# Image Classification on ImageNet 1K (val) (Top-1, ReaL, Multilabel)

Source: [Image Classification on ImageNet 1K (val) (Top-1, ReaL, Multilabel)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-val-top-1-real-multilabel>)

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

Data updated: 2026-02-26T06:06:39.197Z

Evaluation of Image Classification performance on the validation split of ImageNet 1K, reporting Top-1 accuracy alongside ReaL and Multilabel metrics.

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 1K (val) |
| Metric | Top-1 Acc |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 7 of 7 |
| 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

- [Top-1 Acc](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-val-top-1-real-multilabel.md?metric=Top-1%20Acc>): Higher is better; unit: PERCENT.

- [ReaL Score](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-val-top-1-real-multilabel.md?metric=ReaL%20Score>): Higher is better; unit: PERCENT.

- [Multilabel Acc](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-val-top-1-real-multilabel.md?metric=Multilabel%20Acc>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Top-1 Acc | Result date | Report paper |
| --- | --- | --- | --- |
| Greedy soup | 90.94 | Not recorded | 68fa6ce11e209131cea3679c |
| Greedy ensemble | 90.93 | Not recorded | 68fa6ce11e209131cea3679c |
| CoAtNet-7 | 90.88 | Not recorded | 68fa6ce11e209131cea3679c |
| Best model on each test set (oracle) | 90.78 | Not recorded | 68fa6ce11e209131cea3679c |
| Best model on held out val set | 90.72 | Not recorded | 68fa6ce11e209131cea3679c |
| ViT/G-14 | 90.47 | Not recorded | 68fa6ce11e209131cea3679c |
| ViT/G-14 | 90.45 | Not recorded | 68fa6ce11e209131cea3679c |

### Greedy soup

Top-1 Acc: 90.94. Result ID: 698f48c63a6c314849f7ab73.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Implementation](<https://github.com/mlfoundations/model-soups>)

Recorded evaluation configuration: {"Backbone":"ViT-G/14"}. This record does not assert independent reproduction.

### Greedy ensemble

Top-1 Acc: 90.93. Result ID: 698f48c63a6c314849f7ab72.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Implementation](<https://github.com/mlfoundations/model-soups>)

Recorded evaluation configuration: {"Backbone":"ViT-G/14"}. This record does not assert independent reproduction.

### CoAtNet-7

Top-1 Acc: 90.88. Result ID: 698f48c63a6c314849f7ab6e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ad11e209131cea2a845.md>)

Recorded evaluation configuration: {"Backbone":"CoAtNet-7"}. This record does not assert independent reproduction.

### Best model on each test set (oracle)

Top-1 Acc: 90.78. Result ID: 698f48c63a6c314849f7ab71.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Implementation](<https://github.com/mlfoundations/model-soups>)

Recorded evaluation configuration: {"Backbone":"ViT-G/14","Selection":"Oracle (each test set)"}. This record does not assert independent reproduction.

### Best model on held out val set

Top-1 Acc: 90.72. Result ID: 698f48c63a6c314849f7ab70.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Implementation](<https://github.com/mlfoundations/model-soups>)

Recorded evaluation configuration: {"Backbone":"ViT-G/14","Selection":"Held out validation set"}. This record does not assert independent reproduction.

### ViT/G-14

Top-1 Acc: 90.47. Result ID: 698f48c63a6c314849f7ab6f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cd91e209131cea363c2.md>)

[Implementation](<https://github.com/google-research/big_vision>)

Recorded evaluation configuration: {"Backbone":"ViT-G/14","Selection":"Reevaluated"}. This record does not assert independent reproduction.

### ViT/G-14

Top-1 Acc: 90.45. Result ID: 698f48c63a6c314849f7ab6d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6ce11e209131cea3679c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cd91e209131cea363c2.md>)

[Implementation](<https://github.com/google-research/big_vision>)

Recorded evaluation configuration: {"Backbone":"ViT-G/14","Selection":"Original (Zhai et al., 2021)"}. This record does not assert independent reproduction.

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