Leon Hetzel

Leon Hetzel

PhD Student in Machine Learning and Computational Biology

Helmholtz Munich

TUM

Hi ­čĹő

I pursue my doctoral studies as part of MUDS programme at the Computational Health Center at Helmholtz Munich and the Technical University Munich (TUM), supervised by Fabian Theis and Stephan G├╝nnemann. I am interested in generative models, in particular for graph-structued data, and their application to the sciences, mostly in the domain of single-cell genomics and drug discovery, i.e. small molecules or proteins.

Prior to my PhD, I have interned at the Bosch Center for AI and worked on explanation methods in computer vision and co-authored the paper “Interpretable and fine-grained visual explanations for convolutional neural networks”.

In addition to my academic pursuits, I am passionate about addressing the societal dimensions of AI. I have co-authored the book “Wie Maschinen lernen” to make machine learning accessible to a wide audience and actively engage in teaching computer science to both elementary and high school students through initiatives like KI macht Schule. I firmly believe in the importance of introducing young learners to contemporary computer science topics from an early age.

If you are confused by my GitHub handle: To be anonymous, go generic. In the English language, “John Doe” is about as generic as it gets and the German pendant is “Max Musterman”. Strip away the vowels and you have it: MxMstrmn.

Interests
  • Generative Modelling
  • Single-Cell Genomics
  • Drug Discovery
  • AI for Science
Education & Research
  • PhD in Machine Learning & Computational Biology

    Technical University Munich

    2020 - now

  • Associated PhD student

    Helmholtz Munich

    2020 - now

  • MSc Mathematical Sciences

    University of Oxford

    2018 - 2019

  • MSc Computational Physics

    Heidelberg University

    2018 - 2020

  • Research Scientist Intern

    Bosch Center for AI, Renningen

    Spring 2018

  • BSc Physics

    Bremen University

    2014 - 2017

News

This section will be extended soon

2023

  • 27/10 Our work on flattening statisical manifolds of VAEs got an oral at the “AI4Science” workshop at Neurips'23
  • 15/09 chemCPA was awarded the “Best Paper of the Year 2022” at the MDSI Opening Ceremony
  • 20/08 I presented our “MAGNet” paper at an MDSI student event

2022

  • 01/12 I presented chemCPA at Neurips'22