Mitigating Reward Hacking in RLHF via Bayesian Non-negative Reward Modeling

Add code
Feb 11, 2026

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: