Reward Modeling on Weighted-sum (test)
79Pairwise AccuracyGemma-2-2B
Evaluation results
| Method | Links | |
|---|---|---|
| Gemma-2-2BTraining data setup=Weighted-sum & Targetedness ↑2026.03 | 79 | |
| Gemma-2-2BTraining data setup=Combined2026.03 | 76 | |
| Gemma-2-2BTraining data setup=Weighted-sum & Mistake ID & Location ↑2026.03 | 76 | |
| Qwen2.5-1.5BTraining data setup=Combined2026.03 | 74 | |
| Gemma-2-2BTraining data setup=Weighted-sum2026.03 | 74 | |
| Qwen2.5-0.5BTraining data setup=Combined2026.03 | 73 | |
| Qwen2.5-0.5BTraining data setup=Weighted-sum & Mistake ID & Location ↑2026.03 | 72 | |
| Qwen2.5-0.5BTraining data setup=Weighted-sum & Targetedness ↑2026.03 | 71 | |
| Qwen2.5-1.5BTraining data setup=Weighted-sum & Targetedness ↑2026.03 | 71 | |
| Qwen2.5-1.5BTraining data setup=Weighted-sum & Mistake ID & Location ↑2026.03 | 71 | |
| Qwen2.5-0.5BTraining data setup=Weighted-sum2026.03 | 70 | |
| Qwen2.5-1.5BTraining data setup=Weighted-sum2026.03 | 70 | |
| Gemma-2-2BTraining data setup=Synthetic2026.03 | 70 | |
| Qwen2.5-0.5BTraining data setup=Synthetic2026.03 | 66 | |
| Qwen2.5-1.5BTraining data setup=Synthetic2026.03 | 66 |