Adversarial and Jailbreaking Attack Detection on MaliciousInstruct
0.8825AUROC
OpenAI Omni
Updated 4d ago
Evaluation Results
| Method | Links | ||
|---|---|---|---|
2026.02 | 0.8825 | 0.92 | |
2026.02 | 0.8718 | 1 | |
2026.02 | 0.8617 | 0.9982 | |
2026.02 | 0.851 | 0.71 | |
2026.02 | 0.8507 | 1 | |
2026.02 | 0.8501 | 0.78 | |
2026.02 | 0.828 | 0.59 | |
2026.02 | 0.7957 | 0.97 | |
2026.02 | 0.7745 | 1 | |
2026.02 | 0.7586 | 0.68 | |
2026.02 | 0.6828 | 1 | |
2026.02 | 0.5954 | 1 | |
2026.02 | 0.5791 | 0.93 | |
2026.02 | 0.3869 | 1 | |
2026.02 | 0.2994 | 1 | |
2026.02 | 0.258 | 1 | |
2026.02 | 0.2252 | 1 | |
2026.02 | 0.1913 | 1 | |
2026.02 | 0.1622 | 1 | |
2026.02 | 0.1441 | 1 |