Navigation on D4RL AntMaze medium-play v2
81Normalized D4RL ScorePEX
Evaluation Results
| Method | Links | |
|---|---|---|
| PEXLearning Stage=Online fine-tuning2024.12 | 81 | |
| O2PPOLearning Stage=Online fine-tuning, Environment steps=250,0002024.12 | 80.3 | |
| O2SACLearning Stage=Online fine-tuning, Environment steps=200,000, Double Q networks=false, Threshold tau max step=100,0002024.12 | 78.8 | |
| IQLLearning Stage=Online fine-tuning2024.12 | 78 | |
| PEXLearning Stage=Offline2024.12 | 73.6 | |
| O2SACLearning Stage=Offline2024.12 | 67 | |
| O2PPOLearning Stage=Offline2024.12 | 66.5 | |
| IQLLearning Stage=Offline2024.12 | 64.8 | |
| AWACLearning Stage=Offline2024.12 | 0 | |
| AWACLearning Stage=Online fine-tuning2024.12 | 0 |