Goal-conditioned Reinforcement Learning on PandaReachDense Modified Joint space, +N-G v3
-5.03Avg RewardP2S
Evaluation results
| Method | Links | |
|---|---|---|
| P2SAction Space=Joint, Initial Joint Randomization=true, Goal Randomization=false2025.04 | -5.03 | |
| O2SAction Space=Joint, Initial Joint Randomization=true, Goal Randomization=false2025.04 | -5.48 | |
| Q2SAction Space=Joint, Initial Joint Randomization=true, Goal Randomization=false2025.04 | -6.19 | |
| RSAction Space=Joint, Initial Joint Randomization=true, Goal Randomization=false2025.04 | -7.3 |