# Image Classification on ImageNet OOD Suite (test)

Source: [Image Classification on ImageNet OOD Suite (test)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-suite-test>)

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

Data updated: 2026-05-29T10:13:47.158Z

Evaluation of Image Classification performance on the test split of the ImageNet Out-of-Distribution (OOD) Suite, reporting accuracy across specific subsets including ImageNet-V2, ImageNet-R, ImageNet-A, ImageNet Sketch, and ObjectNet.

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 OOD Suite (test) |
| Metric | Accuracy (ImageNet-V2) |
| 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

- [Accuracy (ImageNet-V2)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-suite-test.md?metric=Accuracy%20(ImageNet-V2)>): Higher is better; unit: PERCENT.

- [Accuracy (ImageNet-R)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-suite-test.md?metric=Accuracy%20(ImageNet-R)>): Higher is better; unit: PERCENT.

- [Accuracy (ImageNet-A)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-suite-test.md?metric=Accuracy%20(ImageNet-A)>): Higher is better; unit: PERCENT.

- [Accuracy (ImageNet Sketch)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-suite-test.md?metric=Accuracy%20(ImageNet%20Sketch)>): Higher is better; unit: PERCENT.

- [Accuracy (ObjectNet)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-suite-test.md?metric=Accuracy%20(ObjectNet)>): Higher is better; unit: PERCENT.

- [Average Accuracy (OOD Suite)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-suite-test.md?metric=Average%20Accuracy%20(OOD%20Suite)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (ImageNet-V2) | Result date | Report paper |
| --- | --- | --- | --- |
| CLIP ViT-B/16 + ED | 72.04 | Not recorded | 6a191dd6dbd7150506f82af7 |
| CLIP ViT-B/16 | 70.89 | Not recorded | 6a191dd6dbd7150506f82af7 |
| CLIP ViT-B/32 + ED | 65.37 | Not recorded | 6a191dd6dbd7150506f82af7 |
| CLIP ViT-B/32 | 64.21 | Not recorded | 6a191dd6dbd7150506f82af7 |

### CLIP ViT-B/16 + ED

Accuracy (ImageNet-V2): 72.04. Result ID: 6a1966befbc373592c144f1a.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dd6dbd7150506f82af7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a191dd6dbd7150506f82af7.md>)

Recorded evaluation configuration: {"Backbone":"ViT-B/16","Regularization":"Effective Degree (ED)"}. This record does not assert independent reproduction.

### CLIP ViT-B/16

Accuracy (ImageNet-V2): 70.89. Result ID: 6a1966befbc373592c144f19.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dd6dbd7150506f82af7.md>)

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

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

Recorded evaluation configuration: {"Backbone":"ViT-B/16"}. This record does not assert independent reproduction.

### CLIP ViT-B/32 + ED

Accuracy (ImageNet-V2): 65.37. Result ID: 6a1966befbc373592c144f18.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dd6dbd7150506f82af7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a191dd6dbd7150506f82af7.md>)

Recorded evaluation configuration: {"Backbone":"ViT-B/32","Regularization":"Effective Degree (ED)"}. This record does not assert independent reproduction.

### CLIP ViT-B/32

Accuracy (ImageNet-V2): 64.21. Result ID: 6a1966befbc373592c144f17.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dd6dbd7150506f82af7.md>)

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

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

Recorded evaluation configuration: {"Backbone":"ViT-B/32"}. This record does not assert independent reproduction.

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