Language models for cognitive science

Language models offer new tools to address longstanding challenges in cognitive science. 

We outline key use cases, including helping reduce research silos, supporting formalization, consolidating measurement taxonomies, enabling integrative modeling frameworks, and capturing contextual variation.

The promise of language models is contingent on their responsible use; they should complement human expertise and be deployed with safeguards against risks such as oversimplification, opacity, deskilling, and bias.

https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(26)00158-0

https://sciencemission.com/cognitive-science-with-language-models