Author of the publication

REBEL: A Regularization-Based Solution for Reward Overoptimization in Reinforcement Learning from Human Feedback.

, , , , , and . CoRR, (2023)