# Image Classification on ImageNet (V2/S/R/A Accuracies)

Source: [Image Classification on ImageNet (V2/S/R/A Accuracies)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-v2-s-r-a-accuracies>)

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

Data updated: 2026-04-08T09:13:47.764Z

Evaluation of Image Classification performance on the standard ImageNet dataset, reporting accuracies under four specific conditions or variants denoted as V2, S, R, and A.

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 |
| Metric | Accuracy (V2) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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 (V2)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-v2-s-r-a-accuracies.md?metric=Accuracy%20(V2)>): Higher is better; unit: PERCENT.

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

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

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

## Recorded results

| Method | Accuracy (V2) | Result date | Report paper |
| --- | --- | --- | --- |
| InfluenceFunction | 65.6 | Not recorded | 69004e123ef7ccd74e5e2e1d |
| Random | 65.5 | Not recorded | 69004e123ef7ccd74e5e2e1d |
| Data-OOB | 59.22 | Not recorded | 69004e123ef7ccd74e5e2e1d |
| EV + Data-OOB | 54.76 | Not recorded | 69004e123ef7ccd74e5e2e1d |
| KNN Shapley | 40.39 | Not recorded | 69004e123ef7ccd74e5e2e1d |
| EV + KNN Shapley | 40.34 | Not recorded | 69004e123ef7ccd74e5e2e1d |

### InfluenceFunction

Accuracy (V2): 65.6. Result ID: 69d61c3a47f34ce924bee70a.

[Reporting paper](<https://www.sota2.com/research/paper/69004e123ef7ccd74e5e2e1d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68261e209131cea1a47e.md>)

Recorded evaluation configuration: {"Data Removal":"50%"}. This record does not assert independent reproduction.

### Random

Accuracy (V2): 65.5. Result ID: 69d61c3a47f34ce924bee709.

[Reporting paper](<https://www.sota2.com/research/paper/69004e123ef7ccd74e5e2e1d.md>)

Recorded evaluation configuration: {"Data Removal":"50%"}. This record does not assert independent reproduction.

### Data-OOB

Accuracy (V2): 59.22. Result ID: 69d61c3a47f34ce924bee70c.

[Reporting paper](<https://www.sota2.com/research/paper/69004e123ef7ccd74e5e2e1d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f8c1e209131cea4622f.md>)

Recorded evaluation configuration: {"Data Removal":"50%"}. This record does not assert independent reproduction.

### EV + Data-OOB

Accuracy (V2): 54.76. Result ID: 69d61c3a47f34ce924bee70e.

[Reporting paper](<https://www.sota2.com/research/paper/69004e123ef7ccd74e5e2e1d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69004e123ef7ccd74e5e2e1d.md>)

Recorded evaluation configuration: {"Data Removal":"50%"}. This record does not assert independent reproduction.

### KNN Shapley

Accuracy (V2): 40.39. Result ID: 69d61c3a47f34ce924bee70b.

[Reporting paper](<https://www.sota2.com/research/paper/69004e123ef7ccd74e5e2e1d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa673c1e209131cea15138.md>)

Recorded evaluation configuration: {"Data Removal":"50%"}. This record does not assert independent reproduction.

### EV + KNN Shapley

Accuracy (V2): 40.34. Result ID: 69d61c3a47f34ce924bee70d.

[Reporting paper](<https://www.sota2.com/research/paper/69004e123ef7ccd74e5e2e1d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69004e123ef7ccd74e5e2e1d.md>)

Recorded evaluation configuration: {"Data Removal":"50%"}. This record does not assert independent reproduction.

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