# Image Classification on CIFAR-10 IR=100 long-tail (test) (Adversarial Robustness)

Source: [Image Classification on CIFAR-10 IR=100 long-tail (test) (Adversarial Robustness)](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-ir-100-long-tail-test-adversarial-robustness>)

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

Data updated: 2026-02-26T05:39:19.043Z

Evaluation of Image Classification performance and adversarial robustness (Clean, FGSM, PGD-20, CW, AA) on the test split of the CIFAR-10 dataset configured with an imbalance ratio (IR) of 100.

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 IR=100 long-tail (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-ir-100-long-tail-test-adversarial-robustness.md?metric=Clean%20Accuracy>): Higher is better; unit: PERCENT.

- [FGSM Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-ir-100-long-tail-test-adversarial-robustness.md?metric=FGSM%20Accuracy>): Higher is better; unit: PERCENT.

- [PGD-20 Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-ir-100-long-tail-test-adversarial-robustness.md?metric=PGD-20%20Accuracy>): Higher is better; unit: PERCENT.

- [CW Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-ir-100-long-tail-test-adversarial-robustness.md?metric=CW%20Accuracy>): Higher is better; unit: PERCENT.

- [AA Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-ir-100-long-tail-test-adversarial-robustness.md?metric=AA%20Accuracy>): Higher is better; unit: PERCENT.

- [MIM Accuracy](<https://www.sota2.com/research/sota/image-classification-on-cifar-10-ir-100-long-tail-test-adversarial-robustness.md?metric=MIM%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Clean Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| RoBal | 68.07 | Not recorded | 68fa6a8c1e209131cea28b65 |
| AT | 56.72 | Not recorded | 68fa6a8c1e209131cea28b65 |
| AT-BSL | 54.73 | Not recorded | 68fa77941e209131cea7419b |
| TAET | 52.68 | Not recorded | 68fa77941e209131cea7419b |
| TRADES | 45.67 | Not recorded | 68fa6a8c1e209131cea28b65 |

### RoBal

Clean Accuracy: 68.07. Result ID: 699ab8405698665ec4463281.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a8c1e209131cea28b65.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a8c1e209131cea28b65.md>)

[Implementation](<https://github.com/wutong16/Adversarial_Long-Tail>)

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

### AT

Clean Accuracy: 56.72. Result ID: 699ab8405698665ec446327f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a8c1e209131cea28b65.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa665a1e209131cea0fbae.md>)

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

### AT-BSL

Clean Accuracy: 54.73. Result ID: 6995459c3badfaffe6334589.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77941e209131cea7419b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72801e209131cea56cd7.md>)

[Implementation](<https://github.com/NISPLab/AT-BSL>)

Recorded evaluation configuration: {"Backbone":"ResNet-18","IR":"100"}. This record does not assert independent reproduction.

### TAET

Clean Accuracy: 52.68. Result ID: 6995459c3badfaffe633458a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77941e209131cea7419b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77941e209131cea7419b.md>)

[Implementation](<https://github.com/BuhuiOK/TAET-Two-Stage-Adversarial-Equalization-Training-on-Long-Tailed-Distributions>)

Recorded evaluation configuration: {"Backbone":"ResNet-18","IR":"100"}. This record does not assert independent reproduction.

### TRADES

Clean Accuracy: 45.67. Result ID: 699ab8405698665ec4463280.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a8c1e209131cea28b65.md>)

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

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

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