# Offline-to-online Reinforcement Learning on Robomimic and OGBench

Source: [Offline-to-online Reinforcement Learning on Robomimic and OGBench](<https://www.sota2.com/research/sota/offline-to-online-reinforcement-learning-on-robomimic-and-ogbench>)

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

Data updated: 2026-05-11T10:09:44.152Z

Evaluation of Offline-to-online Reinforcement Learning performance spanning both the Robomimic and OGBench datasets, measured by average success rates.

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 | Offline-to-online Reinforcement Learning |
| Dataset (including any stated split) | Robomimic and OGBench |
| Metric | Average Success Rate |
| Unit | PERCENT |
| Direction | Higher 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

- [Average Success Rate](<https://www.sota2.com/research/sota/offline-to-online-reinforcement-learning-on-robomimic-and-ogbench.md?metric=Average%20Success%20Rate>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Average Success Rate | Result date | Report paper |
| --- | --- | --- | --- |
| DFP | 95.8 | Not recorded | 6a0162badbc836b87d823326 |
| QC-FQL | 86.8 | Not recorded | 6a0162badbc836b87d823326 |
| QC-BFN | 83.8 | Not recorded | 6a0162badbc836b87d823326 |
| MVP | 80.3 | Not recorded | 6a0162badbc836b87d823326 |
| BFN | 38.4 | Not recorded | 6a0162badbc836b87d823326 |
| FQL | 30.3 | Not recorded | 6a0162badbc836b87d823326 |

### DFP

Average Success Rate: 95.8. Result ID: 6a01aae147f34ce924e41944.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162badbc836b87d823326.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0162badbc836b87d823326.md>)

Recorded evaluation configuration: {"inference\_type":"teacher-free one-step","action\_chunking":"true"}. This record does not assert independent reproduction.

### QC-FQL

Average Success Rate: 86.8. Result ID: 6a01aae147f34ce924e41942.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162badbc836b87d823326.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a0f1e209131cea80d8d.md>)

Recorded evaluation configuration: {"inference\_type":"distilled one-step","action\_chunking":"true"}. This record does not assert independent reproduction.

### QC-BFN

Average Success Rate: 83.8. Result ID: 6a01aae147f34ce924e41940.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162badbc836b87d823326.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a0f1e209131cea80d8d.md>)

Recorded evaluation configuration: {"inference\_type":"multi-step","action\_chunking":"true"}. This record does not assert independent reproduction.

### MVP

Average Success Rate: 80.3. Result ID: 6a01aae147f34ce924e41943.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162badbc836b87d823326.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f5f69ebc499c974099c3.md>)

Recorded evaluation configuration: {"inference\_type":"teacher-free one-step","action\_chunking":"true"}. This record does not assert independent reproduction.

### BFN

Average Success Rate: 38.4. Result ID: 6a01aae147f34ce924e4193f.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162badbc836b87d823326.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa691a1e209131cea203dd.md>)

Recorded evaluation configuration: {"inference\_type":"multi-step","action\_chunking":"false"}. This record does not assert independent reproduction.

### FQL

Average Success Rate: 30.3. Result ID: 6a01aae147f34ce924e41941.

[Reporting paper](<https://www.sota2.com/research/paper/6a0162badbc836b87d823326.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78c51e209131cea7a67a.md>)

Recorded evaluation configuration: {"inference\_type":"distilled one-step","action\_chunking":"false"}. This record does not assert independent reproduction.

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