# Image Classification on CIFAR-100 clean (test) (Epsilon-Dependent Accuracy) Source: [Image Classification on CIFAR-100 clean (test) (Epsilon-Dependent Accuracy)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:34:34.601Z Evaluation of Image Classification performance on the clean test split of the CIFAR-100 dataset, measuring accuracy under varying levels of perturbation defined by epsilon (ε=0.2, 0.4, 0.6). 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-100 clean (test) | | Metric | Accuracy (ε=0.2) | | 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 - [Accuracy (ε=0.2)](): Higher is better; unit: PERCENT. - [Accuracy (ε=0.4)](): Higher is better; unit: PERCENT. - [Accuracy (ε=0.6)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy (ε=0.2) | Result date | Report paper | | --- | --- | --- | --- | | CORES2\* | 72.91 | Not recorded | 68fa698f1e209131cea22a5b | | CORES2 | 61.25 | Not recorded | 68fa698f1e209131cea22a5b | | Peer Loss | 59.92 | Not recorded | 68fa698f1e209131cea22a5b | | SCE | 59.87 | Not recorded | 68fa698f1e209131cea22a5b | | Forward T | 58.95 | Not recorded | 68fa698f1e209131cea22a5b | | Cross Entropy | 58.72 | Not recorded | 68fa698f1e209131cea22a5b | | LDMI | 58.66 | Not recorded | 68fa698f1e209131cea22a5b | | Lq | 58.18 | Not recorded | 68fa698f1e209131cea22a5b | | JoCoR | 44.28 | Not recorded | 68fa698f1e209131cea22a5b | | Co-teaching | 43.03 | Not recorded | 68fa698f1e209131cea22a5b | | Co-teaching+ | 41.84 | Not recorded | 68fa698f1e209131cea22a5b | ### CORES2\* Accuracy (ε=0.2): 72.91. Result ID: 698db6708b517eb43323440f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Noise type":"Instance-based","Variant":"Star"}. This record does not assert independent reproduction. ### CORES2 Accuracy (ε=0.2): 61.25. Result ID: 698db6708b517eb43323440e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Noise type":"Instance-based"}. This record does not assert independent reproduction. ### Peer Loss Accuracy (ε=0.2): 59.92. Result ID: 698db6708b517eb43323440d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Noise type":"Instance-based"}. This record does not assert independent reproduction. ### SCE Accuracy (ε=0.2): 59.87. Result ID: 698db6708b517eb433234409. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Noise type":"Instance-based"}. This record does not assert independent reproduction. ### Forward T Accuracy (ε=0.2): 58.95. Result ID: 698db6708b517eb433234406. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Noise type":"Instance-based"}. This record does not assert independent reproduction. ### Cross Entropy Accuracy (ε=0.2): 58.72. Result ID: 698db6708b517eb433234405. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Noise type":"Instance-based"}. This record does not assert independent reproduction. ### LDMI Accuracy (ε=0.2): 58.66. Result ID: 698db6708b517eb433234407. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Noise type":"Instance-based"}. This record does not assert independent reproduction. ### Lq Accuracy (ε=0.2): 58.18. Result ID: 698db6708b517eb433234408. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Noise type":"Instance-based"}. This record does not assert independent reproduction. ### JoCoR Accuracy (ε=0.2): 44.28. Result ID: 698db6708b517eb43323440c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Noise type":"Instance-based"}. This record does not assert independent reproduction. ### Co-teaching Accuracy (ε=0.2): 43.03. Result ID: 698db6708b517eb43323440a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Noise type":"Instance-based"}. This record does not assert independent reproduction. ### Co-teaching+ Accuracy (ε=0.2): 41.84. Result ID: 698db6708b517eb43323440b. [Reporting paper]() Recorded evaluation configuration: {"Noise type":"Instance-based"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.