My research work lies between statistics, natural language processing and machine learning. Since 2023, I am a post-doctoral researcher at EPFL, in the NLP lab, under the supervision of Antoine Bosselut. Among other topics, I work on vision-language models towards Knowledge-guided Explanations for Satellite Images and on commonsense reasoning.

In 2022, I was an invited researcher at Jozef Stefan Institute (JSI) in Ljubjana, Slovenia; working on multi-task learning and causal discovery. Before JSI, I did a post-doc at INRIA Paris, in the Almanach Team, designing models to detect hateful content in social media posts and memes; the focus was on multimodal models, cross-lingual transfer and temporal graph modeling for social networks.

I did my PhD at LISN and University Paris-Saclay, under the supervision of Alexandre Allauzen, on computational models of diachronic semantic change. I worked mostly on historical text data, news and social media, and focused on low-resource corpora and cross-lingual semantic change. Despite moving to other topics with my successive post-docs, I’m still active in the sematic change community, co-organizing the LChange workshop on Historical Language Change since two years.

News:

• 01/01/2023: Started working as a post-doctoral researcher in EPFL’s NLP group.
• 01/06/2022: Started my contract as an invited researcher at Jozef Stefan Institute.
• 29/05/2022: Attented ACL 2022 in Dublin, Ireland, where I was co-organizing the LChange’22 workshop.
• 09/12/2021: Gave seminar at INRIA Lille on Models of diachronic semantic change.
• 06/10/2021: Gave seminar at Josef Stefan Institute, Ljubjana (online) on Detecting Omissions of Risk Factors in Company Annual Reports.
• 23/06/2021: Gave seminar at Hong Kong Polytechnic University (online) on Models of diachronic semantic change.