reward-bench

• Spaceallenai/reward-bench
• SDKgradio
• Likes436
• Licenseapache-2.0
• Updated2025-08-12T07:58:08.000Z

Open on Hugging Face

Reward Bench Leaderboard — text generation demo

• Availabilityrunning
• Last checked2026-09-28T22:18:40.993853+00:00
• Licenseapache-2.0
• Publisher (host)allenai
• Formatstext
• RequirementsNone detected
• Example inputPrompt / chat message
• Example outputGenerated text

• allenai/Llama-3.1-Tulu-3-70B underlying model
• allenai/Llama-3.1-Tulu-3-8B underlying model
• allenai/OLMoE-1B-7B-0924 underlying model
• allenai/open_instruct_dev underlying model
• allenai/tulu-2-dpo-70b underlying model
• ContextualAI/LMUnit-llama3.1-70b underlying model
• ContextualAI/LMUnit-qwen2.5-72b underlying model
• deepseek-ai/deepseek-math-7b-rl underlying model
• facebook/Self-taught-evaluator-llama3.1-70B underlying model
• general-preference/GPM-Llama-3.1-8B underlying model
• HuggingFaceH4/zephyr-7b-beta underlying model
• LxzGordon/URM-LLaMa-3.1-8B underlying model