i am currently not taking this at face value. frontier models seem to be doing something like "reward hacking in conversation" (that they weren't doing before intensive RLVR) that i don't yet…

Posters: QiaochuYuan. QiaochuYuan

3 likes recorded by the source corpus.

Topics: AI & LLMs, AI safety

View Post: https://x.com/QiaochuYuan/status/2053554419460096465