Laser Learning Environment: A new environment for coordination-critical multi-agent tasks

arXiv:2404.03596v1 Announce Type: new
Abstract: We introduce the Laser Learning Environment (LLE), a collaborative multi-agent reinforcement learning environment in which coordination is central. In LLE, agents depend on each other to make progress (interdependence), must jointly take specific sequences of actions to succeed (perfect coordination), and accomplishing those joint actions does not yield any intermediate reward (zero-incentive dynamics). The challenge of such problems lies in the difficulty of escaping state space bottlenecks caused by interdependence steps since escaping those bottlenecks …

abstract agent agents arxiv collaborative cs.ai cs.lg cs.ma environment multi-agent progress reinforcement reinforcement learning tasks type