RobustBench
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Tool Use | RobustBench-TC Clean 1.0 (test) | Clean Accuracy77.9 | 21 | |
| Robust Image Classification | RobustBench (test) | RA90 | 12 | |
| Adversarial Classification ($W_2$ distributional threat model) | RobustBench | Adv Acc (W-PGD-CE)74.62 | 2 | |
| Adversarial Classification ($W_{\infty}$ distributional threat model) | RobustBench | Adversarial Accuracy (AutoAttack)75.78 | 2 |
Showing 4 of 4 rows