Optimal-Agent-Selection: State-Aware Routing Framework for Efficient Multi-Agent Collaboration

2025-11-04

AI Summary

To address the limitations of fixed scheduling and inefficient coordination in multi-agent systems when adapting to dynamic task requirements, this paper proposes the State-Aware Routing (SAR) framework. SAR decouples interaction history encoding from agent knowledge modeling, enabling adaptive, step-wise selection of the optimal single agent for each decision. It introduces a novel state-aware router and integrates a self-evolving data generation mechanism to efficiently construct high-quality execution-path training datasets. Evaluated on challenging collaborative reasoning benchmarks, SAR achieves a 23.8% performance improvement over baseline methods while reducing data collection overhead by 90.1% compared to exhaustive search. This work significantly enhances the generalization capability and coordination efficiency of LLM-driven multi-agent systems, establishing a scalable new paradigm for agent collaboration in dynamic task environments.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Multiagent Systems: Adversarial Agents
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

The emergence of multi-agent systems powered by large language models (LLMs) has unlocked new frontiers in complex task-solving, enabling diverse agents to integrate unique expertise, collaborate flexibly, and address challenges unattainable for individual models. However, the full potential of such systems is hindered by rigid agent scheduling and inefficient coordination strategies that fail to adapt to evolving task requirements. In this paper, we propose STRMAC, a state-aware routing framework designed for efficient collaboration in multi-agent systems. Our method separately encodes interaction history and agent knowledge to power the router, which adaptively selects the most suitable single agent at each step for efficient and effective collaboration. Furthermore, we introduce a self-evolving data generation approach that accelerates the collection of high-quality execution paths for efficient system training. Experiments on challenging collaborative reasoning benchmarks demonstrate that our method achieves state-of-the-art performance, achieving up to 23.8% improvement over baselines and reducing data collection overhead by up to 90.1% compared to exhaustive search.

Problem

• Addresses rigid agent scheduling in multi-agent LLM systems
• Improves inefficient coordination strategies for dynamic tasks
• Reduces data collection overhead for system training

Innovation

• State-aware routing framework for multi-agent collaboration
• Separately encodes history and knowledge for adaptive routing
• Self-evolving data generation reduces collection overhead