Publications
2024
- Ghostbuster: Detecting Text Ghostwritten by Large Language ModelsProceedings of NAACL2024. [paper] [bib]
- Autonomous Evaluation and Refinement of Digital AgentsProceedings of COLM2024. [paper] [bib]
- Decision-Oriented Dialogue for Human-AI CollaborationTransactions of the ACL2024. [paper] [bib]
- Re-evaluating the Need for Visual Signals in Unsupervised Grammar InductionProceedings of NAACL, Findings2024. [paper] [bib]
- The Cortical Representation of Language Timescales is Shared between Reading and ListeningCommunications Biology2024. [paper] [bib]
2023
- DOC: Improving Long Story Coherence With Detailed Outline ControlProceedings of ACL2023. [paper] [bib]
- Modular Visual Question Answering via Code GenerationProceedings of ACL2023. [paper] [bib]
- Neural Unsupervised Reconstruction of Protolanguage Word FormsProceedings of ACL2023. [paper] [bib]
- InCoder: A Generative Model for Code Infilling and SynthesisProceedings of ICLR2023. [paper] [bib]
- Are Layout-Infused Language Models Robust to Layout Distribution Shifts? A Case Study with Scientific DocumentsProceedings of ACL, Findings2023. [paper] [bib]
- Extracting Training Data from Diffusion ModelsUSENIX Security Symposium2023. [paper] [bib]
- Large Language Models Struggle to Learn Long-Tail KnowledgeInternational Conference on Machine Learning2023. [paper] [bib]
- Poisoning Language Models During Instruction TuningInternational Conference on Machine Learning2023. [paper] [bib]
- Revisiting Entropy Rate Constancy in TextProceedings of EMNLP, Findings2023. [paper] [bib]
2022
- Re3: Generating Longer Stories With Recursive Reprompting and RevisionProceedings of EMNLP2022. [paper] [bib]
- Automated Crossword SolvingProceedings of ACL2022. [paper] [bib]
- Inferring Rewards from Language in ContextProceedings of ACL2022. [paper] [bib]
- Learned Incremental Representations for ParsingProceedings of ACL2022. Best Paper Award. [paper] [bib]
- Meta-learning via Language Model In-context TuningProceedings of ACL2022. [paper] [bib]
- ReCLIP: A Strong Zero-Shot Baseline for Referring Expression ComprehensionProceedings of ACL2022. [paper] [bib]
- Understanding Game-Playing Agents with Natural Language AnnotationsProceedings of ACL2022. [paper] [bib]
- Voxel-informed Language GroundingProceedings of ACL2022. [paper] [bib]
- Addressing Resource and Privacy Constraints in Semantic Parsing Through Data AugmentationProceedings of ACL, Findings2022. [paper] [bib]
- Attention weights accurately predict language representations in the brainProceedings of EMNLP, Findings2022. [paper] [bib]
- Describing Differences between Text Distributions with Natural LanguageInternational Conference on Machine Learning2022. [paper] [bib]
2021
- Reference-Centric Models for Grounded Collaborative DialogueProceedings of EMNLP2021. [paper] [bib]
- An Improved Model for Voicing Silent SpeechProceedings of ACL2021. [paper] [bib]
- Concealed Data Poisoning Attacks on NLP ModelsProceedings of NAACL2021. [paper] [bib]
- Constructing Taxonomies from Pretrained Language ModelsProceedings of NAACL2021. [paper] [bib]
- Detoxifying Language Models Risks Marginalizing Minority VoicesProceedings of NAACL2021. [paper] [bib]
- FUDGE: Controlled Text Generation With Future DiscriminatorsProceedings of NAACL2021. [paper] [bib]
- Low-Complexity Probing via Finding SubnetworksProceedings of NAACL2021. [paper] [bib]
- Modular Networks for Compositional Instruction FollowingProceedings of NAACL2021. [paper] [bib]
- Adapting Language Models for Zero-shot Learning by Meta-tuning on Dataset and Prompt CollectionsProceedings of EMNLP, Findings2021. [paper] [bib]
- Are Larger Pretrained Language Models Uniformly Better? Comparing Performance at the Instance LevelProceedings of ACL, Findings2021. [paper] [bib]
- Interactive Assignments for Teaching Structured Neural NLPProceedings of the Teaching NLP Workshop at NAACL2021. [paper] [bib]
2020
- A Streaming Approach For Efficient Batched Beam SearchProceedings of EMNLP2020. [paper] [bib]
- Digital Voicing of Silent SpeechProceedings of EMNLP2020. Best Paper Award. [paper] [bib]
- Imitation Attacks and Defenses for Black-box Machine Translation SystemsProceedings of EMNLP2020. [paper] [software] [bib]
- Semantic Evaluation for Text-to-SQL with Distilled Test SuitesProceedings of EMNLP2020. [paper] [software] [bib]
- Unsupervised Parsing via Constituency TestsProceedings of EMNLP2020. [paper] [bib]
- Pretrained Transformers Improve Out-of-Distribution RobustnessProceedings of ACL2020. [paper] [bib]
- Semantic Scaffolds for Pseudocode-to-Code GenerationProceedings of ACL2020. [paper] [bib]
- Tetra-Tagging: Word-Synchronous Parsing with Linear-Time InferenceProceedings of ACL2020. [paper] [software] [bib]
- Train Large, Then Compress: Rethinking Model Size for Efficient Training and Inference of TransformersProceedings of ICML2020. [paper] [bib]
- Multilingual Alignment of Contextual Word RepresentationsProceedings of ICLR2020. [paper] [bib]
2019
- A Deep Factorization of Style and Structure in FontsProceedings of EMNLP2019. [paper] [bib]
- Are You Looking? Grounding to Multiple Modalities in Vision-and-Language NavigationProceedings of ACL2019. [paper] [bib]
- Cross-Domain Generalization of Neural Constituency ParsersProceedings of ACL2019. [paper] [software] [bib]
- Multilingual Constituency Parsing with Self-Attention and Pre-TrainingProceedings of ACL2019. [paper] [software] [bib]
- Pre-Learning Environment Representations for Data-Efficient Neural Instruction FollowingProceedings of ACL2019. [paper] [bib]
- Pragmatically Informative Text GenerationProceedings of NAACL2019. [paper] [bib]
2018
- Constituency Parsing with a Self-Attentive EncoderProceedings of ACL2018. [paper] [software] [bib]
- Policy Gradient as a Proxy for Dynamic Oracles in Constituency ParsingProceedings of ACL2018. [paper] [software] [bib]
- Learning with Latent LanguageProceedings of NAACL2018. [paper] [bib]
- Unified Pragmatic Models for Generating and Following InstructionsProceedings of NAACL2018. [paper] [software] [bib]
- What's Going On in Neural Constituency Parsers? An AnalysisProceedings of NAACL2018. [paper] [bib]
- Speaker-Follower Models for Vision-and-Language NavigationProceedings of NIPS2018. [paper] [software] [bib]
2017
- Analogs of Linguistic Structure in Deep RepresentationsProceedings of EMNLP2017. [paper] [bib]
- Effective Inference for Generative Neural ParsingProceedings of EMNLP2017. [paper] [bib]
- Where is Misty? Interpreting Spatial Descriptors by Modeling Regions in SpaceProceedings of EMNLP2017. [paper] [bib]
- A Minimal Span-Based Neural Constituency ParserProceedings of ACL2017. [paper] [bib]
- Abstract Syntax Networks for Code Generation and Semantic ParsingProceedings of ACL2017. Outstanding Paper. [paper] [bib]
- Fine-Grained Entity Typing with High-Multiplicity AssignmentsProceedings of ACL2017. [paper] [bib]
- Improving Neural Parsing by Disentangling Model Combination and Reranking EffectsProceedings of ACL2017. [paper] [software] [bib]
- Translating NeuraleseProceedings of ACL2017. [paper] [bib]
- Modular Multitask Reinforcement Learning with Policy SketchesProceedings of ICML2017. Best Paper Honorable Mention. [paper] [bib]
- Parsing with Traces: An O(n^4) Algorithm and a Structural RepresentationTransactions of the ACL2017. [paper] [bib]
2016
- Reasoning About Pragmatics with Neural Listeners and SpeakersProceedings of EMNLP2016. [paper] [bib]
- Learning-Based Single-Document Summarization with Compression and Anaphoricity ConstraintsProceedings of ACL2016. [paper] [bib]
- Capturing Semantic Similarity for Entity Linking with Convolutional Neural NetworksProceedings of NAACL2016. [paper] [bib]
- Learning to Compose Neural Networks for Question AnsweringProceedings of NAACL2016. Best Paper Award. [paper] [bib]
- Neural Module NetworksProceedings of CVPR2016. [paper] [bib]
2015
- Alignment-based Compositional Semantics for Instruction FollowingProceedings of EMNLP2015. [paper] [bib]
- An Empirical Analysis of Optimization for Max-Margin NLPProceedings of EMNLP2015. [paper] [software] [bib]
- Neural CRF ParsingProceedings of ACL2015. [paper] [bib]
- Disfluency Detection with a Semi-Markov Model and Prosodic Features Proceedings of NAACL2015. [paper] [bib]
- When and Why are Log-Linear Models Self-Normalizing?Proceedings of NAACL2015. [paper] [bib]
- On the Accuracy of Self-Normalized Log-Linear ModelsProceedings of NIPS2015. [paper] [bib]
2014
- How much do word embeddings encode about syntax?Proceedings of ACL2014. [paper] [bib]
- Improved Typesetting Models for Historical OCRProceedings of ACL2014. [paper] [bib]
- Less Grammar, More FeaturesProceedings of ACL2014. [paper] [bib]
- Sparser, Better, Faster GPU ParsingProceedings of ACL2014. [paper] [bib]
- Structured Learning for Taxonomy Induction with Belief PropagationProceedings of ACL2014. Best Paper Honorable Mention. [paper] [bib]
- Grounding language with points and paths in continuous spacesProceedings of CoNLL2014. [paper] [bib]
- Unsupervised Transcription of Piano MusicProceedings of NIPS2014. [paper] [bib]
- A Joint Model for Entity Analysis: Coreference, Typing, and LinkingTransactions of the ACL2014. [paper] [bib]
2013
- Easy Victories and Uphill Battles in Coreference ResolutionProceedings of EMNLP2013. Best Paper Finalist. [paper] [bib]
- A Multi-Teraflop Constituency Parser using GPUsProceedings of EMNLP2013. [paper] [bib]
- Decipherment with a Million Random RestartsProceedings of EMNLP2013. [paper] [bib]
- Error-Driven Analysis of Challenges in Coreference ResolutionProceedings of EMNLP2013. [paper] [software] [bib]
- Decentralized Entity-Level Modeling for Coreference ResolutionProceedings of ACL2013. [paper] [bib]
- Unsupervised Transcription of Historical DocumentsProceedings of ACL2013. [paper] [bib]
- An Empirical Examination of Challenges in Chinese ParsingProceedings of ACL (Short Papers)2013. [paper] [software] [bib]
- Supervised Learning of Complete Morphological ParadigmsProceedings of NAACL2013. [paper] [bib]
- Automated reconstruction of ancient languages using probabilistic models of sound changeProceedings of the National Academy of Sciences2013. [paper] [url] [eprint] [bib]
- Faster Optimal Planning with Partial-Order PruningProceedings of the International Conference on Automated Planning Systems2013. [paper] [bib]
- Good, Great, Excellent: Global Inference of Semantic IntensitiesTransactions of the ACL2013. [paper] [bib]
- Grounding Spatial Relations for Human-Robot InteractionIEEE/RSJ International Conference on Intelligent Robots and Systems2013. [paper] [bib]
2012
- An Empirical Investigation of Statistical Significance in NLPProceedings of EMNLP2012. [paper] [bib]
- Parser Showdown at the Wall Street Corral: An Empirical Investigation of Error Types in Parser OutputProceedings of EMNLP2012. [slides] [keynote] [paper] [software] [bib]
- Syntactic Transfer Using a Bilingual LexiconProceedings of EMNLP2012. [paper] [bib]
- Training Factored PCFGs with Expectation PropagationProceedings of EMNLP2012. Distinguished Paper. [paper] [bib]
- Transforming Trees to Improve Syntactic ConvergenceProceedings of EMNLP2012. [paper] [bib]
- Coreference Semantics from Web FeaturesProceedings of ACL2012. [paper] [bib]
- Large-Scale Syntactic Language Modeling with TreeletsProceedings of ACL2012. [paper] [software] [bib]
- A Feature-Rich Constituent Context Model for Grammar InductionProceedings of ACL (Short Papers)2012. [paper] [bib]
- Robust Conversion of CCG Derivations to Phrase Structure TreesProceedings of ACL (Short Papers)2012. [paper] [software] [bib]
- Fast Inference in Phrase Extraction Models with Belief PropagationProceedings of NAACL2012. [paper] [bib]
- Unsupervised Translation Sense ClusteringProceedings of NAACL2012. [paper] [bib]
2011
- Large-Scale Cognate RecoveryProceedings of EMNLP2011. [paper] [bib]
- Simple Effective Decipherment via Combinatorial OptimizationProceedings of EMNLP2011. [paper] [bib]
- Faster and Smaller N-Gram Language ModelsProceedings of ACL2011. [paper] [bib]
- Gappy Phrasal Alignment By AgreementProceedings of ACL2011. [paper] [bib]
- Jointly Learning to Extract and CompressProceedings of ACL2011. [paper] [bib]
- Learning Dependency-Based Compositional SemanticsProceedings of ACL2011. [paper] [bib]
- Web-Scale Features for Full-Scale ParsingProceedings of ACL2011. [paper] [bib]
- An Empirical Investigation of Discounting in Cross-Domain Language ModelsProceedings of ACL (Short Papers)2011. [paper] [bib]
- The Surprising Variance in Shortest-Derivation ParsingProceedings of ACL (Short Papers)2011. [paper] [bib]
- Mention Detection: Heuristics for the OntoNotes annotationsProceedings of CoNLL2011. [paper] [bib]
- Iterative Monotonically Bounded A*Proceedings of AAAI2011. [paper] [bib]
2010
- A Game-Theoretic Approach to Generating Spatial DescriptionsProceedings of EMNLP2010. [paper] [bib]
- A Simple Domain-Independent Probabilistic Approach to GenerationProceedings of EMNLP2010. [paper] [bib]
- An Entity-Level Approach to Information ExtractionProceedings of ACL2010. [paper] [url] [bib]
- Discriminative Modeling of Extraction Sets for Machine TranslationProceedings of ACL2010. [paper] [bib]
- Finding Cognate Groups using PhylogeniesProceedings of ACL2010. [paper] [bib]
- Hierarchical A* Parsing with Bridge Outside ScoresProceedings of ACL2010. [paper] [bib]
- Phylogenetic Grammar InductionProceedings of ACL2010. [paper] [bib]
- Simple, Accurate Parsing with an All-Fragments GrammarProceedings of ACL2010. [paper] [bib]
- Top-Down K-Best A* ParsingProceedings of ACL2010. [paper] [bib]
- Coreference Resolution in a Modular, Entity-Centered ModelProceedings of NAACL2010. Best Paper Award. [paper] [url] [bib]
- Joint Parsing and Alignment with Weakly Synchronized GrammarsProceedings of NAACL2010. [paper] [bib]
- Painless Unsupervised Learning with FeaturesProceedings of NAACL2010. [slides] [paper] [bib]
- Type-Based MCMCProceedings of NAACL2010. [paper] [bib]
- Unsupervised Syntactic Alignment with Inversion Transduction GrammarsProceedings of NAACL2010. [paper] [bib]
- Learning Better Monolingual Models with Unannotated Bilingual TextProceedings of CoNLL2010. [paper] [url] [bib]
- Learning Programs: A Hierarchical Bayesian ApproachProceedings of ICML2010. [paper] [bib]
- Teaching Introductory Artificial Intelligence with Pac-ManProceedings of the Symposium on Educational Advances in Artificial Intelligence (EAAI)2010. [paper] [software] [bib]
- The Pac-Man Projects Software Package for Introductory Artificial IntelligenceProceedings of the Symposium on Educational Advances in Artificial Intelligence, Model Assignments Track2010. [software] [bib]
- Iterated Learning of Multiple Languages from Multiple TeachersProceedings of Evolang2010. [paper] [bib]
2009
- Consensus Training for Consensus Decoding in Machine TranslationProceedings of EMNLP2009. [paper] [bib]
- Simple Coreference Resolution with Rich Syntactic and Semantic FeaturesProceedings of EMNLP2009. [paper] [url] [bib]
- Better Word Alignments with Supervised ITG ModelsProceedings of ACL2009. [slides] [paper] [url] [bib]
- K-Best A* ParsingProceedings of ACL2009. Best Paper Award. [paper] [bib]
- Learning Semantic Correspondences with Less SupervisionProceedings of ACL2009. [slides] [paper] [bib]
- Efficient Parsing for Transducer GrammarsProceedings of NAACL2009. [paper] [bib]
- Hierarchical Search for ParsingProceedings of NAACL2009. [slides] [paper] [bib]
- Improved Reconstruction of Protolanguage Word FormsProceedings of NAACL2009. [paper] [bib]
- Online EM for Unsupervised ModelsProceedings of NAACL2009. [slides] [paper] [bib]
- Learning from Measurements in Exponential FamiliesProceedings of ICML2009. [slides] [paper] [bib]
- Efficient Inference in Phylogenetic InDel TreesProceedings of NIPS2009. [paper] [bib]
- Convergence Bounds for Language Evolution by Iterated LearningProceedings of the 31st Annual Conference of the Cognitive Science Society2009. [paper] [bib]
- Probabilistic grammars and hierarchical Dirichlet processesThe Oxford Handbook of Applied Bayesian Analysis2009. [chapter] [bib]
- Asynchronous Binarization for Synchronous GrammarsProceedings of ACL-IJCNLP (Short Papers)2009. [paper] [bib]
2008
- Coarse-to-Fine Syntactic Machine Translation using Language ProjectionsProceedings of EMNLP2008. [slides] [paper] [url] [bib]
- Sampling Alignment Structure under a Bayesian Translation ModelProceedings of EMNLP2008. [paper] [bib]
- Sparse Multi-Scale Grammars for Discriminative Latent Variable ParsingProceedings of EMNLP2008. [paper] [url] [bib]
- Two Languages are Better than One (for Syntactic Parsing)Proceedings of EMNLP2008. [paper] [bib]
- Analyzing the Errors of Unsupervised InductionProceedings of ACL2008. [slides] [paper] [bib]
- Learning Bilingual Lexicons from Monolingual CorporaProceedings of ACL2008. [paper] [url] [bib]
- The Complexity of Phrase Alignment ModelsProceedings of ACL (Short Papers)2008. [slides] [paper] [bib]
- Fully Distributed EM for Very Large DatasetsProceedings of ICML2008. [slides] [paper] [bib]
- Structured Compilation: Trading off Structure for FeaturesProceedings of ICML2008. [slides] [paper] [bib]
- A Probabilistic Approach to Language ChangeProceedings of NIPS2008. [slides] [paper] [bib]
- Agreement-Based LearningProceedings of NIPS2008. [poster] [paper] [bib]
- Discriminative Log-Linear Grammars with Latent VariablesProceedings of NIPS2008. [slides] [paper] [url] [bib]
- Efficient Sentence Segmentation using Syntactic FeaturesProceedings of Spoken Language Technologies (SLT)2008. [poster] [paper] [url] [bib]
2007
- Learning Structured Models for Phone RecognitionProceedings of EMNLP-CoNLL2007. [slides] [paper] [url] [bib]
- A Probabilistic Approach to Diachronic PhonologyProceedings of EMNLP2007. [slides] [paper] [bib]
- The Infinite PCFG using Hierarchical Dirichlet ProcessesProceedings of EMNLP2007. [slides] [paper] [url] [bib]
- Tailoring Word Alignments to Syntactic Machine TranslationProceedings of ACL2007. [slides] [paper] [bib]
- Unsupervised Coreference Resolution in a Nonparametric Bayesian ModelProceedings of ACL2007. [slides] [paper] [url] [bib]
- Approximate Factoring for A* SearchProceedings of HLT-NAACL2007. [slides] [paper] [url] [bib]
- Improved Inference for Unlexicalized ParsingProceedings of HLT-NAACL2007. [slides] [paper] [url] [software] [bib]
- A* Search via Approximate FactoringProceedings of AAAI (Nectar Track)2007. [paper] [bib]
- Learning and Inference for Hierarchically Split PCFGsProceedings of AAAI (Nectar Track)2007. [slides] [poster] [paper] [url] [software] [bib]
- Mixture-of-Parents Maximum Entropy Markov ModelsProceedings of Uncertainty in Artificial Intelligence (UAI)2007. [paper] [bib]
2006
- An End-to-End Discriminative Approach to Machine TranslationProceedings of COLING-ACL2006. [slides] [paper] [url] [bib]
- Learning Accurate, Compact, and Interpretable Tree AnnotationProceedings of COLING-ACL2006. [slides] [paper] [url] [software] [bib]
- Prototype-Driven Grammar InductionProceedings of COLING-ACL2006. [slides] [paper] [url] [bib]
- Alignment by AgreementProceedings of HLT-NAACL2006. [slides] [paper] [url] [software] [bib]
- Prototype-Driven Learning for Sequence ModelsProceedings of HLT-NAACL2006. Best Student Paper Award. [slides] [paper] [url] [bib]
- Word Alignment via Quadratic AssignmentProceedings of HLT-NAACL2006. [paper] [url] [bib]
- Non-Local Modeling with a Mixture of PCFGsProceedings of CoNLL2006. [slides] [paper] [url] [bib]
- Why Generative Phrase Models Underperform Surface HeuristicsWorkshop on Statistical Machine Translation at HLT-NAACL2006. [slides] [paper] [url] [bib]
2005
- A Discriminative Matching Approach to Word AlignmentProceedings of EMNLP2005. [paper] [url] [bib]
- Robust Textual Inference via Graph MatchingProceedings of HLT-EMNLP2005. [paper] [url] [bib]
- Unsupervised Learning of Field Segmentation Models for Information ExtractionProceedings of ACL2005. [paper] [bib]
- The Unsupervised Learning of Natural Language Structure2005. [thesis] [bib]
2004
- Max-Margin ParsingProceedings of EMNLP2004. Best Paper Award. [paper] [bib]
- Corpus-Based Induction of Syntactic Structure: Models of Dependency and ConstituencyProceedings of ACL2004. [paper] [bib]
- Review of Data-Oriented ParsingComputational Linguistics2004. [bib]
2003
- Accurate Unlexicalized ParsingProceedings of ACL2003. Best Paper Award. [paper] [bib]
- A* Parsing: Fast Exact Viterbi Parse SelectionProceedings of NAACL2003. [paper] [bib]
- Feature-Rich Part-of-Speech Tagging with a Cyclic Dependency NetworkProceedings of NAACL2003. [paper] [bib]
- Named Entity Recognition with Character-Level ModelsProceedings of CoNLL2003. [paper] [bib]
- Factored A* Search for Models over Sequences and TreesProceedings of the International Joint Conference on Artificial Intelligence (IJCAI)2003. [paper] [bib]
- Spectral LearningProceedings of the International Joint Conference on Artificial Intelligence (IJCAI)2003. [paper] [bib]
2002
- Conditional Structure versus Conditional Estimation in NLP ModelsProceedings of EMNLP2002. [paper] [bib]
- A Generative Constituent-Context Model for Improved Grammar InductionProceedings of ACL2002. [paper] [bib]
- From Instance-level Constraints to Space-Level Constraints: Making the Most of Prior Knowledge in Data ClusteringProceedings of ICML2002. [paper] [bib]
- Interpreting and Extending Classical Agglomerative Clustering Algorithms using a Model-Based ApproachProceedings of ICML2002. [paper] [bib]
- Fast Exact Inference with a Factored Model for Natural Language ProcessingProceedings of NIPS2002. [paper] [bib]
- Combining Heterogeneous Classifiers for Word-Sense DisambiguationProceedings of the ACL Workshop on Word Sense Disambiguation2002. [paper] [bib]
- Evaluating Strategies for Similarity Search on the WebProceedings of the International World Wide Web Conference (WWW)2002. [paper] [bib]
- Parsing and HypergraphsNew Developments in Parsing Technology2002. [bib]
2001
- Parsing with Treebank Grammars: Empirical Bounds, Theoretical Models, and the Structure of the Penn TreebankProceedings of ACL2001. [paper] [bib]
- Distributional Phrase Structure InductionProceedings of CoNLL2001. [paper] [bib]
- Natural Language Grammar Induction Using a Constituent-Context ModelProceedings of NIPS2001. [paper] [bib]
- Parsing and HypergraphsProceedings of the International Workshop on Parsing Technologies (IWPT)2001. [paper] [bib]
- An O(n^3) Agenda-Based Chart Parser for Arbitrary Probabilistic Context-Free Grammars2001. [paper] [bib]