Generating Symmetric Materials using Latent Flow Matching
Published in Fortieth Annual Conference on Neural Information Processing Systems (NeurIPS 2026), 2026
Abstract:
Tackling the task of materials generation, we aim to enhance the previously proposed All-atom Diffusion Transformer (ADiT) by introducing SymADiT, a symmetry-aware variant. To do so, we use a representation of materials based on Wyckoff positions. We follow ADiT and perform generative modelling in latent space, adapted to our symmetry-aware representation. By forcing the output of the generative model to adhere to the symmetry restrictions imposed by the generated crystal’s space group and each atom’s Wyckoff-position, the generated materials exhibit more realistic symmetry properties. We benchmark our method against both symmetry-aware and symmetry-agnostic models for materials generation and show competitive performance, generating stable, symmetric materials with a simple Transformer architecture.
Recommended citation: Karmush, A., Brandenburg, C.M., Ershadrad, S., Rosén, J., Felsberg, M., Ekström Kelvinius, F. (2026). Generating Symmetric Materials using Latent Flow Matching Advances in Neural Information Processing Systems, 40