Christian Koke

Christian Koke is a PhD student and AI Researcher at the Technical University of Munich (TUM), affiliated with the Computer Vision Group under Informatics 9 (Department of Informatics). He is jointly supervised by Prof. Daniel Cremers (TUM) and Prof. Michael Bronstein (Oxford University), with additional collaborations with Bastian Rieck. His academic background includes M.Sc. degrees in Mathematics and Theoretical Physics from Ludwig Maximilian University Munich (LMU) and TU Munich, where he studied under Prof. Gitta Kutyniok on graph neural networks.

Education:
• PhD in Computer Science (ongoing), TUM
• M.S. in Mathematics, Ludwig Maximilian University Munich
• M.S. in Theoretical Physics, LMU & TUM
• B.S. in Physics, Heidelberg University

His research focuses on the mathematical foundations of machine learning, particularly in graph neural networks, directed graphs, and spectral methods. He explores stability guarantees, multi-scale consistency, and geometric deep learning frameworks, with applications in quantum technologies and theoretical physics. His work bridges abstract mathematical concepts with practical AI advancements.

Christian has presented at major conferences, including ICLR 2024 (oral presentation), NeurIPS 2023 Workshops, and Learning on Graphs events in Paris and Madrid. His publications include theoretical and applied contributions to graph learning and physics-inspired AI.

Awards:
• Outstanding Extended Abstract, ICML 2024 Workshop on Geometry-grounded Representation Learning

He is an ELLIS PhD Student and collaborates with labs at TUM, Oxford, and Ecole Polytechnique. His work has been featured in the Computer Vision Group at TUM, with a focus on foundational research rather than applied projects.