You Jiacheng
@YouJiacheng
加入 August 2015
2.4K 正在關注 12.6K 粉絲
this is good
RL with LLMs is very unstable when training and sampling policies differ. Standard fixes (matching numerics, importance sampling) work around the problem.
We find the root cause of this instability from first principles and propose a way to directly cancel it. Score Centering is competitive and compatible with existing approaches — while simple to implement!
🧵 [1/6]
顯示更多
