---
id: 20260624-T0-09
title: "PEAR：解决多智能体辩论中的位置偏差问题"
title_en: "PEAR: Solving Position Bias in Multi-Agent Debates"
url: https://ai.daily.yangsir.net/daily/20260624-T0-09
issue_date: 2026-06-24
publish_date: 2026-06-23T04:00:00.000Z
category: research
source_name: "arXiv cs.AI"
source_url: https://arxiv.org/abs/2606.20621
---

# PEAR：解决多智能体辩论中的位置偏差问题

arXiv 论文提出了一种名为 PEAR 的新框架，旨在解决多智能体辩论中的固定拓扑偏差问题。通过引入置换等变自适应路由机制，PEAR 防止了不可靠智能体的观点被放大，并减少了对角色分配的敏感性。测试显示，该方法显著提升了 LLM 在辩论中的推理可靠性。

## English Version

**PEAR: Solving Position Bias in Multi-Agent Debates**

A new arXiv paper introduces PEAR, a framework designed to resolve fixed topology biases in multi-agent debates. By utilizing permutation-equivariant adaptive routing, PEAR prevents the amplification of unreliable agents and reduces sensitivity to role assignment, significantly boosting LLM reasoning reliability.

---

**来源**：[arXiv cs.AI](https://arxiv.org/abs/2606.20621)

**详情页**：https://ai.daily.yangsir.net/daily/20260624-T0-09

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*