<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="guo-etal-2026-graph">
    <titleInfo>
        <title>Graph Evidence Is Not Enough: Diagnosing Native Decoder Use in Graph-Augmented LLMs</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Xiaoyu</namePart>
        <namePart type="family">Guo</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Pengcheng</namePart>
        <namePart type="family">Chen</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Y</namePart>
        <namePart type="given">U</namePart>
        <namePart type="family">Jiong</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yi</namePart>
        <namePart type="family">Lu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yaohua</namePart>
        <namePart type="family">Wang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Ziyang</namePart>
        <namePart type="family">Li</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>Graph-augmented large language models often assume that graph evidence produced by external computation and placed in the input can be used by the native decoder. We test this assumption with HopQA, a deliberately bounded diagnostic that asks for the shortest-hop distance between two query nodes. Because the answer is a small integer and the target is purely topological, failure cannot be dismissed as open-ended generation or ambiguous evaluation. Yet existing graph-augmented baselines still fail on this setting, showing that providing graph evidence is not the same as making it usable. We introduce an intervention triangle with three matched conditions: readable graph evidence, shuffled graph evidence, and no-graph input. This separates evidence inclusion, structural readability, and decoder-usable topology. Guided by this diagnosis, we present S²GE as an instance showing that diagnosis-driven interface design can improve native decoder usability. S²GE uses query-aware sampling, endpoint- and proximity-based ordering, and structure-preserving alignment. Across DBLP, Biomedical, GoodReads, and PubMed, S²GE achieves strict exact-match scores of 36.5%, 57.8%, 76.6%, and 52.0%, improving over the strongest native-generation baseline by 53.5 points on average. The interventions further reveal harmful-shuffle, shuffle-robust, and no-graph-saturated regimes.</abstract>
    <identifier type="citekey">guo-etal-2026-graph</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.647/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>14900</start>
            <end>14917</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Graph Evidence Is Not Enough: Diagnosing Native Decoder Use in Graph-Augmented LLMs
%A Guo, Xiaoyu
%A Chen, Pengcheng
%A Jiong, Y. U.
%A Lu, Yi
%A Wang, Yaohua
%A Li, Ziyang
%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 guo-etal-2026-graph
%X Graph-augmented large language models often assume that graph evidence produced by external computation and placed in the input can be used by the native decoder. We test this assumption with HopQA, a deliberately bounded diagnostic that asks for the shortest-hop distance between two query nodes. Because the answer is a small integer and the target is purely topological, failure cannot be dismissed as open-ended generation or ambiguous evaluation. Yet existing graph-augmented baselines still fail on this setting, showing that providing graph evidence is not the same as making it usable. We introduce an intervention triangle with three matched conditions: readable graph evidence, shuffled graph evidence, and no-graph input. This separates evidence inclusion, structural readability, and decoder-usable topology. Guided by this diagnosis, we present S²GE as an instance showing that diagnosis-driven interface design can improve native decoder usability. S²GE uses query-aware sampling, endpoint- and proximity-based ordering, and structure-preserving alignment. Across DBLP, Biomedical, GoodReads, and PubMed, S²GE achieves strict exact-match scores of 36.5%, 57.8%, 76.6%, and 52.0%, improving over the strongest native-generation baseline by 53.5 points on average. The interventions further reveal harmful-shuffle, shuffle-robust, and no-graph-saturated regimes.
%U https://aclanthology.org/2026.emnlp-main.647/
%P 14900-14917

Markdown (Informal)

[Graph Evidence Is Not Enough: Diagnosing Native Decoder Use in Graph-Augmented LLMs](https://aclanthology.org/2026.emnlp-main.647/) (Guo et al., EMNLP 2026)

• Graph Evidence Is Not Enough: Diagnosing Native Decoder Use in Graph-Augmented LLMs (Guo et al., EMNLP 2026)

ACL

• Xiaoyu Guo, Pengcheng Chen, YU Jiong, Yi Lu, Yaohua Wang, and Ziyang Li. 2026. Graph Evidence Is Not Enough: Diagnosing Native Decoder Use in Graph-Augmented LLMs. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 14900–14917, Budapest, Hungary. Association for Computational Linguistics.