注册并分享邀请链接,可获得视频播放与邀请奖励。

Jake Ward
@_jake_ward
i'm trying to figure out the computer
加入 February 2019
363 正在关注    1.2K 粉丝
Models love reward hacking. Maybe we should just let them! In our new paper, we let models reward hack during RL but we dictate which parameters learn this behavior. When we disable these parameters for deployment, the model keeps its new capabilities but doesn't hack!
显示更多
0
106
3.5K
237
转发到社区