Michael Hahn

@mhahn29

LLM theory and interpretability. Professor at Saarland University @LstSaar @SIC_Saar. Previously PhD at Stanford @stanfordnlp. Hiring PhD students and Postdocs

following
followers

Can transformer LLMs hide reasoning from chain-of-thought monitors?

A post argues they cannot hide complex reasoning, but might cryptographically obscure their chain-of-thought traces.

2 Sources, 20d ago, first seen 20d ago

Featured in

Can transformer LLMs hide reasoning from chain-of-thought monitors?

Featured in

Can transformer LLMs hide reasoning from chain-of-thought monitors?