# Image Classification on CIFAR-10 (Adversarial Robustness Metrics)

Source: [Image Classification on CIFAR-10 (Adversarial Robustness Metrics)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-adversarial-robustness-metrics>)

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

Data updated: 2026-04-15T09:12:47.589Z

Evaluation of Image Classification performance on the CIFAR-10 dataset, reporting specific adversarial robustness metrics including PGD-cw, Logit Flipping, and Worst Case analysis.

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 |
| Metric | Logit Flipping 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

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

- [SIA-bb Robustness Score](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-adversarial-robustness-metrics.md?metric=SIA-bb%20Robustness%20Score>): Higher is better; unit: PERCENT.

- [LMA Robustness Score](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-adversarial-robustness-metrics.md?metric=LMA%20Robustness%20Score>): Higher is better; unit: PERCENT.

- [CPA Robustness Score](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-adversarial-robustness-metrics.md?metric=CPA%20Robustness%20Score>): Higher is better; unit: PERCENT.

- [SIA Robustness Score](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-adversarial-robustness-metrics.md?metric=SIA%20Robustness%20Score>): Higher is better; unit: PERCENT.

- [PGD-cw Robustness Score](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-adversarial-robustness-metrics.md?metric=PGD-cw%20Robustness%20Score>): Higher is better; unit: PERCENT.

- [Worst Case Robustness Score](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-adversarial-robustness-metrics.md?metric=Worst%20Case%20Robustness%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Logit Flipping Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| DeepSet-TM | 67.6 | Not recorded | 68fa7c421e209131cea8d392 |
| Mean | 65.4 | Not recorded | 68fa7c421e209131cea8d392 |
| GM | 63.9 | Not recorded | 68fa7c421e209131cea8d392 |
| CWTM | 63.3 | Not recorded | 68fa7c421e209131cea8d392 |
| CWMed | 56.7 | Not recorded | 68fa7c421e209131cea8d392 |

### DeepSet-TM

Logit Flipping Accuracy: 67.6. Result ID: 69df567f47f34ce924c84d68.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c421e209131cea8d392.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c421e209131cea8d392.md>)

Recorded evaluation configuration: {"Heterogeneity (α)":"0.5","Clients (n)":"17","Adversaries (f)":"4"}. This record does not assert independent reproduction.

### Mean

Logit Flipping Accuracy: 65.4. Result ID: 69df567f47f34ce924c84d64.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c421e209131cea8d392.md>)

Recorded evaluation configuration: {"Heterogeneity (α)":"0.5","Clients (n)":"17","Adversaries (f)":"4"}. This record does not assert independent reproduction.

### GM

Logit Flipping Accuracy: 63.9. Result ID: 69df567f47f34ce924c84d66.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c421e209131cea8d392.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa703e1e209131cea49e91.md>)

Recorded evaluation configuration: {"Heterogeneity (α)":"0.5","Clients (n)":"17","Adversaries (f)":"4"}. This record does not assert independent reproduction.

### CWTM

Logit Flipping Accuracy: 63.3. Result ID: 69df567f47f34ce924c84d67.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c421e209131cea8d392.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69531e209131cea218fb.md>)

Recorded evaluation configuration: {"Heterogeneity (α)":"0.5","Clients (n)":"17","Adversaries (f)":"4"}. This record does not assert independent reproduction.

### CWMed

Logit Flipping Accuracy: 56.7. Result ID: 69df567f47f34ce924c84d65.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7c421e209131cea8d392.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69531e209131cea218fb.md>)

Recorded evaluation configuration: {"Heterogeneity (α)":"0.5","Clients (n)":"17","Adversaries (f)":"4"}. This record does not assert independent reproduction.

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