Offline Reinforcement Learning on Android Door Human v2
9.9Normalized ScoreCQL
Evaluation results
| Method | Links | |
|---|---|---|
| CQLevaluations=final 10, random seeds=52024.10 | 9.9 | |
| OTR+IQLevaluations=final 10, random seeds=52024.10 | 5.9 | |
| IQLevaluations=final 10, random seeds=52024.10 | 3.1 | |
| IQLevaluations=final 10, random seeds=52024.10 | 1.053 | |
| PPLevaluations=final 10, random seeds=52024.10 | 1.0466 | |
| ReBRACevaluations=final 10, random seeds=52024.10 | 1.046 | |
| CQLevaluations=final 10, random seeds=52024.10 | 1.015 | |
| PPLevaluations=final 10, random seeds=52024.10 | 1 | |
| TD3+BCevaluations=final 10, random seeds=52024.10 | 0.864 | |
| BCevaluations=final 10, random seeds=52024.10 | 0.5 | |
| BCevaluations=final 10, random seeds=52024.10 | 0.349 | |
| ReBRACevaluations=final 10, random seeds=52024.10 | 0 | |
| TD3+BCevaluations=final 10, random seeds=52024.10 | -0.1 |