Offline Reinforcement Learning on OGBench antmaze-large-singletask 5 tasks
90Normalized ScoreDROL
Evaluation results
| Method | Links | |
|---|---|---|
| DROLPolicy Class=One-step Method, K=162026.04 | 90 | |
| SORLPolicy Class=Multi-step / Iterative Policy2026.04 | 89 | |
| DROLPolicy Class=One-step Method, K-selection protocol=benchmark-specific2026.04 | 89 | |
| ReBRACPolicy Class=Gaussian Policy2026.04 | 81 | |
| DeFlowPolicy Class=Multi-step / Iterative Policy2026.04 | 81 | |
| FQLPolicy Class=One-step Method2026.04 | 79 | |
| FBRACPolicy Class=Multi-step / Iterative Policy2026.04 | 60 | |
| IQLPolicy Class=Gaussian Policy2026.04 | 53 | |
| IFQLPolicy Class=Multi-step / Iterative Policy2026.04 | 28 | |
| BCPolicy Class=Gaussian Policy2026.04 | 11 | |
| FAWACPolicy Class=Multi-step / Iterative Policy2026.04 | 6 |