Over-refusal Evaluation on CoCoNot Contrast
1.58Over-refusal Rate
Categorical Steering
Updated 1mo ago
Evaluation Results
| Method | Links | |
|---|---|---|
2026.03 | 1.58 | |
2026.03 | 2.37 | |
2026.03 | 3.69 | |
2026.03 | 7.12 | |
2026.03 | 9.5 | |
2026.03 | 11.87 | |
2026.03 | 12.66 |
Categorical Steering
| Method | Links | |
|---|---|---|
2026.03 | 1.58 | |
2026.03 | 2.37 | |
2026.03 | 3.69 | |
2026.03 | 7.12 | |
2026.03 | 9.5 | |
2026.03 | 11.87 | |
2026.03 | 12.66 |