#ModelScorePaper
1PaTaRM: Bridging Pairwise and Pointwise Signals via Preference-Aware Task-Adaptive Reward Modeling86.20link
2Skywork/Skywork-Reward-V2-Llama-3.1-8B84.13link
3ContextualAI/LMUnit-qwen2.5-72b82.08link
4ContextualAI/LMUnit-llama3.1-70b80.54link
5Databricks-Mosaic-Research/PGRM80.02link
6google/gemini-2.5-pro79.48link
7Skywork/Skywork-Reward-V2-Qwen3-8B78.37link
8google/gemini-2.5-flash77.67link
9nicolinho/QRM-Gemma-2-27B76.67link
10infly/INF-ORM-Llama3.1-70B76.48link
11anthropic/claude-opus-4-2025051476.48link
12allenai/Llama-3.1-70B-Instruct-RM-RB276.06link
13Skywork/Skywork-Reward-Gemma-2-27B75.76link
14Skywork/Skywork-Reward-V2-Qwen3-4B75.51link
15anthropic/claude-3-7-sonnet-2025021975.39link
16Skywork/Skywork-Reward-Gemma-2-27B-v0.275.31link
17Skywork/Skywork-Reward-V2-Llama-3.2-3B74.66link
18LxzGordon/URM-LLaMa-3.1-8B73.94link
19Schrieffer/Llama-SARM-4B73.79link
20Skywork/Skywork-Reward-Llama-3.1-8B73.14link
21allenai/Llama-3.1-8B-Instruct-RM-RB272.85link
22ShikaiChen/LDL-Reward-Gemma-2-27B-v0.172.49link
23openai/gpt-4.1-2025-04-1472.32link
24allenai/Llama-3.1-Tulu-3-70B-SFT-RM-RB272.20link
25Skywork/Skywork-Reward-Llama-3.1-8B-v0.271.75link
26anthropic/claude-sonnet-4-2025051471.17link
27nicolinho/QRM-Llama3.1-8B-v270.74link
28HFXM/RAMO-Llama3.1-8B69.17link
29Skywork/Skywork-VL-Reward-7B68.85link
30allenai/Llama-3.1-Tulu-3-8B-RL-RM-RB268.71link
31allenai/Llama-3.1-Tulu-3-8B-DPO-RM-RB268.70link
32Personalized RewardBench: Evaluating Reward Models with Human Aligned Personalization0.66link
33Your Mouse and Eyes Secretly Leak Your Preference: LLM Alignment using Implicit Feedback from Users0.53link
RewardBench 2 rewardbench-2 Leaderboard