When it comes to machine learning, Bastian is interested in all things geometrical and topological.

Julius is particularly interested in the development and investigation of robust topological representation learning procedures. This includes the construction of new feature extraction algorithms, as well as the investigation of limitations and performance of these techniques.

Jeremy is interested developing practical and usable topology-based methods for improving our understanding of various generative models. Additionally, he enjoys applying topological methods in collaborative projects that focus on predictive medicine and/or climate change.

Katharina is interested in using and developing topological machine learning methods to analyse complex structures in biomedical data, in particular single cell data.

Ernst is interested in building scalable topological descriptors that can be used as machine learning layers. Furthermore, he is interested in the application of abstract algebra and sheaves to understand data structures and machine learning algorithms.

Salome is interested in unsupervised, self-supervised, and weakly supervised deep models, focusing on trustworthy representation and manifold learning to advance AI in health applications.

Daniel is interested in gaining a deeper understanding in the intrinsic mechanics of machine learning models, with a particular focus on statistical learning theory and symbolic deep learning. He also has a keen interest in building topological datasets to study the expressivity of topological deep learning models and graph neural networks. In his spare time he enjoys creating music and its intersection with generative AI.

Rubén is interested in developing the foundations of topological and categorical deep learning models. On the topological side, Rubén enjoys working with models that leverage the topological and combinatorial structures of the input. On the categorical side, Rubén is especially interested in studying the fundamental blocks of deep learning by means of category theory and in developing logic and symbolic neural networks backed by categorical methods.

Emily recently received her Bachelor’s degree in mathematics from Bowdoin College in Maine, USA, and is originally from the Washington, D.C. area. Funded by a Fulbright Study/Research grant, she will be joining the AIDOS Lab as a visiting student until July 2025, during which she will be based in Munich. Outside the office, Emily enjoys spending her time hiking and rock climbing, solving and writing crossword puzzles, cooking (and eating) good food, and improving her German. She is always keen to play some soccer and is hoping to pick up the flute again.

Richard is interested in topological and geometric ideas to understand high dimensional data and seeks to bring these ideas into action, forming methods to understand data. He’s in particular excited in applications in machine learning, improving and making learning more interpretable.

Mathieu seeks to extend machine learning to topological and combinatorial domains, including manifolds, graphs, and simplicial complexes. This exciting adventure inspires him to explore tools in disciplines as diverse as differential geometry, algebraic topology, quantum mechanics, differential equations, and signal processing. Currently, his research focuses on developing Gaussian processes that leverage the richness of non-Euclidean data.