# Image Classification on MNIST (retained set D\\Du)

Source: [Image Classification on MNIST (retained set D\\Du)](<https://www.sota2.com/research/sota/image-classification-on-mnist-retained-set-d-du>)

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

Data updated: 2026-04-23T09:41:31.848Z

Evaluation of Image Classification performance on the specific retained subset (D\\Du) of the MNIST dataset, reporting Macro F1-score.

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) | MNIST (retained set D\\Du) |
| Metric | Macro F1-score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| 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

- [Macro F1-score](<https://www.sota2.com/research/sota/image-classification-on-mnist-retained-set-d-du.md?metric=Macro%20F1-score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Macro F1-score | Result date | Report paper |
| --- | --- | --- | --- |
| Original | 99.76 | Not recorded | 68fa74501e209131cea60fc7 |
| Adv-R (Sr) | 99.71 | Not recorded | 68fa74501e209131cea60fc7 |
| Retrain | 99.52 | Not recorded | 68fa74501e209131cea60fc7 |
| Adv-R (Sr) | 99.33 | Not recorded | 68fa74501e209131cea60fc7 |
| Adv-F | 99.33 | Not recorded | 68fa74501e209131cea60fc7 |
| Adv-R (Sn) | 98.65 | Not recorded | 68fa74501e209131cea60fc7 |
| Adv-R (Sn) | 97.96 | Not recorded | 68fa74501e209131cea60fc7 |
| Adv-F | 97.28 | Not recorded | 68fa74501e209131cea60fc7 |
| Original | 97 | Not recorded | 68fa74501e209131cea60fc7 |
| Retrain | 95.86 | Not recorded | 68fa74501e209131cea60fc7 |

### Original

Macro F1-score: 99.76. Result ID: 69e9e93b47f34ce924d178d6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

Recorded evaluation configuration: {"Backbone":"MLP","Class Setting":"Balanced"}. This record does not assert independent reproduction.

### Adv-R (Sr)

Macro F1-score: 99.71. Result ID: 69e9e93b47f34ce924d178dd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

Recorded evaluation configuration: {"Backbone":"MLP","Class Setting":"Imbalanced","Selection Strategy":"Random Sampling"}. This record does not assert independent reproduction.

### Retrain

Macro F1-score: 99.52. Result ID: 69e9e93b47f34ce924d178d7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

Recorded evaluation configuration: {"Backbone":"MLP","Class Setting":"Balanced"}. This record does not assert independent reproduction.

### Adv-R (Sr)

Macro F1-score: 99.33. Result ID: 69e9e93b47f34ce924d178d8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

Recorded evaluation configuration: {"Backbone":"MLP","Class Setting":"Balanced","Selection Strategy":"Random Sampling"}. This record does not assert independent reproduction.

### Adv-F

Macro F1-score: 99.33. Result ID: 69e9e93b47f34ce924d178da.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

Recorded evaluation configuration: {"Backbone":"MLP","Class Setting":"Balanced"}. This record does not assert independent reproduction.

### Adv-R (Sn)

Macro F1-score: 98.65. Result ID: 69e9e93b47f34ce924d178d9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

Recorded evaluation configuration: {"Backbone":"MLP","Class Setting":"Balanced","Selection Strategy":"Nearest Neighbor"}. This record does not assert independent reproduction.

### Adv-R (Sn)

Macro F1-score: 97.96. Result ID: 69e9e93b47f34ce924d178de.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

Recorded evaluation configuration: {"Backbone":"MLP","Class Setting":"Imbalanced","Selection Strategy":"Nearest Neighbor"}. This record does not assert independent reproduction.

### Adv-F

Macro F1-score: 97.28. Result ID: 69e9e93b47f34ce924d178df.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

Recorded evaluation configuration: {"Backbone":"MLP","Class Setting":"Imbalanced"}. This record does not assert independent reproduction.

### Original

Macro F1-score: 97. Result ID: 69e9e93b47f34ce924d178db.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

Recorded evaluation configuration: {"Backbone":"MLP","Class Setting":"Imbalanced"}. This record does not assert independent reproduction.

### Retrain

Macro F1-score: 95.86. Result ID: 69e9e93b47f34ce924d178dc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74501e209131cea60fc7.md>)

Recorded evaluation configuration: {"Backbone":"MLP","Class Setting":"Imbalanced"}. This record does not assert independent reproduction.

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