# Image Classification on ImageNet 1k (val) (Standard, Last, and Oracle Accuracy)

Source: [Image Classification on ImageNet 1k (val) (Standard, Last, and Oracle Accuracy)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-val-standard-last-and-oracle-accuracy>)

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

Data updated: 2026-02-26T05:28:52.363Z

Evaluation of Image Classification performance on the validation split of the ImageNet 1k dataset, specifically reporting standard accuracy alongside last-step and oracle accuracy 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 | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-val-standard-last-and-oracle-accuracy.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [Last Step Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-val-standard-last-and-oracle-accuracy.md?metric=Last%20Step%20Accuracy>): Higher is better; unit: PERCENT.

- [Oracle Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-val-standard-last-and-oracle-accuracy.md?metric=Oracle%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Lion | 90.52 | Not recorded | 68fa6f421e209131cea448e3 |
| Adafactor | 90.45 | Not recorded | 68fa6f421e209131cea448e3 |
| Adafactor | 90.25 | Not recorded | 68fa6f421e209131cea448e3 |
| Lion | 89.09 | Not recorded | 68fa6f421e209131cea448e3 |
| AdamW | 88.55 | Not recorded | 68fa6f421e209131cea448e3 |
| Lion | 88.5 | Not recorded | 68fa6f421e209131cea448e3 |
| AdamW | 87.72 | Not recorded | 68fa6f421e209131cea448e3 |

### Lion

Accuracy: 90.52. Result ID: 6990b0ff3a6c314849027b3e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6f421e209131cea448e3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f421e209131cea448e3.md>)

[Implementation](<https://github.com/google/automl/tree/master/lion>)

Recorded evaluation configuration: {"Model":"ViT-g/14","Input Resolution":"518x518","#Params":"1.04B","Pre-training Dataset":"JFT-3B"}. This record does not assert independent reproduction.

### Adafactor

Accuracy: 90.45. Result ID: 6990b0ff3a6c314849027b3f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6f421e209131cea448e3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64471e209131cea03a91.md>)

Recorded evaluation configuration: {"Model":"ViT-G/14","Input Resolution":"518x518","#Params":"1.88B","Pre-training Dataset":"JFT-3B"}. This record does not assert independent reproduction.

### Adafactor

Accuracy: 90.25. Result ID: 6990b0ff3a6c314849027b3d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6f421e209131cea448e3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64471e209131cea03a91.md>)

Recorded evaluation configuration: {"Model":"ViT-g/14","Input Resolution":"518x518","#Params":"1.04B","Pre-training Dataset":"JFT-3B"}. This record does not assert independent reproduction.

### Lion

Accuracy: 89.09. Result ID: 6990b0ff3a6c314849027b3c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6f421e209131cea448e3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f421e209131cea448e3.md>)

[Implementation](<https://github.com/google/automl/tree/master/lion>)

Recorded evaluation configuration: {"Model":"ViT-H/14","Input Resolution":"518x518","#Params":"633.47M","Pre-training Dataset":"JFT-300M"}. This record does not assert independent reproduction.

### AdamW

Accuracy: 88.55. Result ID: 6990b0ff3a6c314849027b3b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6f421e209131cea448e3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa55761e209131ce9946db.md>)

Recorded evaluation configuration: {"Model":"ViT-H/14","Input Resolution":"518x518","#Params":"633.47M","Pre-training Dataset":"JFT-300M"}. This record does not assert independent reproduction.

### Lion

Accuracy: 88.5. Result ID: 6990b0ff3a6c314849027b3a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6f421e209131cea448e3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f421e209131cea448e3.md>)

[Implementation](<https://github.com/google/automl/tree/master/lion>)

Recorded evaluation configuration: {"Model":"ViT-L/16","Input Resolution":"512x512","#Params":"305.18M","Pre-training Dataset":"JFT-300M"}. This record does not assert independent reproduction.

### AdamW

Accuracy: 87.72. Result ID: 6990b0ff3a6c314849027b39.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6f421e209131cea448e3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa55761e209131ce9946db.md>)

Recorded evaluation configuration: {"Model":"ViT-L/16","Input Resolution":"512x512","#Params":"305.18M","Pre-training Dataset":"JFT-300M"}. This record does not assert independent reproduction.

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