# Offline Reinforcement Learning on D4RL Gym halfcheetah (full-replay) Source: [Offline Reinforcement Learning on D4RL Gym halfcheetah (full-replay)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:53:45.680Z Evaluation of Offline Reinforcement Learning performance on the D4RL HalfCheetah environment using the full-replay dataset configuration, typically measured by normalized 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) | D4RL Gym halfcheetah (full-replay) | | Metric | Normalized Return | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 17 of 17 | | 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 - [Normalized Return](): Higher is better; unit: SCALAR. ## Recorded results | Method | Normalized Return | Result date | Report paper | | --- | --- | --- | --- | | SPQR | 88.5 | Not recorded | 68fa71b41e209131cea52a13 | | SAC | 86.8 | Not recorded | 68fa6aec1e209131cea2b133 | | SPQR | 86.8 | Not recorded | 68fa71b41e209131cea52a13 | | EDAC | 84.6 | Not recorded | 68fa6aec1e209131cea2b133 | | EDAC | 84.6 | Not recorded | 68fa71b41e209131cea52a13 | | SAC-N | 84.5 | Not recorded | 68fa6aec1e209131cea2b133 | | SAC-Min | 84.5 | Not recorded | 68fa71b41e209131cea52a13 | | CQL-Min | 82.07 | Not recorded | 68fa71b41e209131cea52a13 | | BRAC | 78 | Not recorded | 68fa6aec1e209131cea2b133 | | CQL | 76.9 | Not recorded | 68fa6aec1e209131cea2b133 | | MOREL | 70.1 | Not recorded | 68fa6aec1e209131cea2b133 | | BCQ | 69.5 | Not recorded | 68fa6aec1e209131cea2b133 | | UWAC | 65.1 | Not recorded | 68fa6aec1e209131cea2b133 | | BC | 62.9 | Not recorded | 68fa6aec1e209131cea2b133 | | BC | 62.9 | Not recorded | 68fa71b41e209131cea52a13 | | BEAR | 60.1 | Not recorded | 68fa6aec1e209131cea2b133 | | REM | 27.8 | Not recorded | 68fa6aec1e209131cea2b133 | ### SPQR Normalized Return: 88.5. Result ID: 6991dcb93a6c3148490b0f9a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Base algorithm":"SAC-Min"}. This record does not assert independent reproduction. ### SAC Normalized Return: 86.8. Result ID: 698e41bbffc3901de09c1263. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Implementation":"Reproduced"}. This record does not assert independent reproduction. ### SPQR Normalized Return: 86.8. Result ID: 6991dcb93a6c3148490b0f9b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Base algorithm":"EDAC"}. This record does not assert independent reproduction. ### EDAC Normalized Return: 84.6. Result ID: 698e41bbffc3901de09c126c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Implementation":"Ours"}. This record does not assert independent reproduction. ### EDAC Normalized Return: 84.6. Result ID: 6991dcb93a6c3148490b0f98. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### SAC-N Normalized Return: 84.5. Result ID: 698e41bbffc3901de09c126b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Implementation":"Ours"}. This record does not assert independent reproduction. ### SAC-Min Normalized Return: 84.5. Result ID: 6991dcb93a6c3148490b0f99. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### CQL-Min Normalized Return: 82.07. Result ID: 6991dcb93a6c3148490b0f97. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### BRAC Normalized Return: 78. Result ID: 698e41bbffc3901de09c1267. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Implementation":"Reproduced"}. This record does not assert independent reproduction. ### CQL Normalized Return: 76.9. Result ID: 698e41bbffc3901de09c126a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Implementation":"Reproduced"}. This record does not assert independent reproduction. ### MOREL Normalized Return: 70.1. Result ID: 698e41bbffc3901de09c1268. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Implementation":"Reproduced"}. This record does not assert independent reproduction. ### BCQ Normalized Return: 69.5. Result ID: 698e41bbffc3901de09c1265. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Implementation":"Reproduced"}. This record does not assert independent reproduction. ### UWAC Normalized Return: 65.1. Result ID: 698e41bbffc3901de09c1269. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Implementation":"Reproduced"}. This record does not assert independent reproduction. ### BC Normalized Return: 62.9. Result ID: 698e41bbffc3901de09c1262. [Reporting paper]() Recorded evaluation configuration: {"Implementation":"Reproduced"}. This record does not assert independent reproduction. ### BC Normalized Return: 62.9. Result ID: 6991dcb93a6c3148490b0f96. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### BEAR Normalized Return: 60.1. Result ID: 698e41bbffc3901de09c1266. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Implementation":"Reproduced"}. This record does not assert independent reproduction. ### REM Normalized Return: 27.8. Result ID: 698e41bbffc3901de09c1264. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Implementation":"Reproduced"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.