| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| CIFAR-10 (test) | FGSM | Attack Success Rate (ASR)47.7 | 76 | 2d ago | |
| CIFAR-100 (test) | FAT | Natural Acc66.74 | 46 | 4d ago | |
| ImageNet 1k (test) | RVT-B* | FGSM Robustness53 | 34 | 3d ago | |
| CIFAR100 1000 images (val) | SA | Clean Acc70.25 | 24 | 4d ago | |
| CIFAR10 1000 images (test) | Square Attack (SA) | Robust Accuracy72.3 | 24 | 4d ago | |
| GCG | GCG Rate0.13 | 21 | 4d ago | ||
| CIFAR-100 LT (test) | AT-BSL-AuA | Clean Accuracy50.66 | 20 | 4d ago | |
| GTSRB (test) | FGSM-Neuro-Symbolic | FGSM Accuracy74 | 18 | 4d ago | |
| AutoDAN | ASR0 | 18 | 4d ago | ||
| PAIR | ASR26 | 18 | 4d ago | ||
| ImageNet 1k (val) | BG_Random | Accuracy13.4 | 18 | 4d ago | |
| ImageNet sr=90% (val) | AT | Clean Accuracy83.29 | 14 | 4d ago | |
| HarmBench | DR56.25 | 12 | 4d ago | ||
| CIFAR-100 | Cutout | Final Auto-Attack Accuracy98.5 | 12 | 4d ago | |
| Adversarial Robustness Prompts | GCG ASR92 | 11 | 4d ago | ||
| NSFW | AdvUnlearn | ASR4.69 | 11 | 4d ago | |
| CIFAR-10 L2, epsilon=8/255 | Stutz et al. (2020) | Robust Acc (AA)77.64 | 11 | 4d ago | |
| STL10 (test) | ZePAD | Baseline Acc82.07 | 10 | 4d ago | |
| CIFAR-10 L-infinity, epsilon=4/255 | Xiao et al. (2020) ADV | Robust Accuracy (AA)64.91 | 10 | 4d ago | |
| Church | SafeRedir | ASR3.12 | 9 | 4d ago | |
| Van Gogh | SafeRedir | ASR50.16 | 9 | 4d ago | |
| CIFAR-100 sr=90% (test) | AT | Clean Accuracy64.37 | 9 | 4d ago | |
| CIFAR-100 l_inf, epsilon=8/255 (test) | IKL-AT | Clean Accuracy73.85 | 9 | 4d ago | |
| CIFAR10 | AAA | Accuracy97.7 | 8 | 3d ago | |
| ImageNet-LT | FGSM-Our | Clean Accuracy30.02 | 8 | 4d ago |
