Papers by Marta Villegas
Building a Data Infrastructure for a Mid-Resource Language: The Case of Catalan (2024.lrec-main)
Copied to clipboard
Aitor Gonzalez-Agirre, Montserrat Marimon, Carlos Rodriguez-Penagos, Javier Aula-Blasco, Irene Baucells, Carme Armentano-Oller, Jorge Palomar-Giner, Baybars Kulebi, Marta Villegas
| Challenge: | Aina Project aims to provide Catalan with the resources needed to keep its relevance in AI/NLP applications. |
| Approach: | They propose a set of strategies to consider when improving technology support for a mid- or low-resource language . they propose annotated datasets and a framework to make models ready to use . |
| Outcome: | The Aina Project aims to provide Catalan with the necessary resources to keep its relevance in AI/NLP-related industry and research. |
FLOR: On the Effectiveness of Language Adaptation (2024.lrec-main)
Copied to clipboard
Severino Da Dalt, Joan Llop, Irene Baucells, Marc Pamies, Yishi Xu, Aitor Gonzalez-Agirre, Marta Villegas
| Challenge: | Large language models have amply proven their capabilities, but low- and mid-resource languages do not have access to the necessary means to train such models from scratch. |
| Approach: | They use a 26B tokens corpus to further pre-train BLOOM, giving rise to FLOR models. |
| Outcome: | The proposed model achieves consistent gains across Catalan and Spanish tasks. |
VeritasQA: A Truthfulness Benchmark Aimed at Multilingual Transferability (2025.coling-main)
Copied to clipboard
Javier Aula-Blasco, Júlia Falcão, Susana Sotelo, Silvia Paniagua, Aitor Gonzalez-Agirre, Marta Villegas
| Challenge: | Large Language Models (LLMs) struggle with falsehoods and model hallucination . many efforts struggle to surpass 50% accuracy, with only targeted techniques reaching around 65% . |
| Approach: | They propose a truthfulness benchmark that focuses on imitative falsehoods . they use a set of 353 questions and answers inspired by common misconceptions based on the language . |
| Outcome: | The benchmark is available in Spanish, Catalan, Galician and English . it measures the truthfulness of multilingual LLMs using 353 questions and answers . |
Mass-Editing Memory with Attention in Transformers: A cross-lingual exploration of knowledge (2024.findings-acl)
Copied to clipboard
| Challenge: | Recent studies have explored methods for updating and modifying factual knowledge in large language models, often focusing on specific multi-layer perceptron blocks. |
| Approach: | They propose a method that allows users to edit factual associations without catastrophic forgetting. |
| Outcome: | The proposed method achieves 10% increase in magnitude metrics while requiring minimal parameter modifications. |
Multi-LMentry: Can Multilingual LLMs Solve Elementary Tasks Across Languages? (2025.emnlp-main)
Copied to clipboard
Luca Moroni, Javier Aula-Blasco, Simone Conia, Irene Baucells, Naiara Perez, Silvia Paniagua Suárez, Anna Sallés, Malte Ostendorff, Júlia Falcão, Guijin Son, Aitor Gonzalez-Agirre, Roberto Navigli, Marta Villegas
| Challenge: | a recent study focused on complex, high-level tasks, but LMentry is limited to English . a multilingual evaluation of large language models is needed to address this gap, authors say . |
| Approach: | They propose a compact benchmark that enables systematic evaluation of large language models . they propose to use tasks that are trivial for humans but remain surprisingly difficult for LLMs . |
| Outcome: | The proposed benchmark is limited to English, leaving its insights linguistically narrow. |
Becoming a High-Resource Language in Speech: The Catalan Case in the Common Voice Corpus (2024.lrec-main)
Copied to clipboard
| Challenge: | a project to create a publicly available voice dataset for speech recognition systems in Catalan is a multifaceted challenge. |
| Approach: | They propose to create a publicly available voice dataset for future speech technologies in Catalan using the Mozilla Common Voice crowd-sourcing platform. |
| Outcome: | The proposed dataset shows that Catalan ranks as the most prominent language in the corpus. |
A weakly supervised textual entailment approach to zero-shot text classification (2023.eacl-main)
Copied to clipboard
Marc Pàmies, Joan Llop, Francesco Multari, Nicolau Duran-Silva, César Parra-Rojas, Aitor Gonzalez-Agirre, Francesco Alessandro Massucci, Marta Villegas
| Challenge: | Existing methods to train on weakly supervised datasets are expensive due to the computational cost of pre-training. |
| Approach: | They propose a method that trains on a weakly supervised dataset that is used as a proxy for a textual entailment problem and a target zero-shot text classification task. |
| Outcome: | The proposed model achieves state-of-the-art performance in the scientific domain and competitive results in other areas. |
PharmaCoNER: Pharmacological Substances, Compounds and proteins Named Entity Recognition track (D19-57)
Copied to clipboard
Aitor Gonzalez-Agirre, Montserrat Marimon, Ander Intxaurrondo, Obdulia Rabal, Marta Villegas, Martin Krallinger
| Challenge: | Biomedical text mining is one of the most prolific application domains of natural language processing technologies. |
| Approach: | They propose to share a task on detecting drug and chemical entities in medical documents in Spanish with other languages to improve access to biomedical text mining. |
| Outcome: | The first task on detecting drug and chemical entities in Spanish medical documents yielded competitive results with F-measures above 0.91. |
Are Multilingual Models the Best Choice for Moderately Under-resourced Languages? A Comprehensive Assessment for Catalan (2021.findings-acl)
Copied to clipboard
Jordi Armengol-Estapé, Casimiro Pio Carrino, Carlos Rodriguez-Penagos, Ona de Gibert Bonet, Carme Armentano-Oller, Aitor Gonzalez-Agirre, Maite Melero, Marta Villegas
| Challenge: | Multilingual language models have been a crucial breakthrough for under-resourced languages . however, the superiority of language-specific models has already been proven for underresourced ones . |
| Approach: | They propose to build a monolingual monolingual model that is comparable to state-of-the-art large multilingual models. |
| Outcome: | The proposed model consistently outperforms state-of-the-art models across tasks and settings. |
IberoBench: A Benchmark for LLM Evaluation in Iberian Languages (2025.coling-main)
Copied to clipboard
Irene Baucells, Javier Aula-Blasco, Iria de-Dios-Flores, Silvia Paniagua Suárez, Naiara Perez, Anna Salles, Susana Sotelo Docio, Júlia Falcão, Jose Javier Saiz, Robiert Sepulveda Torres, Jeremy Barnes, Pablo Gamallo, Aitor Gonzalez-Agirre, German Rigau, Marta Villegas
| Challenge: | Existing multi-task benchmarks for Large Language Models are limited to English . a new benchmark is needed to evaluate models on a range of tasks . |
| Approach: | They propose a multilingual, multi-task benchmark for Iberian languages built on the LM Evaluation Harness framework. |
| Outcome: | The proposed benchmark covers 62 tasks divided into 179 subtasks and is available in Iberian, Basque, Catalan, Galician, European Spanish and European Portuguese. |
A CURATEd CATalog: Rethinking the Extraction of Pretraining Corpora for Mid-Resourced Languages (2024.lrec-main)
Copied to clipboard
Jorge Palomar-Giner, Jose Javier Saiz, Ferran Espuña, Mario Mina, Severino Da Dalt, Joan Llop, Malte Ostendorff, Pedro Ortiz Suarez, Georg Rehm, Aitor Gonzalez-Agirre, Marta Villegas
| Challenge: | CATalog 1.0 is the largest text corpus in Catalan to date . CURATE is a pipeline that can be parallelizable to run in high performance clusters . |
| Approach: | They propose a data pipeline that uses binary filters to filter documents based on text quality . they optimised the pipeline to run in high performance clusters . |
| Outcome: | The proposed pipeline is optimized for high performance cluster environments and runs in high performance. |