# Image Classification on Imagenette (L-inf Robustness)

Source: [Image Classification on Imagenette (L-inf Robustness)](<https://www.sota2.com/research/sota/image-classification-on-imagenette-l-inf-robustness>)

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

Data updated: 2026-02-26T05:39:19.118Z

Evaluation of Image Classification performance on the Imagenette dataset, specifically reporting clean accuracy alongside robust accuracy under L-infinity adversarial attacks at epsilon values of 2/255 and 4/255.

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 | Classification |
| Dataset (including any stated split) | Imagenette |
| 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-imagenette-l-inf-robustness.md?metric=Clean%20Accuracy>): Higher is better; unit: PERCENT.

- [Robust Accuracy (L-inf, eps=2/255)](<https://www.sota2.com/research/sota/image-classification-on-imagenette-l-inf-robustness.md?metric=Robust%20Accuracy%20(L-inf%2C%20eps%3D2%2F255)>): Higher is better; unit: PERCENT.

- [Robust Accuracy (L-inf, eps=4/255)](<https://www.sota2.com/research/sota/image-classification-on-imagenette-l-inf-robustness.md?metric=Robust%20Accuracy%20(L-inf%2C%20eps%3D4%2F255)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Clean Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| TiTok-S128 | 86.2 | Not recorded | 699bdef109c33900fcd8041f |
| TiTok-S128 | 86 | Not recorded | 699bdef109c33900fcd8041f |
| TiTok-S128 | 84.2 | Not recorded | 699bdef109c33900fcd8041f |
| TiTok-S128 | 82.8 | Not recorded | 699bdef109c33900fcd8041f |
| TiTok-S128 | 79.8 | Not recorded | 699bdef109c33900fcd8041f |

### TiTok-S128

Clean Accuracy: 86.2. Result ID: 699d09a65698665ec4a9ad7b.

[Reporting paper](<https://www.sota2.com/research/paper/699bdef109c33900fcd8041f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/699bdef109c33900fcd8041f.md>)

Recorded evaluation configuration: {"Setup":"Adversarial Training","Radius":"4/255"}. This record does not assert independent reproduction.

### TiTok-S128

Clean Accuracy: 86. Result ID: 699d09a65698665ec4a9ad7a.

[Reporting paper](<https://www.sota2.com/research/paper/699bdef109c33900fcd8041f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/699bdef109c33900fcd8041f.md>)

Recorded evaluation configuration: {"Setup":"Clean/Original"}. This record does not assert independent reproduction.

### TiTok-S128

Clean Accuracy: 84.2. Result ID: 699d09a65698665ec4a9ad7c.

[Reporting paper](<https://www.sota2.com/research/paper/699bdef109c33900fcd8041f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/699bdef109c33900fcd8041f.md>)

Recorded evaluation configuration: {"Setup":"Adversarial Training","Radius":"8/255"}. This record does not assert independent reproduction.

### TiTok-S128

Clean Accuracy: 82.8. Result ID: 699d09a65698665ec4a9ad7d.

[Reporting paper](<https://www.sota2.com/research/paper/699bdef109c33900fcd8041f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/699bdef109c33900fcd8041f.md>)

Recorded evaluation configuration: {"Setup":"Adversarial Training","Radius":"12/255"}. This record does not assert independent reproduction.

### TiTok-S128

Clean Accuracy: 79.8. Result ID: 699d09a65698665ec4a9ad7e.

[Reporting paper](<https://www.sota2.com/research/paper/699bdef109c33900fcd8041f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/699bdef109c33900fcd8041f.md>)

Recorded evaluation configuration: {"Setup":"Adversarial Training","Radius":"16/255"}. This record does not assert independent reproduction.

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