Certified Robustness on CIFAR-10 (L1 Certified Accuracy Profile)
86.1Certified Accuracy (L1, R=0.5)Uniform, LBS
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Uniform, LBSNoise Distribution=Uniform2026.04 | 86.1 | 85 | 83.8 | 82.7 | 81.9 | 81 | |
| Laplace, LBSNoise Distribution=Laplace2026.04 | 84.3 | 82 | 80.2 | 78.4 | 77 | 76.1 | |
| Uniform (w/), Yang et al. (2020)Noise Distribution=Uniform, Noise-Augmented Training=true2026.04 | 69.8 | 59.4 | 50.3 | 43.7 | 31.9 | 29.2 | |
| Laplace (w/), Teng et al. (2020)Noise Distribution=Laplace, Noise-Augmented Training=true2026.04 | 60.4 | 38.8 | 24.1 | 16.3 | 10.7 | 7.1 | |
| Laplace (w/), Lecuyer et al. (2019)Noise Distribution=Laplace, Noise-Augmented Training=true2026.04 | 55.7 | 33.2 | 21.6 | 17 | 10.5 | 6.4 | |
| Laplace (w/o), Teng et al. (2020)Noise Distribution=Laplace, Noise-Augmented Training=false2026.04 | 12.8 | 12.3 | 12.2 | 11.2 | 10.4 | 10.2 | |
| Laplace (w/o), Lecuyer et al. (2019)Noise Distribution=Laplace, Noise-Augmented Training=false2026.04 | 10.6 | 10 | 9.4 | 8.3 | 7.8 | 7.1 | |
| Uniform (w/o), Yang et al. (2020)Noise Distribution=Uniform, Noise-Augmented Training=false2026.04 | 10.2 | 10.4 | 9.2 | 10.6 | 10.1 | 9.1 |