arXiv cs.MA· Mingxiu Cai, Haoyang Xian, Xinyu Ma, Yutong Wang, Zhenxi Song, Jing Li, Xuebo Liu·· 12 小时前AI 评分31
UniMASO：面向异构 LLM 多智能体系统的提示词与模型联合优化框架
Not Every Agent Needs the Strongest Model: Unified Prompt-Model Optimization for Heterogeneous LLM-based Multi-Agent Systems
AI 导读
UniMASO 通过联合优化提示词与模型分配，让异构 LLM 多智能体系统在更低部署成本下保持任务表现。在五个基准和五种执行拓扑上，集中式执行时平均分从 69.79 提升至 70.40，同时相对 DeepSeek-V4-Pro 搭配 MASPO 将 token 加权部署成本降低 85.2%。该框架结合成本感知的 Pareto 搜索、拓扑感知协同进化和跨层轨迹记忆。
正文 · 原文
Abstract:Large language model (LLM)-based multi-agent systems (MAS) rely on role-specific prompts and execution models to solve complex tasks collaboratively. However, prompt optimization typically keeps model assignments fixed, leaving differences in role requirements and their associated inference costs outside the search. We present UniMASO, a Unified framework for heterogeneous Multi-Agent System Optimization through the joint adaptation of prompts and model assignments. UniMASO evaluates candidate updates through local, downstream, and global feedback, using cost-aware Pareto search to retain useful quality-cost trade-offs. Topology-aware co-evolution then coordinates these updates according to agent roles and dependencies. Cross-level trajectory memory further supplies historical successes and failures to guide subsequent optimization. Experiments on five benchmarks and five execution topologies show that UniMASO achieves competitive task performance at substantially lower deployment cost than the strongest homogeneous configurations. Under centralized execution, it improves the average score from 69.79 to 70.40 while reducing token-weighted deployment cost by 85.2% relative to DeepSeek-V4-Pro with MASPO. These results support joint prompt-model adaptation as an effective approach to efficient multi-agent collaboration. The code is available at this https URL.
| Subjects: | Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2610.31743 [cs.MA] |
| (or arXiv:2610.31743v1 [cs.MA] for this version) | |
| https://doi.org/10.48550/arXiv.2610.31743 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Mingxiu Cai [view email]
[v1]
Thu, 22 Oct 2026 17:24:23 UTC (5,864 KB)
来源：arXiv cs.MA · arxiv.org