Refusal Evaluation on OR-Bench Toxic
94.66Refusal Rate
Categorical Steering
Updated 1mo ago
Evaluation Results
| Method | Links | |
|---|---|---|
2026.03 | 94.66 | |
2026.03 | 94.5 | |
2026.03 | 90.69 | |
2026.03 | 90.53 | |
2026.03 | 86.56 | |
2026.03 | 86.26 | |
2026.03 | 85.95 |
Categorical Steering
| Method | Links | |
|---|---|---|
2026.03 | 94.66 | |
2026.03 | 94.5 | |
2026.03 | 90.69 | |
2026.03 | 90.53 | |
2026.03 | 86.56 | |
2026.03 | 86.26 | |
2026.03 | 85.95 |