# Image Classification on CIFAR-10 (test) (Uniform & Class Dependent Robustness Metrics) Source: [Image Classification on CIFAR-10 (test) (Uniform & Class Dependent Robustness Metrics)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:26:52.735Z Evaluation of Image Classification model robustness on the CIFAR-10 test set using a suite of metrics measuring accuracy under varying levels of Uniform and Class Dependent noise perturbations (eta). 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 (test) | | Metric | Acc (Uniform η=0.2) | | 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 - [Acc (Uniform η=0.2)](): Higher is better; unit: PERCENT. - [Acc (Uniform η=0.4)](): Higher is better; unit: PERCENT. - [Acc (Uniform η=0.6)](): Higher is better; unit: PERCENT. - [Acc (Uniform η=0.8)](): Higher is better; unit: PERCENT. - [Acc (Class Dependent η=0.1)](): Higher is better; unit: PERCENT. - [Acc (Class Dependent η=0.2)](): Higher is better; unit: PERCENT. - [Acc (Class Dependent η=0.3)](): Higher is better; unit: PERCENT. - [Acc (Class Dependent η=0.4)](): Higher is better; unit: PERCENT. - [Acc (20% Noise)](): Higher is better; unit: PERCENT. - [Acc (40% Noise)](): Higher is better; unit: PERCENT. - [Acc (60% Noise)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Acc (Uniform η=0.2) | Result date | Report paper | | --- | --- | --- | --- | | Lq loss | 89.83 | Not recorded | 68fa651f1e209131cea088eb | | Truncated Lq loss | 89.7 | Not recorded | 68fa651f1e209131cea088eb | | Forward T | 88.63 | Not recorded | 68fa651f1e209131cea088eb | | Forward ^T | 87.99 | Not recorded | 68fa651f1e209131cea088eb | | CCE | 86.98 | Not recorded | 68fa651f1e209131cea088eb | | MAE | 83.72 | Not recorded | 68fa651f1e209131cea088eb | ### Lq loss Acc (Uniform η=0.2): 89.83. Result ID: 698c9f100b614d12cff203ab. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-18","q":0.7}. This record does not assert independent reproduction. ### Truncated Lq loss Acc (Uniform η=0.2): 89.7. Result ID: 698c9f100b614d12cff203ac. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-18","q":0.7}. This record does not assert independent reproduction. ### Forward T Acc (Uniform η=0.2): 88.63. Result ID: 698c9f100b614d12cff203a9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-18"}. This record does not assert independent reproduction. ### Forward ^T Acc (Uniform η=0.2): 87.99. Result ID: 698c9f100b614d12cff203aa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-18"}. This record does not assert independent reproduction. ### CCE Acc (Uniform η=0.2): 86.98. Result ID: 698c9f100b614d12cff203a7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-18"}. This record does not assert independent reproduction. ### MAE Acc (Uniform η=0.2): 83.72. Result ID: 698c9f100b614d12cff203a8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-18"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.