Ryan Cotterell

Latest

• On Efficiently Representing Regular Languages as RNNs
• Do Language Models Exhibit the Same Cognitive Biases in Problem Solving as Human Learners?
• Transformers Can Represent n-gram Language Models
• A Transformer with Stack Attention
• Context versus Prior Knowledge in Language Models
• Lower Bounds on the Expressivity of Recurrent Neural Language Models
• On the Representational Capacity of Neural Language Models with Chain-of-Thought Reasoning
• The Foundations of Tokenization: Statistical and Computational Concerns
• Towards Explainability in Legal Outcome Prediction Models
• What Do Language Models Learn in Context? The Structured Task Hypothesis.
• What Languages are Easy to Language-Model? A Perspective from Learning Probabilistic Regular Languages
• An Exploration of Left-Corner Transformations
• Efficient Algorithms for Recognizing Weighted Tree-Adjoining Languages
• Findings of the BabyLM Challenge: Sample-Efficient Pretraining on Developmentally Plausible Corpora
• Language Model Quality Correlates with Psychometric Predictive Power in Multiple Languages
• Linear-Time Modeling of Linguistic Structure: An Order-Theoretic Perspective
• On the Optimality of Word Lengths
• On the Representational Capacity of Recurrent Neural Language Models
• Quantifying the redundancy between prosody and text
• Recurrent Neural Language Models as Probabilistic Finite-state Automata
• Revisiting the Optimality of Word Lengths
• Structured Voronoi Sampling
• The Ethics of Automating Legal Actors
• A Fast Algorithm for Computing Prefix Probabilities
• A Formal Perspective on Byte-Pair Encoding
• A Measure-theoretic Characterization of Tight Language Model
• An Ordinal Latent Variable Model of Conflict Intensity
• Convergence and Diversity in the Control Hierarchy
• Discourse-Centric Evaluation of Document-level Machine Translation with a New Densely Annotated Parallel Corpus of Novels
• Discourse-Centric Evaluation of Document-level Machine Translation with a New Densely Annotated Parallel Corpus of Novels
• Efficient Semiring-Weighted Earley Parsing
• Efficient Semiring-Weighted Earley Parsing
• Generalizing Backpropagation for Gradient-Based Interpretability
• Generalizing Backpropagation for Gradient-Based Interpretability
• Hexatagging: Projective Dependency Parsing as Tagging
• Locally Typical Sampling
• Log-Linear Guardedness and Its Implications
• Naturalistic Causal Probing for Morpho-Syntax
• On the Effect of Anticipation on Reading Times
• On the Efficacy of Sampling Adapters
• Testing the Predictions of Surprisal Theory in 11 Languages
• Tokenization and the Noiseless Channel
• On the Intersection of Context-Free and Regular Languages
• On the Usefulness of Embeddings, Clusters and Strings for Text Generation Evaluation
• Sentiment as an Ordinal Latent Variable
• The Ordered Matrix Dirichlet for State-Space Models
• A Cross-Linguistic Pressure for Uniform Information Density in Word Order
• A Latent-Variable Model for Intrinsic Probing
• Controlled Text Generation with Natural Language Instructions
• Quantifying Gender Bias Towards Politicians in Cross-Lingual Language Models
• Algorithms for Weighted Finite-State Automata with Failure Arcs
• Algorithms for Weighted Pushdown Automata
• Autoregressive Structure Prediction with Language Models
• Kernelized Concept Erasure
• Mutual Information and Hallucinations in Abstractive Summarization
• On Parsing as Tagging
• The Architectural Bottleneck Principle
• Benchmarking Compositionality with Formal Languages
• Equivariant Transduction through Invariant Alignment
• A Structured Span Selector
• BlonDe: An Automatic Evaluation Metric for Document-level Machine Translation
• Exact Paired-Permutation Testing for Structured Test Statistics
• Linear Adversarial Concept Erasure
• On the Machine Learning of Ethical Judgments from Natural Language
• Probing via Prompting
• Same Neurons, Different Languages: Probing Morphosyntax in Multilingual Pre-trained Models
• SIGMORPHON--UniMorph 2022 Shared Task 0: Generalization and Typologically Diverse Morphological Inflection
• The SIGMORPHON 2022 Shared Task on Morpheme Segmentation
• The SIGTYP 2022 Shared Task on the Prediction of Cognate Reflexes
• Analyzing Wrap-Up Effects through an Information-Theoretic Lens
• Estimating the Entropy of Linguistic Distributions
• Probing as Quantifying the Inductive Bias of Pre-trained Representations
• Probing for the Usage of Grammatical Number
• An Ordinal Latent Variable Model of Conflict Intensity
• Cluster-based Evaluation of Automatically Generated Text
• On Decoding Strategies for Neural Text Generators
• On the Intersection of Context-Free and Regular Languages
• State-of-the-art generalisation research in NLP: a taxonomy and review
• Visual Comparison of Language Model Adaptation
• A Bayesian Framework for Information-Theoretic Probing
• A Bayesian Framework for Information-Theoretic Probing
• A Bayesian Framework for Information-Theoretic Probing
• A Plug-and-Play Method for Controlled Text Generation
• A Plug-and-Play Method for Controlled Text Generation
• A surprisal--duration trade-off across and within the world's languages
• A surprisal--duration trade-off across and within the world’s languages
• Adjusting the Conflict-Cooperation Scale for Armed Conflict Assessment
• Classifying Dyads for Militarized Conflict Analysis
• Classifying Dyads for Militarized Conflict Analysis
• Conditional Poisson Stochastic Beam Search
• Conditional Poisson Stochastic Beam Search
• Conditional Poisson Stochastic Beams
• Efficient Sampling of Dependency Structure
• Efficient Sampling of Dependency Structure
• Efficient Sampling of Dependency Structure
• Equivariant Transduction through Invariant Alignment
• Keyword2Text: A Plug-and-Play Method for Controlled Text Generation
• On Homophony and Rényi Entropy
• On Homophony and Rényi Entropy
• On Homophony and Rényi Entropy
• Phone-level Uniform Information Density across and within Languages
• Revisiting the Uniform Information Density Hypothesis
• Revisiting the Uniform Information Density Hypothesis
• Revisiting the Uniform Information Density Hypothesis
• Searching for More Efficient Dynamic Programs
• Searching for More Efficient Dynamic Programs
• Searching for More Efficient Dynamic Programs
• Text or Topology? Classifying Ally-Enemy Pairs in Militarised Conflict
• A cognitive regularizer for language modeling
• A cognitive regularizer for language modeling
• Determinantal Beam Search
• Determinantal Beam Search
• Examining the Inductive Bias of Neural Language Models with Artificial Languages
• Examining the Inductive Bias of Neural Language Models with Artificial Languages
• Higher-order Derivatives of Weighted Finite-state Machines
• Is Sparse Attention more Interpretable?
• Is Sparse Attention more Interpretable?
• Language Model Evaluation Beyond Perplexity
• Language Model Evaluation Beyond Perplexity
• Modeling the Unigram Distribution
• Modelling the Unigram Distribution
• On Finding the $K$-best Non-projective Dependency Trees
• On Finding the $K$-best Non-projective Dependency Trees
• SIGMORPHON 2021 Shared Task on Morphological Reinflection: Generalization Across Languages
• A Non-Linear Structural Probe
• A Non-Linear Structural Probe
• Do Syntactic Probes Probe Syntax? Experiments with Jabberwocky Probing
• Do Syntactic Probes Probe Syntax? Experiments with Jabberwocky Probing
• Finding Concept-specific Biases in Form--Meaning Associations
• Finding Concept-specific Biases in Form--Meaning Associations
• How (Non-)Optimal is the Lexicon?
• How (Non-)Optimal is the Lexicon?
• SIGTYP 2021 Shared Task: Robust Spoken Language Identification
• What About the Precedent: An Information-Theoretic Analysis of Common Law
• What About the Precedent: An Information-Theoretic Analysis of Common Law
• Applying the Transformer to Character-level Transduction
• Applying the Transformer to Character-level Transduction
• Disambiguatory signals are stronger in word initial positions
• Searching for Search Errors in Neural Morphological Inflection
• Searching for Search Errors in Neural Morphological Inflection
• A Word on Machine Ethics: A Response to Jiang et al. (2021)
• Differentiable Subset Pruning of Transformer Heads
• Efficient Computation of Expectations under Spanning Tree Distributions
• Multimodal Pretraining Unmasked: A Meta-Analysis and a Unified Framework of Vision-and-Language BERTs
• On the Relationships Between the Grammatical Genders of Inanimate Nouns and Their Co-Occurring Adjectives and Verbs
• Parameter Space Factorization for Zero-Shot Learning across Tasks and Languages
• Morphologically Aware Word-Level Translation
• Exploring the Linear Subspace Hypothesis in Gender Bias Mitigation
• Exploring the Linear Subspace Hypothesis in Gender Bias Mitigation
• Finding Concept-specific Biases in Form–Meaning Associations
• If Beam Search is the Answer, What was the Question?
• Intrinsic Probing through Dimension Selection
• Intrinsic Probing through Dimension Selection
• Investigating Cross-Linguistic Adjective Ordering Tendencies with a Latent-Variable Model
• Measuring the Similarity of Grammatical Gender Systems by Comparing Partitions
• Pareto Probing: Trading Off Accuracy for Simplicity
• Please Mind the Root: Decoding Arborescences for Dependency Parsing
• SIGTYP 2020 Shared Task: Prediction of Typological Features
• A Corpus for Large-Scale Phonetic Typology
• A Tale of a Probe and a Parser
• Generalized Entropy Regularization or: There's Nothing Special about Label Smoothing
• Information-Theoretic Probing for Linguistic Structure
• It's Easier to Translate out of English than into it: Measuring Neural Translation Difficulty by Cross-Mutual Information
• Metaphor Detection Using Context and Concreteness
• Predicting Declension Class from Form and Meaning
• SIGMORPHON 2020 Shared Task 0: Typologically Diverse Morphological Inflection
• The Paradigm Discovery Problem
• Best-First Beam Search
• Efficient Computation of Expectations under Spanning Tree Distributions
• On the Relationships Between the Grammatical Genders of Inanimate Nouns and Their Co-Occurring Adjectives and Verbs
• Parameter Space Factorization for Zero-Shot Learning across Tasks and Languages
• Phonotactic Complexity and its Trade-offs
• UniMorph 3.0: Universal Morphology
• Don’t Forget the Long Tail! A Comprehensive Analysis of Morphological Generalization in Bilingual Lexicon Induction
• Examining Gender Bias in Languages with Grammatical Gender
• It’s All in the Name: Mitigating Gender Bias with Name-Based Counterfactual Data Substitution
• Quantifying the Semantic Core of Gender Systems
• Towards Zero-Shot Language Modeling
• The SIGMORPHON 2019 Shared Task: Morphological Analysis in Context and Cross-Lingual Transfer for Inflection
• Counterfactual Data Augmentation for Mitigating Gender Bias in Languages with Rich Morphology
• Meaning to Form: Measuring Systematicity as Information
• Measuring Morphological Irregularity
• On the distribution of deep clausal embeddings: A large cross-linguistic study
• Uncovering Typological Implications with Belief Nets
• Unsupervised Discovery of Gendered Language through Latent-Variable Modeling
• What Kind of Language Is Hard to Language-Model?
• A Probabilistic Generative Model of Linguistic Typology
• A Simple Joint Model for Improved Contextual Neural Lemmatization
• Combining Sentiment Lexica with a Multi-View Variational Autoencoder
• Contextualization of Morphological Inflection
• Gender Bias in Contextualized Word Embeddings
• On the Idiosyncrasies of the Mandarin Chinese Classifier System
• On the Complexity and Typology of Inflectional Morphological Systems
• Recurrent Neural Networks in Linguistic Theory: Revisiting Pinker and Prince (1988) and the Past Tense Debate
• A Discriminative Latent-Variable Model for Bilingual Lexicon Induction
• Generalizing Procrustes Analysis for Better Bilingual Dictionary Induction
• Hard Non-Monotonic Attention for Character-Level Transduction
• Marrying Universal Dependencies and Universal Morphology
• The CoNLL--SIGMORPHON 2018 Shared Task: Universal Morphological Reinflection
• A Structured Variational Autoencoder for Contextual Morphological Inflection
• A Deep Generative Model of Vowel Formant Typology
• Are All Languages Equally Hard to Language-Model?
• Unsupervised Disambiguation of Syncretism in Inflected Lexicons
• UniMorph 2.0: Universal Morphology
• Explaining and Generalizing Back-Translation through Wake-Sleep
• Joint Semantic Synthesis and Morphological Analysis of the Derived Word
• On the Diachronic Stability of Irregularity in Inflectional Morphology
• Recurrent Neural Networks in Linguistic Theory: Revisiting Pinker and Prince (1988) and the Past Tense Debate
• Low-Resource Named Entity Recognition with Cross-lingual, Character-Level Neural Conditional Random Fields
• Cross-lingual, Character-Level Neural Morphological Tagging
• Paradigm Completion for Derivational Morphology
• CoNLL--SIGMORPHON 2017 Shared Task: Universal Morphological Reinflection in 52 Languages
• Frame-Based Continuous Lexical Semantics through Exponential Family Tensor Factorization and Semantic Proto-Roles
• Frame-Based Continuous Lexical Semantics through Exponential Family Tensor Factorization and Semantic Proto-Roles
• One-Shot Neural Cross-Lingual Transfer for Paradigm Completion
• Probabilistic Typology: Deep Generative Models of Vowel Inventories
• A Rich Morphological Tagger for English: Exploring the Cross-Linguistic Tradeoff Between Morphology and Syntax
• Context-Aware Prediction of Derivational Word-forms
• Explaining and Generalizing Skip-Gram through Exponential Family Principal Component Analysis
• Morphological Analysis of the Dravidian Language Family
• Neural Graphical Models over Strings for Principal Parts Morphological Paradigm Completion
• Neural Multi-Source Morphological Reinflection
• Morphological Segmentation Inside-Out
• Neural Morphological Analysis: Encoding-Decoding Canonical Segments
• Speed-Accuracy Tradeoffs in Tagging with Variable-Order CRFs and Structured Sparsity
• Morphological Smoothing and Extrapolation of Word Embeddings
• The SIGMORPHON 2016 Shared Task—Morphological Reinflection
• A Joint Model of Orthography and Morphological Segmentation
• Weighting Finite-State Transductions With Neural Context
• Analysis of Morphology in Topic Modeling
• Contrastive Morphological Typology and Logical Hierarchies
• Dual Decomposition Inference for Graphical Models over Strings
• Joint Lemmatization and Morphological Tagging with Lemming
• Labeled Morphological Segmentation with Semi-Markov Models
• Morphological Word Embeddings
• Penalized Expectation Propagation for Graphical Models over Strings
• Modeling Word Forms Using Latent Underlying Morphs and Phonology
• Stochastic Contextual Edit Distance and Probabilistic FSTs
• A Multi-Dialect, Multi-Genre Corpus of Informal Written Arabic
• An Algerian Arabic-French Code-Switched Corpus
• Translation of the CALLHOME Egyptian Arabic Corpus For Conversational Speech Translation