My name is Dieuwke, I am a PhD student at the Institute for Logic, Language and Computation, under supervision of Willem Zuidema.
I am employed within the Language in Interaction project since July 2015.
I started my journey in academia studying physics, but at some point I had to admit that although I think physics is very interesting and I love abstract systems, my true passion is really the beautifully complex system that is natural language. While doing a master in Logic at the University of Amsterdam I was very lucky to discover the field of computational linguistics (and in particular syntactic parsing). I am still very glad that I got the chance to have a job that allows me to try to find answers to the questions that intrigue me most.
Currently, in my PhD, I focus on understanding how recurrent neural networks can understand and learn the structures that occur in natural language, for me a problem that touches on the core of understanding natural language.
Although artificial neural networks are of course nothing like the real brain, I hope that understanding the principles by which they can encode processes can still teach us something that will lead to a better understanding of language.
In my free time, I spend a lot of time training for pole dancing competitions with my duo partner Suzy. In fact, last year (2018), we won the national championships! If you are not convinced yet that it is a serious sport, have a look at our Instagram, or check our performance at the World’s in 2018!
November 15, 2019 – Today I gave a presentation at the University of Utrecht. Thanks Denis Paperno for inviting me, and thanks to the audience for asking so many interesting questions!
November 4, 2019 – My paper with Jaap Jumelet and Willem Zuidema got an honourable mention at CoNLL2019!
October 25, 2019 – There will be a next edition of BlackboxNLP next year at EMNLP’2020!
October 8, 2019 – On my way to Baltimore for a visit to John’s Hopkins! Thanks to Tal Linzen and Chris Honey for inviting me :).
October 4, 2019 – Together with Willem Zuidema, I’m interviewing Alexandru Baltag today at the ILLC open Day!
October 1, 2019 – I decided to make a news section on my homepage! First piece of news: Today I’m speaking at the Computational Cognition workshop in Instabruck!
Hupkes D., Dankers V., Mul M. and Bruni E. (2019) The compositionality of neural networks: integrating symbolism and connectionism.. Submitted.
Giulianelli, M., Harding, J., Mohnert, F., Hupkes, D. and Zuidema, W. (2018). Under the hood: using diagnostic classifiers to investigate and improve how language models track agreement information. BlackboxNLP 2018, ACL.
Hupkes D., Veldhoen S., and Zuidema W. (2018). Visualisation and ‘diagnostic classifiers’ reveal how recurrent and recursive neural networks process hierarchical structure. JAIR.