Artificial neural network language models predict human brain responses to language even after a developmentally realistic amount of training
Eghbal A. Hosseini, Martin Schrimpf, Yian Zhang, Samuel R. Bowman, Noga Zaslavsky, Evelina G. Fedorenko
The paper shows that language models trained on developmentally realistic data can predict human brain responses to language, achieving near-maximal performance with 100 million words of training.
Neurobiology of Language · Jan 1, 2024Peer-reviewedNeuroscience