Reproduce “CrystalAR: Autoregressive Modeling for Symmetric Crystal Generation” 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

Ziqiao Meng, Rongsheng Chen, Zhi Zheng, Nikita Kazeev, Zhenkun Wang, Andrey Ustyuzhanin, Wee Sun Lee

CrystalAR jointly generates crystal symmetry templates and their geometric realizations, first selecting element-decorated Wyckoff sites and a compatible lattice, then refining site coordinates through symmetry-constrained spatial subdivisions. On MP-20, it produced 161 novel unique templates and a 0.28% S.S.U.N. rate, versus 215 and 0.09% for SymmCD; highly novel templates showed reduced stability.

Summary by alphaXiv.

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