Agent Safety Auditing on ATBench
85.5Accuracy
TRACES
Updated 6d ago
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
2026.05 | 85.5 | 86.3 | 91.9 | 41.4 | 77.9 | |
2026.05 | 83 | 84.5 | 93.9 | 47.5 | 82.2 | |
2026.05 | 82.5 | 80.7 | 73.7 | 54.5 | 62.9 | |
2026.05 | 78 | 68.9 | 56.4 | 16.4 | 56 | |
2026.05 | 77 | 76.5 | 75.8 | 22.2 | 58.8 | |
2026.05 | 74.8 | 77 | 86.8 | 34.9 | 57.1 | |
2026.05 | 72.8 | 74.5 | 77.9 | 35.8 | 52.2 | |
2026.05 | 57.5 | 47.2 | 38.4 | 20.2 | 53.2 | |
2026.05 | 55.6 | 44.3 | 36.1 | 15.2 | 54.3 | |
2026.05 | 53.5 | 26.8 | 17.1 | 12.1 | 55.9 | |
2026.05 | 52.5 | 9.5 | 5.1 | 1 | 50.4 | |
2026.05 | 51.5 | 62 | 80 | 54.5 | 55 | |
2026.05 | 49.5 | 66.2 | 100 | 53.7 | 50 |