Reward Modeling on Anthropic HH-helpful core250
0.292Delta RMTEA
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| TEAN=64, Reward Model=Skywork/Skywork-Reward-V2-Qwen3-8B2026.05 | 0.292 | 60.8 | 0.8 | 38.4 | |
| TEAN=32, Reward Model=Skywork/Skywork-Reward-V2-Qwen3-8B2026.05 | 0.249 | 63.6 | 0.8 | 35.6 | |
| TEAN=128, Reward Model=Skywork/Skywork-Reward-V2-Qwen3-8B2026.05 | 0.248 | 58.8 | 1.6 | 39.6 | |
| TEAN=16, Reward Model=Skywork/Skywork-Reward-V2-Qwen3-8B2026.05 | 0.212 | 67.6 | 0 | 32.4 | |
| TEAN=256, Reward Model=Skywork/Skywork-Reward-V2-Qwen3-8B2026.05 | 0.209 | 56 | 3.2 | 40.8 | |
| TEAN=8, Reward Model=Skywork/Skywork-Reward-V2-Qwen3-8B2026.05 | 0.196 | 69.6 | 0 | 30.4 |