Papers by Alex Warstadt
Are Natural Language Inference Models IMPPRESsive? Learning IMPlicature and PRESupposition (2020.acl-main)
Copied to clipboard
| Challenge: | Natural language inference (NLI) is an increasingly important task for natural language understanding . however, the ability of NLI models to make pragmatic inferences remains understudied . |
| Approach: | They use semi-automatically generated sentence pairs to evaluate whether NLI models make pragmatic inferences. |
| Outcome: | The proposed model trains on multiNLI and shows that it learns to draw pragmatic inferences. |
Investigating Critical Period Effects in Language Acquisition through Neural Language Models (2025.tacl-1)
Copied to clipboard
| Challenge: | Scholars of human development have long debated whether these phenomena are predetermined by innately encoded developmental changes in the maturing brain or natural consequences of increased experience. |
| Approach: | They use language models to test whether CP effects are peculiar to humans . they find that LMs do not show CP when L2 exposure is delayed . scholars have long debated whether innate maturation changes predetermine CP . |
| Outcome: | The proposed model does not show CP effects when the age of exposure of L2 is delayed. |
CLiMP: A Benchmark for Chinese Language Model Evaluation (2021.eacl-main)
Copied to clipboard
| Challenge: | Linguistically informed analyses of language models (LMs) contribute to understanding and improvement of such models. |
| Approach: | They introduce a corpus of Chinese linguistic minimal pairs (CLiMP) to investigate what knowledge Chinese LMs acquire. |
| Outcome: | The proposed corpus of Chinese linguistic minimal pairs (CLiMP) covers 9 major Chinese linguist phenomena. |
It’s Not What You Say, It’s How You Say It: Evaluating LLM Responses to Expressions of Belief (2026.acl-long)
Copied to clipboard
| Challenge: | a typology is grounded in four linguistically motivated dimensions: form, evidentiality, epistemic stance, and tone. |
| Approach: | They propose a typology to evaluate how different EoBs affect whether models follow context versus prior knowledge. |
| Outcome: | The proposed model systematically evaluates 16 LLMs that differ in architecture, scale, and training stages . human listeners subconsciously interpret the belief based on how it is expressed, i.e., its explicitness, tone, or contextual cues. |
BLiMP: The Benchmark of Linguistic Minimal Pairs for English (2020.tacl-1)
Copied to clipboard
Alex Warstadt, Alicia Parrish, Haokun Liu, Anhad Mohananey, Wei Peng, Sheng-Fu Wang, Samuel R. Bowman
| Challenge: | Recent studies have examined how linguistic knowledge of language models (LMs) varies across English phenomena. |
| Approach: | They propose a benchmark to evaluate linguistic knowledge of language models on major grammatical phenomena in English. |
| Outcome: | The proposed benchmark evaluates the linguistic knowledge of language models on major grammatical phenomena in English. |
What Makes Reading Comprehension Questions Difficult? (2022.acl-long)
Copied to clipboard
| Challenge: | a recent study shows that natural language understanding benchmarks are not able to measure future progress . a crowdsourcing approach is needed to collect diverse examples without sacrificing diversity or coverage. |
| Approach: | They crowdsource multiple-choice reading comprehension questions for passages from seven sources . they find passage source, length, and readability measures do not significantly affect question difficulty . |
| Outcome: | The results show that passage source, length, and readability measures do not significantly affect question difficulty. |
Surprise! Uniform Information Density Isn’t the Whole Story: Predicting Surprisal Contours in Long-form Discourse (2024.emnlp-main)
Copied to clipboard
| Challenge: | Uniform Information Density (UID) hypothesis posits that speakers tend to distribute information evenly across linguistic units to achieve efficient communication. |
| Approach: | They propose a functional pressure that speakers modulate information rate based on location within a hierarchically-structured model of discourse. |
| Outcome: | The proposed hypothesis posits that speakers tend to distribute information evenly across linguistic units to achieve efficient communication. |
The time scale of redundancy between prosody and linguistic context (2025.acl-long)
Copied to clipboard
Tamar I Regev, Chiebuka Ohams, Shaylee Xie, Lukas Wolf, Evelina Fedorenko, Alex Warstadt, Ethan Wilcox, Tiago Pimentel
| Challenge: | Prior work has shown that the information carried by prosodic features is substantially redundant with that carried by the surrounding words. |
| Approach: | They examine the time scale of this relationship, studying how it varies with the length of past and future contexts. |
| Outcome: | The results show that prosody features show some redundancy with future words, but only with a short scale of 1-2 words, consistent with reports of incremental short-term planning in language production. |
A Distributional Perspective on Word Learning in Neural Language Models (2025.naacl-long)
Copied to clipboard
| Challenge: | Language models are increasingly being studied as models of human language learners. |
| Approach: | They propose a distributional approach to word learning that captures distributional knowledge and gradient preferences for the word’s appropriateness. |
| Outcome: | The proposed signatures capture knowledge of where the target word can and cannot occur as well as gradient preferences about the word’s appropriateness. |
The Harmonic Structure of Information Contours (2025.acl-long)
Copied to clipboard
Eleftheria Tsipidi, Samuel Kiegeland, Franz Nowak, Tianyang Xu, Ethan Wilcox, Alex Warstadt, Ryan Cotterell, Mario Giulianelli
| Challenge: | Language typically does not maintain a uniform information rate, but it fluctuates around a global average . a new study suggests periodicity may be a factor in information rate oscillations . |
| Approach: | They propose a hypothesis that language does not maintain a uniform information rate . they apply harmonic regression and introduce a new extension to detect periodicity . |
| Outcome: | The proposed method reveals that language oscillates at periodic intervals across frequencies . it also offers a framework for uncovering structural pressures at various levels of linguistic granularity. |
Investigating BERT’s Knowledge of Language: Five Analysis Methods with NPIs (D19-1)
Copied to clipboard
Alex Warstadt, Yu Cao, Ioana Grosu, Wei Peng, Hagen Blix, Yining Nie, Anna Alsop, Shikha Bordia, Haokun Liu, Alicia Parrish, Sheng-Fu Wang, Jason Phang, Anhad Mohananey, Phu Mon Htut, Paloma Jeretic, Samuel R. Bowman
| Challenge: | Recent work evaluating sentence representation models' knowledge of grammar has been slower to emerge. |
| Approach: | They propose five experimental methods inspired by prior work evaluating pretrained sentence representation models to examine their grammatical knowledge. |
| Outcome: | The proposed methods show that the model has significant knowledge of the licensing environment but its success varies widely across different methods. |
What Ingredients Make for an Effective Crowdsourcing Protocol for Difficult NLU Data Collection Tasks? (2021.acl-long)
Copied to clipboard
| Challenge: | Despite the importance of datasets for natural language understanding, there has been little attention on crowdsourcing methods for collecting datasets. |
| Approach: | They compare the effectiveness of crowdsourcing methods for boosting NLU example difficulty with training crowdworkers instead of expert judgments. |
| Outcome: | The proposed method is ineffective for boosting NLU example difficulty, but it is not effective for training crowdworkers and qualifying workers based on expert judgments. |
Can Language Models Learn Typologically Implausible Languages? (2026.tacl-1)
Copied to clipboard
| Challenge: | Language models provide a naturalistic framework for studying artificial language learning . authors: typological universals and tendencies are thought to be caused by a learning bias . |
| Approach: | They propose to train LMs on highly naturalistic counterfactual versions of English and Japanese . they show that LM learn subtly implausible languages more slowly . |
| Outcome: | The proposed language models learn subtly implausible languages more slowly compared to human models . the findings suggest that LMs exhibit typologically aligned learning preferences . |
Using Information Theory to Characterize Prosodic Typology: The Case of Tone, Pitch-Accent and Stress-Accent (2025.acl-long)
Copied to clipboard
| Challenge: | lexical identity and prosody are well-studied parameters of linguistic variation, but they are difficult to predict in tonal languages. |
| Approach: | They propose to characterize the relationship between lexical identity and prosody using information theory to estimate mutual information between the text and pitch curves. |
| Outcome: | The proposed hypothesis supports perspectives that view linguistic typology as gradient, rather than categorical. |
Quantifying the redundancy between prosody and text (2023.emnlp-main)
Copied to clipboard
Lukas Wolf, Tiago Pimentel, Evelina Fedorenko, Ryan Cotterell, Alex Warstadt, Ethan Wilcox, Tamar Regev
| Challenge: | Existing studies suggest partial redundancy between prosody and linguistic information. |
| Approach: | They use large language models to estimate how much information is redundant between prosody and the words themselves. |
| Outcome: | The proposed model can predict prosodic features across prosodic features, including intensity, duration, pauses, and pitch contours. |
What Do Prosody and Text Convey? Characterizing How Meaningful Information is Distributed Across Multiple Channels (2026.acl-long)
Copied to clipboard
| Challenge: | Prosody—the melody of speech—conveys critical information often not captured by the words or text of a message. |
| Approach: | They propose an information-theoretic approach to quantify how much is conveyed by prosody that is not recoverable from text alone. |
| Outcome: | The proposed framework can quantify how much is conveyed by prosody that is not recoverable from text alone and crucially, what prosody conveys. |
Learning Which Features Matter: RoBERTa Acquires a Preference for Linguistic Generalizations (Eventually) (2020.emnlp-main)
Copied to clipboard
| Challenge: | Pretraining on self-supervised linguistic tasks is effective for learning features helpful for language understanding, but it requires more data to learn to prefer linguistic generalizations over surface ones. |
| Approach: | They propose a set of 20 ambiguous binary classification tasks to test whether a pretrained model prefers linguistic or surface generalizations. |
| Outcome: | The proposed model can learn to represent linguistic features with little pretraining data, but requires far more data to learn to prefer linguistic generalizations over surface ones. |
Dual Alignment Between Language Model Layers and Human Sentence Processing (2026.acl-long)
Copied to clipboard
| Challenge: | Existing studies have demonstrated both the successes and limitations of accurate predictability estimation by modern LMs in cognitive modeling. |
| Approach: | They propose to use internal layers to better estimate human cognitive effort observed in syntactic ambiguity processing in English. |
| Outcome: | The proposed models can be modeled using surprisal from early layers of large language models (LLMs) this raises the question whether such advantages extend to more syntactically challenging constructions, where surprised estimates underestimate human cognitive effort. |
When Do You Need Billions of Words of Pretraining Data? (2021.acl-long)
Copied to clipboard
| Challenge: | Pretrained language models (LMs) are dominated by models that can encode billions of words. |
| Approach: | They use classifier probing, information-theoretic probing and unsupervised relative acceptability judgments to evaluate model ability. |
| Outcome: | The proposed models require only about 10M to 100M words to learn to encode most syntactic and semantic features. |
Generalizing Backpropagation for Gradient-Based Interpretability (2023.acl-long)
Copied to clipboard
| Challenge: | Several feature-attribution methods for interpreting deep neural networks rely on computing the gradients of a model’s output with respect to its inputs, but they reveal little about the inner workings of the model itself. |
| Approach: | They propose a generalized backpropagation algorithm that generalizes the gradient computation of a model to efficiently compute other interpretable statistics about the gradient graph of neural networks. |
| Outcome: | The proposed generalized algorithm can be used to compute other interpretable statistics about the gradient graph of a neural network, such as the highest-weighted path and entropy. |
Reconstruction Probing (2023.findings-acl)
Copied to clipboard
| Challenge: | a new analysis method for contextualized representations is proposed . contextualization boosts reconstructability of tokens close to the token being reconstructed . |
| Approach: | They propose a method for contextualized representations based on reconstruction probabilities in masked language models. |
| Outcome: | The proposed method compares reconstruction probabilities of tokens in masked language models . it finds that contextualization boosts reconstructability of token that are close to the token being reconstructed . |
BabyBabelLM: A Multilingual Benchmark of Developmentally Plausible Training Data (2026.eacl-long)
Copied to clipboard
Jaap Jumelet, Abdellah Fourtassi, Akari Haga, Bastian Bunzeck, Bhargav Shandilya, Diana Galvan-Sosa, Faiz Ghifari Haznitrama, Francesca Padovani, Francois Meyer, Hai Hu, Julen Etxaniz, Laurent Prevot, Linyang He, María Grandury, Mila Marcheva, Negar Foroutan, Nikitas Theodoropoulos, Pouya Sadeghi, Siyuan Song, Suchir Salhan, Susana Zhou, Yurii Paniv, Ziyin Zhang, Arianna Bisazza, Alex Warstadt, Leshem Choshen
| Challenge: | prevailing trend in language modeling research is to prioritize scaling, authors say . from infancy to maturity, English learners acquire language through exposure to less than 100M words . |
| Approach: | They propose a multilingual collection of datasets modeling the language a person observes from birth until they acquire a native language. |
| Outcome: | The proposed models outperform models trained on a fixed, developmentally plausible English corpus on various benchmarks. |
Automatic Annotation of Grammaticality in Child-Caregiver Conversations (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing methods for analyzing child language acquisition have been tedious and inconsistent. |
| Approach: | They propose a coding scheme for context-dependent grammaticality in child-caregiver conversations and annotate 4,000 utterances from a large corpus of transcribed conversations. |
| Outcome: | The proposed method achieves human inter-annotation agreement levels and is faster and reproducible than manual methods. |