John Doe
PhD student at Example University. I work on graph learning and topological deep learning: neural networks for data where relations join more than two things at once.
82 edges51 triangles
Research
Most machine learning on graphs treats relations as pairs. Many systems are not built from pairs: a chemical reaction joins several molecules, a coauthored paper joins several people, a mesh face joins several vertices. My work asks how neural networks can use that higher-order structure directly, on simplicial complexes, cell complexes, and sheaves.
Example statement. Replace with your own.
Selected publications
2026
Example title: Higher-order message passing on cell complexes
Example Venue 2026
Example title: sheaf diffusion with learned restriction maps
Example Workshop 2026