Reinforcement Learning on DeepMind Control Finger Spin (Time to 75% Return)
89,100Steps to 75% ReturnDreamer V3
Evaluation results
| Method | Links | |
|---|---|---|
| Dreamer V32026.05 | 89,100 | |
| Mind DreamerHorizon=152026.05 | 92,600 | |
| Mind DreamerHorizon=52026.05 | 114,500 | |
| Mind DreamerHorizon=102026.05 | 165,000 | |
| Dreamer V22026.05 | 361,700 | |
| Plan2Explore2026.05 | 1,163,300 |