Tal

LLMs are poor models of human language due to being "too good," researchers argue

A recent perspective argues that Large Language Models (LLMs) are not effective models of human brain linguistic processes because they are "too good." The authors suggest that to better model human linguistic predictio…