Over-refusal Evaluation on OR-Bench (boundary cases)
1.7OR-FPR
Mistral-v2-7B
Updated 8d ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2026.05 | 1.7 | 56.4 | |
2026.05 | 4.6 | 74.8 | |
2026.05 | 7.4 | 56.5 | |
| 16.6 | 46.5 | ||
| 20.7 | 36.6 | ||
| 23.3 | 53.5 | ||
2026.05 | 24.4 | 37.9 | |
2026.05 | 25.5 | 49.8 | |
2026.05 | 25.6 | 31.4 | |
| 28.3 | 46.8 | ||
2026.05 | 31.4 | 30.1 | |
2026.05 | 31.8 | 42.2 | |
| 32.3 | 24.5 | ||
2026.05 | 35.5 | 40.3 | |
2026.05 | 35.7 | 20.7 | |
| 38.4 | 23.5 | ||
2026.05 | 49 | 14 | |
2026.05 | 83.3 | -26 |