Misalignment Detection on Agent-SafetyBench underspecified queries n=148
0.8Error Rate (ER)
ProvePrompt
Updated 22d ago
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
2026.05 | 0.8 | - | - | |
2026.05 | 1.1 | - | - | |
2026.05 | 1.1 | - | - | |
2026.05 | 1.2 | - | - | |
2026.05 | 1.4 | - | - | |
2026.05 | 1.4 | - | - | |
2026.05 | 1.4 | - | - | |
2026.05 | 1.8 | - | - | |
2026.05 | 2 | - | - | |
2026.05 | 2.2 | - | - | |
2026.05 | 2.6 | - | - | |
2026.05 | 4.6 | - | - | |
2026.05 | 5 | - | - | |
2026.05 | 6.8 | - | - | |
2026.05 | 6.8 | - | - | |
2026.05 | 8.5 | - | - | |
2026.05 | 11.5 | - | - | |
2026.05 | 12.3 | - | - | |
2026.05 | 17.8 | - | - | |
2026.05 | 17.8 | - | - | |
2026.05 | 27.3 | - | - | |
2026.05 | 34.6 | - | - | |
2026.05 | 38 | - | - | |
2026.05 | 40 | - | - | |
2026.05 | 45.9 | - | - | |
2026.05 | 46.4 | - | - | |
2026.05 | 47 | - | - | |
2026.05 | 56.4 | - | - | |
2026.05 | 60.8 | - | - | |
2026.05 | 70.9 | - | - |