# Neighbor Embedding on MNIST (train)

Source: [Neighbor Embedding on MNIST (train)](<https://www.sota2.com/research/sota/neighbor-embedding-on-mnist-train>)

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

Data updated: 2026-06-04T11:59:48.277Z

Evaluation of the quality of learned embeddings on the MNIST training set, specifically assessing local neighborhood preservation using metrics like T5, T30, and T100.

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 | Neighbor Embedding |
| Dataset (including any stated split) | MNIST (train) |
| Metric | T5 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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

- [T5](<https://www.sota2.com/research/sota/neighbor-embedding-on-mnist-train.md?metric=T5>): Higher is better; unit: PERCENT.

- [T30](<https://www.sota2.com/research/sota/neighbor-embedding-on-mnist-train.md?metric=T30>): Higher is better; unit: PERCENT.

- [T100](<https://www.sota2.com/research/sota/neighbor-embedding-on-mnist-train.md?metric=T100>): Higher is better; unit: PERCENT.

## Recorded results

| Method | T5 | Result date | Report paper |
| --- | --- | --- | --- |
| P-UMAP-CE | 95.55 | Not recorded | 6a2106f8c10484bac23f06ab |
| P-UMAP-CEMSE | 95.5 | Not recorded | 6a2106f8c10484bac23f06ab |
| UMAP (Train+Test) | 95.37 | Not recorded | 6a2106f8c10484bac23f06ab |
| UMAP | 95.23 | Not recorded | 6a2106f8c10484bac23f06ab |
| P-UMAP-MSE | 95.23 | Not recorded | 6a2106f8c10484bac23f06ab |

### P-UMAP-CE

T5: 95.55. Result ID: 6a21689afbc373592c1e5d38.

[Reporting paper](<https://www.sota2.com/research/paper/6a2106f8c10484bac23f06ab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a2106f8c10484bac23f06ab.md>)

Recorded evaluation configuration: {"loss":"CE"}. This record does not assert independent reproduction.

### P-UMAP-CEMSE

T5: 95.5. Result ID: 6a21689afbc373592c1e5d37.

[Reporting paper](<https://www.sota2.com/research/paper/6a2106f8c10484bac23f06ab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a2106f8c10484bac23f06ab.md>)

Recorded evaluation configuration: {"loss":"CEMSE"}. This record does not assert independent reproduction.

### UMAP (Train+Test)

T5: 95.37. Result ID: 6a21689afbc373592c1e5d35.

[Reporting paper](<https://www.sota2.com/research/paper/6a2106f8c10484bac23f06ab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68501e209131cea1b5d2.md>)

[Implementation](<https://github.com/lmcinnes/umap>)

Recorded evaluation configuration: {"evaluation\_protocol":"combined training and test data"}. This record does not assert independent reproduction.

### UMAP

T5: 95.23. Result ID: 6a21689afbc373592c1e5d34.

[Reporting paper](<https://www.sota2.com/research/paper/6a2106f8c10484bac23f06ab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68501e209131cea1b5d2.md>)

[Implementation](<https://github.com/lmcinnes/umap>)

Recorded evaluation configuration: {"ns":"5"}. This record does not assert independent reproduction.

### P-UMAP-MSE

T5: 95.23. Result ID: 6a21689afbc373592c1e5d36.

[Reporting paper](<https://www.sota2.com/research/paper/6a2106f8c10484bac23f06ab.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a2106f8c10484bac23f06ab.md>)

Recorded evaluation configuration: {"loss":"MSE"}. This record does not assert independent reproduction.

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