GRPO_Qwen
| Method | Links | |||
|---|---|---|---|---|
2026.04 | 76.1 | 20.2 | 9.4 | |
2026.04 | 75.3 | 23.1 | 10.4 | |
2026.04 | 74.3 | 28.9 | 15.4 | |
2026.04 | 73.4 | 25.8 | 11.8 | |
2026.04 | 73.2 | 24.8 | 12.2 | |
2026.04 | 72.4 | 21.4 | 11 | |
2026.04 | 71.2 | 28.1 | 14.1 | |
2026.04 | 70.3 | 57.7 | 14.1 | |
2026.04 | 69.1 | 34.9 | 17.6 | |
2026.04 | 67.9 | 55.6 | 18.3 | |
2026.04 | 66 | 42.5 | 19.9 | |
2026.04 | 55.4 | 66 | 22.9 | |
2026.04 | 55.3 | 53.5 | 22.8 | |
2026.04 | 52.6 | 57.6 | 29.1 | |
2026.04 | 40.2 | 70.5 | 29 |