Unconventional Crystal
Unconventional crystals, possessing unusual structures or properties, are a focus of intense research aimed at predicting and designing new materials. Current efforts leverage advanced machine learning models, including transformer-based networks and generative models like variational autoencoders and diffusion models, often incorporating geometrically equivariant neural networks to capture the complex symmetries and long-range interactions within these materials. These computational approaches are improving the ability to predict material properties and even design novel crystal structures with desired characteristics, accelerating materials discovery and impacting diverse fields from energy to medicine.
Papers
October 30, 2024
July 23, 2024
May 28, 2024
December 20, 2023