# RBM learning on MNIST (test)

Source: [RBM learning on MNIST (test)](<https://www.sota2.com/research/sota/rbm-learning-on-mnist-test>)

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

Data updated: 2026-02-26T05:48:41.191Z

Evaluation of Restricted Boltzmann Machine (RBM) learning performance on the test split of the MNIST dataset, typically measured using Log Likelihood.

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 | RBM learning |
| Dataset (including any stated split) | MNIST (test) |
| Metric | Log Likelihood (AIS) |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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

- [Log Likelihood (AIS)](<https://www.sota2.com/research/sota/rbm-learning-on-mnist-test.md?metric=Log%20Likelihood%20(AIS)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Log Likelihood (AIS) | Result date | Report paper |
| --- | --- | --- | --- |
| GB | -191.98 | Not recorded | 68fa75581e209131cea67590 |
| ACS | -249.55 | Not recorded | 68fa75581e209131cea67590 |
| DMALA | -278.35 | Not recorded | 68fa75581e209131cea67590 |
| GWG | -387.34 | Not recorded | 68fa75581e209131cea67590 |

### GB

Log Likelihood (AIS): -191.98. Result ID: 6993a9f63a6c31484917a463.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75581e209131cea67590.md>)

Recorded evaluation configuration: {"Sampling method":"Block-Gibbs"}. This record does not assert independent reproduction.

### ACS

Log Likelihood (AIS): -249.55. Result ID: 6993a9f63a6c31484917a466.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75581e209131cea67590.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75581e209131cea67590.md>)

[Implementation](<https://github.com/patrickpynadath1/automatic_cyclical_sampling>)

Recorded evaluation configuration: {"Sampling method":"Automatic Cyclical Sampler"}. This record does not assert independent reproduction.

### DMALA

Log Likelihood (AIS): -278.35. Result ID: 6993a9f63a6c31484917a465.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75581e209131cea67590.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cdc1e209131cea3658f.md>)

Recorded evaluation configuration: {"Sampling method":"Discrete Metropolis-Adjusted Langevin Algorithm"}. This record does not assert independent reproduction.

### GWG

Log Likelihood (AIS): -387.34. Result ID: 6993a9f63a6c31484917a464.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75581e209131cea67590.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a211e209131cea26321.md>)

Recorded evaluation configuration: {"Sampling method":"Gibbs-With-Gradients"}. This record does not assert independent reproduction.

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