The First Few Tokens Matter: Understanding Chain-of-Thought Controllability

arXiv:2610.40939v1 Announce Type: new Abstract: Chain-of-thought (CoT) monitoring is a promising way to oversee increasingly capable agents. It succeeds in part because current models appear limited in their ability to control what they verbalize in their reasoning traces. We investigate the source

arXiv cs.AIMontag 06:00 · 02.11.1 Min. Lesezeit