GLaRef@CRAC2025: Should we transform coreference resolution into a text generation task? Article Swipe

View

Association for Computational Linguistics 2025 ·

YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48448/ckvh-1d46

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.

Related Topics

Computer Science

Artificial Intelligence

Generative Grammar

Computational Linguistics

Machine Learning

Training, Validation, And Test Data Sets

Deep Learning

Concepts

Coreference Computer science Anaphora (linguistics) Artificial intelligence Natural language processing Resolution (logic) Generative grammar Point (geometry) Computational linguistics Computational model Machine learning Computational complexity theory Training set Generative model Track (disk drive) Deep learning Conjunction (astronomy) Test (biology)

Metadata

Type: other
Landing Page: https://doi.org/10.48448/ckvh-1d46
OA Status: green
OpenAlex ID: https://openalex.org/W7106842815

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W7106842815

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.48448/ckvh-1d46

Digital Object Identifier
Title: GLaRef@CRAC2025: Should we transform coreference resolution into a text generation task?

Work title
Type: other

OpenAlex work type
Publication year: 2025

Year of publication
Publication date: 2025-10-25

Full publication date if available
Authors: Association for Computational Linguistics 2025

List of authors in order
Landing page: https://doi.org/10.48448/ckvh-1d46

Publisher landing page
Open access: Yes

Whether a free full text is available
OA status: green

Open access status per OpenAlex
OA URL: https://doi.org/10.48448/ckvh-1d46

Direct OA link when available
Concepts: Coreference, Computer science, Anaphora (linguistics), Artificial intelligence, Natural language processing, Resolution (logic), Generative grammar, Point (geometry), Computational linguistics, Computational model, Machine learning, Computational complexity theory, Training set, Generative model, Track (disk drive), Deep learning, Conjunction (astronomy), Test (biology)

Top concepts (fields/topics) attached by OpenAlex
Cited by: 0

Total citation count in OpenAlex

Full payload

id	https://openalex.org/W7106842815
doi	https://doi.org/10.48448/ckvh-1d46
ids.doi	https://doi.org/10.48448/ckvh-1d46
ids.openalex	https://openalex.org/W7106842815
fwci
type	other
title	GLaRef@CRAC2025: Should we transform coreference resolution into a text generation task?
biblio.issue
biblio.volume
biblio.last_page
biblio.first_page
is_xpac	False
apc_list
apc_paid
concepts[0].id	https://openalex.org/C28076734
concepts[0].level	3
concepts[0].score	0.9766700863838196
concepts[0].wikidata	https://www.wikidata.org/wiki/Q63087
concepts[0].display_name	Coreference
concepts[1].id	https://openalex.org/C41008148
concepts[1].level	0
concepts[1].score	0.8018482327461243
concepts[1].wikidata	https://www.wikidata.org/wiki/Q21198
concepts[1].display_name	Computer science
concepts[2].id	https://openalex.org/C2781449363
concepts[2].level	3
concepts[2].score	0.6962345242500305
concepts[2].wikidata	https://www.wikidata.org/wiki/Q156751
concepts[2].display_name	Anaphora (linguistics)
concepts[3].id	https://openalex.org/C154945302
concepts[3].level	1
concepts[3].score	0.6918583512306213
concepts[3].wikidata	https://www.wikidata.org/wiki/Q11660
concepts[3].display_name	Artificial intelligence
concepts[4].id	https://openalex.org/C204321447
concepts[4].level	1
concepts[4].score	0.6292918920516968
concepts[4].wikidata	https://www.wikidata.org/wiki/Q30642
concepts[4].display_name	Natural language processing
concepts[5].id	https://openalex.org/C138268822
concepts[5].level	2
concepts[5].score	0.6168367862701416
concepts[5].wikidata	https://www.wikidata.org/wiki/Q1051925
concepts[5].display_name	Resolution (logic)
concepts[6].id	https://openalex.org/C39890363
concepts[6].level	2
concepts[6].score	0.6137760877609253
concepts[6].wikidata	https://www.wikidata.org/wiki/Q36108
concepts[6].display_name	Generative grammar
concepts[7].id	https://openalex.org/C28719098
concepts[7].level	2
concepts[7].score	0.5771011710166931
concepts[7].wikidata	https://www.wikidata.org/wiki/Q44946
concepts[7].display_name	Point (geometry)
concepts[8].id	https://openalex.org/C155092808
concepts[8].level	2
concepts[8].score	0.4060477018356323
concepts[8].wikidata	https://www.wikidata.org/wiki/Q182557
concepts[8].display_name	Computational linguistics
concepts[9].id	https://openalex.org/C66024118
concepts[9].level	2
concepts[9].score	0.3592160940170288
concepts[9].wikidata	https://www.wikidata.org/wiki/Q1122506
concepts[9].display_name	Computational model
concepts[10].id	https://openalex.org/C119857082
concepts[10].level	1
concepts[10].score	0.335171639919281
concepts[10].wikidata	https://www.wikidata.org/wiki/Q2539
concepts[10].display_name	Machine learning
concepts[11].id	https://openalex.org/C179799912
concepts[11].level	2
concepts[11].score	0.32631734013557434
concepts[11].wikidata	https://www.wikidata.org/wiki/Q205084
concepts[11].display_name	Computational complexity theory
concepts[12].id	https://openalex.org/C51632099
concepts[12].level	2
concepts[12].score	0.31184956431388855
concepts[12].wikidata	https://www.wikidata.org/wiki/Q3985153
concepts[12].display_name	Training set
concepts[13].id	https://openalex.org/C167966045
concepts[13].level	3
concepts[13].score	0.2932296693325043
concepts[13].wikidata	https://www.wikidata.org/wiki/Q5532625
concepts[13].display_name	Generative model
concepts[14].id	https://openalex.org/C89992363
concepts[14].level	2
concepts[14].score	0.2767502963542938
concepts[14].wikidata	https://www.wikidata.org/wiki/Q5961558
concepts[14].display_name	Track (disk drive)
concepts[15].id	https://openalex.org/C108583219
concepts[15].level	2
concepts[15].score	0.2664281725883484
concepts[15].wikidata	https://www.wikidata.org/wiki/Q197536
concepts[15].display_name	Deep learning
concepts[16].id	https://openalex.org/C59656382
concepts[16].level	2
concepts[16].score	0.26617926359176636
concepts[16].wikidata	https://www.wikidata.org/wiki/Q191536
concepts[16].display_name	Conjunction (astronomy)
concepts[17].id	https://openalex.org/C2777267654
concepts[17].level	2
concepts[17].score	0.2523273229598999
concepts[17].wikidata	https://www.wikidata.org/wiki/Q3519023
concepts[17].display_name	Test (biology)
keywords[0].id	https://openalex.org/keywords/coreference
keywords[0].score	0.9766700863838196
keywords[0].display_name	Coreference
keywords[1].id	https://openalex.org/keywords/anaphora
keywords[1].score	0.6962345242500305
keywords[1].display_name	Anaphora (linguistics)
keywords[2].id	https://openalex.org/keywords/resolution
keywords[2].score	0.6168367862701416
keywords[2].display_name	Resolution (logic)
keywords[3].id	https://openalex.org/keywords/generative-grammar
keywords[3].score	0.6137760877609253
keywords[3].display_name	Generative grammar
keywords[4].id	https://openalex.org/keywords/point
keywords[4].score	0.5771011710166931
keywords[4].display_name	Point (geometry)
keywords[5].id	https://openalex.org/keywords/computational-linguistics
keywords[5].score	0.4060477018356323
keywords[5].display_name	Computational linguistics
keywords[6].id	https://openalex.org/keywords/computational-model
keywords[6].score	0.3592160940170288
keywords[6].display_name	Computational model
language
locations[0].id	doi:10.48448/ckvh-1d46
locations[0].is_oa	True
locations[0].source.id	https://openalex.org/S7407053148
locations[0].source.type	repository
locations[0].source.is_oa	False
locations[0].source.issn_l
locations[0].source.is_core	False
locations[0].source.is_in_doaj	False
locations[0].source.display_name	Underline Science Inc.
locations[0].source.host_organization
locations[0].source.host_organization_name
locations[0].license
locations[0].pdf_url
locations[0].version
locations[0].raw_type	article
locations[0].license_id
locations[0].is_accepted	False
locations[0].is_published
locations[0].raw_source_name
locations[0].landing_page_url	https://doi.org/10.48448/ckvh-1d46
indexed_in	datacite
authorships[0].author.id
authorships[0].author.orcid
authorships[0].author.display_name	Association for Computational Linguistics 2025
authorships[0].author_position	first
authorships[0].raw_author_name	Association for Computational Linguistics 2025
authorships[0].is_corresponding	True
has_content.pdf	False
has_content.grobid_xml	False
is_paratext	False
open_access.is_oa	True
open_access.oa_url	https://doi.org/10.48448/ckvh-1d46
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-11-28T00:00:00
display_name	GLaRef@CRAC2025: Should we transform coreference resolution into a text generation task?
has_fulltext	False
is_retracted	False
updated_date	2025-11-28T02:12:24.556248
primary_topic
cited_by_count	0
locations_count	1
best_oa_location.id	doi:10.48448/ckvh-1d46
best_oa_location.is_oa	True
best_oa_location.source.id	https://openalex.org/S7407053148
best_oa_location.source.type	repository
best_oa_location.source.is_oa	False
best_oa_location.source.issn_l
best_oa_location.source.is_core	False
best_oa_location.source.is_in_doaj	False
best_oa_location.source.display_name	Underline Science Inc.
best_oa_location.source.host_organization
best_oa_location.source.host_organization_name
best_oa_location.license
best_oa_location.pdf_url
best_oa_location.version
best_oa_location.raw_type	article
best_oa_location.license_id
best_oa_location.is_accepted	False
best_oa_location.is_published	False
best_oa_location.raw_source_name
best_oa_location.landing_page_url	https://doi.org/10.48448/ckvh-1d46
primary_location.id	doi:10.48448/ckvh-1d46
primary_location.is_oa	True
primary_location.source.id	https://openalex.org/S7407053148
primary_location.source.type	repository
primary_location.source.is_oa	False
primary_location.source.issn_l
primary_location.source.is_core	False
primary_location.source.is_in_doaj	False
primary_location.source.display_name	Underline Science Inc.
primary_location.source.host_organization
primary_location.source.host_organization_name
primary_location.license
primary_location.pdf_url
primary_location.version
primary_location.raw_type	article
primary_location.license_id
primary_location.is_accepted	False
primary_location.is_published	False
primary_location.raw_source_name
primary_location.landing_page_url	https://doi.org/10.48448/ckvh-1d46
publication_date	2025-10-25
publication_year	2025
referenced_works_count	0
abstract_inverted_index.a	115
abstract_inverted_index.AI	97
abstract_inverted_index.We	0, 102
abstract_inverted_index.at	134
abstract_inverted_index.in	29, 57, 111
abstract_inverted_index.of	4, 13, 17, 44, 79, 107, 114, 128, 137
abstract_inverted_index.on	48, 84
abstract_inverted_index.to	7, 67, 98
abstract_inverted_index.we	26, 87, 92, 121, 131
abstract_inverted_index.LLM	11
abstract_inverted_index.and	10, 20, 32, 46
abstract_inverted_index.are	124
abstract_inverted_index.but	36, 52, 74
abstract_inverted_index.end	136
abstract_inverted_index.how	104
abstract_inverted_index.our	5, 129
abstract_inverted_index.out	89
abstract_inverted_index.ran	93
abstract_inverted_index.the	2, 8, 14, 30, 33, 49, 61, 68, 80, 105, 112, 135
abstract_inverted_index.way	113
abstract_inverted_index.10\%	78
abstract_inverted_index.into	94
abstract_inverted_index.less	76
abstract_inverted_index.many	125
abstract_inverted_index.set,	51
abstract_inverted_index.team	6
abstract_inverted_index.test	50
abstract_inverted_index.text	108
abstract_inverted_index.than	77
abstract_inverted_index.that	91
abstract_inverted_index.them	133
abstract_inverted_index.this	85, 138
abstract_inverted_index.very	54
abstract_inverted_index.with	38, 53, 75
abstract_inverted_index.61.57	45
abstract_inverted_index.62.96	47
abstract_inverted_index.cost.	59, 82
abstract_inverted_index.ended	27
abstract_inverted_index.fact,	86
abstract_inverted_index.fifth	31
abstract_inverted_index.first	34
abstract_inverted_index.large	55
abstract_inverted_index.point	88
abstract_inverted_index.task,	24
abstract_inverted_index.there	123
abstract_inverted_index.track	70
abstract_inverted_index.using	95
abstract_inverted_index.where	25
abstract_inverted_index.Models	16
abstract_inverted_index.place,	35
abstract_inverted_index.scores	43
abstract_inverted_index.shared	23
abstract_inverted_index.stands	110
abstract_inverted_index.system	65
abstract_inverted_index.tracks	12
abstract_inverted_index.Indeed,	60
abstract_inverted_index.average	41
abstract_inverted_index.discuss	132
abstract_inverted_index.explain	103
abstract_inverted_index.perform	99
abstract_inverted_index.present	1
abstract_inverted_index.realize	122
abstract_inverted_index.scores:	40
abstract_inverted_index.similar	39, 72
abstract_inverted_index.Anaphora	19
abstract_inverted_index.CoNLL-F1	42
abstract_inverted_index.article.	139
abstract_inverted_index.obtained	71
abstract_inverted_index.problems	90
abstract_inverted_index.reliable	116
abstract_inverted_index.classical	62
abstract_inverted_index.framework	106
abstract_inverted_index.pair-wise	63
abstract_inverted_index.potential	126
abstract_inverted_index.submitted	66
abstract_inverted_index.(CRAC2025)	22
abstract_inverted_index.Reference,	18
abstract_inverted_index.Reflecting	83
abstract_inverted_index.generation	109
abstract_inverted_index.generative	96
abstract_inverted_index.resolution	64
abstract_inverted_index.Coreference	21
abstract_inverted_index.LLM-system;	130
abstract_inverted_index.coreference	100, 118
abstract_inverted_index.differences	56
abstract_inverted_index.performance	73
abstract_inverted_index.resolution.	101
abstract_inverted_index.submissions	3
abstract_inverted_index.text-global	117
abstract_inverted_index.Nonetheless,	120
abstract_inverted_index.improvements	127
abstract_inverted_index.nevertheless	37
abstract_inverted_index.respectively	28
abstract_inverted_index.Computational	15
abstract_inverted_index.Unconstrained	9, 69
abstract_inverted_index.computational	58, 81
abstract_inverted_index.representation.	119
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	1
citation_normalized_percentile