Reproduce “Evaluating Graph Generative Models with Contrastively Learned Features” 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
Hamed Shirzad, Kaveh Hassani, Danica J. Sutherland
The authors evaluate graph generators by training a Graph Isomorphism Network (GIN) with self-supervised contrastive learning on domain graphs, then comparing generated and held-out real graphs in learned representations. They also test local structural features. On Community graphs under mixing-random, GraphCL with MMD Linear reaches 0.99, versus 0.48 for GIN-Random; this is one controlled perturbation condition, not a universal ranking.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.