# Offline Reinforcement Learning on D4RL HalfCheetah Medium-Replay

Source: [Offline Reinforcement Learning on D4RL HalfCheetah Medium-Replay](<https://www.sota2.com/research/sota/offline-reinforcement-learning-on-d4rl-halfcheetah-medium-replay>)

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

Data updated: 2026-02-26T06:02:17.219Z

Evaluation of Offline Reinforcement Learning performance on the D4RL HalfCheetah environment using the Medium-Replay dataset configuration, measured by Reward.

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 HalfCheetah Medium-Replay |
| Metric | Reward |
| 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

- [Reward](<https://www.sota2.com/research/sota/offline-reinforcement-learning-on-d4rl-halfcheetah-medium-replay.md?metric=Reward>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Reward | Result date | Report paper |
| --- | --- | --- | --- |
| RTB | 48.11 | Not recorded | 68fa76891e209131cea6d825 |
| QGPO | 47.6 | Not recorded | 68fa76891e209131cea6d825 |
| D-QL | 47.5 | Not recorded | 68fa76891e209131cea6d825 |
| IDQL | 45.8 | Not recorded | 68fa76891e209131cea6d825 |
| CQL | 45.5 | Not recorded | 68fa76891e209131cea6d825 |
| IQL | 44.2 | Not recorded | 68fa76891e209131cea6d825 |
| D | 42.2 | Not recorded | 68fa76891e209131cea6d825 |
| AWAC | 40.5 | Not recorded | 68fa6cbf1e209131cea35c2d |
| DD | 39.3 | Not recorded | 68fa76891e209131cea6d825 |
| BC | 36.6 | Not recorded | 68fa6cbf1e209131cea35c2d |
| BC | 36.6 | Not recorded | 68fa76891e209131cea6d825 |
| FLOWFORMER | 34.7 | Not recorded | 68fa6cbf1e209131cea35c2d |
| DT | 34.6 | Not recorded | 68fa6cbf1e209131cea35c2d |
| Reformer | 33.6 | Not recorded | 68fa6cbf1e209131cea35c2d |
| cosFormer | 32.8 | Not recorded | 68fa6cbf1e209131cea35c2d |
| Linear Trans. | 32.1 | Not recorded | 68fa6cbf1e209131cea35c2d |
| Performer | 31.7 | Not recorded | 68fa6cbf1e209131cea35c2d |

### RTB

Reward: 48.11. Result ID: 69943f7a3a6c3148491c0d22.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76891e209131cea6d825.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76891e209131cea6d825.md>)

[Implementation](<https://github.com/GFNOrg/diffusion-finetuning>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### QGPO

Reward: 47.6. Result ID: 69943f7a3a6c3148491c0d21.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76891e209131cea6d825.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f931e209131cea465bc.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### D-QL

Reward: 47.5. Result ID: 69943f7a3a6c3148491c0d1f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76891e209131cea6d825.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70791e209131cea4b56e.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### IDQL

Reward: 45.8. Result ID: 69943f7a3a6c3148491c0d20.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76891e209131cea6d825.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f731e209131cea45a93.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### CQL

Reward: 45.5. Result ID: 69943f7a3a6c3148491c0d1b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76891e209131cea6d825.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa682a1e209131cea1a6c2.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### IQL

Reward: 44.2. Result ID: 69943f7a3a6c3148491c0d1c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76891e209131cea6d825.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6af71e209131cea2b67f.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### D

Reward: 42.2. Result ID: 69943f7a3a6c3148491c0d1d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76891e209131cea6d825.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e3b1e209131cea3e868.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### AWAC

Reward: 40.5. Result ID: 698f3a813a6c314849f7383a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cbf1e209131cea35c2d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69cc1e209131cea24226.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### DD

Reward: 39.3. Result ID: 69943f7a3a6c3148491c0d1e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76891e209131cea6d825.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ffa1e209131cea4849f.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### BC

Reward: 36.6. Result ID: 698f3a813a6c314849f73839.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cbf1e209131cea35c2d.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### BC

Reward: 36.6. Result ID: 69943f7a3a6c3148491c0d1a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa76891e209131cea6d825.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### FLOWFORMER

Reward: 34.7. Result ID: 698f3a813a6c314849f73840.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cbf1e209131cea35c2d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cbf1e209131cea35c2d.md>)

[Implementation](<https://github.com/thuml/Flowformer>)

Recorded evaluation configuration: {"Layers":3,"Hidden Channels":256,"Heads":4}. This record does not assert independent reproduction.

### DT

Reward: 34.6. Result ID: 698f3a813a6c314849f7383b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cbf1e209131cea35c2d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a381e209131cea26b46.md>)

Recorded evaluation configuration: {"Layers":3,"Hidden Channels":256,"Heads":4}. This record does not assert independent reproduction.

### Reformer

Reward: 33.6. Result ID: 698f3a813a6c314849f7383d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cbf1e209131cea35c2d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa671b1e209131cea143ed.md>)

[Implementation](<https://github.com/google/trax/tree/master/trax/models/reformer>)

Recorded evaluation configuration: {"Layers":3,"Hidden Channels":256,"Heads":4}. This record does not assert independent reproduction.

### cosFormer

Reward: 32.8. Result ID: 698f3a813a6c314849f7383f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cbf1e209131cea35c2d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6bd21e209131cea304be.md>)

[Implementation](<https://www.sota2.com/cosFormer>)

Recorded evaluation configuration: {"Layers":3,"Hidden Channels":256,"Heads":4}. This record does not assert independent reproduction.

### Linear Trans.

Reward: 32.1. Result ID: 698f3a813a6c314849f7383c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cbf1e209131cea35c2d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa68471e209131cea1b069.md>)

[Implementation](<https://linear-transformers.com/>)

Recorded evaluation configuration: {"Layers":3,"Hidden Channels":256,"Heads":4}. This record does not assert independent reproduction.

### Performer

Reward: 31.7. Result ID: 698f3a813a6c314849f7383e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6cbf1e209131cea35c2d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e031e209131cea3d699.md>)

[Implementation](<https://github.com/google-research/google-research/tree/master/performer>)

Recorded evaluation configuration: {"Layers":3,"Hidden Channels":256,"Heads":4}. This record does not assert independent reproduction.

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