Refusal Evaluation on HarmfulQA
85.31Refusal Rate
Categorical Steering
Updated 1mo ago
Evaluation Results
| Method | Links | |
|---|---|---|
2026.03 | 85.31 | |
2026.03 | 75.87 | |
2026.03 | 75.61 | |
2026.03 | 71.99 | |
2026.03 | 66.07 | |
2026.03 | 61.58 | |
2026.03 | 60.15 |
Categorical Steering
| Method | Links | |
|---|---|---|
2026.03 | 85.31 | |
2026.03 | 75.87 | |
2026.03 | 75.61 | |
2026.03 | 71.99 | |
2026.03 | 66.07 | |
2026.03 | 61.58 | |
2026.03 | 60.15 |