# Few-shot Image Classification on ImageNet 5-shot

Source: [Few-shot Image Classification on ImageNet 5-shot](<https://www.sota2.com/research/sota/few-shot-image-classification-on-imagenet-5-shot>)

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

Data updated: 2026-02-26T05:34:02.387Z

Evaluation of few-shot image classification performance on the 5-shot configuration of the standard ImageNet dataset.

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 | Few-shot Image Classification |
| Dataset (including any stated split) | ImageNet 5-shot |
| Metric | Accuracy (5-shot) |
| 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 (5-shot)](<https://www.sota2.com/research/sota/few-shot-image-classification-on-imagenet-5-shot.md?metric=Accuracy%20(5-shot)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy (5-shot) | Result date | Report paper |
| --- | --- | --- | --- |
| V-MoE-H/14 | 78.21 | Not recorded | 68fa6a201e209131cea2621c |
| V-MoE-H/14 | 78.08 | Not recorded | 68fa6a201e209131cea2621c |
| V-MoE-L/16 | 77.1 | Not recorded | 68fa6a201e209131cea2621c |
| ViT-H/14 | 76.95 | Not recorded | 68fa6a201e209131cea2621c |
| V-MoE-L/16 | 76.53 | Not recorded | 68fa6a201e209131cea2621c |
| ViT-L/16 | 74.36 | Not recorded | 68fa6a201e209131cea2621c |

### V-MoE-H/14

Accuracy (5-shot): 78.21. Result ID: 698e11dcffc3901de09a9064.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a201e209131cea2621c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a201e209131cea2621c.md>)

Recorded evaluation configuration: {"Experts":"Every-2","Routing":"Vanilla","Time \[%\]":"164.89","FLOPS \[%\]":"135.59"}. This record does not assert independent reproduction.

### V-MoE-H/14

Accuracy (5-shot): 78.08. Result ID: 698e11dcffc3901de09a9063.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a201e209131cea2621c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a201e209131cea2621c.md>)

Recorded evaluation configuration: {"Experts":"Last-5","Routing":"Vanilla","Time \[%\]":"120.22","FLOPS \[%\]":"111.12"}. This record does not assert independent reproduction.

### V-MoE-L/16

Accuracy (5-shot): 77.1. Result ID: 698e11dcffc3901de09a9062.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a201e209131cea2621c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a201e209131cea2621c.md>)

Recorded evaluation configuration: {"Experts":"Every-2","Routing":"Vanilla","Time \[%\]":"57.40","FLOPS \[%\]":"49.95"}. This record does not assert independent reproduction.

### ViT-H/14

Accuracy (5-shot): 76.95. Result ID: 698e11dcffc3901de09a905f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a201e209131cea2621c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a0b1e209131cea25b04.md>)

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

Recorded evaluation configuration: {"Experts":"null","Routing":"null","Time \[%\]":"100.00","FLOPS \[%\]":"100.00"}. This record does not assert independent reproduction.

### V-MoE-L/16

Accuracy (5-shot): 76.53. Result ID: 698e11dcffc3901de09a9061.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a201e209131cea2621c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a201e209131cea2621c.md>)

Recorded evaluation configuration: {"Experts":"Last-2","Routing":"Vanilla","Time \[%\]":"32.56","FLOPS \[%\]":"39.02"}. This record does not assert independent reproduction.

### ViT-L/16

Accuracy (5-shot): 74.36. Result ID: 698e11dcffc3901de09a9060.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6a201e209131cea2621c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a0b1e209131cea25b04.md>)

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

Recorded evaluation configuration: {"Experts":"null","Routing":"null","Time \[%\]":"27.58","FLOPS \[%\]":"36.83"}. This record does not assert independent reproduction.

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