# Where Reasoning Diverges: Localized Multi-Agent Debate for Multi-Hop Question Answering

Source: [Where Reasoning Diverges: Localized Multi-Agent Debate for Multi-Hop Question Answering](<https://www.sota2.com/research/paper/6a71728dbc6f0f0705846061>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-08-06T05:00:10.068Z

Authors: Weijun Gao, Xiang Ding, Haoyang Liu, Tiancheng Xing

Published: 2026-08-02T19:47:03.000Z

## Abstract

Multi-agent debate commonly exchanges complete rationales even when disagreements concern only a few intermediate claims. We introduce Localized Multi-Agent Debate (LMAD), an inference-time protocol that represents agent rationales as nodes, locates their earliest conflict, and restricts debate to the corresponding local segments. Guarded resolution extends a shared committed state so that later conflicts can be addressed without reopening accepted steps. We evaluate LMAD on four multi-hop question-answering benchmarks using ten backbones from four model families. Our method achieves the highest macro-averaged judge accuracy across all ten backbones, outperforming the strongest conventional baseline by up to 7.20 percentage points.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2608.01463>)

[PDF](<https://arxiv.org/pdf/2608.01463.pdf>)