About

Welcome to my website! I’m a PhD student in the Machine Learning group of Matthias Hein at the University of Tübingen and a member of the ELLIS-program, with Gergely Neu as my co-supervisor.

My research interests lie broadly in the area of safe and trustworthy Machine Learning: Neural Networks, despite being tremendously successful at a plethora of tasks, are known to make overly confident predictons. This is, they are unable to communicate low confidence when confronted with data that is different from the data seen during training. Designing safe systems capable of flagging inputs they do not know how to process properly, is therefore crucial for safety-critical areas, like e.g. the medical domain. My work typically includes topics from out-of-distribution detection and adversarial robustness, but I also like to think about more general problems, like e.g. generalization and its relation to loss surface properties. More recently, I got interested in LLMs and VLMs and their robustness, and contributed to various red-teaming efforts for OpenAI.

Before the start of my PhD in Tübingen, I received a M.Sc. in Physics from LMU Munich. In my Masters project, I worked with Bayesian Neural Networks for turbulent fluid simulations at the Max Planck Institute for Plasma Physics. I also hold a M.Sc. in Data Science from the Barcelona GSE, where I worked on computational methods for scalable Bayesian Inference. You can find my CV here.