Research

Publications and research projects in natural language processing, focusing on fairness, bias mitigation, and cross-lingual transfer.

For a full list of publications, see Google Scholar.

2024 9 papers

Metrics for What, Metrics for Whom: Assessing Actionability of Bias Evaluation Metrics in NLP

★EMNLP · Pieter Delobelle, Giuseppe Attanasio, Debora Nozza, Su Lin Blodgett and Zeerak Talat

Paper Blog Post Video
BPE-knockout: Pruning Pre-existing BPE Tokenisers with Backwards-compatible Morphological Semi-supervision

★NAACL · Thomas Bauwens and Pieter Delobelle

Paper Blog Post Code Video
Trans-Tokenization and Cross-lingual Vocabulary Transfers: Language Adaptation of LLMs for Low-Resource NLP

COLM · François Remy, Pieter Delobelle, Hayastan Avetisyan, Alfiya Khabibullina, Miryam de Lhoneux and Thomas Demeester

Blog Post
Whispering Experts: Neural Interventions for Toxicity Mitigation in Language Models

★ICML · Xavier Suau, Pieter Delobelle, Katherine Metcalf, Armand Joulin, Nicholas Apostoloff, Luca Zappella and Pau Rodriguez

RobBERT-2023: Keeping Dutch Language Models Up-To-Date at a Lower Cost Thanks to Model Conversion

CLIN · Pieter Delobelle and François Remy

Blog Post
OneLove beyond the field--A few-shot pipeline for topic and sentiment analysis during the FIFA World Cup in Qatar

KONVENS · Christoph Rauchegger, Sonja Mei Wang and Pieter Delobelle

ItaEval and TweetyIta: A New Extensive Benchmark and Efficiency-First Language Model for Italian

CLiC-it · Giuseppe Attanasio, Pieter Delobelle, Moreno La Quatra, Andrea Santilli and Beatrice Savoldi

High stakes for LLMs: Analysis of Bias in BERT-based Recommender Systems

BNAIC/BeNeLearn · Nisse Degelin, Pieter Delobelle, Kristen Scott and Bettina Berendt

ChocoLlama: Lessons Learned From Teaching Llamas Dutch

arXiv · Matthieu Meeus, Anthony Rathé, François Remy, Pieter Delobelle, Jens-Joris Decorte and Thomas Demeester

arXiv

2023 5 papers

How Far Can It Go? On Intrinsic Gender Bias Mitigation for Text Classification

★EACL · Ewoenam Tokpo, Pieter Delobelle, Bettina Berendt and Toon Calders

Blog Post
ResumeTailor: Improving Resume Quality Through Co-Creative Tools

HHAI · Pieter Delobelle, Sonja Mei Wang and Bettina Berendt

Inferring Missing CV Skills using PU Learning and Variational Inference

AI4HR @ ECML-PKDD · Victor Verreet, Lennert De Smet, Robin Manhaeve, Pieter Delobelle and Jessa Bekker

Tik-to-Tok: Translating Language Models One Token at a Time

arXiv · François Remy, Pieter Delobelle, Bettina Berendt, Kris Demuynck and Thomas Demeester

Towards Fairer Foundation Models: On Measures and Mitigation Strategies

PhD thesis · Pieter Delobelle

Blog Post

2022 5 papers

RobBERTje: A Distilled Dutch BERT Model

CLIN · Pieter Delobelle, Thomas Winters and Bettina Berendt

Algorithmic tools in public employment services: Towards a jobseeker-centric perspective

FAccT · Kristen Scott, Sonja Mei Wang, Milagros Miceli, Pieter Delobelle, Karolina Sztandar-Sztanderska and Bettina Berendt Best Student Paper Award

Measuring Fairness with Biased Rulers: A Comparative Study on Bias Metrics for Pre-trained Language Models

★NAACL · Pieter Delobelle, Ewoenam Tokpo, Toon Calders and Bettina Berendt

Blog Post
FairDistillation: Mitigating Stereotyping in Language Models

ECML-PKDD · Pieter Delobelle and Bettina Berendt

Blog Post
RobBERT-2022: Updating a Dutch Language Model to Account for Evolving Language Use

arXiv · Pieter Delobelle, Thomas Winters and Bettina Berendt

Blog Post arXiv

2021 4 papers

Time to Question if We Should: Data-Driven and Algorithmic Tools in Public Employment Services

FEAST @ ECML-PKDD · Pieter Delobelle, Kristen Scott, Sonja Mei Wang, Milagros Miceli, David Hartmann, Tianling Yang, Elena Murasso, Karolina Sztandar-Sztanderska and Bettina Berendt

Learning to Rank Generated Portmanteaus

ICCC · Lara Pollet, Thomas Winters and Pieter Delobelle

Survival of the Wittiest: Evolving Satire with Language Models

ICCC · Thomas Winters and Pieter Delobelle

Measuring Shifts in Attitudes Towards COVID-19 Measures in Belgium

CLIN Journal · Kristen Scott, Pieter Delobelle and Bettina Berendt

Blog Post

2020 3 papers

RobBERT: a Dutch RoBERTa-based Language Model

EMNLP Findings · Pieter Delobelle, Thomas Winters and Bettina Berendt

Ethical Adversaries: Towards Mitigating Unfairness with Adversarial Machine Learning

ACM SIGKDD Explorations Newsletter · Pieter Delobelle, Paul Temple, Gilles Perrouin, Benoît Frénay, Patrick Heymans and Bettina Berendt

Dutch Humor Detection by Generating Negative Examples

Benelearn · Thomas Winters and Pieter Delobelle Best video award

2019 2 papers

Computational Ad Hominem Detection

ACL Student Research Workshop · Pieter Delobelle, Murilo Cunha, Eric Massip Cano, Jeroen Peperkamp and Bettina Berendt

Time to Take Emoji Seriously: They Vastly Improve Casual Conversational Models

Benelearn · Pieter Delobelle and Bettina Berendt