# Image Classification on CIFAR10 (Standard and Lp Robustness Metrics)

Source: [Image Classification on CIFAR10 (Standard and Lp Robustness Metrics)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-standard-and-lp-robustness-metrics>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-27T08:38:37.542Z

Evaluation of Image Classification performance on the CIFAR-10 dataset, reporting standard accuracy alongside robustness metrics defined by L-infinity (l∞) and L2 norms, and standard adversarial performance (StAdv).

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) | CIFAR10 |
| Metric | Standard Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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](<https://www.sota2.com/research/sota/image-classification-on-cifar10-standard-and-lp-robustness-metrics.md?metric=Standard%20Accuracy>): Higher is better; unit: PERCENT.

- [Robustness (L-inf)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-standard-and-lp-robustness-metrics.md?metric=Robustness%20(L-inf)>): Higher is better; unit: PERCENT.

- [Robustness (L2)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-standard-and-lp-robustness-metrics.md?metric=Robustness%20(L2)>): Higher is better; unit: PERCENT.

- [Standard Adversarial Accuracy (StAdv)](<https://www.sota2.com/research/sota/image-classification-on-cifar10-standard-and-lp-robustness-metrics.md?metric=Standard%20Adversarial%20Accuracy%20(StAdv)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Standard Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Wang et al. (Trained with ℓ2) | 96.09 | Not recorded | 68fa71031e209131cea4e879 |
| Base WRN-70-16 | 95.31 | Not recorded | 68fa71031e209131cea4e879 |
| ScoreOpt | 95.18 | Not recorded | 68fa71031e209131cea4e879 |
| Wang et al. (Trained with ℓ∞) | 92.97 | Not recorded | 68fa71031e209131cea4e879 |

### Wang et al. (Trained with ℓ2)

Standard Accuracy: 96.09. Result ID: 69a157f535dfcabb00e6691f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71031e209131cea4e879.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f8b1e209131cea4620c.md>)

Recorded evaluation configuration: {"Backbone":"WRN-70-16","Training Threat Model":"ℓ2"}. This record does not assert independent reproduction.

### Base WRN-70-16

Standard Accuracy: 95.31. Result ID: 69a157f535dfcabb00e6691d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71031e209131cea4e879.md>)

Recorded evaluation configuration: {"Backbone":"WRN-70-16"}. This record does not assert independent reproduction.

### ScoreOpt

Standard Accuracy: 95.18. Result ID: 69a157f535dfcabb00e66920.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71031e209131cea4e879.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71031e209131cea4e879.md>)

[Implementation](<https://github.com/zzzhangboya/ScoreOpt.git>)

Recorded evaluation configuration: {"Backbone":"WRN-70-16"}. This record does not assert independent reproduction.

### Wang et al. (Trained with ℓ∞)

Standard Accuracy: 92.97. Result ID: 69a157f535dfcabb00e6691e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71031e209131cea4e879.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f8b1e209131cea4620c.md>)

Recorded evaluation configuration: {"Backbone":"WRN-70-16","Training Threat Model":"ℓ∞"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.