# Image Classification on CIFAR-10 (test) (Cl-Acc/Rob-Acc)

Source: [Image Classification on CIFAR-10 (test) (Cl-Acc/Rob-Acc)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-cl-acc-rob-acc>)

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

Data updated: 2026-02-26T05:39:11.231Z

Evaluation of Image Classification performance on the CIFAR-10 test set, focusing on Classification Accuracy (Cl-Acc) and Robust Accuracy (Rob-Acc).

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 | Classification Accuracy (Cl-Acc) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 12 of 12 |
| 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

- [Classification Accuracy (Cl-Acc)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-cl-acc-rob-acc.md?metric=Classification%20Accuracy%20(Cl-Acc)>): Higher is better; unit: PERCENT.

- [Robust Accuracy (Rob-Acc)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-cl-acc-rob-acc.md?metric=Robust%20Accuracy%20(Rob-Acc)>): Higher is better; unit: PERCENT.

- [Metric Difference](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-cl-acc-rob-acc.md?metric=Metric%20Difference>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Classification Accuracy (Cl-Acc) | Result date | Report paper |
| --- | --- | --- | --- |
| AdjAttack | 92.68 | Not recorded | 68fa71ba1e209131cea52cf8 |
| DiffAttack | 92.68 | Not recorded | 68fa71ba1e209131cea52cf8 |
| AdjAttack | 91.03 | Not recorded | 68fa71ba1e209131cea52cf8 |
| DiffAttack | 91.03 | Not recorded | 68fa71ba1e209131cea52cf8 |
| AdjAttack | 90.07 | Not recorded | 68fa71ba1e209131cea52cf8 |
| DiffAttack | 90.07 | Not recorded | 68fa71ba1e209131cea52cf8 |
| AdjAttack | 90.07 | Not recorded | 68fa71ba1e209131cea52cf8 |
| DiffAttack | 90.07 | Not recorded | 68fa71ba1e209131cea52cf8 |
| AdjAttack | 89.02 | Not recorded | 68fa71ba1e209131cea52cf8 |
| DiffAttack | 89.02 | Not recorded | 68fa71ba1e209131cea52cf8 |
| AdjAttack | 89.02 | Not recorded | 68fa71ba1e209131cea52cf8 |
| DiffAttack | 89.02 | Not recorded | 68fa71ba1e209131cea52cf8 |

### AdjAttack

Classification Accuracy (Cl-Acc): 92.68. Result ID: 6991f8833a6c3148490bca2b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c841e209131cea3472d.md>)

Recorded evaluation configuration: {"Model":"WideResNet-70-16","T":"0.075","lp Attack":"l\_2","epsilon":"0.5"}. This record does not assert independent reproduction.

### DiffAttack

Classification Accuracy (Cl-Acc): 92.68. Result ID: 6991f8833a6c3148490bca2c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Implementation](<https://github.com/kangmintong/DiffAttack>)

Recorded evaluation configuration: {"Model":"WideResNet-70-16","T":"0.075","lp Attack":"l\_2","epsilon":"0.5"}. This record does not assert independent reproduction.

### AdjAttack

Classification Accuracy (Cl-Acc): 91.03. Result ID: 6991f8833a6c3148490bca25.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c841e209131cea3472d.md>)

Recorded evaluation configuration: {"Model":"WideResNet-28-10","T":"0.075","lp Attack":"l\_2","epsilon":"0.5"}. This record does not assert independent reproduction.

### DiffAttack

Classification Accuracy (Cl-Acc): 91.03. Result ID: 6991f8833a6c3148490bca26.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Implementation](<https://github.com/kangmintong/DiffAttack>)

Recorded evaluation configuration: {"Model":"WideResNet-28-10","T":"0.075","lp Attack":"l\_2","epsilon":"0.5"}. This record does not assert independent reproduction.

### AdjAttack

Classification Accuracy (Cl-Acc): 90.07. Result ID: 6991f8833a6c3148490bca27.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c841e209131cea3472d.md>)

Recorded evaluation configuration: {"Model":"WideResNet-70-16","T":"0.1","lp Attack":"l\_inf","epsilon":"8/255"}. This record does not assert independent reproduction.

### DiffAttack

Classification Accuracy (Cl-Acc): 90.07. Result ID: 6991f8833a6c3148490bca28.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Implementation](<https://github.com/kangmintong/DiffAttack>)

Recorded evaluation configuration: {"Model":"WideResNet-70-16","T":"0.1","lp Attack":"l\_inf","epsilon":"8/255"}. This record does not assert independent reproduction.

### AdjAttack

Classification Accuracy (Cl-Acc): 90.07. Result ID: 6991f8833a6c3148490bca29.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c841e209131cea3472d.md>)

Recorded evaluation configuration: {"Model":"WideResNet-70-16","T":"0.1","lp Attack":"l\_inf","epsilon":"4/255"}. This record does not assert independent reproduction.

### DiffAttack

Classification Accuracy (Cl-Acc): 90.07. Result ID: 6991f8833a6c3148490bca2a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Implementation](<https://github.com/kangmintong/DiffAttack>)

Recorded evaluation configuration: {"Model":"WideResNet-70-16","T":"0.1","lp Attack":"l\_inf","epsilon":"4/255"}. This record does not assert independent reproduction.

### AdjAttack

Classification Accuracy (Cl-Acc): 89.02. Result ID: 6991f8833a6c3148490bca21.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c841e209131cea3472d.md>)

Recorded evaluation configuration: {"Model":"WideResNet-28-10","T":"0.1","lp Attack":"l\_inf","epsilon":"8/255"}. This record does not assert independent reproduction.

### DiffAttack

Classification Accuracy (Cl-Acc): 89.02. Result ID: 6991f8833a6c3148490bca22.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Implementation](<https://github.com/kangmintong/DiffAttack>)

Recorded evaluation configuration: {"Model":"WideResNet-28-10","T":"0.1","lp Attack":"l\_inf","epsilon":"8/255"}. This record does not assert independent reproduction.

### AdjAttack

Classification Accuracy (Cl-Acc): 89.02. Result ID: 6991f8833a6c3148490bca23.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c841e209131cea3472d.md>)

Recorded evaluation configuration: {"Model":"WideResNet-28-10","T":"0.1","lp Attack":"l\_inf","epsilon":"4/255"}. This record does not assert independent reproduction.

### DiffAttack

Classification Accuracy (Cl-Acc): 89.02. Result ID: 6991f8833a6c3148490bca24.

[Reporting paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71ba1e209131cea52cf8.md>)

[Implementation](<https://github.com/kangmintong/DiffAttack>)

Recorded evaluation configuration: {"Model":"WideResNet-28-10","T":"0.1","lp Attack":"l\_inf","epsilon":"4/255"}. This record does not assert independent reproduction.

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