Reinforcement Learning on Cheetah-Bodyparts Start OpenAI Gym Mujoco (initialization)
110.74Average RewardVBLRL
Evaluation results
| Method | Links | |
|---|---|---|
| VBLRLsamples per iteration=1x2022.10 | 110.74 | |
| T-HiP-MDPsamples per iteration=1x2022.10 | 78.95 | |
| Single-task MBRLmode=train from scratch2022.10 | -101.02 | |
| LPG-FTWsamples per iteration=2x2022.10 | -263.84 | |
| EWCsamples per iteration=2x2022.10 | -5,016.93 |