New GNNs? On Tuesday in the #GraphML reading group James Rowbottom and @b_p_chamberlain present their "GRAND: Graph Neural Diffusion" + their #NeurIPS2021 paper "Beltrami Flow and Neural Diffusion on Graphs" out of @mmbronstein's group @Twitter! Zoom: 1/2 Tweet added by Hannes Stärk @HannesStaerk

Hannes Stärk

3 years

New GNNs? On Tuesday in the #GraphML reading group James Rowbottom and @b_p_chamberlain present their "GRAND: Graph Neural Diffusion" + their #NeurIPS2021 paper "Beltrami Flow and Neural Diffusion on Graphs" out of @mmbronstein 's group @Twitter ! Zoom: 1/2

155

Replies

Hannes Stärk

@HannesStaerk

3 years

GRAND is a new GNN based on discretizations of a learned diffusion equation on graphs: BLEND: Over-smoothing is solved? Join the discussion! @migorinova @stefan_webb @emaros96 @aittalam Francesco Giovanni @epomqo @mmbronstein

Beltrami Flow and Neural Diffusion on Graphs

We propose a novel class of graph neural networks based on the discretised Beltrami flow, a non-Euclidean diffusion PDE. In our model, node features are supplemented with positional encodings...

arxiv.org