# Image Classification on MNIST (Convergence Time Metrics)

Source: [Image Classification on MNIST (Convergence Time Metrics)](<https://www.sota2.com/research/sota/image-classification-on-mnist-convergence-time-metrics>)

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

Data updated: 2026-05-12T11:47:56.264Z

Evaluation of Image Classification performance on the MNIST dataset, focusing on convergence efficiency by reporting the wall-clock time required to achieve specific accuracy milestones (e.g., 80%, 85%) and the final maximum accuracy achieved.

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 |
| Metric | Time to 80% Acc (s) |
| Unit | TIME |
| Direction | Lower 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

- [Time to 80% Acc (s)](<https://www.sota2.com/research/sota/image-classification-on-mnist-convergence-time-metrics.md?metric=Time%20to%2080%25%20Acc%20(s)>): Lower is better; unit: TIME.

- [Time to 85% Acc (s)](<https://www.sota2.com/research/sota/image-classification-on-mnist-convergence-time-metrics.md?metric=Time%20to%2085%25%20Acc%20(s)>): Lower is better; unit: TIME.

- [Max Accuracy](<https://www.sota2.com/research/sota/image-classification-on-mnist-convergence-time-metrics.md?metric=Max%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Time to 80% Acc (s) | Result date | Report paper |
| --- | --- | --- | --- |
| FAR-SIGN | 19 | Not recorded | 6a02b465dbc836b87d824ab7 |
| Trimmed Mean | 266 | Not recorded | 6a02b465dbc836b87d824ab7 |
| Multi-Krum | 268 | Not recorded | 6a02b465dbc836b87d824ab7 |
| Median | 280 | Not recorded | 6a02b465dbc836b87d824ab7 |
| RFA | 285 | Not recorded | 6a02b465dbc836b87d824ab7 |
| Bulyan | 408 | Not recorded | 6a02b465dbc836b87d824ab7 |
| Krum | 643 | Not recorded | 6a02b465dbc836b87d824ab7 |

### FAR-SIGN

Time to 80% Acc (s): 19. Result ID: 6a03135a47f34ce924e700e2.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b465dbc836b87d824ab7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a02b465dbc836b87d824ab7.md>)

Recorded evaluation configuration: {"Model":"MLP","Byzantine workers":"24%","Clients":"51"}. This record does not assert independent reproduction.

### Trimmed Mean

Time to 80% Acc (s): 266. Result ID: 6a03135a47f34ce924e700e6.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b465dbc836b87d824ab7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69531e209131cea218fb.md>)

Recorded evaluation configuration: {"Model":"MLP","Byzantine workers":"24%","Clients":"51"}. This record does not assert independent reproduction.

### Multi-Krum

Time to 80% Acc (s): 268. Result ID: 6a03135a47f34ce924e700e4.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b465dbc836b87d824ab7.md>)

Recorded evaluation configuration: {"Model":"MLP","Byzantine workers":"24%","Clients":"51"}. This record does not assert independent reproduction.

### Median

Time to 80% Acc (s): 280. Result ID: 6a03135a47f34ce924e700e5.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b465dbc836b87d824ab7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69531e209131cea218fb.md>)

Recorded evaluation configuration: {"Model":"MLP","Byzantine workers":"24%","Clients":"51"}. This record does not assert independent reproduction.

### RFA

Time to 80% Acc (s): 285. Result ID: 6a03135a47f34ce924e700e3.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b465dbc836b87d824ab7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa703e1e209131cea49e91.md>)

Recorded evaluation configuration: {"Model":"MLP","Byzantine workers":"24%","Clients":"51"}. This record does not assert independent reproduction.

### Bulyan

Time to 80% Acc (s): 408. Result ID: 6a03135a47f34ce924e700e8.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b465dbc836b87d824ab7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa649f1e209131cea05b21.md>)

Recorded evaluation configuration: {"Model":"MLP","Byzantine workers":"24%","Clients":"51"}. This record does not assert independent reproduction.

### Krum

Time to 80% Acc (s): 643. Result ID: 6a03135a47f34ce924e700e7.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b465dbc836b87d824ab7.md>)

Recorded evaluation configuration: {"Model":"MLP","Byzantine workers":"24%","Clients":"51"}. This record does not assert independent reproduction.

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