Academic staff

PhD position at TUM Medical AI Lab to advance medical deep learning foundations and develop foundation-model architectures for precision medicine. You will explore self-supervised and multimodal representations, focusing on information content, geometry, and learning from missing data. The role combines theory and large-scale experiments in a well-supported project with access to clinical data and international collaboration. You will contribute to top ML/vision venues and medical journals, shaping next-generation medical AI.

• fully funded position (TV-L E13)
• international conference participation
• access to large-scale clinical datasets
• high-end GPU and storage infrastructure
• international collaborations
• embedded at TUM University Hospital

• Develop new learning objectives and architectures for medical deep learning
• Investigate information content and geometry of learned representations
• Work on multimodal representation learning and information decomposition
• Address learning with missing, partial, and irregular observations
• Explore temporal and longitudinal representation learning
• Contribute to method development and large-scale experiments for high-impact publications
• Collaborate with the TUM University Hospital and international partners

• Bachelor and Master in computer science, machine learning, mathematics, physics, engineering, or related field
• Strong foundations in machine learning and deep learning
• Experience with PyTorch or similar frameworks
• Interest in representation learning, self-supervised learning, multimodal learning, foundation models, or ML theory
• Strong programming and analytical skills
• Excellent written and spoken English

• PyTorch
• Representation learning
• Self-supervised learning
• Multimodal learning
• Foundation models
• ML theory