Publications

Research on efficient foundation models, controllable model behaviour, and cross-lingual NLP.

2026

arXiv · 2026 (preprint)

Safety for Whom? Boundary-Aware Self-Distillation for Controlled LLM Safety Refusal

Alejo López-Ávila, Iker García-Ferrero, Jezabel Garcia, Antonio Tiene, Román Orús.

Research summary

Training models to refuse harmful requests within a topic while preserving useful answers to benign ones. We study how self-generated data and harmful-benign prompt pairs shape this boundary.

arXiv · 2026 (preprint)

Quantization-Aware Healing: A Practical Recipe for Recovering Compressed, 4-Bit LLMs

Bakbergen Ryskulov, Iker García-Ferrero, David Montero, David Jansen, Ali Hashemi, Jezabel R. Garcia, Antonio Tiene, Román Orús.

Paper

arXiv · 2026 (preprint)

Efficient Knowledge Distillation for LLMs: Offline Top-K Logits and a Fused Chunked KL Loss

Bakbergen Ryskulov, Iker García-Ferrero, David Montero, David Jansen, Ali Hashemi, Jezabel R. Garcia, Antonio Tiene, Román Orús.

Paper Code

Research summary

Offline top-K teacher logits and a fused, chunked KL loss reduce the memory and computation needed for LLM distillation.

LREC · 2026

Refusal Steering: Fine-grained Control over LLM Refusal Behaviour for Sensitive Topics

Iker García-Ferrero, David Montero, Roman Orus.

Paper Code Model

LREC · 2026

FactOReS: Fact-checking with an Evidence-based Open Resource in Spanish

Nagore Bravo, Jaione Bengoetxea, Iker García-Ferrero, Alba Bonet Jover, Estela Saquete, Rodrigo Agerri.

PLN · 2026

Discriminative Benchmarking of Spanish Language Models: Findings from the ODESIA Challenge 2024

Alejandro Benito Santos, Roser Morante, Adrián Ghajari Espinosa, Iker García-Ferrero, Robiert Sepúlveda Torres, German Rigau, Rodrigo Agerri, et al.

Paper

2025

Krea Blog · 2025

Releasing Open Weights for FLUX.1 Krea

Sangwu Lee, Titus Ebbecke, Erwann Millon, Will Beddow, Le Zhuo, Iker García-Ferrero, Liam Esparraguera, Mihai Petrescu, Gian Saß, Gabriel Menezes, Victor Perez.

Blog Model

EMNLP · 2025

Instructing LLMs for Low-Resource Languages: A Systematic Study for Basque

Oscar Sainz, Naiara Perez, Julen Etxaniz, Joseba Fernandez de Landa, Itziar Aldabe, Iker García-Ferrero, Aimar Zabala, Ekhi Azurmendi, German Rigau, Eneko Agirre, Mikel Artetxe, Aitor Soroa.

Paper Models

Findings of ACL · 2025

GuideX: Guided Synthetic Data Generation for Zero-Shot IE

Neil De La Fuente, Oscar Sainz, Iker García-Ferrero, Eneko Agirre.

Paper Code

Doctoral thesis · 2025

Cross-Lingual Transfer for Low-Resource Natural Language Processing

Iker García-Ferrero

Thesis Slides

2024

CONDA Workshop · 2024

Data Contamination Report from the 2024 CONDA Shared Task

Oscar Sainz, Iker García-Ferrero, Alon Jacovi, Jon Ander Campos, Yanai Elazar, Eneko Agirre, Yoav Goldberg.

Paper

SEPLN-CEDI-PD · 2024

HiTZ@Disargue: Few-shot Learning and Argumentation to Detect and Fight Misinformation in Social Media

Rodrigo Agerri, Jeremy Barnes, Jaione Bengoetxea, Blanca Calvo Figueras, Joseba Fernandez de Landa, Iker García-Ferrero, Olia Toporkov, Irune Zubiaga.

Paper

SIGUL · 2024

Uncovering Social Changes of the Basque Twitter Community during COVID-19

Joseba Fernandez de Landa, Iker García-Ferrero, Ander Salaberria, Jon Ander Campos.

Paper

PLN · 2024

NoticIA: A Clickbait Article Summarization Dataset in Spanish

Iker García-Ferrero, Begoña Altuna.

Paper Dataset

CLARIN Annual Conference · 2024

Prepare to be Amazed: NoticIA, the Spanish Clickbait Dataset Transforming the Way We Read News

Begoña Altuna, Iker García-Ferrero.

Paper

LREC-COLING · 2024

MedicalMT5: An Open-Source Multilingual Text-to-Text LLM for the Medical Domain

Iker García-Ferrero, Rodrigo Agerri, Aitziber Atutxa Salazar, Elena Cabrio, Iker de la Iglesia, Alberto Lavelli, Bernardo Magnini, Benjamin Molinet, Johana Ramirez-Romero, German Rigau, Jose Maria Villa-Gonzalez, Serena Villata, Andrea Zaninello.

Paper Models

ICLR · 2024

GoLLIE: Annotation Guidelines improve Zero-Shot IE

Oscar Sainz, Iker García-Ferrero, Rodrigo Agerri, Oier Lopez de Lacalle, German Rigau, Eneko Agirre.

Paper Code

Research summary

A language model that follows annotation guidelines to extract information, including schemas it has not seen during training.

2023

EMNLP · 2023

This is not a Dataset: A Large Negation Benchmark to Challenge LLMs

Iker García-Ferrero, Begoña Altuna, Javier Alvez, Itziar Gonzalez-Dios, German Rigau.

Paper Code

Findings of EMNLP · 2023

T-Projection: High Quality Annotation Projection for Sequence Labeling Tasks

Iker García-Ferrero, Rodrigo Agerri, German Rigau.

Paper Code

Findings of EMNLP · 2023

NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination

Oscar Sainz, Jon Ander Campos, Iker García-Ferrero, Julen Etxaniz, Oier Lopez de Lacalle, Eneko Agirre.

Paper Code

SemEval · 2023

IXA/Cogcomp at SemEval-2023 Task 2: Context-enriched Multilingual NER using KBs

Iker García-Ferrero, Jon Ander Campos, Oscar Sainz, Ander Salaberria, Dan Roth.

Paper Code

SEPLN · 2023

HiTZ@ Antidote: Explainable AI for Digital Medicine

Rodrigo Agerri, Iñigo Alonso, Aitziber Atutxa, Ander Berrondo, Ainara Estarrona, Iker García-Ferrero, Iakes Goenaga, Koldo Gojenola, Maite Oronoz, Igor Perez-Tejedor, German Rigau, Anar Yeginbergenova.

Paper