Premium content
Access to this content requires a subscription. You must be a premium user to view this content.
workshop paper
How can large language models become more human?
keywords:
sheaf theory
garden-path sentences
large language models
psycholinguistics
Psycholinguistic experiments reveal that efficiency of human language use is founded on predictions at both syntactic and lexical levels. Previous models of human prediction exploiting LLMs have used an information theoretic measure called \emph{surprisal}, with success on naturalistic text in a wide variety of languages, but under-performance on challenging text such as garden path sentences. This paper introduces a novel framework that combines the lexical predictions of an LLM with the syntactic structures provided by a dependency parser. The framework gives rise to an \emph{Incompatibility Fraction}. When tested on two garden path datasets, it correlated well with human reading times, distinguished between easy and hard garden path, and outperformed surprisal.