Multi-agent reinforcement learning on Multi-soft-robot coordination benchmark leaderboard · SOTA2 Research

Multi-agent reinforcement learning on Multi-soft-robot coordination

108,000Convergence Episodes (x10^5)TD-MAPPO

Evaluation Results

Method	Links
TD-MAPPO2026.05		108,000	2.28	92
TD-MASAC2026.05		157,000	1.95	89
HAPPO2026.05		162,000	1.24	76
MAAC2026.05		176,000	1.52	72
FACMAC2026.05		181,000	1.21	76
TD-MADDPG2026.05		200,000	1.73	86
MASAC2026.05		217,000	1.28	74
MADDPG2026.05		262,000	0.95	67
MAPPO2026.05		270,000	1.08	71
QDDPG2026.05		318,000	0.84	63