RAG-RewardBench: Benchmarking Reward Models in Retrieval Augmented Generation for Preference Alignment Paper • 2412.13746 • Published 7 days ago • 8 • 2