用自回归方法快速生成精确晶体结构，支持多种材料属性条件控制。
Materium: An Autoregressive Approach for Material Generation
- 将晶体结构转为符号序列，通过自回归模型逐点生成原子位置。
- 单卡几小时训练完成，生成速度远超扩散模型，支持GPU/CPU加速。
- 可按密度、能隙、磁性等属性精准控制生成，适配材料设计需求。
我们提出Materium：一种用于生成晶体结构的自回归Transformer模型，将三维材料表示转化为包含元素氧化态、分数坐标和晶格参数的符号序列。与需要多步去噪迭代的扩散方法不同，Materium直接在精确的分数坐标上放置原子，实现快速、可扩展的生成。该设计使得模型可在单个GPU上数小时内完成训练，并在GPU和CPU上比基于扩散的方法生成更快。模型使用多种性质作为条件进行训练与评估，包括密度、空间群等基础属性，以及能隙和磁密度等实际目标。在单一或组合条件下，模型均表现稳定，生成结果与设定条件高度一致。
原文摘要 · Abstract (English)
We present Materium: an autoregressive transformer for generating crystal structures that converts 3D material representations into token sequences. These sequences include elements with oxidation states, fractional coordinates and lattice parameters. Unlike diffusion approaches, which refine atomic positions iteratively through many denoising steps, Materium places atoms at precise fractional coordinates, enabling fast, scalable generation. With this design, the model can be trained in a few hours on a single GPU and generate samples much faster on GPUs and CPUs than diffusion-based approaches. The model was trained and evaluated using multiple properties as conditions, including fundamental properties, such as density and space group, as well as more practical targets, such as band gap and magnetic density. In both single and combined conditions, the model performs consistently well, producing candidates that align with the requested inputs.
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。