Reinforcement Learning on Walker-Gravity Start OpenAI Gym Mujoco (initialization)
281.99Average RewardLPG-FTW
Evaluation results
| Method | Links | |
|---|---|---|
| LPG-FTWsamples per iteration=2x2022.10 | 281.99 | |
| VBLRLsamples per iteration=1x2022.10 | 248.97 | |
| T-HiP-MDPsamples per iteration=1x2022.10 | 213.93 | |
| EWCsamples per iteration=2x2022.10 | 163.61 | |
| Single-task MBRLmode=train from scratch2022.10 | 4.32 |