Reproduce “CFL-aware edge design for graph neural solvers of hyperbolic transport” 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
Jinyoung Jeong, Joseph B. Choi, Xinlun Cheng, H. S. Udaykumar, Sanghun Choi, Stephen S. Baek
The authors test whether edge-local wave-travel information improves graph neural network predictions of shocks at larger time steps. They compare raw, geometry-organized, and flux-organized edge features, with and without the Courant–Friedrichs–Lewy (CFL) scale. In a Burgers benchmark, Geometry plus absolute CFL lowers final-step error by 10.4% versus its no-CFL counterpart, across six settings; this does not establish general stability.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.