# Image Classification on ImageNet-1K (PGD, FGSM, SPSA Robustness)

Source: [Image Classification on ImageNet-1K (PGD, FGSM, SPSA Robustness)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-pgd-fgsm-spsa-robustness>)

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

Data updated: 2026-06-01T07:46:24.930Z

Evaluation of Image Classification performance on the ImageNet-1K dataset, specifically measuring adversarial robustness using PGD, FGSM, and SPSA attack metrics.

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) | ImageNet-1K |
| Metric | PGD 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

- [PGD Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-pgd-fgsm-spsa-robustness.md?metric=PGD%20Accuracy>): Higher is better; unit: PERCENT.

- [FGSM Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-pgd-fgsm-spsa-robustness.md?metric=FGSM%20Accuracy>): Higher is better; unit: PERCENT.

- [SPSA Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-pgd-fgsm-spsa-robustness.md?metric=SPSA%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | PGD Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| USMoE | 24 | Not recorded | 68fa77be1e209131cea7510f |
| EC | 22.4 | Not recorded | 68fa77be1e209131cea7510f |
| SoftMoE | 13.3 | Not recorded | 68fa77be1e209131cea7510f |
| TC | 1.3 | Not recorded | 68fa77be1e209131cea7510f |

### USMoE

PGD Accuracy: 24. Result ID: 6a1d38bffbc373592c161da0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77be1e209131cea7510f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77be1e209131cea7510f.md>)

Recorded evaluation configuration: {"Backbone":"ViT-MoE (10M)"}. This record does not assert independent reproduction.

### EC

PGD Accuracy: 22.4. Result ID: 6a1d38bffbc373592c161da2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77be1e209131cea7510f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6da71e209131cea3b2d8.md>)

Recorded evaluation configuration: {"Backbone":"ViT-MoE (10M)"}. This record does not assert independent reproduction.

### SoftMoE

PGD Accuracy: 13.3. Result ID: 6a1d38bffbc373592c161da3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77be1e209131cea7510f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73521e209131cea5bd78.md>)

Recorded evaluation configuration: {"Backbone":"ViT-MoE (10M)"}. This record does not assert independent reproduction.

### TC

PGD Accuracy: 1.3. Result ID: 6a1d38bffbc373592c161da1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77be1e209131cea7510f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63171e209131ce9fc941.md>)

Recorded evaluation configuration: {"Backbone":"ViT-MoE (10M)"}. This record does not assert independent reproduction.

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