News for hackers. Stuff that matters.

A comment on OpenAI Misled You on RLHF.

Re:OpenAI Misled You on RLHF by byyoung3 on Sunday August 17, 2025 @07:39AM (#44930903)
this seems to disagree with a lot of research showing RL is not necessary for reasoning -- im not sure about alignment
[ Parent | on HN ]

There are no comments on this comment yet.