Continuous Control on Multimodal continuous control environment (final 80 episodes)
1,295.7Mean RewardQuantFPFlow
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| QuantFPFlowtraining_episodes=400, averaging_window=final 80 episodes2026.05 | 1,295.7 | 2,235.5 | 33.9 | 5.877 | |
| SACtraining_episodes=400, averaging_window=final 80 episodes2026.05 | 1,284 | 2,378.5 | 30.7 | 6.576 | |
| Randomtraining_episodes=400, averaging_window=final 80 episodes2026.05 | 670.4 | 355.2 | 0.1 | 3.319 | |
| DDPGtraining_episodes=400, averaging_window=final 80 episodes2026.05 | 628.4 | 2,274.5 | 22.3 | 5.969 |