I am pursuing my MSc in Machine Learning at the University of Tübingen. Currently, I am working on a research project about the loss landscape of CLIP models during fine-tuning, supervised by Ankit Sonthalia, Arnas Uselis and Seong Joon Oh. I am interested in parameter spaces, particularly in how connecting multiple models in the parameter space can improve in- and out-of-distribution generalization.

Before coming to Tübingen, I received my BSc in Computer Science from the Karlsruhe Institute of Technology (KIT). I was a research assistant at the Fraunhofer Institute, working on anomaly detection with Variational Autoencoders. Currently, I am also a research assistant at the Mathematical Methods in Computer Science group, working on merging neural networks.