I'm a machine learning researcher with a PhD from ETH Zurich, focused on multimodal generative models, self-supervised representation learning, and foundation models for real-world applications in healthcare.
Most recently, I’ve been working on scalable multimodal learning as part of the Swiss AI initiative, and developing new approaches to representation learning across healthcare and biology datasets.
- Multimodal & cross-modal generative modeling
- Vision-language and structure-function learning
- Probabilistic modeling and self-supervised representation learning
- Applications in healthcare, computational biology, and scientific discovery
A scalable VAE for heterogeneous data sources — improves representation learning across modalities. Featured at NeurIPS 2024.
Training and evaluating large-scale representation models across clinical, imaging, and omics data. Focus on missing data robustness and downstream task generalization.
A probabilistic approach to modeling group structure and importance. Published at ICLR 2023 (Spotlight) and NeurIPS 2023.
- Unity by Diversity: Improved Representation Learning in Multimodal VAEs – NeurIPS 2024
- Learning Group Importance using the Differentiable Hypergeometric Distribution – ICLR 2023 (Spotlight)
- Generalized Multimodal ELBO – ICLR 2021
- Full publication list →
Languages: Python, C++, R, Matlab
Frameworks: PyTorch, TensorFlow, Hugging Face, Scikit-learn, OpenCV, Pandas
Tools: Weights & Biases, Docker, HPC, Git
Other: Probabilistic modeling, multimodal generative models, multimodal representation learning, vision-language architectures
- Generative AI for Good, Panelist – ICLR 2024
- Unity by Diversity: VAEs in Biomedicine – UCSF, Abbasi Lab
- Scalable Multimodal VAEs – Freiburg Young Scientist AI Network
Thanks for stopping by!