Spoofing Attack Traceability on RealToxicityPrompts (test)
90.11AUC
DualGuard
Updated 1mo ago
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
2025.12 | 90.11 | 52 | 66 | 87.62 | |
2025.12 | 85.13 | 37.5 | 57 | 87.62 | |
2025.12 | 64.45 | 22.5 | 30.5 | 58.9 | |
2025.12 | 59.61 | 6.5 | 15.5 | 50.75 | |
2025.12 | 59.18 | 10.5 | 17.5 | 54.29 | |
2025.12 | 55.41 | 5 | 15.5 | 49.97 | |
2025.12 | 54.65 | 8.5 | 18 | 48.15 | |
2025.12 | 54.62 | 5.88 | 11.75 | 54.21 | |
2025.12 | 54.51 | 5.5 | 11.5 | 48.19 | |
2025.12 | 54.41 | 9.5 | 14.5 | 52.77 | |
2025.12 | 53.79 | 6.17 | 12.33 | 54.21 | |
2025.12 | 53.36 | 9.5 | 20 | 58.9 | |
2025.12 | 51.14 | 4.5 | 10.5 | 52.77 | |
2025.12 | 49.48 | 5 | 8.5 | 47.43 | |
2025.12 | 49.4 | 5 | 9 | 54.29 | |
2025.12 | 45.39 | 1.01 | 3.54 | 47.43 | |
2025.12 | 44.54 | 3.5 | 5 | 49.97 | |
2025.12 | 41.9 | 2.5 | 4.5 | 50.75 | |
2025.12 | 41.88 | 1.5 | 4.5 | 48.19 | |
2025.12 | 41.66 | 1.5 | 9 | 48.15 |