Tabular Classification on URL (test)
99.5Clean AccuracyTabNet
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| TabNetRobustification=Madry adversarial training2025.06 | 99.5 | 93.8 | 91.8 | |
| TabNetRobustification=Madry adversarial training2025.06 | 99.5 | 93.8 | 91.8 | |
| TabICLRobustification=AICL (CAA)2025.06 | 97.3 | 97 | 75.6 | |
| TabICLRobustification=AICL, Hardening attack=CAA2025.06 | 97.3 | 97 | 75.6 | |
| TabICLRobustification=AICL (CAPGD)2025.06 | 96.4 | 95.8 | 31 | |
| TabICLRobustification=AICL, Hardening attack=CAPGD2025.06 | 96.4 | 95.8 | 31 | |
| TabPFNv2Robustification=CAA2025.06 | 96.2 | 96.1 | 69.4 | |
| TabPFNv2Robustification=AICL, Hardening attack=CAA2025.06 | 96.2 | 96.1 | 69.4 | |
| TabPFNv2Robustification=CAPGD2025.06 | 95.6 | 94.7 | 25.9 | |
| TabPFNv2Robustification=AICL, Hardening attack=CAPGD2025.06 | 95.6 | 94.7 | 25.9 | |
| RLNRobustification=Madry adversarial training2025.06 | 95.2 | 81.3 | 52.3 | |
| RLNRobustification=Madry adversarial training2025.06 | 95.2 | 81.3 | 52.3 | |
| STGRobustification=Madry adversarial training2025.06 | 94.3 | 91.6 | 89.8 | |
| STGRobustification=Madry adversarial training2025.06 | 94.3 | 91.6 | 89.8 | |
| TabTr.Robustification=Madry adversarial training2025.06 | 93.9 | 81 | 52.9 | |
| TabTr.Robustification=Madry adversarial training2025.06 | 93.9 | 81 | 52.9 | |
| VIMERobustification=Madry adversarial training2025.06 | 93.4 | 78.8 | 67.8 | |
| VIMERobustification=Madry adversarial training2025.06 | 93.4 | 78.8 | 67.8 |