Reproduce “Score-based neural ordinary differential equations for reaction-diffusions” 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

Jianguo Hou, Wuchen Li

The authors introduce a deterministic particle method for reaction–diffusion equations. Particles carry position, log-density, score, and physical weight; adaptive training weights emphasize regions predicted to gain mass. In a four-mode WFR test, with sampling fixed, adaptive weights reduce relative mass error from 3.6% to 0.74% and component-share error from 0.049 to 0.013. The theoretical bound does not apply to this example.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.