Backdoor Detection on StrongREJECT prompts with triggers
100TPR
WeightWatch
Updated 1mo ago
Evaluation Results
| Method | Links | |
|---|---|---|
2025.07 | 100 | |
2025.07 | 100 | |
2025.07 | 100 | |
2025.07 | 100 | |
2025.07 | 98.72 | |
2025.07 | 97.76 | |
2025.07 | 97.44 | |
2025.07 | 96.49 | |
2025.07 | 63.58 | |
2025.07 | 51.76 | |
2025.07 | 43.45 | |
2025.07 | 31.95 | |
2025.07 | 26.84 | |
2025.07 | 19.81 | |
2025.07 | 11.5 | |
2025.07 | 7.99 | |
2025.07 | 2.88 | |
2025.07 | 1.92 | |
2025.07 | 1.92 | |
2025.07 | 0.32 |