<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="you-etal-2026-messagesieve">
    <titleInfo>
        <title>MessageSieve: Runtime Message Pruning for Accurate and Efficient LLM-Based Multi-Agent Communication</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Nianhao</namePart>
        <namePart type="family">You</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yubo</namePart>
        <namePart type="family">Dong</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jun</namePart>
        <namePart type="family">Wang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Hehe</namePart>
        <namePart type="family">Fan</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yi</namePart>
        <namePart type="family">Yang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-10</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Hung-Yi</namePart>
            <namePart type="family">Lee</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Sunipa</namePart>
            <namePart type="family">Dev</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jessy</namePart>
            <namePart type="family">Li</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Desmond</namePart>
            <namePart type="family">Elliott</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Budapest, Hungary</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-459-0</identifier>
    </relatedItem>
    <abstract>Large language model (LLM)-based multi-agent systems (MAS) improve reasoning through specialized collaboration, but dense multi-round communication can increase prompt cost and expose agents to outdated, redundant, or noisy context. A prominent line of MAS communication-efficiency work reduces communication through agent-, edge-, or topology-level structural decisions, which operate on graph structure rather than the realized text entering a specific receiver prompt. Such structural decisions are too coarse for runtime communication: messages sent over the same edge may carry useful evidence in one query, redundant content in another, or only a few useful spans. We introduce MessageSieve, a training-free runtime message pruning policy for fixed-graph MAS that operates on message instances rather than agents or edges. Before prompt assembly, it uses lightweight state cards and local message comparisons to preserve original text, retain compact evidence, or omit messages, without a trained selector or additional LLM call. We evaluate MessageSieve on two backbone models and five benchmarks covering expert QA, math reasoning, and code generation. Experiments show that MessageSieve improves the macro-average score over dense Full MAS by 4.10 points on Qwen3-8B and 2.59 points on Gemini 3.1 Flash-Lite, while reducing evaluation tokens by 19–49%.</abstract>
    <identifier type="citekey">you-etal-2026-messagesieve</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.2179/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>49544</start>
            <end>49566</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T MessageSieve: Runtime Message Pruning for Accurate and Efficient LLM-Based Multi-Agent Communication
%A You, Nianhao
%A Dong, Yubo
%A Wang, Jun
%A Fan, Hehe
%A Yang, Yi
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F you-etal-2026-messagesieve
%X Large language model (LLM)-based multi-agent systems (MAS) improve reasoning through specialized collaboration, but dense multi-round communication can increase prompt cost and expose agents to outdated, redundant, or noisy context. A prominent line of MAS communication-efficiency work reduces communication through agent-, edge-, or topology-level structural decisions, which operate on graph structure rather than the realized text entering a specific receiver prompt. Such structural decisions are too coarse for runtime communication: messages sent over the same edge may carry useful evidence in one query, redundant content in another, or only a few useful spans. We introduce MessageSieve, a training-free runtime message pruning policy for fixed-graph MAS that operates on message instances rather than agents or edges. Before prompt assembly, it uses lightweight state cards and local message comparisons to preserve original text, retain compact evidence, or omit messages, without a trained selector or additional LLM call. We evaluate MessageSieve on two backbone models and five benchmarks covering expert QA, math reasoning, and code generation. Experiments show that MessageSieve improves the macro-average score over dense Full MAS by 4.10 points on Qwen3-8B and 2.59 points on Gemini 3.1 Flash-Lite, while reducing evaluation tokens by 19–49%.
%U https://aclanthology.org/2026.emnlp-main.2179/
%P 49544-49566

Markdown (Informal)

[MessageSieve: Runtime Message Pruning for Accurate and Efficient LLM-Based Multi-Agent Communication](https://aclanthology.org/2026.emnlp-main.2179/) (You et al., EMNLP 2026)

• MessageSieve: Runtime Message Pruning for Accurate and Efficient LLM-Based Multi-Agent Communication (You et al., EMNLP 2026)

ACL

• Nianhao You, Yubo Dong, Jun Wang, Hehe Fan, and Yi Yang. 2026. MessageSieve: Runtime Message Pruning for Accurate and Efficient LLM-Based Multi-Agent Communication. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 49544–49566, Budapest, Hungary. Association for Computational Linguistics.