2024 ICML ICML 2024

RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedback