Reproduce “Spectral graph clustering with inhomogeneous latent geometry” 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
Konstantin Avrachenkov, Lucas S. Sibemberg, Alexander Van Werde
DBSPEC combines several eigenvectors near an approximately located community signal and clusters node embeddings using density-based clustering. In a simulated soft geometric block model on a circle with 5,000 nodes, it achieves 100% community-recovery accuracy, compared with 64% for sign-clustering on one eigenvector. The authors’ guarantee is almost-exact recovery for a two-community model under assumptions, not arbitrary networks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.