# Image Classification on CIFAR-10 (test) (FGSM/PGD Robustness Profile) Source: [Image Classification on CIFAR-10 (test) (FGSM/PGD Robustness Profile)]() Collection index: [llms.txt]() Data updated: 2026-02-26T06:04:45.552Z Evaluation of Image Classification model robustness on the CIFAR-10 test set, reporting standard accuracy (SA) and robustness accuracy (RA) against various strengths of FGSM and PGD adversarial 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 | Image Classification | | Dataset (including any stated split) | CIFAR-10 (test) | | Metric | Standard Accuracy (SA) | | 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 (SA)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - FGSM (0.01)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - FGSM (0.02)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - FGSM (0.03)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - FGSM (0.04)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - FGSM (0.05)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - FGSM (0.06)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - FGSM (0.07)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - FGSM (0.08)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - FGSM (0.09)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - PGD (10 Iterations)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - PGD (20 Iterations)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - PGD (30 Iterations)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - PGD (40 Iterations)](): Higher is better; unit: PERCENT. - [Robustness Accuracy (RA) - PGD (50 Iterations)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Standard Accuracy (SA) | Result date | Report paper | | --- | --- | --- | --- | | ResNet-50 | 95.29 | Not recorded | 694cc50c53fdd5c36173a90c | | DenseNet-169 | 95.1 | Not recorded | 694cc50c53fdd5c36173a90c | | GoogLeNet | 95.01 | Not recorded | 694cc50c53fdd5c36173a90c | | Inception-V3 | 94.78 | Not recorded | 694cc50c53fdd5c36173a90c | | MobileNet-V2 | 94.2 | Not recorded | 694cc50c53fdd5c36173a90c | | Xception | 93.48 | Not recorded | 694cc50c53fdd5c36173a90c | | VGG-19 | 93.18 | Not recorded | 694cc50c53fdd5c36173a90c | | ResNet-18 | 92.35 | Not recorded | 694cc50c53fdd5c36173a90c | | DWF | 91.08 | Not recorded | 694cc50c53fdd5c36173a90c | | ADVMoE | 88.27 | Not recorded | 694cc50c53fdd5c36173a90c | | SoE | 81.69 | Not recorded | 694cc50c53fdd5c36173a90c | ### ResNet-50 Standard Accuracy (SA): 95.29. Result ID: 6999418d5698665ec4ab9911. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-50"}. This record does not assert independent reproduction. ### DenseNet-169 Standard Accuracy (SA): 95.1. Result ID: 6999418d5698665ec4ab9912. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"DenseNet-169"}. This record does not assert independent reproduction. ### GoogLeNet Standard Accuracy (SA): 95.01. Result ID: 6999418d5698665ec4ab9913. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"GoogLeNet"}. This record does not assert independent reproduction. ### Inception-V3 Standard Accuracy (SA): 94.78. Result ID: 6999418d5698665ec4ab9914. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Inception-V3"}. This record does not assert independent reproduction. ### MobileNet-V2 Standard Accuracy (SA): 94.2. Result ID: 6999418d5698665ec4ab9916. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"MobileNet-V2"}. This record does not assert independent reproduction. ### Xception Standard Accuracy (SA): 93.48. Result ID: 6999418d5698665ec4ab9915. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Xception"}. This record does not assert independent reproduction. ### VGG-19 Standard Accuracy (SA): 93.18. Result ID: 6999418d5698665ec4ab990f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"VGG-19"}. This record does not assert independent reproduction. ### ResNet-18 Standard Accuracy (SA): 92.35. Result ID: 6999418d5698665ec4ab9910. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-18"}. This record does not assert independent reproduction. ### DWF Standard Accuracy (SA): 91.08. Result ID: 6999418d5698665ec4ab9919. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-18"}. This record does not assert independent reproduction. ### ADVMoE Standard Accuracy (SA): 88.27. Result ID: 6999418d5698665ec4ab9918. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"ResNet-18"}. This record does not assert independent reproduction. ### SoE Standard Accuracy (SA): 81.69. Result ID: 6999418d5698665ec4ab9917. [Reporting 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.