Publications

On Over-Squashing in Message Passing Neural Networks: The Impact of Width, Depth, and Topology [PDF]
Francesco Di Giovanni, Lorenzo Giusti, Federico Barbero, Giulia Luise, Pietro Lio, Michael Bronstein
Preprint, Feb. 2023

Graph Neural Networks as Gradient Flows [PDF]
Francesco Di Giovanni*, James Rowbottom*, Benjamin Chamberlain, Thomas Markovich, Michael Bronstein
Preprint, Jun. 2022

Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs [PDF] [Blog] [Video by Aleksa Gordić]
Cristian Bodnar, Francesco Di Giovanni, Benjamin Chamberlain, Pietro Liò, Michael Bronstein
NeurIPS2022

Heterogeneous manifolds for curvature-aware graph embedding [PDF]
Francesco Di Giovanni*, Giulia Luise*, Michael Bronstein
Preprint, Feb. 2022

Understanding over-squashing and bottlenecks on graphs via curvature [PDF] [Blog] [Video by Aleksa Gordić]
Jake Topping*, Francesco Di Giovanni*, Benjamin Chamberlain, Xiaowen Dong, Michael Bronstein
Outstanding paper Honorable mention , ICLR 2022 (oral)