Michael Beukman

PhD Student | RL Researcher

Hi there! I'm a third year PhD student at the University of Oxford, supervised by Jakob Foerster and I've just finished a student researchership with Clare Lyle at Google DeepMind. Previously I was an MSc student in the RAIL lab at the University of the Witwatersrand, South Africa, advised by Benjamin Rosman, Steven James, Devon Jarvis and Richard Klein.

I'm currently interested in open-ended reinforcement learning (RL), including unsupervised environment design, which helps to make RL agents be more robust and learn faster by adaptively proposing environments to train on. If you are interested, have a look at some of my publications, projects or get in touch!