Flow matching enables accurate prediction of partially occupied crystal structures
CatAstro_Piyush · x · 2026-10-09
A materials science paper introduces a new structure-prediction method: whereas conventional generative models can only handle fully occupied (occupancy = 1) ordered structures, this work represents occupancy as a matrix and learns it with flow matching, accurately predicting partially occupied, disordered structures from composition and site count. An important capability for battery and catalyst materials research.
More from Research
- AI-designed nanoparticles stabilize mRNA vaccines at high temperatures, Nature Biotech reports — emmanuelvivier · 2026-10-09
- LLM agent polarization study from Tsinghua-UChicago team published in Nature Communications — profjamesevans · 2026-10-09
- Forced contact moves LLM agent opinions just 2%, polarization study finds — profjamesevans · 2026-10-09
- LLM agents build human-like social networks — and polarize anyway — profjamesevans · 2026-10-09
- Mathematician downplays Navier-Stokes 'mistranslation' uproar: no doubt about the proof — basedjensen · 2026-10-09
- Prima Mente Launches Alzheimer's AI Challenge With 150M-Cell Perturbation Atlas — katieruthmishra · 2026-10-09