# Adversarial Robustness on CIFAR-10 l\_inf, epsilon=8/255 (test) Source: [Adversarial Robustness on CIFAR-10 l\_inf, epsilon=8/255 (test)]() Collection index: [llms.txt]() Data updated: 2026-06-18T07:12:17.280Z Evaluation of model adversarial robustness on the CIFAR-10 test set under L-infinity perturbation constraints with epsilon set to 8/255. Metrics include clean accuracy and robustness against AA attacks. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Adversarial Robustness | | Dataset (including any stated split) | CIFAR-10 l\_inf, epsilon=8/255 (test) | | Metric | Clean Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 10 of 10 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Clean Accuracy](): Higher is better; unit: PERCENT. - [Adversarial Accuracy (AA)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Clean Accuracy | Result date | Report paper | | --- | --- | --- | --- | | DM-AT | 92.44 | Not recorded | 68fa775a1e209131cea72a2d | | GKL-AT | 92.16 | Not recorded | 68fa775a1e209131cea72a2d | | LBGAT | 88.7 | Not recorded | 68fa775a1e209131cea72a2d | | ACAT | 88.64 | Not recorded | 68fa775a1e209131cea72a2d | | LAS-AT | 87.74 | Not recorded | 68fa775a1e209131cea72a2d | | AWP | 85.36 | Not recorded | 68fa775a1e209131cea72a2d | | Rice et al. | 85.34 | Not recorded | 68fa775a1e209131cea72a2d | | GKL-AT | 85.2 | Not recorded | 68fa775a1e209131cea72a2d | | GKL-AT | 84.8 | Not recorded | 68fa775a1e209131cea72a2d | | ACAT | 82.41 | Not recorded | 68fa775a1e209131cea72a2d | ### DM-AT Clean Accuracy: 92.44. Result ID: 6a339a4bfbc373592c3065fa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Architecture":"WRN-28-10","Augmentation Type":"20M Generated Data"}. This record does not assert independent reproduction. ### GKL-AT Clean Accuracy: 92.16. Result ID: 6a339a4bfbc373592c3065fb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Architecture":"WRN-28-10","Augmentation Type":"20M Generated Data"}. This record does not assert independent reproduction. ### LBGAT Clean Accuracy: 88.7. Result ID: 6a339a4bfbc373592c3065f3. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Architecture":"WRN-34-20","Augmentation Type":"Basic"}. This record does not assert independent reproduction. ### ACAT Clean Accuracy: 88.64. Result ID: 6a339a4bfbc373592c3065f8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Architecture":"WRN-34-10","Augmentation Type":"AutoAug"}. This record does not assert independent reproduction. ### LAS-AT Clean Accuracy: 87.74. Result ID: 6a339a4bfbc373592c3065f5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Architecture":"WRN-34-10","Augmentation Type":"Basic"}. This record does not assert independent reproduction. ### AWP Clean Accuracy: 85.36. Result ID: 6a339a4bfbc373592c3065f4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Architecture":"WRN-34-10","Augmentation Type":"Basic"}. This record does not assert independent reproduction. ### Rice et al. Clean Accuracy: 85.34. Result ID: 6a339a4bfbc373592c3065f2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Architecture":"WRN-34-20","Augmentation Type":"Basic"}. This record does not assert independent reproduction. ### GKL-AT Clean Accuracy: 85.2. Result ID: 6a339a4bfbc373592c3065f9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Architecture":"WRN-34-10","Augmentation Type":"AutoAug"}. This record does not assert independent reproduction. ### GKL-AT Clean Accuracy: 84.8. Result ID: 6a339a4bfbc373592c3065f7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Architecture":"WRN-34-10","Augmentation Type":"Basic"}. This record does not assert independent reproduction. ### ACAT Clean Accuracy: 82.41. Result ID: 6a339a4bfbc373592c3065f6. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Architecture":"WRN-34-10","Augmentation Type":"Basic"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.