Puzzle Solving on OGBench puzzle-4x4-task4
26Success RateIDQL
Evaluation results
| Method | Links | |
|---|---|---|
| IDQLPolicy Type=Diffusion Policies, Evaluation Protocol=Offline RL2026.05 | 26 | |
| IFQLPolicy Type=Flow Policies, Evaluation Protocol=Offline RL2026.05 | 21 | |
| DFPPolicy Type=Ours, Evaluation Protocol=Offline RL2026.05 | 20 | |
| FQLPolicy Type=Flow Policies, Evaluation Protocol=Offline RL2026.05 | 11 | |
| ReBRACPolicy Type=Gaussian Policies, Evaluation Protocol=Offline RL2026.05 | 10 | |
| SRPOPolicy Type=Diffusion Policies, Evaluation Protocol=Offline RL2026.05 | 7 | |
| FBRACPolicy Type=Flow Policies, Evaluation Protocol=Offline RL2026.05 | 5 | |
| IQLPolicy Type=Gaussian Policies, Evaluation Protocol=Offline RL2026.05 | 4 | |
| CACPolicy Type=Flow Policies, Evaluation Protocol=Offline RL2026.05 | 1 | |
| BCPolicy Type=Gaussian Policies, Evaluation Protocol=Offline RL2026.05 | 0 | |
| FAWACPolicy Type=Flow Policies, Evaluation Protocol=Offline RL2026.05 | 0 |