Metric: Average Reward (higher is better)
| # | Model↕ | Average Reward▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | PMDB | 88.2 | No | Model-Based Offline Reinforcement Learning with ... | 2022-10-13 | Code |
| 2 | KFC | 81.8 | No | Koopman Q-learning: Offline Reinforcement Learni... | 2021-11-02 | - |
| 3 | Primal.+DT | 77.5 | No | Primal-Attention: Self-attention through Asymmet... | 2023-05-31 | Code |
| 4 | Flowformer | 73.5 | No | Flowformer: Linearizing Transformers with Conser... | 2022-02-13 | Code |
| 5 | Decision Transformer (DT) | 72.2 | No | Decision Transformer: Reinforcement Learning via... | 2021-06-02 | Code |
| 6 | cosFormer | 67.8 | No | cosFormer: Rethinking Softmax in Attention | 2022-02-17 | Code |
| 7 | Linear Transformer | 64.4 | No | Transformers are RNNs: Fast Autoregressive Trans... | 2020-06-29 | Code |
| 8 | Reformer | 63.9 | No | Reformer: The Efficient Transformer | 2020-01-13 | Code |
| 9 | Performer | 63.8 | No | Rethinking Attention with Performers | 2020-09-30 | Code |