RLHFlow-PRM-Deepseek-8B
| Method | Links | |||
|---|---|---|---|---|
2025.05 | 24.2 | 98.4 | 38.8 | |
2025.05 | 32.4 | 91.7 | 47.9 | |
2025.05 | 33.8 | 99 | 50.4 | |
2025.05 | 46.9 | 42 | 44.3 | |
2025.05 | 50.2 | 71.5 | 59 | |
2025.05 | 51.2 | 44 | 47.3 | |
2025.05 | 53.1 | 95.3 | 68.2 | |
2025.05 | 53.6 | 87.6 | 66.5 | |
2025.05 | 56.5 | 83.9 | 67.6 | |
2025.05 | 56.5 | 90.2 | 69.5 | |
2025.05 | 59.9 | 86 | 70.6 | |
| 61.4 | 85.5 | 71.4 | ||
2025.05 | 61.8 | 82.9 | 70.8 |