Reproduce “Learning Spectrally Optimised Mesh-Free Discretisations” 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
Lucas Gerken Starepravo, Henry Broadley, Steven Lind, Jack R. C. King
SpeND constructs mesh-free differential operators by predicting stencil weights from local point geometry and projecting them onto exact polynomial moment constraints. It trains the remaining weight freedom against Fourier-wave responses over a chosen band, without reference solution data. In tests, SpeND reaches matched error faster than RBF-FD on a Poisson problem with a circular obstacle; evidence is limited to these settings.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.