Selective Refusal Editing on HarmBench (Gemma-3-4B-IT held-out split)
88.6Edit Refusal Rate
Base model (no intervention)
Updated 13d ago
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
2026.05 | 88.6 | 100 | 100 | 81.6 | 0 | |
| 4 | 95.5 | 87.3 | 65.3 | -16.3 | ||
2026.05 | 0.2 | 100 | 100 | 81.6 | 0 |
Base model (no intervention)
| Method | Links | |||||
|---|---|---|---|---|---|---|
2026.05 | 88.6 | 100 | 100 | 81.6 | 0 | |
| 4 | 95.5 | 87.3 | 65.3 | -16.3 | ||
2026.05 | 0.2 | 100 | 100 | 81.6 | 0 |