Image Classification on CIFAR-10 (Empirical Adversarial Robustness - FGSM/PGD)
93.32Clean AccuracyVanilla
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| VanillaModel=WRN-16, Train=BPTT2026.02 | 93.32 | 14.05 | 31.21 | 0 | 0 | 0 | 0 | |
| RATModel=WRN-16, Train=RAT2026.02 | 91.44 | 42.02 | 75.89 | 23.9 | 19.81 | 16.24 | 14.18 | |
| ATModel=WRN-16, Train=AT2026.02 | 91.32 | 39.14 | 74.31 | 21.15 | 17.45 | 14.41 | 12.93 | |
| TGOModel=WRN-16, Train=BPTT2026.02 | 88.79 | 51.4 | 71.38 | 11.67 | 6.14 | 1.52 | 0.45 | |
| TGOModel=WRN-16, Train=AT2026.02 | 88.16 | 63.03 | 79.69 | 42.52 | 35.01 | 24.76 | 20.11 | |
| TGOModel=WRN-16, Train=RAT2026.02 | 87.33 | 69.16 | 79.28 | 54.59 | 47.69 | 38.07 | 33.13 |