Self-Adapting Group of Experts for Multi-Agent Reasoning

原文：Self-Adapting Group of Experts for Multi-Agent Reasoning

作者：Mohammad Atif Quamar, Nurbek Tastan, Karthik Nandakumar, Junpei Komiyama

来源：arXiv cs.CL（自然语言）

正文

Computer Science > Artificial Intelligence

arXiv:2609.35412v1 (cs)

[Submitted on 28 Sep 2026]

Title:Self-Adapting Group of Experts for Multi-Agent Reasoning

Authors:Mohammad Atif Quamar, Nurbek Tastan, Karthik Nandakumar, Junpei Komiyama

View a PDF of the paper titled Self-Adapting Group of Experts for Multi-Agent Reasoning, by Mohammad Atif Quamar and 3 other authors

View PDF

HTML (experimental)

Abstract:Multi-agent systems bring together language model agents with different roles to propose, review, and refine solutions. Each agent’s response depends on its model’s capabilities, the reasoning strategy defined by its system prompt, and the information in its input context. Existing frameworks often adapt communication by changing this context while leaving individual prompts fixed, even when a problem calls for different skills. We study whether agents’ initial responses can identify a strategy better suited to the current problem and guide its transfer to other agents. To address this, we introduce SAGE (Self-Adapting Group of Experts), a training-free framework that uses answer agreement, prefix consistency, and reciprocal peer review to select a strategy donor. SAGE transfers the selected donor’s reasoning strategy to the other agents while preserving their original roles. This transfer uses only the agents’ original system prompts, without access to the problem or generated solutions. After strategy adaptation, agents exchange responses through a dynamic, sparse directed acyclic graph that routes information from higher-scoring agents to lower-scoring agents. Experiments across multiple agent backbones and reasoning benchmarks show that SAGE achieves higher average accuracy than the evaluated baselines. Our code is available at this https URL.

Subjects:

Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Multiagent Systems (cs.MA)

Cite as:

arXiv:2609.35412 [cs.AI]

(or

arXiv:2609.35412v1 [cs.AI] for this version)

https://doi.org/10.48550/arXiv.2609.35412

Focus to learn more

arXiv-issued DOI via DataCite

主题

人工智能 · 自然语言处理 · 多智能体

由「前沿雷达」于 2026-09-28 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。