Local Work, Global Progress: A Convergence Theory of AI Agent Collaboration

原文：Local Work, Global Progress: A Convergence Theory of AI Agent Collaboration

作者：Ivan Ilin, Peter Richtárik

来源：arXiv cs.LG（机器学习）

正文

Computer Science > Multiagent Systems

arXiv:2610.28286v1 (cs)

[Submitted on 20 Oct 2026]

Title:Local Work, Global Progress: A Convergence Theory of AI Agent Collaboration

Authors:Ivan Ilin, Peter Richtárik

View a PDF of the paper titled Local Work, Global Progress: A Convergence Theory of AI Agent Collaboration, by Ivan Ilin and 1 other authors

View PDF

HTML (experimental)

Abstract:Teams of large language model (LLM) agents are increasingly applied to complex tasks, but their collaboration protocols are mostly designed empirically, without convergence guarantees. We formalize collaborative problem solving as block optimization: each agent improves one part of a shared solution, guided by its local objective plus a correction for how its changes affect the other parts. A coordinator checks every proposal against the global objective and merges proposals that do not interfere with each other. We prove linear convergence up to an error floor that vanishes when the corrections are exact. Interfering proposals can also be merged once their combination has been checked; the rate then depends on how well the proposals actually combine, not on worst-case interactions. With GPT-OSS-120B agents on robot routing, a three-agent team finds collision-free plans in every run, better than those of a single agent that controls all robots. On a larger routing task, a six-agent team completes its search about twice as fast as the single agent (in median). Merging the best checked combination of proposals, rather than only proposals that do not interfere, cuts the team’s search time by about 30%. With constants measured on the runs, our convergence bound holds in every round and follows the observed progress. We also test the theory on query planning in SQLite and on a synthetic task.

Comments:

59 pages, 15 figures, 28 tables. Code: this https URL

Subjects:

Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Optimization and Control (math.OC)

MSC classes:

68T42, 90C27

ACM classes:

I.2.11; G.1.6

Cite as:

arXiv:2610.28286 [cs.MA]

(or

arXiv:2610.28286v1 [cs.MA] for this version)

https://doi.org/10.48550/arXiv.2610.28286

Focus to learn more

arXiv-issued DOI via DataCite

主题

多智能体 · 人工智能 · 机器学习 · 优化与控制

由「前沿雷达」于 2026-10-20 采集。正文取自原文页面，已保留出处链接。标题与正文版权归原作者所有。