Tag
This paper investigates how language model representations predict neural activity during naturalistic language comprehension across MEG, ECoG, and other recordings. The findings demonstrate that language model features serve as useful neural predictors, but caution against overinterpreting predictive success as evidence for shared neural organization.