Long-Horizon Control
20.7ThroughputSAC
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SACAgent=SAC, Training=online fine-tuning, Reference=FIFO2026.06 | 20.7 | 17.5 | 13.8 | |
| DQLAgent=DQL, Training=online fine-tuning, Reference=FIFO2026.06 | 19.7 | 16.6 | 13.2 | |
| PPOAgent=PPO, Training=online fine-tuning, Reference=FIFO2026.06 | 16.3 | 15.3 | 12.1 | |
| SPTAgent=SPT, Reference=FIFO2026.06 | 13.9 | 13.3 | 10.7 | |
| RandomAgent=Random, Reference=FIFO2026.06 | -21.2 | -12.3 | -10.8 |