Tool-use Policy Auditing on τ2-BENCH Telecom (test)
30PASS^4
Baseline
Updated 26d ago
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
2026.06 | 30 | 47.6 | 10.5 | |
2026.06 | 30 | 42.9 | 15.8 | |
2026.06 | 25 | 47.6 | 0 |
Baseline
| Method | Links | |||
|---|---|---|---|---|
2026.06 | 30 | 47.6 | 10.5 | |
2026.06 | 30 | 42.9 | 15.8 | |
2026.06 | 25 | 47.6 | 0 |