He is interested in making large-scale machine learning models safe and robust. Specifically, he works towards evaluating and enhancing the robustness of classification, segmentation, and vision-language models. Previously, he completed his M.Sc in Machine Learning at the Universtity of Tübingen.