# Image Classification on CIFAR-10 (test) (Certified Robustness)

Source: [Image Classification on CIFAR-10 (test) (Certified Robustness)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-certified-robustness>)

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

Data updated: 2026-05-26T10:49:46.970Z

Evaluation of Image Classification model performance on the CIFAR-10 test set, reporting certified accuracy under various perturbation bounds (epsilon).

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 | Clean Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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

- [Clean Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-certified-robustness.md?metric=Clean%20Accuracy>): Higher is better; unit: PERCENT.

- [Certified Accuracy (eps=0.25)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-certified-robustness.md?metric=Certified%20Accuracy%20(eps%3D0.25)>): Higher is better; unit: PERCENT.

- [Certified Accuracy (eps=0.5)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-certified-robustness.md?metric=Certified%20Accuracy%20(eps%3D0.5)>): Higher is better; unit: PERCENT.

- [Certified Accuracy (eps=1.0)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-test-certified-robustness.md?metric=Certified%20Accuracy%20(eps%3D1.0)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Clean Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| GENELLIPS | 90.14 | Not recorded | 6a152965dbd7150506f7e325 |
| SmoothAdv | 86.2 | Not recorded | 6a152965dbd7150506f7e325 |
| DRT + MME (Gaussian) | 81.4 | Not recorded | 6a152965dbd7150506f7e325 |
| DRT + MME (SmoothAdv) | 72.6 | Not recorded | 6a152965dbd7150506f7e325 |
| DRT + WE (SmoothAdv) | 72.6 | Not recorded | 6a152965dbd7150506f7e325 |

### GENELLIPS

Clean Accuracy: 90.14. Result ID: 6a157ac9fbc373592c0dae6c.

[Reporting paper](<https://www.sota2.com/research/paper/6a152965dbd7150506f7e325.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a152965dbd7150506f7e325.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### SmoothAdv

Clean Accuracy: 86.2. Result ID: 6a157ac9fbc373592c0dae68.

[Reporting paper](<https://www.sota2.com/research/paper/6a152965dbd7150506f7e325.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66e41e209131cea12df7.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### DRT + MME (Gaussian)

Clean Accuracy: 81.4. Result ID: 6a157ac9fbc373592c0dae69.

[Reporting paper](<https://www.sota2.com/research/paper/6a152965dbd7150506f7e325.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c541e209131cea33522.md>)

Recorded evaluation configuration: {"MME mode":"Gaussian"}. This record does not assert independent reproduction.

### DRT + MME (SmoothAdv)

Clean Accuracy: 72.6. Result ID: 6a157ac9fbc373592c0dae6a.

[Reporting paper](<https://www.sota2.com/research/paper/6a152965dbd7150506f7e325.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c541e209131cea33522.md>)

Recorded evaluation configuration: {"MME mode":"SmoothAdv"}. This record does not assert independent reproduction.

### DRT + WE (SmoothAdv)

Clean Accuracy: 72.6. Result ID: 6a157ac9fbc373592c0dae6b.

[Reporting paper](<https://www.sota2.com/research/paper/6a152965dbd7150506f7e325.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c541e209131cea33522.md>)

Recorded evaluation configuration: {"WE mode":"SmoothAdv"}. This record does not assert independent reproduction.

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