<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="mao-etal-2026-controllable">
    <titleInfo>
        <title>Controllable Text-Attributed Graph Generation via LLM-Based Substructure Modeling</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Ruijun</namePart>
        <namePart type="family">Mao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jiaxuan</namePart>
        <namePart type="family">Chen</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Boyan</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Zixing</namePart>
        <namePart type="family">Zhang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yuhang</namePart>
        <namePart type="family">Zhang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Fangxin</namePart>
        <namePart type="family">Wang</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>Text-attributed graphs (TAGs) underpin many NLP applications, but real-world TAGs are often limited in scale, class balance, and diversity, motivating realistic and controllable TAG generation. Existing generators either replace node text with precomputed embeddings, reducing the task to vector-attributed graph generation, or rely on LLM prompting and agent simulation; neither learns a joint distribution over raw node text, edges, and labels. We propose a controllable substructure generator built from a lightweight causal LM. The model generates connected TAG substructures rather than whole graphs: each training substructure is serialized into lossless token sequences from multiple start nodes. Averaging next-token loss over these traversals treats traversal order as a latent variable and optimizes a variational lower bound on the permutation-invariant substructure marginal likelihood. Training combines continual pretraining for the unconditional distribution with control-prefix fine-tuning that conditions generation on target structural and label descriptors. At inference, a single autoregressive pass per prefix emits a parseable substructure, and the same generator supports distribution-aligned synthesis and boundary-aware augmentation through target-side adaptation alone. Across five real-world TAGs, it produces losslessly parseable graphs, matches or exceeds structure-only generators in structural fidelity, and improves cross-graph GNN generalization.</abstract>
    <identifier type="citekey">mao-etal-2026-controllable</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.996/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>22911</start>
            <end>22930</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Controllable Text-Attributed Graph Generation via LLM-Based Substructure Modeling
%A Mao, Ruijun
%A Chen, Jiaxuan
%A Li, Boyan
%A Zhang, Zixing
%A Zhang, Yuhang
%A Wang, Fangxin
%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 mao-etal-2026-controllable
%X Text-attributed graphs (TAGs) underpin many NLP applications, but real-world TAGs are often limited in scale, class balance, and diversity, motivating realistic and controllable TAG generation. Existing generators either replace node text with precomputed embeddings, reducing the task to vector-attributed graph generation, or rely on LLM prompting and agent simulation; neither learns a joint distribution over raw node text, edges, and labels. We propose a controllable substructure generator built from a lightweight causal LM. The model generates connected TAG substructures rather than whole graphs: each training substructure is serialized into lossless token sequences from multiple start nodes. Averaging next-token loss over these traversals treats traversal order as a latent variable and optimizes a variational lower bound on the permutation-invariant substructure marginal likelihood. Training combines continual pretraining for the unconditional distribution with control-prefix fine-tuning that conditions generation on target structural and label descriptors. At inference, a single autoregressive pass per prefix emits a parseable substructure, and the same generator supports distribution-aligned synthesis and boundary-aware augmentation through target-side adaptation alone. Across five real-world TAGs, it produces losslessly parseable graphs, matches or exceeds structure-only generators in structural fidelity, and improves cross-graph GNN generalization.
%U https://aclanthology.org/2026.emnlp-main.996/
%P 22911-22930

Markdown (Informal)

[Controllable Text-Attributed Graph Generation via LLM-Based Substructure Modeling](https://aclanthology.org/2026.emnlp-main.996/) (Mao et al., EMNLP 2026)

• Controllable Text-Attributed Graph Generation via LLM-Based Substructure Modeling (Mao et al., EMNLP 2026)

ACL

• Ruijun Mao, Jiaxuan Chen, Boyan Li, Zixing Zhang, Yuhang Zhang, and Fangxin Wang. 2026. Controllable Text-Attributed Graph Generation via LLM-Based Substructure Modeling. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 22911–22930, Budapest, Hungary. Association for Computational Linguistics.