# Image Classification on ImageNet OOD Variants V2, Sketch, A, R (test val)

Source: [Image Classification on ImageNet OOD Variants V2, Sketch, A, R (test val)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-variants-v2-sketch-a-r-test-val>)

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

Data updated: 2026-05-26T10:10:42.416Z

Evaluation of Image Classification performance across the base ImageNet test/validation set and the specific Out-of-Distribution variants: V2, Sketch, A, and R, reporting individual scores and an overall average.

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 Variants V2 Sketch A R (test val) |
| Metric | ImageNet Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| 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

- [ImageNet Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-variants-v2-sketch-a-r-test-val.md?metric=ImageNet%20Accuracy>): Higher is better; unit: PERCENT.

- [ImageNet-V2 Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-variants-v2-sketch-a-r-test-val.md?metric=ImageNet-V2%20Accuracy>): Higher is better; unit: PERCENT.

- [ImageNet-Sketch Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-variants-v2-sketch-a-r-test-val.md?metric=ImageNet-Sketch%20Accuracy>): Higher is better; unit: PERCENT.

- [ImageNet-A Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-variants-v2-sketch-a-r-test-val.md?metric=ImageNet-A%20Accuracy>): Higher is better; unit: PERCENT.

- [ImageNet-R Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-variants-v2-sketch-a-r-test-val.md?metric=ImageNet-R%20Accuracy>): Higher is better; unit: PERCENT.

- [Average Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-ood-variants-v2-sketch-a-r-test-val.md?metric=Average%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | ImageNet Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| SCAN | 69.92 | Not recorded | 68fa78971e209131cea792d5 |
| SimNL | 66.52 | Not recorded | 68fa78971e209131cea792d5 |
| GraphAdapter | 64.94 | Not recorded | 68fa78971e209131cea792d5 |
| TaskRes | 64.75 | Not recorded | 68fa78971e209131cea792d5 |
| CoOp | 62.95 | Not recorded | 68fa78971e209131cea792d5 |
| CoCoOp | 62.71 | Not recorded | 68fa78971e209131cea792d5 |
| ProGrad | 62.17 | Not recorded | 68fa78971e209131cea792d5 |
| TPT | 60.74 | Not recorded | 68fa78971e209131cea792d5 |
| Zero-Shot CLIP | 60.33 | Not recorded | 68fa78971e209131cea792d5 |
| LP CLIP | 56.13 | Not recorded | 68fa78971e209131cea792d5 |

### SCAN

ImageNet Accuracy: 69.92. Result ID: 6a15718dfbc373592c0d6aba.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

Recorded evaluation configuration: {"Shots":"16"}. This record does not assert independent reproduction.

### SimNL

ImageNet Accuracy: 66.52. Result ID: 6a15718dfbc373592c0d6ab9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75a01e209131cea690e0.md>)

Recorded evaluation configuration: {"Shots":"16"}. This record does not assert independent reproduction.

### GraphAdapter

ImageNet Accuracy: 64.94. Result ID: 6a15718dfbc373592c0d6ab8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70ab1e209131cea4c58c.md>)

[Implementation](<https://github.com/lixinustc/GraphAdapter>)

Recorded evaluation configuration: {"Shots":"16"}. This record does not assert independent reproduction.

### TaskRes

ImageNet Accuracy: 64.75. Result ID: 6a15718dfbc373592c0d6ab7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ee41e209131cea42ac8.md>)

Recorded evaluation configuration: {"Shots":"16"}. This record does not assert independent reproduction.

### CoOp

ImageNet Accuracy: 62.95. Result ID: 6a15718dfbc373592c0d6ab3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d8c1e209131cea3a831.md>)

[Implementation](<https://github.com/KaiyangZhou/CoOp>)

Recorded evaluation configuration: {"Shots":"16"}. This record does not assert independent reproduction.

### CoCoOp

ImageNet Accuracy: 62.71. Result ID: 6a15718dfbc373592c0d6ab4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d8c1e209131cea3a831.md>)

[Implementation](<https://github.com/KaiyangZhou/CoOp>)

Recorded evaluation configuration: {"Shots":"16"}. This record does not assert independent reproduction.

### ProGrad

ImageNet Accuracy: 62.17. Result ID: 6a15718dfbc373592c0d6ab5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a981e209131cea841f7.md>)

Recorded evaluation configuration: {"Shots":"16"}. This record does not assert independent reproduction.

### TPT

ImageNet Accuracy: 60.74. Result ID: 6a15718dfbc373592c0d6ab6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d691e209131cea399f9.md>)

[Implementation](<https://tinyurl.com/yr3zmhma>)

Recorded evaluation configuration: {"Shots":"16"}. This record does not assert independent reproduction.

### Zero-Shot CLIP

ImageNet Accuracy: 60.33. Result ID: 6a15718dfbc373592c0d6ab1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

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

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

Recorded evaluation configuration: {"Setting":"Zero-shot"}. This record does not assert independent reproduction.

### LP CLIP

ImageNet Accuracy: 56.13. Result ID: 6a15718dfbc373592c0d6ab2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78971e209131cea792d5.md>)

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

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

Recorded evaluation configuration: {"Setting":"Linear Probing"}. This record does not assert independent reproduction.

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