Improving LLMs’ Learning of Coreference Resolution

Yujian Gan; Yuan Liang; Yanni Lin; Juntao Yu; Massimo Poesio

Improving LLMs’ Learning of Coreference Resolution

Yujian Gan, Yuan Liang, Yanni Lin, Juntao Yu, Massimo Poesio

Correct Metadata for

Important: The Anthology treat PDFs as authoritative. Please use this form only to correct data that is out of line with the PDF. See our corrections guidelines if you need to change the PDF.

Title Adjust the title. Retain tags such as <fixed-case>.

Authors Adjust author names and order to match the PDF.

Abstract Correct abstract if needed. Retain XML formatting tags such as <tex-math>.

Verification against PDF Ensure that the new title/authors match the snapshot below. (If there is no snapshot or it is too small, consult the PDF.)

Authors concatenated from the text boxes above:

ALL author names match the snapshot above—including middle initials, hyphens, and accents.

Abstract

Coreference Resolution (CR) is crucial for many NLP tasks, but existing LLMs struggle with hallucination and under-performance. In this paper, we investigate the limitations of existing LLM-based approaches to CR—specifically the Question-Answering (QA) Template and Document Template methods—and propose two novel techniques: Reversed Training with Joint Inference and Iterative Document Generation. Our experiments show that Reversed Training improves the QA Template method, while Iterative Document Generation eliminates hallucinations in the generated source text and boosts coreference resolution. Integrating these methods and techniques offers an effective and robust solution to LLM-based coreference resolution

Anthology ID:: 2025.sigdial-1.25
Volume:: Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue
Month:: August
Year:: 2025
Address:: Avignon, France
Editors:: Frédéric Béchet, Fabrice Lefèvre, Nicholas Asher, Seokhwan Kim, Teva Merlin
Venue:: SIGDIAL
SIG:
Publisher:: Association for Computational Linguistics
Note:
Pages:: 311–321
Language:
URL:: https://aclanthology.org/2025.sigdial-1.25/
DOI:
Bibkey:
Cite (ACL):: Yujian Gan, Yuan Liang, Yanni Lin, Juntao Yu, and Massimo Poesio. 2025. Improving LLMs’ Learning of Coreference Resolution. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 311–321, Avignon, France. Association for Computational Linguistics.
Cite (Informal):: Improving LLMs’ Learning of Coreference Resolution (Gan et al., SIGDIAL 2025)
Copy Citation:
PDF:: https://aclanthology.org/2025.sigdial-1.25.pdf

PDF Cite Search Fix data

Export citation

BibTeX
MODS XML
Endnote
Preformatted

@inproceedings{gan-etal-2025-improving,
    title = "Improving {LLM}s' Learning of Coreference Resolution",
    author = "Gan, Yujian  and
      Liang, Yuan  and
      Lin, Yanni  and
      Yu, Juntao  and
      Poesio, Massimo",
    editor = "B{\'e}chet, Fr{\'e}d{\'e}ric  and
      Lef{\`e}vre, Fabrice  and
      Asher, Nicholas  and
      Kim, Seokhwan  and
      Merlin, Teva",
    booktitle = "Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue",
    month = aug,
    year = "2025",
    address = "Avignon, France",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2025.sigdial-1.25/",
    pages = "311--321",
    abstract = "Coreference Resolution (CR) is crucial for many NLP tasks, but existing LLMs struggle with hallucination and under-performance. In this paper, we investigate the limitations of existing LLM-based approaches to CR{---}specifically the Question-Answering (QA) Template and Document Template methods{---}and propose two novel techniques: Reversed Training with Joint Inference and Iterative Document Generation. Our experiments show that Reversed Training improves the QA Template method, while Iterative Document Generation eliminates hallucinations in the generated source text and boosts coreference resolution. Integrating these methods and techniques offers an effective and robust solution to LLM-based coreference resolution"
}

Download as File

<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="gan-etal-2025-improving">
    <titleInfo>
        <title>Improving LLMs’ Learning of Coreference Resolution</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yujian</namePart>
        <namePart type="family">Gan</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yuan</namePart>
        <namePart type="family">Liang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yanni</namePart>
        <namePart type="family">Lin</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Juntao</namePart>
        <namePart type="family">Yu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Massimo</namePart>
        <namePart type="family">Poesio</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-08</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Frédéric</namePart>
            <namePart type="family">Béchet</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Fabrice</namePart>
            <namePart type="family">Lefèvre</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Nicholas</namePart>
            <namePart type="family">Asher</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Seokhwan</namePart>
            <namePart type="family">Kim</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Teva</namePart>
            <namePart type="family">Merlin</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Avignon, France</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>Coreference Resolution (CR) is crucial for many NLP tasks, but existing LLMs struggle with hallucination and under-performance. In this paper, we investigate the limitations of existing LLM-based approaches to CR—specifically the Question-Answering (QA) Template and Document Template methods—and propose two novel techniques: Reversed Training with Joint Inference and Iterative Document Generation. Our experiments show that Reversed Training improves the QA Template method, while Iterative Document Generation eliminates hallucinations in the generated source text and boosts coreference resolution. Integrating these methods and techniques offers an effective and robust solution to LLM-based coreference resolution</abstract>
    <identifier type="citekey">gan-etal-2025-improving</identifier>
    <location>
        <url>https://aclanthology.org/2025.sigdial-1.25/</url>
    </location>
    <part>
        <date>2025-08</date>
        <extent unit="page">
            <start>311</start>
            <end>321</end>
        </extent>
    </part>
</mods>
</modsCollection>

Download as File

%0 Conference Proceedings
%T Improving LLMs’ Learning of Coreference Resolution
%A Gan, Yujian
%A Liang, Yuan
%A Lin, Yanni
%A Yu, Juntao
%A Poesio, Massimo
%Y Béchet, Frédéric
%Y Lefèvre, Fabrice
%Y Asher, Nicholas
%Y Kim, Seokhwan
%Y Merlin, Teva
%S Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue
%D 2025
%8 August
%I Association for Computational Linguistics
%C Avignon, France
%F gan-etal-2025-improving
%X Coreference Resolution (CR) is crucial for many NLP tasks, but existing LLMs struggle with hallucination and under-performance. In this paper, we investigate the limitations of existing LLM-based approaches to CR—specifically the Question-Answering (QA) Template and Document Template methods—and propose two novel techniques: Reversed Training with Joint Inference and Iterative Document Generation. Our experiments show that Reversed Training improves the QA Template method, while Iterative Document Generation eliminates hallucinations in the generated source text and boosts coreference resolution. Integrating these methods and techniques offers an effective and robust solution to LLM-based coreference resolution
%U https://aclanthology.org/2025.sigdial-1.25/
%P 311-321

Download as File

Markdown (Informal)

[Improving LLMs’ Learning of Coreference Resolution](https://aclanthology.org/2025.sigdial-1.25/) (Gan et al., SIGDIAL 2025)

Improving LLMs’ Learning of Coreference Resolution (Gan et al., SIGDIAL 2025)

ACL

Yujian Gan, Yuan Liang, Yanni Lin, Juntao Yu, and Massimo Poesio. 2025. Improving LLMs’ Learning of Coreference Resolution. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 311–321, Avignon, France. Association for Computational Linguistics.