CrystalDiT: A Diffusion Transformer for Crystal Generation
Xiaohan Yi, Guikun Xu, Xi Xiao, Zhong Zhang, Liu Liu, Yatao Bian, Peilin Zhao · 2025
Read CrystalDiT: A Diffusion Transformer for Crystal Generation by Xiaohan Yi, Guikun Xu, Xi Xiao, Zhong Zhang, Liu Liu, Yatao Bian, Peilin Zhao (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-16614.txt.
Description
Yi and co-authors present CrystalDiT, a diffusion transformer that generates new crystal structures with a deliberately simple design. Recent generators have grown into intricate multi-stream systems. This one instead treats the lattice and the atoms as a single interdependent input to one unified transformer, represents atoms through the periodic table, and balances its training data. The model was tested on the standard MP-20 benchmark for materials generation.
It produced stable, unique, and novel crystals at a rate of 8.78 percent, well above two recent rivals at about 4 and 3.7 percent. Nearly two-thirds of its outputs were both unique and novel, while stability held up. The authors argue that for data-limited science problems, a simple architecture with the right built-in assumptions beats complex designs that overfit.
Catalog record
| CREATOR | Xiaohan Yi, Guikun Xu, Xi Xiao, Zhong Zhang, Liu Liu, Yatao Bian, Peilin Zhao |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-16614