1 paper
Yuko Nakagi, Keigo Tada, Sota Yoshino +2
Large language models (LLMs) often exhibit abrupt emergent behavior, whereby new abilities arise at certain points during their training. This phenomenon, commonly referred to as a…