Understanding over-squashing and bottlenecks on graphs via curvature J Topping*, F Di Giovanni*, BP Chamberlain, X Dong, MM Bronstein
ICLR 2022, outstanding paper honorable mention, 2021
511 2021 Neural sheaf diffusion: A topological perspective on heterophily and oversmoothing in gnns C Bodnar, F Di Giovanni, BP Chamberlain, P Liò, MM Bronstein
NeurIPS 2022, 2022
189 2022 On Over-Squashing in Message Passing Neural Networks: The Impact of Width, Depth, and Topology F Di Giovanni, L Giusti, F Barbero, G Luise, P Lio, M Bronstein
ICML23, 2023
134 2023 Understanding convolution on graphs via energies F Di Giovanni, J Rowbottom, BP Chamberlain, T Markovich, MM Bronstein
Transactions on Machine Learning Research, 2022
97 * 2022 Beltrami flow and neural diffusion on graphs B Chamberlain, J Rowbottom, D Eynard, F Di Giovanni, X Dong, ...
Advances in Neural Information Processing Systems 34, 1594-1609, 2021
91 2021 Drew: Dynamically rewired message passing with delay B Gutteridge, X Dong, MM Bronstein, F Di Giovanni
International Conference on Machine Learning, 12252-12267, 2023
71 2023 Edge directionality improves learning on heterophilic graphs E Rossi, B Charpentier, F Di Giovanni, F Frasca, S Günnemann, ...
Learning on Graphs Conference, 25: 1-25: 27, 2024
69 2024 How does over-squashing affect the power of GNNs? F Di Giovanni, TK Rusch, MM Bronstein, A Deac, M Lackenby, S Mishra, ...
Transactions on Machine Learning Research, 2024
35 * 2024 Heterogeneous manifolds for curvature-aware graph embedding F Di Giovanni*, G Luise*, M Bronstein
ICLR 2022 GTRL Workshop, 2022
29 2022 Locality-Aware Graph-Rewiring in GNNs F Barbero, A Velingker, A Saberi, M Bronstein, F Di Giovanni
ICLR 2024, 2024
22 2024 Rotationally symmetric Ricci flow on F Di Giovanni
Advances in Mathematics 381, 2019
8 * 2019 Metric flow matching for smooth interpolations on the data manifold K Kapuśniak, P Potaptchik, T Reu, L Zhang, A Tong, M Bronstein, AJ Bose, ...
arXiv preprint arXiv:2405.14780, 2024
5 2024 Ricci flow of warped Berger metrics on R4 F Di Giovanni
Calculus of Variations and Partial Differential Equations 59 (162), 2019
5 2019 Understanding Virtual Nodes: Oversmoothing, Oversquashing, and Node Heterogeneity J Southern*, F Di Giovanni*, M Bronstein, JF Lutzeyer
arXiv preprint arXiv:2405.13526, 2024
2 2024 Advancing Graph Convolutional Networks via General Spectral Wavelets N Liu, X He, T Laurent, F Di Giovanni, MM Bronstein, X Bresson
arXiv preprint arXiv:2405.13806, 2024
2 2024 Can strong structural encoding reduce the importance of Message Passing? F Eijkelboom, EJ Bekkers, MM Bronstein, F Di Giovanni
Topological, Algebraic and Geometric Learning Workshops 2023, 278-288, 2023
1 2023 Over-squashing and over-smoothing through the lenses of curvature and multi-particle dynamics F Di Giovanni
1 2022 Graph neural diffusion B Chamberlain, C Bodnar, F Di Giovanni, MM Bronstein
US Patent App. 17/650,219, 2022
1 2022 Convergence of Ricci flow solutions to Taub-NUT F Di Giovanni
Communications in Partial Differential Equations 46 (8), 1521-1568, 2021
1 2021 Relaxed Equivariance via Multitask Learning AA Elhag, TK Rusch, F Di Giovanni, M Bronstein
arXiv preprint arXiv:2410.17878, 2024
2024