Dieuwke Hupkes

Research Scientist

I am a Research Scientist at Facebook AI Research, based in Paris. Before I joined Facebook, I obtained my PhD in computational linguistics at the University of Amsterdam (UvA), at the Institute for Logic, Language and Information. I also worked briefly at a postdoc at the UvA, where I was also the scientific manager of the Amsterdam unit of ELLIS. I have a masters degree in logic and a bachelor degree in (astro)physics.
My main research goal is to understand more about the beautifully complex system that is language, and how it is possible that humans are so good at it. I am particularly interested in syntax, hierarchy, structure and compositionality. The way that I approach this goal, is by studying neural networks tapping using the vast amount of knowledge we have about language from (psycho/neuro/theoretical)linguistics and philosophy and — in the process — use that knowledge to sharpen our understanding of language (processing). Tasks that I find particularly interesting are machine translation, language modelling and all kinds of versions of natural language understanding.


NLP, interpretability, compositionality, recursion, machine translation, language modelling, grammar, cognitive neuroscience of language, natural language inference, parsing

Latest Publications

CoNLL - November 9, 2021

Generalising to German Plural Noun Classes, from the Perspective of a Recurrent Neural Network

Verna Dankers, Anna Langedijk, Kate McCurdy, Adina Williams, Dieuwke Hupkes

EMNLP - October 1, 2021

Masked Language Modeling and the Distributional Hypothesis: Order Word Matters Pre-training for Little

Koustuv Sinha, Robin Jia, Dieuwke Hupkes, Joelle Pineau, Adina Williams, Douwe Kiela

EACL - April 18, 2021

Co-evolution of language and agents in referential games

Gautier Dagan, Dieuwke Hupkes, Elia Bruni

EACL - April 18, 2021

Language Modelling as a Multi-Task Problem

Lucas Weber, Jaap Jumelet, Elia Bruni, Dieuwke Hupkes

NAACL - June 2, 2019

The emergence of number and syntax units in LSTM language models

Yair Lakretz, Germán Kruszewski, Theo Desbordes, Dieuwke Hupkes, Stanislas Dehaene, Marco Baroni