation vector for the entire text. In this paper we present the Pivot Based Language Model (PBLM), a representation learning model that marries together pivot-based and NN modeling in a structure aware manner. Particularly, our model processes the information in the text with a sequential NN (LSTM) and its output consists of a representation vector for every input word. Unlike most previous representation learning models in domain adaptation, PBLM can naturally feed structure aware text classifiers such as LSTM and CNN. We experiment with the task of cross-domain sentiment classification on 20 domain pairs and show substantial improvements over strong baselines.

We present a deep neural network that leverages images to improve bilingual text embeddings. Relying on bilingual image tags and descriptions, our approach conditions text embedding induction on the shared visual information for both languages, producing highly correlated bilingual embeddings. In particular, we propose a novel model based on Partial Canonical Correlation Analysis (PCCA). While the original PCCA finds linear projections of two views in order to maximize their canonical correlation conditioned on a shared third variable, we introduce a non-linear Deep PCCA (DPCCA) model, and develop a new stochastic iterative algorithm for its optimization. We evaluate PCCA and DPCCA on multilingual word similarity and cross-lingual image description retrieval. Our models outperform a large variety of previous methods, despite not having access to any visual signal during test time inference.

Statistical significance testing is a standard statistical tool designed to ensure that experimental results are not coincidental. In this opinion/ theoretical paper we discuss the role of statistical significance testing in Natural Language Processing (NLP) research. We establish the fundamental concepts of significance testing and discuss the specific aspects of NLP tasks, experimental setups and evaluation measures that affect the choice of significance tests in NLP research. Based on this discussion we propose a simple practical protocol for statistical significance test selection in NLP setups and accompany this protocol with a brief survey of the most relevant tests. We then survey recent empirical papers published in ACL and TACL during 2017 and show that while our community assigns great value to experimental results, statistical significance testing is often ignored or misused. We conclude with a brief discussion of open issues that should be properly addressed so that this important tool can be applied. in NLP research in a statistically sound manner.

The transfer or share of knowledge between languages is a potential solution to resource scarcity in NLP. However, the effectiveness of cross-lingual transfer can be challenged by variation in syntactic structures. Frameworks such as Universal Dependencies (UD) are designed to be cross-lingually consistent, but even in carefully designed resources trees representing equivalent sentences may not always overlap. In this paper, we measure cross-lingual syntactic variation, or anisomorphism, in the UD treebank collection, considering both morphological and structural properties. We show that reducing the level of anisomorphism yields consistent gains in cross-lingual transfer tasks. We introduce a source language selection procedure that facilitates effective cross-lingual parser transfer, and propose a typologically driven method for syntactic tree processing which reduces anisomorphism. Our results show the effectiveness of this method for both machine translation and cross-lingual sentence similarity, demonstrating the importance of syntactic structure compatibility for boosting cross-lingual transfer in NLP.

Neural architectures are prominent in the construction of language models (LMs). However, word-level prediction is typically agnostic of subword-level information (characters and character sequences) and operates over a closed vocabulary, consisting of a limited word set. Indeed, while subword-aware models boost performance across a variety of NLP tasks, previous work did not evaluate the ability of these models to assist next-word prediction in language modeling tasks. Such subword-level informed models should be particularly effective for morphologically-rich languages (MRLs) that exhibit high type-to-token ratios. In this work, we present a large-scale LM study on 50 typologically diverse languages covering a wide variety of morphological systems, and offer new LM benchmarks to the community, while considering subword-level information. The main technical contribution of our work is a novel method for injecting subword-level information into semantic word vectors, integrated into the neural language modeling training, to facilitate word-level prediction. We conduct experiments in the LM setting where the number of infrequent words is large, and demonstrate strong perplexity gains across our 50 languages, especially for morphologically-rich languages. Our code and data sets are publicly available.

While cross-domain and cross-language transfer have long been prominent topics in NLP research, their combination has hardly been explored. In this work we consider this problem, and propose a framework that builds on pivot-based learning, structure-aware Deep Neural Networks (particularly LSTMs and CNNs) and bilingual word embeddings, with the goal of training a model on labeled data from one (language, domain) pair so that it can be effectively applied to another (language, domain) pair. We consider two setups, differing with respect to the unlabeled data available for model training. In the full setup the model has access to unlabeled data from both pairs, while in the lazy setup, which is more realistic for truly resource-poor languages, unlabeled data is available for both domains but only for the source language. We design our model for the lazy setup so that for a given target domain, it can train once on the source language and then be applied to any target language without re-training. In experiments with nine English-German and nine English-French domain pairs our best model substantially outperforms previous models even when it is trained in the lazy setup and previous models are trained in the full setup.

A key challenge in cross-lingual NLP is developing general language-independent architectures that are equally applicable to any language. However, this ambition is largely hampered by the variation in structural and semantic properties, i.e. the typological profiles of the world’s languages. In this work, we analyse the implications of this variation on the language modeling (LM) task. We present a large-scale study of state-of-the art n-gram based and neural language models on 50 typologically diverse languages covering a wide variety of morphological systems. Operating in the full vocabulary LM setup focused on word-level prediction, we demonstrate that a coarse typology of morphological systems is predictive of absolute LM performance. Moreover, fine-grained typological features such as exponence, flexivity, fusion, and inflectional synthesis are borne out to be responsible for the proliferation of low-frequency phenomena which are organically difficult to model by statistical architectures, or for the meaning ambiguity of character n-grams. Our study strongly suggests that these features have to be taken into consideration during the construction of next-level language-agnostic LM architectures, capable of handling morphologically complex languages such as Tamil or Korean.

Web queries with question intent manifest a complex syntactic structure and the processing of this structure is important for their interpretation. Pinter et al. (2016) has formalized the grammar of these queries and proposed semi-supervised algorithms for the adaptation of parsers originally designed to parse according to the standard dependency grammar, so that they can account for the unique forest grammar of queries. However, their algorithms rely on resources typically not available outside of big web corporates. We propose a new BiLSTM query parser that: (1) Explicitly accounts for the unique grammar of web queries; and (2) Utilizes named entity (NE) information from a BiLSTM NE tagger, that can be jointly trained with the parser. In order to train our model we annotate the query treebank of Pinter et al. (2016) with NEs. When trained on 2500 annotated queries our parser achieves UAS of 83.5% and segmentation F1-score of 84.5, substantially outperforming existing state-of-the-art parsers.

This paper is concerned with identifying contexts useful for training word representation models for different word classes such as adjectives (A), verbs (V), and nouns (N). We introduce a simple yet effective framework for an automatic selection of class-specific context configurations. We construct a context configuration space based on universal dependency relations between words, and efficiently search this space with an adapted beam search algorithm. In word similarity tasks for each word class, we show that our framework is both effective and efficient. Particularly, it improves the Spearman’s rho correlation with human scores on SimLex-999 over the best previously proposed class-specific contexts by 6 (A), 6 (V) and 5 (N) rho points. With our selected context configurations, we train on only 14% (A), 26.2% (V), and 33.6% (N) of all dependency-based contexts, resulting in a reduced training time. Our results generalise: we show that the configurations our algorithm learns for one English training setup outperform previously proposed context types in another training setup for English. Moreover, basing the configuration space on universal dependencies, it is possible to transfer the learned configurations to German and Italian. We also demonstrate improved per-class results over other context types in these two languages..

We introduce a neural network model that marries together ideas from two prominent strands of research on domain adaptation through representation learning: structural correspondence learning (SCL, (Blitzer et al., 2006)) and autoencoder neural networks (NNs). Our model is a three-layer NN that learns to encode the non-pivot features of an input example into a low dimensional representation, so that the existence of pivot features (features that are prominent in both domains and convey useful information for the NLP task) in the example can be decoded from that representation. The low-dimensional representation is then employed in a learning algorithm for the task. Moreover, we show how to inject pre-trained word embeddings into our model in order to improve generalization across examples with similar pivot features. We experiment with the task of cross-domain sentiment classification on 16 domain pairs and show substantial improvements over strong baselines.

We present Attract-Repel, an algorithm for improving the semantic quality of word vectors by injecting constraints extracted from lexical resources. Attract-Repel facilitates the use of constraints from mono- and cross-lingual resources, yielding semantically specialized cross-lingual vector spaces. Our evaluation shows that the method can make use of existing cross-lingual lexicons to construct high-quality vector spaces for a plethora of different languages, facilitating semantic transfer from high- to lower-resource ones. The effectiveness of our approach is demonstrated with state-of-the-art results on semantic similarity datasets in six languages. We next show that Attract-Repel-specialized vectors boost performance in the downstream task of dialogue state tracking (DST) across multiple languages. Finally, we show that cross-lingual vector spaces produced by our algorithm facilitate the training of multilingual DST models, which brings further performance improvements.

With the ever growing amount of textual data from a large variety of languages, domains, and genres, it has become standard to evaluate NLP algorithms on multiple datasets in order to ensure a consistent performance across heterogeneous setups. However, such multiple comparisons pose significant challenges to traditional statistical analysis methods in NLP and can lead to erroneous conclusions. In this paper we propose a Replicability Analysis framework for a statistically sound analysis of multiple comparisons between algorithms for NLP tasks. We discuss the theoretical advantages of this framework over the current, statistically unjustified, practice in the NLP literature, and demonstrate its empirical value across four applications: multi-domain dependency parsing, multilingual POS tagging, cross-domain sentiment classification and word similarity prediction.

Morphologically rich languages accentuate two properties of distributional vector space models: 1) the difficulty of inducing accurate representations for low-frequency word forms; and 2) insensitivity to distinct lexical relations that have similar distributional signatures. These effects are detrimental for language understanding systems, which may infer that ‘inexpensive’ is a rephrasing for ‘expensive’ or may not associate ‘acquire’ with ‘acquires’. In this work, we propose a novel morph-fitting procedure which moves past the use of curated semantic lexicons for improving distributional vector spaces. Instead, our method injects morphological constraints generated using simple language-specific rules, pulling inflectional forms of the same word close together and pushing derivational antonyms far apart. In intrinsic evaluation over four languages, we show that our approach: 1) improves low-frequency word estimates; and 2) boosts the semantic quality of the entire word vector collection. Finally, we show that morph-fitted vectors yield large gains in the downstream task of dialogue state tracking, highlighting the importance of morphology for tackling long-tail phenomena in language understanding tasks.

Sarcasm is a form of speech in which speakers say the opposite of what they truly mean in order to convey a strong sentiment. In other words, “Sarcasm is the giant chasm between what I say, and the person who doesn’t get it.”. In this paper we present the novel task of sarcasm interpretation, defined as the generation of a non-sarcastic utterance conveying the same message as the original sarcastic one. We introduce a novel dataset of 3000 sarcastic tweets, each interpreted by five human judges. Addressing the task as monolingual machine translation (MT), we experiment with MT algorithms and evaluation measures. We then present SIGN: an MT based sarcasm interpretation algorithm that targets sentiment words, a defining element of textual sarcasm. We show that while the scores of n-gram based automatic measures are similar for all interpretation models, SIGN’s interpretations are scored higher by humans for adequacy and sentiment polarity. We conclude with a discussion on future research directions for our new task.

In recent years linguistic typologies, which classify the world’s languages according to their functional and structural properties, have been widely used to support multilingual NLP. While the growing importance of typologies in supporting multilingual tasks has been recognised, no systematic survey of existing typological resources and their use in NLP has been published. This paper provides such a survey as well as discussion which we hope will both inform and inspire future work in the area.

Inferring the information structure of scientific documents is useful for many NLP applications. Existing approaches to this task require substantial human effort. We propose a framework for constraint learning that reduces human involvement considerably. Our model uses topic models to identify latent topics and their key linguistic features in input documents, induces constraints from this information and maps sentences to their dominant information structure categories through a constrained unsupervised model. When the induced constraints are combined with a fully unsupervised model, the resulting model challenges existing lightly supervised feature-based models as well as unsupervised models that use manually constructed declarative knowledge. Our results demonstrate that useful declarative knowledge can be learned from data with very limited human involvement.

Multi-modal models that learn semantic representations from both linguistic and perceptual input outperform language-only models on a range of evaluations, and better reflect human concept acquisition. Most perceptual input to such models corresponds to concrete noun concepts and the superiority of the multi-modal approach has only been established when evaluating on such concepts. We therefore investigate which concepts can be effectively learned by multi-modal models. We show that concreteness determines both which linguistic features are most informative and the impact of perceptual input in such models. We then introduce ridge regression as a means of propagating perceptual information from concrete nouns to more abstract concepts that is more robust than previous approaches. Finally, we present weighted gram matrix combination, a means of combining representations from distinct modalities that outperforms alternatives when both modalities are sufficiently rich.