Matches in SemOpenAlex for { <https://semopenalex.org/work/W3134106875> ?p ?o ?g. }
Showing items 1 to 72 of
72
with 100 items per page.
- W3134106875 abstract "Human reading behavior is tuned to the statistics of natural language: the time it takes human subjects to read a word can be predicted from estimates of the word's probability in context. However, it remains an open question what computational architecture best characterizes the expectations deployed in real time by humans that determine the behavioral signatures of reading. Here we test over two dozen models, independently manipulating computational architecture and training dataset size, on how well their next-word expectations predict human reading time behavior on naturalistic text corpora. We find that across model architectures and training dataset sizes the relationship between word log-probability and reading time is (near-)linear. We next evaluate how features of these models determine their psychometric predictive power, or ability to predict human reading behavior. In general, the better a model's next-word expectations, the better its psychometric predictive power. However, we find nontrivial differences across model architectures. For any given perplexity, deep Transformer models and n-gram models generally show superior psychometric predictive power over LSTM or structurally supervised neural models, especially for eye movement data. Finally, we compare models' psychometric predictive power to the depth of their syntactic knowledge, as measured by a battery of syntactic generalization tests developed using methods from controlled psycholinguistic experiments. Once perplexity is controlled for, we find no significant relationship between syntactic knowledge and predictive power. These results suggest that different approaches may be required to best model human real-time language comprehension behavior in naturalistic reading versus behavior for controlled linguistic materials designed for targeted probing of syntactic knowledge." @default.
- W3134106875 created "2021-03-15" @default.
- W3134106875 creator A5011708753 @default.
- W3134106875 creator A5031400281 @default.
- W3134106875 creator A5045238792 @default.
- W3134106875 creator A5048080768 @default.
- W3134106875 creator A5090215557 @default.
- W3134106875 date "2020-01-01" @default.
- W3134106875 modified "2023-09-24" @default.
- W3134106875 title "On the Predictive Power of Neural Language Models for Human Real-Time Comprehension Behavior." @default.
- W3134106875 hasPublicationYear "2020" @default.
- W3134106875 type Work @default.
- W3134106875 sameAs 3134106875 @default.
- W3134106875 citedByCount "4" @default.
- W3134106875 countsByYear W31341068752021 @default.
- W3134106875 crossrefType "journal-article" @default.
- W3134106875 hasAuthorship W3134106875A5011708753 @default.
- W3134106875 hasAuthorship W3134106875A5031400281 @default.
- W3134106875 hasAuthorship W3134106875A5045238792 @default.
- W3134106875 hasAuthorship W3134106875A5048080768 @default.
- W3134106875 hasAuthorship W3134106875A5090215557 @default.
- W3134106875 hasConcept C100279451 @default.
- W3134106875 hasConcept C111472728 @default.
- W3134106875 hasConcept C119857082 @default.
- W3134106875 hasConcept C137293760 @default.
- W3134106875 hasConcept C138885662 @default.
- W3134106875 hasConcept C154945302 @default.
- W3134106875 hasConcept C204321447 @default.
- W3134106875 hasConcept C2778136018 @default.
- W3134106875 hasConcept C2778780117 @default.
- W3134106875 hasConcept C41008148 @default.
- W3134106875 hasConcept C41895202 @default.
- W3134106875 hasConcept C554936623 @default.
- W3134106875 hasConceptScore W3134106875C100279451 @default.
- W3134106875 hasConceptScore W3134106875C111472728 @default.
- W3134106875 hasConceptScore W3134106875C119857082 @default.
- W3134106875 hasConceptScore W3134106875C137293760 @default.
- W3134106875 hasConceptScore W3134106875C138885662 @default.
- W3134106875 hasConceptScore W3134106875C154945302 @default.
- W3134106875 hasConceptScore W3134106875C204321447 @default.
- W3134106875 hasConceptScore W3134106875C2778136018 @default.
- W3134106875 hasConceptScore W3134106875C2778780117 @default.
- W3134106875 hasConceptScore W3134106875C41008148 @default.
- W3134106875 hasConceptScore W3134106875C41895202 @default.
- W3134106875 hasConceptScore W3134106875C554936623 @default.
- W3134106875 hasLocation W31341068751 @default.
- W3134106875 hasOpenAccess W3134106875 @default.
- W3134106875 hasPrimaryLocation W31341068751 @default.
- W3134106875 hasRelatedWork W1996979899 @default.
- W3134106875 hasRelatedWork W2045461347 @default.
- W3134106875 hasRelatedWork W2068820241 @default.
- W3134106875 hasRelatedWork W2230336798 @default.
- W3134106875 hasRelatedWork W2404397873 @default.
- W3134106875 hasRelatedWork W2606165136 @default.
- W3134106875 hasRelatedWork W2796052949 @default.
- W3134106875 hasRelatedWork W2889024690 @default.
- W3134106875 hasRelatedWork W2957941633 @default.
- W3134106875 hasRelatedWork W2963394326 @default.
- W3134106875 hasRelatedWork W2964303116 @default.
- W3134106875 hasRelatedWork W2981523423 @default.
- W3134106875 hasRelatedWork W3024492259 @default.
- W3134106875 hasRelatedWork W3033254023 @default.
- W3134106875 hasRelatedWork W3045555480 @default.
- W3134106875 hasRelatedWork W3100679627 @default.
- W3134106875 hasRelatedWork W3100748148 @default.
- W3134106875 hasRelatedWork W3131755153 @default.
- W3134106875 hasRelatedWork W3186588815 @default.
- W3134106875 hasRelatedWork W3210616652 @default.
- W3134106875 isParatext "false" @default.
- W3134106875 isRetracted "false" @default.
- W3134106875 magId "3134106875" @default.
- W3134106875 workType "article" @default.