Reinforcement Learning on Door
113Task 1 ScoreHyperCRL
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| HyperCRLTarget Network=MLP (4x200), Hypernetwork=MLP (2x256)2020.09 | 113 | 100 | 99 | 86 | 100 | |
| Multi-taskTarget Network=MLP (4x200), Hypernetwork=MLP (2x256)2020.09 | 37 | 81 | 59 | 93 | 68 | |
| CoresetTarget Network=MLP (4x200), Hypernetwork=MLP (2x256)2020.09 | -5 | 26 | 21 | 45 | 22 | |
| EWCTarget Network=MLP (4x200), Hypernetwork=MLP (2x256)2020.09 | -6 | 17 | 4 | 16 | 7 | |
| FinetuningTarget Network=MLP (4x200), Hypernetwork=MLP (2x256)2020.09 | -14 | 11 | 3 | 16 | 4 | |
| SITarget Network=MLP (4x200), Hypernetwork=MLP (2x256)2020.09 | -17 | 11 | 6 | 16 | 4 |