Across the Levels of Analysis: Explaining Predictive Processing in Humans Requires More Than Machine-Estimated Probabilities
This work addresses theoretical gaps in cognitive science and AI by challenging assumptions about the role of machine learning in explaining human cognition, but it is incremental as it builds on existing critiques without new empirical results.
The paper critiques claims that language models are central to understanding human language processing and that psycholinguistics depends on them, proposing future directions to integrate LLMs with psycholinguistic models.
Under the lens of Marr's levels of analysis, we critique and extend two claims about language models (LMs) and language processing: first, that predicting upcoming linguistic information based on context is central to language processing, and second, that many advances in psycholinguistics would be impossible without large language models (LLMs). We further outline future directions that combine the strengths of LLMs with psycholinguistic models.