<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2026-reprof">
    <titleInfo>
        <title>ReProF: Residual Prototype Fusion for Continual LLM Attribution</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yujie</namePart>
        <namePart type="family">Wang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Zhendong</namePart>
        <namePart type="family">Mao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Quan</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>Continual LLM attribution aims to identify the source of generated text amid the ongoing emergence of new models. While existing training-free, prototype-based methods excel at this task, they rely solely on the final-layer features of a frozen extractor, overlooking rich hierarchical stylistic cues. Moreover, naive multi-layer fusion yields suboptimal results. To address these limitations, we propose ReProF (Residual Prototype Fusion), a training-free framework that partitions the feature extractor’s hierarchy into layer blocks and derives depth-aware analytic prototypes via shallow-to-deep residual fitting: deeper prototypes model the residuals of shallower ones. By fusing predictions across these blocks, ReProF captures multi-granular stylistic cues without requiring parameter tuning or data rehearsal, thereby intrinsically eliminating catastrophic forgetting. Extensive experiments across 19 LLMs and 8 domains demonstrate that ReProF establishes a new state-of-the-art, consistently outperforming existing approaches by a wide margin. Code is available at: https://github.com/wyj-hb/ReProF.</abstract>
    <identifier type="citekey">wang-etal-2026-reprof</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.2051/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>46709</start>
            <end>46724</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T ReProF: Residual Prototype Fusion for Continual LLM Attribution
%A Wang, Yujie
%A Mao, Zhendong
%A Wang, Quan
%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 wang-etal-2026-reprof
%X Continual LLM attribution aims to identify the source of generated text amid the ongoing emergence of new models. While existing training-free, prototype-based methods excel at this task, they rely solely on the final-layer features of a frozen extractor, overlooking rich hierarchical stylistic cues. Moreover, naive multi-layer fusion yields suboptimal results. To address these limitations, we propose ReProF (Residual Prototype Fusion), a training-free framework that partitions the feature extractor’s hierarchy into layer blocks and derives depth-aware analytic prototypes via shallow-to-deep residual fitting: deeper prototypes model the residuals of shallower ones. By fusing predictions across these blocks, ReProF captures multi-granular stylistic cues without requiring parameter tuning or data rehearsal, thereby intrinsically eliminating catastrophic forgetting. Extensive experiments across 19 LLMs and 8 domains demonstrate that ReProF establishes a new state-of-the-art, consistently outperforming existing approaches by a wide margin. Code is available at: https://github.com/wyj-hb/ReProF.
%U https://aclanthology.org/2026.emnlp-main.2051/
%P 46709-46724

Markdown (Informal)

[ReProF: Residual Prototype Fusion for Continual LLM Attribution](https://aclanthology.org/2026.emnlp-main.2051/) (Wang et al., EMNLP 2026)

• ReProF: Residual Prototype Fusion for Continual LLM Attribution (Wang et al., EMNLP 2026)

ACL

• Yujie Wang, Zhendong Mao, and Quan Wang. 2026. ReProF: Residual Prototype Fusion for Continual LLM Attribution. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 46709–46724, Budapest, Hungary. Association for Computational Linguistics.