Dmitry Kobak used to be a group leader in the Department of Data Science at the Hertie Institute for AI in Brain Health at the University of Tübingen, where he developed unsupervised and self-supervised learning methods—especially contrastive learning, manifold learning, and dimensionality reduction—for exploring and visualizing high-dimensional biological data such as single-cell genomics and multimodal brain datasets. His approach combines AI with a strong focus on mechanistic understanding rather than prediction alone. He has also worked on statistical forensics, including analyses of Russian electoral irregularities and COVID-19 excess mortality. Kobak is a member of the ELLIS society and an associated scientist of the International Max Planck Research School for Intelligent Systems (IMPRS-IS). He is now a group leader at VIB.AI, where he applies self-supervised learning to complex genomics, epigenomics, and protein-biology datasets. His work at the Hertie Institute and his connections to Tübingen’s AI ecosystem make him relevant to Cyber Valley’s broader research community in machine learning and data-driven biology.
