Offline Goal-Conditioned Reinforcement Learning on cube-quadruple 100M
Evaluation results
| Method | Links | |
|---|
| DQCMethod=Decoupled Q-chunking2025.12 | | 92 |
|---|
| SHARSA2025.12 | | 64 |
|---|
| IQL2025.12 | | 53 |
|---|
| DQC-naïveExecution=partial action chunk2025.12 | | 40 |
|---|
| HFBC2025.12 | | 37 |
|---|
| QCStrategy=Q-chunking2025.12 | | 35 |
|---|
| FBC2025.12 | | 34 |
|---|
| NSBackup=n-step return backup2025.12 | | 27 |
|---|
| HIQL2025.12 | | 24 |
|---|
| OSBackup=1-step TD-backup2025.12 | | 0 |
|---|