# Accuracy Estimation on ImageNet

Source: [Accuracy Estimation on ImageNet](<https://www.sota2.com/research/sota/accuracy-estimation-on-imagenet>)

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

Data updated: 2026-02-26T05:49:06.427Z

Evaluation of model performance in estimating accuracy on the ImageNet dataset, measured using Mean Absolute Error (MAE).

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 | Accuracy Estimation |
| Dataset (including any stated split) | ImageNet |
| Metric | MAE |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 27 of 27 |
| 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

- [MAE](<https://www.sota2.com/research/sota/accuracy-estimation-on-imagenet.md?metric=MAE>): Lower is better; unit: SCALAR.

## Recorded results

| Method | MAE | Result date | Report paper |
| --- | --- | --- | --- |
| MANO | 0.695 | Not recorded | 68fa75d01e209131cea69ff5 |
| COT | 0.705 | Not recorded | 68fa75d01e209131cea69ff5 |
| MANO | 1.057 | Not recorded | 68fa75d01e209131cea69ff5 |
| MANO | 1.388 | Not recorded | 68fa75d01e209131cea69ff5 |
| COT | 1.796 | Not recorded | 68fa75d01e209131cea69ff5 |
| ATC | 2.503 | Not recorded | 68fa75d01e209131cea69ff5 |
| COT | 2.564 | Not recorded | 68fa75d01e209131cea69ff5 |
| Dispersion | 2.602 | Not recorded | 68fa75d01e209131cea69ff5 |
| Nuclear | 2.679 | Not recorded | 68fa75d01e209131cea69ff5 |
| ATC | 2.911 | Not recorded | 68fa75d01e209131cea69ff5 |
| ConfScore | 2.99 | Not recorded | 68fa75d01e209131cea69ff5 |
| Nuclear | 3.203 | Not recorded | 68fa75d01e209131cea69ff5 |
| Fréchet | 3.242 | Not recorded | 68fa75d01e209131cea69ff5 |
| ConfScore | 3.325 | Not recorded | 68fa75d01e209131cea69ff5 |
| Entropy | 3.357 | Not recorded | 68fa75d01e209131cea69ff5 |
| ATC | 3.388 | Not recorded | 68fa75d01e209131cea69ff5 |
| ConfScore | 3.616 | Not recorded | 68fa75d01e209131cea69ff5 |
| Fréchet | 3.75 | Not recorded | 68fa75d01e209131cea69ff5 |
| Entropy | 3.812 | Not recorded | 68fa75d01e209131cea69ff5 |
| Nuclear | 4.091 | Not recorded | 68fa75d01e209131cea69ff5 |
| Entropy | 4.132 | Not recorded | 68fa75d01e209131cea69ff5 |
| MDE | 4.818 | Not recorded | 68fa75d01e209131cea69ff5 |
| Fréchet | 5.21 | Not recorded | 68fa75d01e209131cea69ff5 |
| MDE | 6.346 | Not recorded | 68fa75d01e209131cea69ff5 |
| MDE | 6.755 | Not recorded | 68fa75d01e209131cea69ff5 |
| Dispersion | 7.318 | Not recorded | 68fa75d01e209131cea69ff5 |
| Dispersion | 10.05 | Not recorded | 68fa75d01e209131cea69ff5 |

### MANO

MAE: 0.695. Result ID: 6993e6c53a6c3148491953bc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

Recorded evaluation configuration: {"Network":"WRN-50-2"}. This record does not assert independent reproduction.

### COT

MAE: 0.705. Result ID: 6993e6c53a6c3148491953a8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71021e209131cea4e815.md>)

Recorded evaluation configuration: {"Network":"ResNet18"}. This record does not assert independent reproduction.

### MANO

MAE: 1.057. Result ID: 6993e6c53a6c3148491953aa.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

Recorded evaluation configuration: {"Network":"ResNet18"}. This record does not assert independent reproduction.

### MANO

MAE: 1.388. Result ID: 6993e6c53a6c3148491953b3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

Recorded evaluation configuration: {"Network":"ResNet50"}. This record does not assert independent reproduction.

### COT

MAE: 1.796. Result ID: 6993e6c53a6c3148491953b1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71021e209131cea4e815.md>)

Recorded evaluation configuration: {"Network":"ResNet50"}. This record does not assert independent reproduction.

### ATC

MAE: 2.503. Result ID: 6993e6c53a6c3148491953a4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dab1e209131cea3b57f.md>)

Recorded evaluation configuration: {"Network":"ResNet18"}. This record does not assert independent reproduction.

### COT

MAE: 2.564. Result ID: 6993e6c53a6c3148491953ba.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71021e209131cea4e815.md>)

Recorded evaluation configuration: {"Network":"WRN-50-2"}. This record does not assert independent reproduction.

### Dispersion

MAE: 2.602. Result ID: 6993e6c53a6c3148491953a6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70f21e209131cea4e42c.md>)

Recorded evaluation configuration: {"Network":"ResNet18"}. This record does not assert independent reproduction.

### Nuclear

MAE: 2.679. Result ID: 6993e6c53a6c3148491953a9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e751e209131cea3ff83.md>)

Recorded evaluation configuration: {"Network":"ResNet18"}. This record does not assert independent reproduction.

### ATC

MAE: 2.911. Result ID: 6993e6c53a6c3148491953ad.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dab1e209131cea3b57f.md>)

Recorded evaluation configuration: {"Network":"ResNet50"}. This record does not assert independent reproduction.

### ConfScore

MAE: 2.99. Result ID: 6993e6c53a6c3148491953b4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64e41e209131cea0731a.md>)

[Implementation](<https://www.sota2.com/github.com/hendrycks/error-detection>)

Recorded evaluation configuration: {"Network":"WRN-50-2"}. This record does not assert independent reproduction.

### Nuclear

MAE: 3.203. Result ID: 6993e6c53a6c3148491953b2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e751e209131cea3ff83.md>)

Recorded evaluation configuration: {"Network":"ResNet50"}. This record does not assert independent reproduction.

### Fréchet

MAE: 3.242. Result ID: 6993e6c53a6c3148491953ae.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a091e209131cea25a4a.md>)

Recorded evaluation configuration: {"Network":"ResNet50"}. This record does not assert independent reproduction.

### ConfScore

MAE: 3.325. Result ID: 6993e6c53a6c3148491953ab.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64e41e209131cea0731a.md>)

[Implementation](<https://www.sota2.com/github.com/hendrycks/error-detection>)

Recorded evaluation configuration: {"Network":"ResNet50"}. This record does not assert independent reproduction.

### Entropy

MAE: 3.357. Result ID: 6993e6c53a6c3148491953ac.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a9b1e209131cea29564.md>)

Recorded evaluation configuration: {"Network":"ResNet50"}. This record does not assert independent reproduction.

### ATC

MAE: 3.388. Result ID: 6993e6c53a6c3148491953b6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6dab1e209131cea3b57f.md>)

Recorded evaluation configuration: {"Network":"WRN-50-2"}. This record does not assert independent reproduction.

### ConfScore

MAE: 3.616. Result ID: 6993e6c53a6c3148491953a2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64e41e209131cea0731a.md>)

[Implementation](<https://www.sota2.com/github.com/hendrycks/error-detection>)

Recorded evaluation configuration: {"Network":"ResNet18"}. This record does not assert independent reproduction.

### Fréchet

MAE: 3.75. Result ID: 6993e6c53a6c3148491953a5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a091e209131cea25a4a.md>)

Recorded evaluation configuration: {"Network":"ResNet18"}. This record does not assert independent reproduction.

### Entropy

MAE: 3.812. Result ID: 6993e6c53a6c3148491953a3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a9b1e209131cea29564.md>)

Recorded evaluation configuration: {"Network":"ResNet18"}. This record does not assert independent reproduction.

### Nuclear

MAE: 4.091. Result ID: 6993e6c53a6c3148491953bb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e751e209131cea3ff83.md>)

Recorded evaluation configuration: {"Network":"WRN-50-2"}. This record does not assert independent reproduction.

### Entropy

MAE: 4.132. Result ID: 6993e6c53a6c3148491953b5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a9b1e209131cea29564.md>)

Recorded evaluation configuration: {"Network":"WRN-50-2"}. This record does not assert independent reproduction.

### MDE

MAE: 4.818. Result ID: 6993e6c53a6c3148491953a7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727f1e209131cea56c3d.md>)

Recorded evaluation configuration: {"Network":"ResNet18"}. This record does not assert independent reproduction.

### Fréchet

MAE: 5.21. Result ID: 6993e6c53a6c3148491953b7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a091e209131cea25a4a.md>)

Recorded evaluation configuration: {"Network":"WRN-50-2"}. This record does not assert independent reproduction.

### MDE

MAE: 6.346. Result ID: 6993e6c53a6c3148491953b0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727f1e209131cea56c3d.md>)

Recorded evaluation configuration: {"Network":"ResNet50"}. This record does not assert independent reproduction.

### MDE

MAE: 6.755. Result ID: 6993e6c53a6c3148491953b9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727f1e209131cea56c3d.md>)

Recorded evaluation configuration: {"Network":"WRN-50-2"}. This record does not assert independent reproduction.

### Dispersion

MAE: 7.318. Result ID: 6993e6c53a6c3148491953af.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70f21e209131cea4e42c.md>)

Recorded evaluation configuration: {"Network":"ResNet50"}. This record does not assert independent reproduction.

### Dispersion

MAE: 10.05. Result ID: 6993e6c53a6c3148491953b8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75d01e209131cea69ff5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70f21e209131cea4e42c.md>)

Recorded evaluation configuration: {"Network":"WRN-50-2"}. This record does not assert independent reproduction.

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