# Reinforcement Learning on HalfCheetah Pure-8-20 Source: [Reinforcement Learning on HalfCheetah Pure-8-20]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:35:45.017Z Evaluation of Reinforcement Learning performance on the HalfCheetah Pure environment variant configured as '8-20', measured by specific reward configurations. 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 | Reinforcement Learning | | Dataset (including any stated split) | HalfCheetah Pure-8-20 | | Metric | Average Reward (0.3/1.7) | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 10 of 10 | | 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 Reward (0.3/1.7)](): Higher is better; unit: SCALAR. - [Average Reward (1.7/0.3)](): Higher is better; unit: SCALAR. - [Average Reward (0.3/0.3)](): Higher is better; unit: SCALAR. ## Recorded results | Method | Average Reward (0.3/1.7) | Result date | Report paper | | --- | --- | --- | --- | | True env+MPC | 7459 | Not recorded | 68fa6b371e209131cea2cf55 | | PerSim | 3186 | Not recorded | 68fa6b371e209131cea2cf55 | | PE-TS CaDM | 1082 | Not recorded | 68fa6b371e209131cea2cf55 | | CQL | 838.7 | Not recorded | 68fa6b371e209131cea2cf55 | | Vanilla CaDM | 412 | Not recorded | 68fa6b371e209131cea2cf55 | | BCQ | 376.8 | Not recorded | 68fa6b371e209131cea2cf55 | | BCQ | 254.6 | Not recorded | 68fa6b371e209131cea2cf55 | | MOREL | 0.6 | Not recorded | 68fa6b371e209131cea2cf55 | | CQL | -15.5 | Not recorded | 68fa6b371e209131cea2cf55 | | MOREL | -781 | Not recorded | 68fa6b371e209131cea2cf55 | ### True env+MPC Average Reward (0.3/1.7): 7459. Result ID: 698e71aaffc3901de09da412. [Reporting paper]() Recorded evaluation configuration: {"Data source (optimality)":"Oracle (True Environment)"}. This record does not assert independent reproduction. ### PerSim Average Reward (0.3/1.7): 3186. Result ID: 698e71aaffc3901de09da409. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Data source (optimality)":"Pure-8-20"}. This record does not assert independent reproduction. ### PE-TS CaDM Average Reward (0.3/1.7): 1082. Result ID: 698e71aaffc3901de09da40b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Data source (optimality)":"Pure-8-20"}. This record does not assert independent reproduction. ### CQL Average Reward (0.3/1.7): 838.7. Result ID: 698e71aaffc3901de09da40e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Data source (optimality)":"Pure-8-20","Algorithm variant":"Planning"}. This record does not assert independent reproduction. ### Vanilla CaDM Average Reward (0.3/1.7): 412. Result ID: 698e71aaffc3901de09da40a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Data source (optimality)":"Pure-8-20"}. This record does not assert independent reproduction. ### BCQ Average Reward (0.3/1.7): 376.8. Result ID: 698e71aaffc3901de09da40d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Data source (optimality)":"Pure-8-20","Algorithm variant":"Acting"}. This record does not assert independent reproduction. ### BCQ Average Reward (0.3/1.7): 254.6. Result ID: 698e71aaffc3901de09da40c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Data source (optimality)":"Pure-8-20","Algorithm variant":"Planning"}. This record does not assert independent reproduction. ### MOREL Average Reward (0.3/1.7): 0.6. Result ID: 698e71aaffc3901de09da410. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Data source (optimality)":"Pure-8-20","Algorithm variant":"Planning"}. This record does not assert independent reproduction. ### CQL Average Reward (0.3/1.7): -15.5. Result ID: 698e71aaffc3901de09da40f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Data source (optimality)":"Pure-8-20","Algorithm variant":"Acting"}. This record does not assert independent reproduction. ### MOREL Average Reward (0.3/1.7): -781. Result ID: 698e71aaffc3901de09da411. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Data source (optimality)":"Pure-8-20","Algorithm variant":"Acting"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.