Virtual Talk with Maximilian Nickel

Dr. Maximilian Nickel, Research Scientist at FAIR / Meta AI (Generative) Models of a Complex World


Some of the most intriguing and challenging problems in machine learning and artificial intelligence are related to the modeling of complex systems, i.e., systems which are characterized by a large number of interacting parts and, potentially, non-trivial emergent properties. These systems are central to a wide range of fields including the social sciences, biology, knowledge representation, as well as economics.

In this talk, I will discuss how the combination of deep learning and geometry can be employed to approach some of these challenges. I will first discuss how networks (in the graph sense) are related to the geometry of a representation space in deep learning and how these insights lead to methods such as Poincare Embeddings, Poincare Maps for single-cell analysis, as well as Hyperbolic CLIP.

In the second part of the talk, I will discuss how to extend these methods to generative modeling. For this purpose, I will first discuss Continuous Normalizing Flows and how they can be extended to Riemannian spaces. I will then connect this work to state-of-the-art methods in generative AI such as Flow Matching and its Riemannian variant.

I will conclude the talk with an outlook on future work and how AI for modeling complex systems can make important contributions to the social sciences as well as responsible AI.


Since 2016, I am a research scientist at FAIR, Meta AI where I have also acted as area lead for Machine Learning and Society & Responsible AI. Before joining FAIR, I was a postdoctoral fellow at MIT where I was with the Laboratory for Computational and Statistical Learning and the Center for Brains, Minds and Machines. I have received my PhD with summa cum laude from the Ludwig Maximilian University Munich as a research assistant at Siemens Corporate Technology.

I am interested in modeling and understanding complex systems via artificial intelligence, including networks, dynamics, emergence, and the interplay of AI with the social systems in which they are situated. For this purpose, I’m combining fundamental research on geometric representation learning, neural dynamical systems, and generative AI for complex systems.

Zoom Registration

Please register with to receive the Zoom login.

Back to the event overview