Reward Modeling on Aggregate RewardBench PPE RMB RM-Bench JudgeBench
82.8Average ScoreSkywork-Reward-V2
Evaluation Results
| Method | Links | |
|---|---|---|
| Skywork-Reward-V2Backbone=gemma-2-9b-it2025.07 | 82.8 | |
| Skywork-Reward-V2Backbone=gemma-3-4b2025.07 | 77.1 | |
| Skywork-Reward-V2Backbone=gemma-3-1b-it2025.07 | 76.3 | |
| Skywork-Reward-V2Backbone=Qwen2.5-7B2025.07 | 75.4 | |
| Skywork-Reward-V2Backbone=gemma-2-2b-it2025.07 | 74 | |
| INF-ORM-Llama3.1-70B2025.07 | 73.5 | |
| CIR-AMS/BTRM_Qwen2_7b_06132025.07 | 67.2 | |
| Ray2333/GRM-gemma2-2B-rewardmodel-ft2025.07 | 65.8 |