@inproceedings{seminck-etal-2025-glaref,
title = "{GL}a{R}ef@{CRAC}2025: Should we transform coreference resolution into a text generation task?",
author = "Seminck, Olga and
Bourgois, Antoine and
Dupont, Yoann and
Dehouck, Mathieu and
Delaborde, Marine",
editor = "Ogrodniczuk, Maciej and
Novak, Michal and
Poesio, Massimo and
Pradhan, Sameer and
Ng, Vincent",
booktitle = "Proceedings of the Eighth Workshop on Computational Models of Reference, Anaphora and Coreference",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.crac-1.10/",
pages = "119--129",
abstract = "We present the submissions of our team to the Unconstrained and LLM tracks of the Computational Models of Reference, Anaphora and Coreference (CRAC2025) shared task, where we ended respectively in the fifth and the first place, but nevertheless with similar scores: average CoNLL-F1 scores of 61.57 and 62.96 on the test set, but with very large differences in computational cost. Indeed, the classical pair-wise resolution system submitted to the Unconstrained track obtained similar performance but with less than 10{\%} of the computational cost. Reflecting on this fact, we point out problems that we ran into using generative AI to perform coreference resolution. We explain how the framework of text generation stands in the way of a reliable text-global coreference representation. Nonetheless, we realize there are many potential improvements of our LLM-system; we discuss them at the end of this article."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="seminck-etal-2025-glaref">
<titleInfo>
<title>GLaRef@CRAC2025: Should we transform coreference resolution into a text generation task?</title>
</titleInfo>
<name type="personal">
<namePart type="given">Olga</namePart>
<namePart type="family">Seminck</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Antoine</namePart>
<namePart type="family">Bourgois</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yoann</namePart>
<namePart type="family">Dupont</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mathieu</namePart>
<namePart type="family">Dehouck</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marine</namePart>
<namePart type="family">Delaborde</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Eighth Workshop on Computational Models of Reference, Anaphora and Coreference</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maciej</namePart>
<namePart type="family">Ogrodniczuk</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Michal</namePart>
<namePart type="family">Novak</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Massimo</namePart>
<namePart type="family">Poesio</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sameer</namePart>
<namePart type="family">Pradhan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vincent</namePart>
<namePart type="family">Ng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>We present the submissions of our team to the Unconstrained and LLM tracks of the Computational Models of Reference, Anaphora and Coreference (CRAC2025) shared task, where we ended respectively in the fifth and the first place, but nevertheless with similar scores: average CoNLL-F1 scores of 61.57 and 62.96 on the test set, but with very large differences in computational cost. Indeed, the classical pair-wise resolution system submitted to the Unconstrained track obtained similar performance but with less than 10% of the computational cost. Reflecting on this fact, we point out problems that we ran into using generative AI to perform coreference resolution. We explain how the framework of text generation stands in the way of a reliable text-global coreference representation. Nonetheless, we realize there are many potential improvements of our LLM-system; we discuss them at the end of this article.</abstract>
<identifier type="citekey">seminck-etal-2025-glaref</identifier>
<location>
<url>https://aclanthology.org/2025.crac-1.10/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>119</start>
<end>129</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T GLaRef@CRAC2025: Should we transform coreference resolution into a text generation task?
%A Seminck, Olga
%A Bourgois, Antoine
%A Dupont, Yoann
%A Dehouck, Mathieu
%A Delaborde, Marine
%Y Ogrodniczuk, Maciej
%Y Novak, Michal
%Y Poesio, Massimo
%Y Pradhan, Sameer
%Y Ng, Vincent
%S Proceedings of the Eighth Workshop on Computational Models of Reference, Anaphora and Coreference
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%F seminck-etal-2025-glaref
%X We present the submissions of our team to the Unconstrained and LLM tracks of the Computational Models of Reference, Anaphora and Coreference (CRAC2025) shared task, where we ended respectively in the fifth and the first place, but nevertheless with similar scores: average CoNLL-F1 scores of 61.57 and 62.96 on the test set, but with very large differences in computational cost. Indeed, the classical pair-wise resolution system submitted to the Unconstrained track obtained similar performance but with less than 10% of the computational cost. Reflecting on this fact, we point out problems that we ran into using generative AI to perform coreference resolution. We explain how the framework of text generation stands in the way of a reliable text-global coreference representation. Nonetheless, we realize there are many potential improvements of our LLM-system; we discuss them at the end of this article.
%U https://aclanthology.org/2025.crac-1.10/
%P 119-129
Markdown (Informal)
[GLaRef@CRAC2025: Should we transform coreference resolution into a text generation task?](https://aclanthology.org/2025.crac-1.10/) (Seminck et al., CRAC 2025)
ACL