Roger P. Levy

Latest

• Dependency locality as an explanatory principle for word order
• Lossy-Context Surprisal: An information-theoretic model of memory effects in sentence processing
• Maze Made Easy: Better and easier measurement of incremental processing difficulty
• Do RNNs learn human-like abstract word order preferences?
• Hierarchical representation in neural language models: Suppression and recovery of expectations
• How efficiency shapes human language
• Neural language models as psycholinguistic subjects: Representations of syntactic state
• Structural supervision improves learning of non-local grammatical dependencies
• What syntactic structures block dependencies in RNN language models?
• A statistical comparison of some theories of NP word order
• What do RNN language models learn about filler--gap dependencies?
• Generalizing dependency distance
• Noisy-context surprisal as a human sentence processing cost model