Saved in:
Bibliographic Details
Main Authors: Wang, Xudong, Zhang, Chaoning, Zhang, Jiaquan, Li, Chenghao, Sun, Qigan, Bae, Sung-Ho, Wang, Peng, Xie, Ning, Zou, Jie, Yang, Yang, Shen, Hengtao
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.12933
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914390997991424
author Wang, Xudong
Zhang, Chaoning
Zhang, Jiaquan
Li, Chenghao
Sun, Qigan
Bae, Sung-Ho
Wang, Peng
Xie, Ning
Zou, Jie
Yang, Yang
Shen, Hengtao
author_facet Wang, Xudong
Zhang, Chaoning
Zhang, Jiaquan
Li, Chenghao
Sun, Qigan
Bae, Sung-Ho
Wang, Peng
Xie, Ning
Zou, Jie
Yang, Yang
Shen, Hengtao
contents Large Language Model (LLM)-driven Multi-Agent Systems (MAS) have demonstrated strong capability in complex reasoning and tool use, and heterogeneous agent pools further broaden the quality--cost trade-off space. Despite these advances, real-world deployment is often constrained by high inference cost, latency, and limited transparency, which hinders scalable and efficient routing. Existing routing strategies typically rely on expensive LLM-based selectors or static policies, and offer limited controllability for semantic-aware routing under dynamic loads and mixed intents, often resulting in unstable performance and inefficient resource utilization. To address these limitations, we propose AMRO-S, an efficient and interpretable routing framework for Multi-Agent Systems (MAS). AMRO-S models MAS routing as a semantic-conditioned path selection problem, enhancing routing performance through three key mechanisms: First, it leverages a supervised fine-tuned (SFT) small language model for intent inference, providing a low-overhead semantic interface for each query; second, it decomposes routing memory into task-specific pheromone specialists, reducing cross-task interference and optimizing path selection under mixed workloads; finally, it employs a quality-gated asynchronous update mechanism to decouple inference from learning, optimizing routing without increasing latency. Extensive experiments on five public benchmarks and high-concurrency stress tests demonstrate that AMRO-S consistently improves the quality--cost trade-off over strong routing baselines, while providing traceable routing evidence through structured pheromone patterns.
format Preprint
id arxiv_https___arxiv_org_abs_2603_12933
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Efficient and Interpretable Multi-Agent LLM Routing via Ant Colony Optimization
Wang, Xudong
Zhang, Chaoning
Zhang, Jiaquan
Li, Chenghao
Sun, Qigan
Bae, Sung-Ho
Wang, Peng
Xie, Ning
Zou, Jie
Yang, Yang
Shen, Hengtao
Artificial Intelligence
Large Language Model (LLM)-driven Multi-Agent Systems (MAS) have demonstrated strong capability in complex reasoning and tool use, and heterogeneous agent pools further broaden the quality--cost trade-off space. Despite these advances, real-world deployment is often constrained by high inference cost, latency, and limited transparency, which hinders scalable and efficient routing. Existing routing strategies typically rely on expensive LLM-based selectors or static policies, and offer limited controllability for semantic-aware routing under dynamic loads and mixed intents, often resulting in unstable performance and inefficient resource utilization. To address these limitations, we propose AMRO-S, an efficient and interpretable routing framework for Multi-Agent Systems (MAS). AMRO-S models MAS routing as a semantic-conditioned path selection problem, enhancing routing performance through three key mechanisms: First, it leverages a supervised fine-tuned (SFT) small language model for intent inference, providing a low-overhead semantic interface for each query; second, it decomposes routing memory into task-specific pheromone specialists, reducing cross-task interference and optimizing path selection under mixed workloads; finally, it employs a quality-gated asynchronous update mechanism to decouple inference from learning, optimizing routing without increasing latency. Extensive experiments on five public benchmarks and high-concurrency stress tests demonstrate that AMRO-S consistently improves the quality--cost trade-off over strong routing baselines, while providing traceable routing evidence through structured pheromone patterns.
title Efficient and Interpretable Multi-Agent LLM Routing via Ant Colony Optimization
topic Artificial Intelligence
url https://arxiv.org/abs/2603.12933