Large language model tokens are psychologically salient.

David A. Haslett, Antoni B. Chan, Janet H. Hsiao

Browse the full CogSci paper archive.