Welcome to the I-Machine-Think website

We are a research group at the University of Amsterdam.

In our group, we consider different research topics, including but not limited to Compositionality, interpretability of neural networks, neural language models and language emergence in referential games. We collaborate with several people from both other universities and industry. Are you a master student and looking for a research project? Have a look at our publications page, check Elia Bruni’s or Dieuwke Hupkes’ personal websites, or get in touch via email!

Also, have a look at our interpretability library for neural networks: https://github.com/i-machine-think/diagnnose!

News

April 20, 2021

Thrilled to announce that Gautier Dagan accepted a PhD position in Edinburgh! He will work under the supervision of Alex Lascarides. Congratulations Gautier!

April 17, 2021

Congratulations to Tom, Hugh Mee, Jaap and Dieuwke for having their paper accepted at the NAACL workshop DeeLIO! Check the preprint here>.

January 11, 2021

Two imt papers accepted at EACL! Congrats to Lucas, Jaap, Gautier, Elia and Dieuwke. For preprints, check the publications page.

November 17, 2020

This is probably the strangest conference presentation ever done in imt. Check the photo in the 'all news' section.

September 15, 2020

Two imt papers accepted for EMNLP!

  • The grammar of emergent languages
  • Internal and External Pressures on Language Emergence: Least effort, Object Constancy and Frequency

August 29, 2020

An article about Dieuwke's work appeared in a national Dutch newspaper. You can read it here (in Dutch).

... see all News