Offline Reinforcement Learning on scene-play OGBench 5 tasks v0
97Average Success RateMAC
Evaluation results
| Method | Links | ||
|---|---|---|---|
| MACModel Paradigm=Model-Based2025.12 | 97 | — | |
| Value FlowsPolicy Type=Flow Policies2025.10 | 59 | — | |
| Value FlowsSeeds=82026.05 | 59 | — | |
| FloQSeeds=82026.05 | 58 | — | |
| FQLModel Paradigm=Model-Free2025.12 | 56 | — | |
| FQLPolicy Type=Flow Policies2025.10 | 56 | — | |
| FQLSeeds=82026.05 | 56 | — | |
| CODACPolicy Type=Flow Policies2025.10 | 55 | — | |
| CODACSeeds=82026.05 | 55 | — | |
| PCBFSeeds=82026.05 | 54 | — | |
| IDQLModel Paradigm=Model-Free2025.12 | 46 | — | |
| FBRACPolicy Type=Flow Policies2025.10 | 45 | — | |
| Value FlowsPolicy Type=Flow Policies2025.10 | 43 | — | |
| ReBRACModel Paradigm=Model-Free2025.12 | 41 | — | |
| ReBRACPolicy Type=Gaussian Policies2025.10 | 41 | — | |
| FQLPolicy Type=Flow Policies2025.10 | 41 | — | |
| IQNPolicy Type=Flow Policies2025.10 | 41 | — | |
| IQNPolicy Type=Flow Policies2025.10 | 40 | — | |
| IQNSeeds=82026.05 | 40 | — | |
| IFQLPolicy Type=Flow Policies2025.10 | 30 | — | |
| IQLModel Paradigm=Model-Free2025.12 | 28 | — | |
| IQLPolicy Type=Gaussian Policies2025.10 | 28 | — | |
| ReBRACPolicy Type=Gaussian Policies2025.10 | 28 | — | |
| IQLSeeds=82026.05 | 28 | — | |
| IQLPolicy Type=Gaussian Policies2025.10 | 26 | — | |
| IFQLPolicy Type=Flow Policies2025.10 | 21 | — | |
| FBRACPolicy Type=Flow Policies2025.10 | 11 | — | |
| MOBILEModel Paradigm=Model-Based2025.12 | 8 | — | |
| MOPOModel Paradigm=Model-Based2025.12 | 6 | — | |
| BCPolicy Type=Gaussian Policies2025.10 | 5 | — | |
| FMPCModel Paradigm=Model-Based2025.12 | 4 | — | |
| C51Policy Type=Flow Policies2025.10 | 4 | — | |
| LEQModel Paradigm=Model-Based2025.12 | 0 | — |