Offline Reinforcement Learning on OGBench cube-double
72Average Task SuccessFlowIQN-R
Evaluation results
| Method | Links | |
|---|---|---|
| FlowIQN-RPolicy Type=Flow Policies2026.05 | 72 | |
| IQNPolicy Type=Gaussian Policies2026.05 | 64 | |
| Value FlowsPolicy Type=Flow Policies2026.05 | 63 | |
| CODACPolicy Type=Gaussian Policies2026.05 | 61 | |
| FlowIQN-FQLPolicy Type=Flow Policies2026.05 | 52 | |
| floqPolicy Type=Flow Policies2026.05 | 47 | |
| FQLPolicy Type=Gaussian Policies2026.05 | 25 | |
| ReBRACPolicy Type=Gaussian Policies2026.05 | 12 | |
| BCPolicy Type=Gaussian Policies2026.05 | 2 |