Published multiple papers at top-tier venues including ICML, ICLR, NeurIPS, and Transactions on Machine Learning Research (TMLR)
Research focuses on flow matching, Gaussian processes, state space models, generative modeling for time series, deep learning with differential equations, and optimal transport
Developed torchode: a parallel ODE solver for PyTorch
Many projects are open-sourced with datasets (e.g., 3D/4D turbulent flow simulation, gesture recognition)
Several papers received Spotlight presentations
Background
Cares about solving problems and improving efficiency with software and machine learning
Values aesthetically pleasing and clean code
Currently a PhD student in the DAML group at Technical University of Munich (TUM), researching machine learning methods
Interested in different programming paradigms offered by various programming languages
Enjoys reading obscure books about programming language theory