# Image Classification Robustness on Cifar100

Source: [Image Classification Robustness on Cifar100](<https://www.sota2.com/research/sota/image-classification-robustness-on-cifar100>)

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

Data updated: 2026-02-27T11:46:09.269Z

Evaluation of Image Classification performance on the Cifar100 dataset under various adversarial attack scenarios, reporting accuracy metrics for Clean data and specific attacks (AutoAttack, PGD).

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) | Cifar100 |
| Metric | Clean Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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-robustness-on-cifar100.md?metric=Clean%20Accuracy>): Higher is better; unit: PERCENT.

- [Robust Accuracy (AutoAttack)](<https://www.sota2.com/research/sota/image-classification-robustness-on-cifar100.md?metric=Robust%20Accuracy%20(AutoAttack)>): Higher is better; unit: PERCENT.

- [Robust Accuracy (PGD)](<https://www.sota2.com/research/sota/image-classification-robustness-on-cifar100.md?metric=Robust%20Accuracy%20(PGD)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Clean Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| CLIP | 60.47 | Not recorded | 6940e7a8456146936284f897 |
| CLIP | 60.45 | Not recorded | 6940e7a8456146936284f897 |
| FARE | 59.85 | Not recorded | 6940e7a8456146936284f897 |
| UCAT | 58.11 | Not recorded | 6940e7a8456146936284f897 |
| UCAT | 54.45 | Not recorded | 6940e7a8456146936284f897 |
| TeCoA | 49.68 | Not recorded | 6940e7a8456146936284f897 |
| DKL | 35.61 | Not recorded | 6940e7a8456146936284f897 |
| ACAT | 34.71 | Not recorded | 6940e7a8456146936284f897 |
| TRADES | 34.38 | Not recorded | 6940e7a8456146936284f897 |

### CLIP

Clean Accuracy: 60.47. Result ID: 69a183f135dfcabb00e78852.

[Reporting paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"Fine-tuning dataset":"TinyImageNet","Adversarial training":"2-step PGD","Epsilon":"1/255","Evaluation protocol":"Zero-shot"}. This record does not assert independent reproduction.

### CLIP

Clean Accuracy: 60.45. Result ID: 69a183f135dfcabb00e788a1.

[Reporting paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"Training Data":"ImageNet-1k","Adversarial Training":"2-step PGD","Epsilon":"1/255"}. This record does not assert independent reproduction.

### FARE

Clean Accuracy: 59.85. Result ID: 69a183f135dfcabb00e788a3.

[Reporting paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa738c1e209131cea5d24a.md>)

Recorded evaluation configuration: {"Training Data":"ImageNet-1k","Adversarial Training":"2-step PGD","Epsilon":"1/255"}. This record does not assert independent reproduction.

### UCAT

Clean Accuracy: 58.11. Result ID: 69a183f135dfcabb00e788a4.

[Reporting paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

Recorded evaluation configuration: {"Training Data":"ImageNet-1k","Adversarial Training":"2-step PGD","Epsilon":"1/255"}. This record does not assert independent reproduction.

### UCAT

Clean Accuracy: 54.45. Result ID: 69a183f135dfcabb00e78856.

[Reporting paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

Recorded evaluation configuration: {"Fine-tuning dataset":"TinyImageNet","Adversarial training":"2-step PGD","Epsilon":"1/255","Evaluation protocol":"Zero-shot"}. This record does not assert independent reproduction.

### TeCoA

Clean Accuracy: 49.68. Result ID: 69a183f135dfcabb00e788a2.

[Reporting paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f271e209131cea44177.md>)

Recorded evaluation configuration: {"Training Data":"ImageNet-1k","Adversarial Training":"2-step PGD","Epsilon":"1/255"}. This record does not assert independent reproduction.

### DKL

Clean Accuracy: 35.61. Result ID: 69a183f135dfcabb00e78855.

[Reporting paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa756d1e209131cea67d56.md>)

[Implementation](<https://github.com/jiequancui/DKL>)

Recorded evaluation configuration: {"Fine-tuning dataset":"TinyImageNet","Adversarial training":"2-step PGD","Epsilon":"1/255","Evaluation protocol":"Zero-shot"}. This record does not assert independent reproduction.

### ACAT

Clean Accuracy: 34.71. Result ID: 69a183f135dfcabb00e78854.

[Reporting paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dc61e209131cea3c0ff.md>)

[Implementation](<https://github.com/val-iisc/DAJAT>)

Recorded evaluation configuration: {"Fine-tuning dataset":"TinyImageNet","Adversarial training":"2-step PGD","Epsilon":"1/255","Evaluation protocol":"Zero-shot"}. This record does not assert independent reproduction.

### TRADES

Clean Accuracy: 34.38. Result ID: 69a183f135dfcabb00e78853.

[Reporting paper](<https://www.sota2.com/research/paper/6940e7a8456146936284f897.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65f71e209131cea0d897.md>)

Recorded evaluation configuration: {"Fine-tuning dataset":"TinyImageNet","Adversarial training":"2-step PGD","Epsilon":"1/255","Evaluation protocol":"Zero-shot"}. This record does not assert independent reproduction.

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