Adversarial Robustness Error Metrics on CIFAR-100 (test)
22.78Adversarial Error RateDenseNet
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| DenseNetTraining=REVa2025.09 | 22.78 | 27.8 | 31.14 | 27.97 | 27.43 | 31.41 | 29.58 | 29.22 | |
| WideResNetTraining=REVa2025.09 | 23.98 | 28.77 | 31.86 | 28.87 | 28.26 | 30.01 | 30.28 | 29.68 | |
| WideResNetTraining=AugMix2025.09 | 24.2 | 34.09 | 41.53 | 38.18 | 34.48 | 52.18 | 38.92 | 39.9 | |
| DenseNetTraining=AugMix2025.09 | 24.5 | 35.93 | 43.5 | 39.55 | 34.94 | 56.09 | 39.74 | 41.63 | |
| AllConvNetTraining=REVa2025.09 | 25.02 | 27.77 | 31.32 | 27.65 | 27.6 | 31.21 | 29.54 | 29.18 | |
| AllConvNetTraining=Standard2025.09 | 25.1 | 54.56 | 71.34 | 70.52 | 56.3 | 87.14 | 60.86 | 66.8 | |
| AllConvNetTraining=AugMix2025.09 | 25.59 | 38.91 | 49.81 | 45.83 | 39.22 | 64.25 | 43.63 | 46.94 | |
| WideResNetTraining=Standard2025.09 | 27.9 | 58.67 | 66.47 | 68.03 | 56.75 | 83.43 | 62.18 | 65.92 | |
| DenseNetTraining=Standard2025.09 | 28.6 | 57.95 | 69.31 | 69.92 | 58.52 | 86.5 | 63.54 | 67.62 |