Physical State Prediction on DeepMind Control Suite Reacher Easy tasks (random policy)
0.0005MSEMoSim
Evaluation results
| Method | Links | |
|---|---|---|
| MoSimHorizon=16, Training data source=random policy data2025.04 | 0.0005 | |
| MoSimHorizon=100, Training data source=random policy data2025.04 | 0.0009 | |
| MoSimHorizon=16, Training data source=experience data2025.04 | 0.0037 | |
| MoSimHorizon=100, Training data source=experience data2025.04 | 0.0053 | |
| DreamerV3Horizon=16, Prediction steps=5-step, Training data source=random policy data2025.04 | 0.0972 | |
| DreamerV3Horizon=100, Prediction steps=5-step, Training data source=random policy data2025.04 | 0.0988 | |
| DreamerV3Horizon=16, Prediction steps=5-step, Training data source=experience data2025.04 | 0.1081 | |
| DreamerV3Horizon=100, Prediction steps=5-step, Training data source=experience data2025.04 | 0.1743 | |
| DreamerV3Horizon=16, Prediction steps=1-step, Training data source=experience data2025.04 | 0.737 | |
| DreamerV3Horizon=100, Prediction steps=1-step, Training data source=experience data2025.04 | 0.7405 | |
| DreamerV3Horizon=16, Prediction steps=1-step, Training data source=random policy data2025.04 | 0.7451 | |
| DreamerV3Horizon=100, Prediction steps=1-step, Training data source=random policy data2025.04 | 0.7532 |