Imitation Learning on MuJoCo
109.46Hopper RewardDACfO
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| DACfOAccess to expert actions=No (LfO), Evaluation trajectories=502026.03 | 109.46 | 61.52 | 45.28 | 113.4 | 82.41 | |
| LWAILAccess to expert actions=No (LfO), Evaluation trajectories=502026.03 | 108.84 | 90.4 | 106.51 | 90.53 | 99.07 | |
| WDAILAccess to expert actions=Yes, Evaluation trajectories=502026.03 | 107.72 | 38.3 | 126.07 | 87.59 | 89.92 | |
| OPOLOAccess to expert actions=No (LfO), Evaluation trajectories=502026.03 | 99.24 | 58.98 | 37.07 | 129.46 | 81.19 | |
| DIFOAccess to expert actions=No (LfO), Evaluation trajectories=502026.03 | 98.5 | 78.62 | 99.43 | 93.47 | 92.51 | |
| DiffAILAccess to expert actions=Yes, Evaluation trajectories=502026.03 | 95.52 | 33.36 | 133.9 | 80.56 | 85.84 | |
| IQlearnAccess to expert actions=No (LfO), Evaluation trajectories=502026.03 | 86.24 | 2.51 | 7.07 | 7.39 | 25.8 | |
| PWILAccess to expert actions=No (LfO), Evaluation trajectories=502026.03 | 78.93 | 20.81 | 84.01 | 105.62 | 72.34 | |
| LS-IQAccess to expert actions=No (LfO), Evaluation trajectories=502026.03 | 76.84 | 41.64 | 102.02 | 132.28 | 88.2 | |
| OLLIEAccess to expert actions=Yes, Evaluation trajectories=502026.03 | 71.1 | 35.5 | 59.8 | 57.1 | 55.87 | |
| GAIfOAccess to expert actions=No (LfO), Evaluation trajectories=502026.03 | 58.74 | 29.79 | 52.73 | 12.99 | 38.56 | |
| BCOAccess to expert actions=No (LfO), Evaluation trajectories=502026.03 | 21.31 | 4.08 | 0.88 | 25.33 | 12.9 | |
| GAILAccess to expert actions=Yes, Evaluation trajectories=502026.03 | 7.78 | -0.33 | 2.14 | -1.98 | 1.9 | |
| BCAccess to expert actions=Yes, Evaluation trajectories=502026.03 | 1.51 | 1.92 | 2.06 | 12.74 | 4.56 | |
| AIRLAccess to expert actions=Yes, Evaluation trajectories=502026.03 | 1.16 | 6.02 | 0.83 | 3.3 | 2.83 |