Complex, non-linear tasks challenge LLM-enhanced multi-agent systems (MAS) due to partial observability and suboptimal coordination. We propose Orchestrator, a novel MAS framework that leverages attention-inspired self-emergent coordination and reflective benchmarking to optimize global task performance. Orchestrator introduces a monitoring mechanism to track agent-environment dynamics, using active inference benchmarks to optimize system behavior. By tracking agent-to-agent and agent-to-environment interaction, Orchestrator mitigates the effects of partial observability and enables agents to approximate global task solutions more efficiently. We evaluate the framework on a series of maze puzzles of increasing complexity, demonstrating its effectiveness in enhancing coordination and performance in dynamic, non-linear environments with long-horizon objectives.
@article{arxiv.2509.05651,
title = {Orchestrator: Active Inference for Multi-Agent Systems in Long-Horizon Tasks},
author = {Lukas Beckenbauer and Johannes-Lucas Loewe and Ge Zheng and Alexandra Brintrup},
journal= {arXiv preprint arXiv:2509.05651},
year = {2025}
}