Distinct patterns of syntactic agreement errors in recurrent networks and humans

Tal Linzen, Brian Leonard

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    Determining the correct form of a verb in context requires an understanding of the syntactic structure of the sentence. Recurrent neural networks have been shown to perform this task with an error rate comparable to humans, despite the fact that they are not designed with explicit syntactic representations. To examine the extent to which the syntactic representations of these networks are similar to those used by humans when processing sentences, we compare the detailed pattern of errors that RNNs and humans make on this task. Despite significant similarities (attraction errors, asymmetry between singular and plural subjects), the error patterns differed in important ways. In particular, in complex sentences with relative clauses error rates increased in RNNs but decreased in humans. Furthermore, RNNs showed a cumulative effect of attractors but humans did not. We conclude that at least in some respects the syntactic representations acquired by RNNs are fundamentally different from those used by humans.

    Original languageEnglish (US)
    Title of host publicationProceedings of the 40th Annual Meeting of the Cognitive Science Society, CogSci 2018
    PublisherThe Cognitive Science Society
    Pages690-695
    Number of pages6
    ISBN (Electronic)9780991196784
    StatePublished - 2018
    Event40th Annual Meeting of the Cognitive Science Society: Changing Minds, CogSci 2018 - Madison, United States
    Duration: Jul 25 2018Jul 28 2018

    Publication series

    NameProceedings of the 40th Annual Meeting of the Cognitive Science Society, CogSci 2018

    Conference

    Conference40th Annual Meeting of the Cognitive Science Society: Changing Minds, CogSci 2018
    Country/TerritoryUnited States
    CityMadison
    Period7/25/187/28/18

    Keywords

    • agreement attraction
    • Psycholinguistics
    • recurrent neural networks
    • syntax

    ASJC Scopus subject areas

    • Artificial Intelligence
    • Computer Science Applications
    • Human-Computer Interaction
    • Cognitive Neuroscience

    Fingerprint

    Dive into the research topics of 'Distinct patterns of syntactic agreement errors in recurrent networks and humans'. Together they form a unique fingerprint.

    Cite this