MAPF-World: דגם עולם של פעולה למציאת דרך עבור סוכנים רבים

תקציר מקורי באנגליתarXiv:2508.12087v3 Announce Type: replace Abstract: Multi-agent path finding (MAPF) studies the problem of planning conflict-free paths for multiple agents from given start locations to designated goals, with applications in robot-assisted logistics and social navigation. Recent decentralized learned solvers have shown promise for large-scale MAPF, particularly when leveraging foundation models and large datasets. However, most existing methods rely on reactive policies, often resulting in congestion, deadlocks, and degraded generalization in high agent-density environments. To address these limitations, we propose MAPF-World, an autoregressive action world model for MAPF that unifies short-horizon local future prediction and action generation, enabling decision-making beyond immediate loc