T. Konstantin Rusch

Postdoctoral researcher at MIT.

My name is T. Konstantin Rusch and I’m a postdoctoral researcher at CSAIL, MIT working with Daniela Rus. Before that I finished a PhD in Machine Learning / Applied Mathematics at ETH Zurich, where I was lucky enough to get advised by Siddhartha Mishra. Moreover, during my PhD studies I had a second affiliation at the EECS Department at UC Berkeley, advised by Michael Mahoney.

My main research interest is in combining physics with machine learning. Thereby, I focus on physics-inspired machine learning, which can be described as leveraging structure from physical systems to construct novel machine learning methods with better inductive biases. Additionally, I work on combining methods from numerical analysis with machine learning to solve problems in computational science and engineering.

Contact: tkrusch [at] mit.edu


May, 2024 New preprint (joint work with UWaterloo, and Oxford): Message-Passing Monte Carlo: Generating low-discrepancy point sets via Graph Neural Networks
Feb, 2024 Paper accepted at TMLR (joint work with Cambridge, DeepMind, Oxford, and MILA): How does over-squashing affect the power of GNNs?
Sep, 2023 Paper accepted at NeurIPS 2023 (joint work with Caltech): Neural Oscillators are Universal
Sep, 2023 I successfully defended my PhD thesis. I am thankful to my advisor Siddhartha Mishra and my examiners Max Welling and Michael Mahoney.
Jan, 2023 Paper accepted at ICLR 2023 (joint work with Berkeley and Oxford): Gradient Gating for Deep Multi-Rate Learning on Graphs