Context Limitations Make Neural Language Models More Human-Like
This work tackles the problem of making language models more cognitively plausible for researchers in computational linguistics and cognitive science, though it is incremental in nature.
The study addressed the discrepancy between neural language models' context access capacities and human reading behavior, finding that constraining context access improved simulation of human reading patterns, with specific syntactic constructions identified as key factors.
Language models (LMs) have been used in cognitive modeling as well as engineering studies -- they compute information-theoretic complexity metrics that simulate humans' cognitive load during reading. This study highlights a limitation of modern neural LMs as the model of choice for this purpose: there is a discrepancy between their context access capacities and that of humans. Our results showed that constraining the LMs' context access improved their simulation of human reading behavior. We also showed that LM-human gaps in context access were associated with specific syntactic constructions; incorporating syntactic biases into LMs' context access might enhance their cognitive plausibility.