用简单统一的Transformer生成晶体结构，效果超越复杂模型。
CrystalDiT: A Diffusion Transformer for Crystal Generation
- 用统一Transformer同时建模晶格与原子属性，引入强先验知识
- 在MP-20数据集上达到8.78%的稳定、唯一、新颖结构率
- 适合材料发现领域研究者，尤其关注架构简洁性与泛化能力
我们提出CrystalDiT，一种用于晶体结构生成的扩散Transformer，通过挑战架构复杂化的趋势，实现了当前最优性能。不同于复杂的多流设计，CrystalDiT采用统一的Transformer架构，将晶格与原子属性视为一个相互依赖的整体，引入强大的归纳偏置。结合基于元素周期表的原子表示和平衡训练策略，在MP-20数据集上达到8.78%的SUN（稳定、唯一、新颖）率，显著优于FlowMM（4.21%）和MatterGen（3.66%）。值得注意的是，CrystalDiT生成了63.28%的唯一且新颖结构，同时保持相近的稳定性，表明在数据有限的科学领域中，精心设计的简单架构比易过拟合的复杂模型更有效。
原文摘要 · Abstract (English)
We present CrystalDiT, a diffusion transformer for crystal structure generation that achieves state-of-the-art performance by challenging the trend of architectural complexity. Instead of intricate, multi-stream designs, CrystalDiT employs a unified transformer that imposes a powerful inductive bias: treating lattice and atomic properties as a single, interdependent system. Combined with a periodic table-based atomic representation and a balanced training strategy, our approach achieves 8.78% SUN (Stable, Unique, Novel) rate on MP-20, substantially outperforming recent methods including FlowMM (4.21%) and MatterGen (3.66%). Notably, CrystalDiT generates 63.28% unique and novel structures while maintaining comparable stability rates, demonstrating that architectural simplicity can be more effective than complexity for materials discovery. Our results suggest that in data-limited scientific domains, carefully designed simple architectures outperform sophisticated alternatives that are prone to overfitting.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。