Imitation Learning on Blockpush
88Success RateLC-DP (OURS)
Evaluation Results
| Method | Links | |
|---|---|---|
| LC-DP (OURS)Policy Architecture=Diffusion2026.07 | 88 | |
| LC-DP-REWARDPolicy Architecture=Diffusion2026.07 | 85.2 | |
| DPPolicy Architecture=Diffusion2026.07 | 83.6 | |
| LC-DP-CLASSIFIERPolicy Architecture=Diffusion2026.07 | 82.8 | |
| LPB-OFFLINEPolicy Architecture=Diffusion2026.07 | 78 | |
| LC-BC (OURS)Policy Architecture=Feedforward2026.07 | 47.2 | |
| LC-BC-CLASSIFIERPolicy Architecture=Feedforward2026.07 | 40.4 | |
| LC-BC-REWARDPolicy Architecture=Feedforward2026.07 | 38.8 | |
| BCPolicy Architecture=Feedforward2026.07 | 34.4 | |
| ILIDPolicy Architecture=Feedforward2026.07 | 28.4 | |
| DWBCPolicy Architecture=Feedforward2026.07 | 26 | |
| DEMODICEPolicy Architecture=Feedforward2026.07 | 3.6 |