New SymADiT model generates symmetric crystals using latent flow matching

Researchers have developed SymADiT, a new variant of the All-atom Diffusion Transformer (ADiT) designed for crystal generation. This symmetry-aware model operates in latent space and enforces symmetry restrictions based on the crystal's space group and atomic Wyckoff positions. Benchmarking shows SymADiT generates stable, symmetric, and physically plausible crystals, performing competitively with existing models using a standard Transformer architecture. AI

IMPACT This research advances generative modeling techniques for materials science, potentially accelerating the discovery of new stable crystalline structures.

RANK_REASON The cluster describes a new scientific paper detailing a novel model for crystal generation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• All-atom Diffusion Transformer (ADiT)
• alphaXiv
• Anmar Karmush
• arXiv
• CatalyzeX
• DagsHub
• Gotit.pub
• Hugging Face
• ScienceCast
• SymADiT
• transformer
• Wyckoff positions

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →