Harmful Fine-tuning Attack on BeaverTails
64.5Attack Success Rate (ASR)
SFT
Updated 25d ago
Evaluation Results
| Method | Links | |
|---|---|---|
2026.06 | 64.5 | |
2026.06 | 63.6 | |
2026.06 | 62.5 | |
2026.06 | 61.4 | |
2026.06 | 60.8 | |
2026.06 | 60.1 | |
2026.06 | 47.7 | |
2026.06 | 46.4 | |
2026.06 | 37.8 | |
2026.06 | 17 | |
2026.06 | 14.3 | |
2026.06 | 8.8 |