Safety Classification on SafeEditBench
100Policy L1 Success Rate
Llama Guard
Updated 1mo ago
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
2026.03 | 100 | 0 | 0 | 42.37 | 0 | 22.73 | |
2026.03 | 65.25 | 15.38 | 24 | 60.87 | 44 | 36.65 | |
2026.03 | 54.24 | 23.08 | 35.9 | 72.58 | 66.17 | 49.43 | |
2026.03 | 49.15 | 16.67 | 31.82 | 66.13 | 64.62 | 44.16 | |
2026.03 | 47.46 | 20.59 | 37.36 | 70.87 | 70.34 | 48.68 | |
2026.03 | 44.92 | 5.63 | 14.12 | 37.29 | 51.06 | 27.5 | |
2026.03 | 44.07 | 20.29 | 35.16 | 70.4 | 64.71 | 45.59 | |
2026.03 | 16.1 | 13.56 | 32.61 | 66.02 | 52.34 | 32.76 | |
2026.03 | 11.02 | 9.68 | 27.94 | 49.37 | 49.7 | 30.52 | |
2026.03 | 0 | 14.68 | 27.54 | 67.95 | 73.37 | 41.16 | |
2026.03 | 0 | 0 | 1.63 | 22.39 | 0 | 5.19 |