# Model-free learning on Stochastic MDPs Bilinear or BE

Source: [Model-free learning on Stochastic MDPs Bilinear or BE](<https://www.sota2.com/research/sota/model-free-learning-on-stochastic-mdps-bilinear-or-be>)

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

Data updated: 2026-03-20T08:03:07.030Z

Evaluation of model-free reinforcement learning algorithms applied to Stochastic MDPs in the Bilinear or BE setting, focusing on the theoretical dependence of the regret bound on the time horizon T.

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 | Model-free learning |
| Dataset (including any stated split) | Stochastic MDPs Bilinear or BE |
| Metric | Regret T Exponent |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 6 of 6 |
| 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

- [Regret T Exponent](<https://www.sota2.com/research/sota/model-free-learning-on-stochastic-mdps-bilinear-or-be.md?metric=Regret%20T%20Exponent>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Regret T Exponent | Result date | Report paper |
| --- | --- | --- | --- |
| Du et al. (2021), Jin et al. (2021), Xie et al. (2023) | 2 | Not recorded | 68fa7bfe1e209131cea8bbc3 |
| Du et al. (2021), Jin et al. (2021), Xie et al. (2023) | 2 | Not recorded | 68fa7bfe1e209131cea8bbc3 |
| Dig-DEC | 2 | Not recorded | 68fa7bfe1e209131cea8bbc3 |
| Foster et al. (2023b) | 3 | Not recorded | 68fa7bfe1e209131cea8bbc3 |
| Foster et al. (2023b) | 5 | Not recorded | 68fa7bfe1e209131cea8bbc3 |
| Dig-DEC | 7 | Not recorded | 68fa7bfe1e209131cea8bbc3 |

### Du et al. (2021), Jin et al. (2021), Xie et al. (2023)

Regret T Exponent: 2. Result ID: 69bcfd7047f34ce924a6d160.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bfe1e209131cea8bbc3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa584b1e209131ce9acfde.md>)

Recorded evaluation configuration: {"Exploration Mechanism":"optimism","Estimation Protocol":"on-policy"}. This record does not assert independent reproduction.

### Du et al. (2021), Jin et al. (2021), Xie et al. (2023)

Regret T Exponent: 2. Result ID: 69bcfd7047f34ce924a6d161.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bfe1e209131cea8bbc3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa584b1e209131ce9acfde.md>)

Recorded evaluation configuration: {"Exploration Mechanism":"optimism","Estimation Protocol":"off-policy"}. This record does not assert independent reproduction.

### Dig-DEC

Regret T Exponent: 2. Result ID: 69bcfd7047f34ce924a6d164.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bfe1e209131cea8bbc3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bfe1e209131cea8bbc3.md>)

Recorded evaluation configuration: {"Exploration Mechanism":"information gain","Estimation Protocol":"on-policy"}. This record does not assert independent reproduction.

### Foster et al. (2023b)

Regret T Exponent: 3. Result ID: 69bcfd7047f34ce924a6d162.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bfe1e209131cea8bbc3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa5aca1e209131ce9c2c4d.md>)

Recorded evaluation configuration: {"Exploration Mechanism":"information gain + optimism","Estimation Protocol":"on-policy"}. This record does not assert independent reproduction.

### Foster et al. (2023b)

Regret T Exponent: 5. Result ID: 69bcfd7047f34ce924a6d163.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bfe1e209131cea8bbc3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa5aca1e209131ce9c2c4d.md>)

Recorded evaluation configuration: {"Exploration Mechanism":"information gain + optimism","Estimation Protocol":"off-policy"}. This record does not assert independent reproduction.

### Dig-DEC

Regret T Exponent: 7. Result ID: 69bcfd7047f34ce924a6d165.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bfe1e209131cea8bbc3.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bfe1e209131cea8bbc3.md>)

Recorded evaluation configuration: {"Exploration Mechanism":"information gain","Estimation Protocol":"off-policy"}. This record does not assert independent reproduction.

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