Offline Multi-Agent Reinforcement Learning on Marine-Easy Medium-Expert
98.8Score (3m)STAIRS-Former
Evaluation Results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| STAIRS-Former2026.03 | 98.8 | 98.8 | 100 | 60 | 94.4 | 96.9 | 84.4 | 100 | 98.1 | 95.6 | 92.7 | |
| HiSSD2026.03 | 81.3 | 80 | 74.4 | 75.6 | 75.6 | 73.1 | 71.9 | 73.1 | 57.5 | 69.4 | 73.2 | |
| ODIS2026.03 | 76.3 | 84.4 | 51.9 | 64.4 | 67.5 | 62.5 | 45.6 | 62.5 | 49.4 | 40.6 | 60.5 | |
| UPDeT-m2026.03 | 48.1 | 66.3 | 60.6 | 24.4 | 46.9 | 37.5 | 10.6 | 48.1 | 55.6 | 36.9 | 43.5 |