跳到内容
onsuggest
登录
文章 en

Baiting monitorable reward hacks

来源: LessWrong ·

This is joint work between Ziqian Zhong, Lawrence Feng, Gaurav Rohit Ghosal, Jiangqi Dai, and Aditi Raghunathan. This work is supported by a BlueDot Rapid Grant. We are unsure of whether this intervention is net positive (see discussions below) despite positive results, but we think similar techniques are worth further studying. Gradient routing can fail to mitigate reward hacks when the models settle on hacks not caught by the monitor. For example, if we swap the monitorable and unmonitorable set of hacks in the Countdown-Code environment, GRAFT still settles on the now-unmonitorable hacks in

什么是 Baiting monitorable reward hacks?

Baiting monitorable reward hacks is a 文章 discovery curated from LessWrong. This is joint work between Ziqian Zhong, Lawrence Feng, Gaurav Rohit Ghosal, Jiangqi Dai, and Aditi Raghunathan. This work is supported by a BlueDot Rapid Grant. We are unsure of…

如何访问或体验此内容?

You can experience this content directly on its original host (LessWrong) by clicking the "Open Original" button.

此内容来源于哪里?

Curated from LessWrong by the OnSuggest team to surface the most interesting and serendipitous content on the web.

键盘快捷键

上一条发现
←P
下一条发现
→N
喜欢
L
不适合我
X
保存
S
显示此帮助
?
关闭
Esc

快捷键在发现页面上有效。