Felix Draxler

Felix Draxler profile image

I am a Computer Science & Math PhD student at Heidelberg University.
My main interest are generative models, in particular methods that are flexible, efficient and high-quality.

Supervisors: Ullrich Köthe and Christoph Schnörr
E-Mail: felix.draxler at iwr.uni-heidelberg.de
CV, LinkedIn, Google Scholar, GitHub

Publications

Current Preprints

  • On the Universality of Coupling-based Normalizing Flows
 (preprint 2024)
    Felix Draxler, Stefan Wahl, Christoph Schnörr, Ullrich Köthe
  • Learning Distributions on Manifolds with Free-form Flows (preprint 2023)
    
Peter Sorrenson*, Felix Draxler*, Armand Rousselot*, Sander Hummerich, Ullrich Köthe

Selected Publications

  • Free-form Flows: Make Any Architecture a Normalizing Flow (AISTATS 2024)
    Felix Draxler*, Peter Sorrenson*, Lea Zimmermann, Armand Rousselot, Ullrich Köthe
  • Whitening Convergence Rate of Coupling-based Normalizing Flows (NeurIPS 2022) Oral (top 1.8%), Scholar Award
    Felix Draxler, Christoph Schnörr, Ullrich Köthe
  • Characterizing The Role of A Single Coupling Layer in Affine Normalizing Flows (GCPR 2020) Best Paper Honorable Mention (top 3)
    Felix Draxler, Jonathan Schwarz, Christoph Schnörr, Ullrich Köthe
  • Essentially No Barriers in Neural Network Energy Landscape (ICML 2018) Long Oral (top 8.6%)
    Felix Draxler, Kambis Veschgini, Manfred Salmhofer, Fred Hamprecht
  • Software library FrEIA (2018-2023)
    Package to easily build invertible neureal networks with PyTorch
    Lynton Ardizzone, Till Bungert, Felix Draxler, Ullrich Köthe, Jakob Kruse, Robert Schmier, Peter Sorrenson

See Google Scholar for full list.

Education

  • Ph.D. Computer Science & Mathematics, Heidelberg University, Germany.
    • Estimated graduation: Summer 2024
    • Supervisors: Ullrich Köthe, Christoph Schnörr
  • M.Sc. Physics, Heidelberg University, Germany. 2018
    • Loss Landscape of Neural Networks
    • Supervisors: Fred Hamprecht, Manfred Salmhofer
    • Scholarship Max Weber-Programm by the State of Bavaria
  • B.Sc. Physics, LMU Munich. 2015
    • Thesis: „Evolutionary Optimization of a Cooling Sequence for the Generation of Ultracold Atoms“
    • Supervisors: Ulrich Schneider and Immanuel Bloch 
    • Semester abroad: winter 2015/16: INP Grenoble, France
    • Scholarship Max Weber-Programm by the State of Bavaria

Funding

Thankfully acknowledge funding by vector Stiftung (TRINN) and STRUCTURES.