Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
RewardBenchgithub.com · Low evidenceRewardBench is a benchmark for evaluating the capabilities and safety of reward models, including those trained with Direct Preference Optimization (DPO). It provides common inference code, dataset formatting, and analysis tools for testing various reward models.
Parse Score