Eyon Jang
@eyonjang
AGI safety researcher (MATS 8)
参加 December 2012
1.2K フォロー中 131 ファン
Can LLMs learn to resist RL training?
We empirically study exploration hacking: models controlling behavior during RL to prevent unwanted capabilities from being reinforced.
Joint work with @GoogleDeepMind and @MATSprogram.
More in the thread below 👇
もっと見る
