# Image Classification on CIFAR-10 L2 Robustness (epsilon=0.5, test) Source: [Image Classification on CIFAR-10 L2 Robustness (epsilon=0.5, test)]() Collection index: [llms.txt]() Data updated: 2026-02-27T08:38:37.588Z Evaluation of Image Classification robustness on the CIFAR-10 test set under an L2 adversarial threat model with a perturbation constraint of epsilon=0.5, reporting standard and robust accuracy. 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 | Image Classification | | Dataset (including any stated split) | CIFAR-10 l2 threat model, epsilon=0.5 (test) | | Metric | Standard Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 11 of 11 | | 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 - [Standard Accuracy](): Higher is better; unit: PERCENT. - [Robust Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Standard Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Wang et al. | 96.09 | Not recorded | 68fa71031e209131cea4e879 | | ScoreOpt | 95.18 | Not recorded | 68fa71031e209131cea4e879 | | ScoreOpt | 94.99 | Not recorded | 68fa71031e209131cea4e879 | | Augustin et al. | 93.96 | Not recorded | 68fa71031e209131cea4e879 | | Rebuffi et al. | 92.41 | Not recorded | 68fa71031e209131cea4e879 | | Nie et al. | 92.15 | Not recorded | 68fa71031e209131cea4e879 | | Rebuffi et al. | 91.79 | Not recorded | 68fa71031e209131cea4e879 | | Nie et al. | 91.41 | Not recorded | 68fa71031e209131cea4e879 | | Sehwag et al. | 90.93 | Not recorded | 68fa71031e209131cea4e879 | | Yoon et al. | 86.76 | Not recorded | 68fa71031e209131cea4e879 | | Yoon et al. | 85.66 | Not recorded | 68fa71031e209131cea4e879 | ### Wang et al. Standard Accuracy: 96.09. Result ID: 69a157f535dfcabb00e66937. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"AT","Architecture":"WideResNet-70-16"}. This record does not assert independent reproduction. ### ScoreOpt Standard Accuracy: 95.18. Result ID: 69a157f535dfcabb00e6693a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Type":"AP","Architecture":"WideResNet-70-16"}. This record does not assert independent reproduction. ### ScoreOpt Standard Accuracy: 94.99. Result ID: 69a157f535dfcabb00e66935. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Type":"AP","Architecture":"WideResNet-28-10"}. This record does not assert independent reproduction. ### Augustin et al. Standard Accuracy: 93.96. Result ID: 69a157f535dfcabb00e66932. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"AT","Architecture":"WideResNet-28-10"}. This record does not assert independent reproduction. ### Rebuffi et al. Standard Accuracy: 92.41. Result ID: 69a157f535dfcabb00e66936. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"AT","Architecture":"WideResNet-70-16"}. This record does not assert independent reproduction. ### Nie et al. Standard Accuracy: 92.15. Result ID: 69a157f535dfcabb00e66939. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"AP","Architecture":"WideResNet-70-16"}. This record does not assert independent reproduction. ### Rebuffi et al. Standard Accuracy: 91.79. Result ID: 69a157f535dfcabb00e66931. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"AT","Architecture":"WideResNet-28-10"}. This record does not assert independent reproduction. ### Nie et al. Standard Accuracy: 91.41. Result ID: 69a157f535dfcabb00e66934. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"AP","Architecture":"WideResNet-28-10"}. This record does not assert independent reproduction. ### Sehwag et al. Standard Accuracy: 90.93. Result ID: 69a157f535dfcabb00e66930. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"AT","Architecture":"WideResNet-28-10"}. This record does not assert independent reproduction. ### Yoon et al. Standard Accuracy: 86.76. Result ID: 69a157f535dfcabb00e66938. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"AP","Architecture":"WideResNet-70-16"}. This record does not assert independent reproduction. ### Yoon et al. Standard Accuracy: 85.66. Result ID: 69a157f535dfcabb00e66933. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"AP","Architecture":"WideResNet-28-10"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.