J. Brandstetter
Papers
1
Total Citations
2
H-Index
1
About
Johannes Brandstetter is a leading researcher at the intersection of machine learning and the physical sciences, with a primary focus on geometric deep learning, scientific machine learning, and generative modeling. His most impactful contributions lie in developing neural network architectures that respect the symmetries and invariances of physical systems, particularly through his pioneering work on equivariant message-passing neural networks. Brandstetter has been instrumental in advancing the application of deep learning to complex scientific domains, including molecular dynamics, particle physics, and climate modeling. His research on message-passing neural PDE solvers has garnered significant attention, with several of his papers accumulating hundreds of citations, establishing him as a key figure in the field of neural operators for partial differential equations. Notably, his work on "Minkowski" and "SE(3)-equivariant" models has set new standards for data efficiency and accuracy in modeling three-dimensional physical systems. His recent foray into large action models for robotics, exemplified by "A Large Recurrent Action Model: xLSTM enables Fast Inference for Robotics Tasks," demonstrates his versatility in applying sequence modeling to real-world control problems. Brandstetter's research continues to bridge the gap between fundamental AI theory and practical scientific discovery.
Research Focus
Key Achievements
Top Papers
- 1