Fine-Tuned Llama for Multilingual Text-to-Text Coreference Resolution

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F26%3AXN3GBKKA" target="_blank" >RIV/00216208:11320/26:XN3GBKKA - isvavai.cz</a>

Výsledek na webu

<a href="https://aclanthology.org/2025.crac-1.12/" target="_blank" >https://aclanthology.org/2025.crac-1.12/</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.18653/v1/2025.crac-1.12" target="_blank" >10.18653/v1/2025.crac-1.12</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Fine-Tuned Llama for Multilingual Text-to-Text Coreference Resolution

Popis výsledku v původním jazyce

This paper describes our approach to the CRAC 2025 Shared Task on Multilingual Coreference Resolution. We compete in the LLM track, where the systems are limited to generative text-to-text approaches. Our system is based on Llama 3.1-8B, fine-tuned to tag the document with coreference annotations. We have made one significant modification to the text format provided by the organizers: The model relies on the syntactic head for mention span representation. Additionally, we use joint pre-training, and we train the model to generate empty nodes. We provide an in-depth analysis of the performance of our models, which reveals several implementation problems. Although our system ended up in last place, we achieved the best performance on 10 datasets out of 22 within the track. By fixing the discovered problems in the post-evaluation phase, we improved our results substantially, outperforming all the systems in the LLM track and even some unconstrained track systems.

Název v anglickém jazyce

Fine-Tuned Llama for Multilingual Text-to-Text Coreference Resolution

Popis výsledku anglicky

This paper describes our approach to the CRAC 2025 Shared Task on Multilingual Coreference Resolution. We compete in the LLM track, where the systems are limited to generative text-to-text approaches. Our system is based on Llama 3.1-8B, fine-tuned to tag the document with coreference annotations. We have made one significant modification to the text format provided by the organizers: The model relies on the syntactic head for mention span representation. Additionally, we use joint pre-training, and we train the model to generate empty nodes. We provide an in-depth analysis of the performance of our models, which reveals several implementation problems. Although our system ended up in last place, we achieved the best performance on 10 datasets out of 22 within the track. By fixing the discovered problems in the post-evaluation phase, we improved our results substantially, outperforming all the systems in the LLM track and even some unconstrained track systems.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

—

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of the Eighth Workshop on Computational Models of Reference, Anaphora and Coreference

ISBN

979-8-89176-342-5

ISSN

—

e-ISSN

—

Počet stran výsledku

9

Strana od-do

140-148

Název nakladatele

—

Místo vydání

—

Místo konání akce

Suzhou, China

Datum konání akce

1. 1. 2026

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—