Offline Multi-Agent Reinforcement Learning on SMAC Marine-Hard Medium-Replay benchmark leaderboard · SOTA2 Research

Offline Multi-Agent Reinforcement Learning on SMAC Marine-Hard Medium-Replay

81.9Win Rate (3m)ODIS

Evaluation Results

Method	Links
ODIS2026.03		81.9	11.9	15	59.4	55.6	51.9	51.3	8.1	3.1	12.5	0	1.3	29.3
STAIRS2026.03		78.1	50.6	78.1	93.8	100	97.5	94.4	23.1	26.9	66.9	3.1	4.4	59.7
HiSSD2026.03		75.6	24.4	45	71.9	71.3	93.1	91.9	12.5	11.3	33.8	0	6.3	44.8
UPDeT-m2026.03		40	0	0.6	46.3	64.4	29.4	25.6	0.6	0.6	0.6	0	0	17.3