Mitigating Reward Over-optimization in Direct Alignment Algorithms with Importance Sampling

Add code
Jun 11, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: