Reproduce “Sheaf-MPNN: A Library for Sheaf Neural Networks as Message Passing” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Alessio Borgi, Mario Severino, Luke Braithwaite, Steve Azzolin, Rémi Bourgerie, Olga Zaghen, Emanuele Mule, Riccardo Ali and 9 more
Sheaf-MPNN expresses sheaf-based graph updates as local message passing rather than assembling a global Laplacian. It transports endpoint features through restriction maps into a shared edge space, computes an edge message there, then pulls it back and aggregates at nodes. The authors derive equivalent formulations for specified architectures, but report no measured speed, memory, or accuracy gains.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.