<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ku-etal-2025-multi">
    <titleInfo>
        <title>Multi-Agent LLM Debate Unveils the Premise Left Unsaid</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Harvey</namePart>
        <namePart type="given">Bonmu</namePart>
        <namePart type="family">Ku</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jeongyeol</namePart>
        <namePart type="family">Shin</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Hyoun</namePart>
        <namePart type="given">Jun</namePart>
        <namePart type="family">Lee</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Seonok</namePart>
        <namePart type="family">Na</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Insu</namePart>
        <namePart type="family">Jeon</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 12th Argument mining Workshop</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Elena</namePart>
            <namePart type="family">Chistova</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Philipp</namePart>
            <namePart type="family">Cimiano</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Shohreh</namePart>
            <namePart type="family">Haddadan</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Gabriella</namePart>
            <namePart type="family">Lapesa</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ramon</namePart>
            <namePart type="family">Ruiz-Dolz</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Vienna, Austria</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-258-9</identifier>
    </relatedItem>
    <abstract>Implicit premise is central to argumentative coherence and faithfulness, yet remain elusive in traditional single-pass computational models. We introduce a multi-agent framework that casts implicit premise recovery as a dialogic reasoning task between two LLM agents. Through structured rounds of debate, agents critically evaluate competing premises and converge on the most contextually appropriate interpretation. Evaluated on a controlled binary classification benchmark for premise selection, our approach achieves state-of-the-art accuracy, outperforming both neural baselines and single-agent LLMs. We find that accuracy gains stem not from repeated generation, but from agents refining their predictions in response to opposing views. Moreover, we show that forcing models to defend assigned stances degrades performance—engendering rhetorical rigidity to flawed reasoning. These results underscore the value of interactive debate in revealing pragmatic components of argument structure.</abstract>
    <identifier type="citekey">ku-etal-2025-multi</identifier>
    <identifier type="doi">10.18653/v1/2025.argmining-1.6</identifier>
    <location>
        <url>https://aclanthology.org/2025.argmining-1.6/</url>
    </location>
    <part>
        <date>2025-07</date>
        <extent unit="page">
            <start>58</start>
            <end>73</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Multi-Agent LLM Debate Unveils the Premise Left Unsaid
%A Ku, Harvey Bonmu
%A Shin, Jeongyeol
%A Lee, Hyoun Jun
%A Na, Seonok
%A Jeon, Insu
%Y Chistova, Elena
%Y Cimiano, Philipp
%Y Haddadan, Shohreh
%Y Lapesa, Gabriella
%Y Ruiz-Dolz, Ramon
%S Proceedings of the 12th Argument mining Workshop
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-258-9
%F ku-etal-2025-multi
%X Implicit premise is central to argumentative coherence and faithfulness, yet remain elusive in traditional single-pass computational models. We introduce a multi-agent framework that casts implicit premise recovery as a dialogic reasoning task between two LLM agents. Through structured rounds of debate, agents critically evaluate competing premises and converge on the most contextually appropriate interpretation. Evaluated on a controlled binary classification benchmark for premise selection, our approach achieves state-of-the-art accuracy, outperforming both neural baselines and single-agent LLMs. We find that accuracy gains stem not from repeated generation, but from agents refining their predictions in response to opposing views. Moreover, we show that forcing models to defend assigned stances degrades performance—engendering rhetorical rigidity to flawed reasoning. These results underscore the value of interactive debate in revealing pragmatic components of argument structure.
%R 10.18653/v1/2025.argmining-1.6
%U https://aclanthology.org/2025.argmining-1.6/
%U https://doi.org/10.18653/v1/2025.argmining-1.6
%P 58-73

Markdown (Informal)

[Multi-Agent LLM Debate Unveils the Premise Left Unsaid](https://aclanthology.org/2025.argmining-1.6/) (Ku et al., ArgMining 2025)

• Multi-Agent LLM Debate Unveils the Premise Left Unsaid (Ku et al., ArgMining 2025)

ACL

• Harvey Bonmu Ku, Jeongyeol Shin, Hyoun Jun Lee, Seonok Na, and Insu Jeon. 2025. Multi-Agent LLM Debate Unveils the Premise Left Unsaid. In Proceedings of the 12th Argument mining Workshop, pages 58–73, Vienna, Austria. Association for Computational Linguistics.