Reproduce “Graph Cascades: Contagion-Based Mesoscopic Rewiring for Structure-Aware Graph Machine Learning” 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
Meher Chaitanya, My Le, Luana Ruiz
Graph Cascades rewires graph neighborhoods by repeatedly activating nodes with support from active neighbors, then linking frequently co-activated pairs with weighted edges. It supplies this sparse intermediate-scale structure to existing graph models. On Chameleon, VCR-Graphormer reaches 60.02% ± 2.87 with MAS rewiring, versus 27.28% ± 2.51 on the original graph, averaged over 20 splits. This result is benchmark- and backbone-specific, not a universal gain.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.