Orchestrator: Active Inference for Multi-Agent Systems in Long-Horizon Tasks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Beckenbauer, Lukas, Loewe, Johannes-Lucas, Zheng, Ge, Brintrup, Alexandra
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912574412423168
author Beckenbauer, Lukas
Loewe, Johannes-Lucas
Zheng, Ge
Brintrup, Alexandra
author_facet Beckenbauer, Lukas
Loewe, Johannes-Lucas
Zheng, Ge
Brintrup, Alexandra
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2509_05651
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Orchestrator: Active Inference for Multi-Agent Systems in Long-Horizon Tasks
Beckenbauer, Lukas
Loewe, Johannes-Lucas
Zheng, Ge
Brintrup, Alexandra
Multiagent Systems
Artificial Intelligence
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.
title Orchestrator: Active Inference for Multi-Agent Systems in Long-Horizon Tasks
topic Multiagent Systems
Artificial Intelligence
url https://arxiv.org/abs/2509.05651