# Offline Reinforcement Learning on HalfCheetah Medium-Expert 1T10S Source: [Offline Reinforcement Learning on HalfCheetah Medium-Expert 1T10S]() Collection index: [llms.txt]() Data updated: 2026-03-04T14:55:29.260Z Evaluation of Offline Reinforcement Learning performance on the HalfCheetah environment using the 1 Trillion transition, 10 Seed (1T10S) dataset configured at the Medium-Expert difficulty level, measured by Average Return. 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 Reinforcement Learning | | Dataset (including any stated split) | HalfCheetah Medium-Expert 1T10S | | Metric | Average Return | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 18 of 18 | | 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 Return](): Higher is better; unit: SCALAR. ## Recorded results | Method | Average Return | Result date | Report paper | | --- | --- | --- | --- | | REAG\*MV | 83.692 | Not recorded | 68fa757c1e209131cea684c9 | | QT | 82.961 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*Dara | 82.148 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*MV | 79.39 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*Dara | 78.981 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*MV | 77.762 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*Dara | 77.751 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*Dara | 76.533 | Not recorded | 68fa757c1e209131cea684c9 | | Reinformer | 76.073 | Not recorded | 68fa757c1e209131cea684c9 | | QT | 71.008 | Not recorded | 68fa757c1e209131cea684c9 | | DT | 70.573 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*MV | 69.819 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*MV | 56.228 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*Dara | 55.818 | Not recorded | 68fa757c1e209131cea684c9 | | DT | 54.981 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*Dara | 51.357 | Not recorded | 68fa757c1e209131cea684c9 | | REAG\*MV | 46.048 | Not recorded | 68fa757c1e209131cea684c9 | | Reinformer | 40.568 | Not recorded | 68fa757c1e209131cea684c9 | ### REAG\*MV Average Return: 83.692. Result ID: 69a847db6dbb7075b94f3672. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"JointNoise (JN)","Base Architecture":"QT"}. This record does not assert independent reproduction. ### QT Average Return: 82.961. Result ID: 69a847db6dbb7075b94f366f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"JointNoise (JN)","Base Architecture":"QT"}. This record does not assert independent reproduction. ### REAG\*Dara Average Return: 82.148. Result ID: 69a847db6dbb7075b94f3675. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"JointNoise (JN)","Base Architecture":"QT"}. This record does not assert independent reproduction. ### REAG\*MV Average Return: 79.39. Result ID: 69a847db6dbb7075b94f3671. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"JointNoise (JN)","Base Architecture":"Reinformer"}. This record does not assert independent reproduction. ### REAG\*Dara Average Return: 78.981. Result ID: 69a847db6dbb7075b94f3674. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"JointNoise (JN)","Base Architecture":"Reinformer"}. This record does not assert independent reproduction. ### REAG\*MV Average Return: 77.762. Result ID: 69a847db6dbb7075b94f3670. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"JointNoise (JN)","Base Architecture":"DT"}. This record does not assert independent reproduction. ### REAG\*Dara Average Return: 77.751. Result ID: 69a847db6dbb7075b94f3673. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"JointNoise (JN)","Base Architecture":"DT"}. This record does not assert independent reproduction. ### REAG\*Dara Average Return: 76.533. Result ID: 69a847db6dbb7075b94f366c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"BodyMass (BM)","Base Architecture":"QT"}. This record does not assert independent reproduction. ### Reinformer Average Return: 76.073. Result ID: 69a847db6dbb7075b94f366e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"JointNoise (JN)","Base Architecture":"Reinformer"}. This record does not assert independent reproduction. ### QT Average Return: 71.008. Result ID: 69a847db6dbb7075b94f3666. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"BodyMass (BM)","Base Architecture":"QT"}. This record does not assert independent reproduction. ### DT Average Return: 70.573. Result ID: 69a847db6dbb7075b94f366d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"JointNoise (JN)","Base Architecture":"DT"}. This record does not assert independent reproduction. ### REAG\*MV Average Return: 69.819. Result ID: 69a847db6dbb7075b94f3669. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"BodyMass (BM)","Base Architecture":"QT"}. This record does not assert independent reproduction. ### REAG\*MV Average Return: 56.228. Result ID: 69a847db6dbb7075b94f3667. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"BodyMass (BM)","Base Architecture":"DT"}. This record does not assert independent reproduction. ### REAG\*Dara Average Return: 55.818. Result ID: 69a847db6dbb7075b94f366b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"BodyMass (BM)","Base Architecture":"Reinformer"}. This record does not assert independent reproduction. ### DT Average Return: 54.981. Result ID: 69a847db6dbb7075b94f3664. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"BodyMass (BM)","Base Architecture":"DT"}. This record does not assert independent reproduction. ### REAG\*Dara Average Return: 51.357. Result ID: 69a847db6dbb7075b94f366a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"BodyMass (BM)","Base Architecture":"DT"}. This record does not assert independent reproduction. ### REAG\*MV Average Return: 46.048. Result ID: 69a847db6dbb7075b94f3668. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"BodyMass (BM)","Base Architecture":"Reinformer"}. This record does not assert independent reproduction. ### Reinformer Average Return: 40.568. Result ID: 69a847db6dbb7075b94f3665. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Distribution Shift":"BodyMass (BM)","Base Architecture":"Reinformer"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.