Domain-Aware Reinforcement Fine-Tuning
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| 24.6 | 29.9 | 20.6 | 29.6 | 27.8 | 18.4 | 21.3 | ||
2026.01 | 22.7 | 27.3 | 18.8 | 27.9 | 25.9 | 17.1 | 19.2 | |
2026.01 | 22.1 | 27.1 | 17.9 | 26.7 | 25.1 | 16.9 | 18.7 | |
| 21 | 27.6 | 18.6 | 17.9 | 25.8 | 16.7 | 19.5 | ||
2026.01 | 19.6 | 25.3 | 17.7 | 16.8 | 24.2 | 15.8 | 17.9 | |
2026.01 | 17.8 | 22.2 | 15.4 | 15.8 | 23.8 | 13.6 | 15.9 | |
2026.01 | 11.3 | 15.6 | 5.71 | 13.2 | 19.8 | 8.66 | 4.6 |