发表机构
The University of Texas at Austin; Lawrence Livermore National Laboratory（德克萨斯大学奥斯汀分校; 劳伦斯利弗莫尔国家实验室）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出Mosaic混合网格-粒子表示，通过锚定区域和原子槽位提供局部空间寻址，支持条件生成、层次化组装及百万原子级结构生成，实现跨尺度可操控原子生成。
AI 中文摘要
生成模型正在为分子发现和材料科学开辟新的可能性。然而，要实现这一潜力，既需要对原子放置进行局部控制，也需要能够组装大型异质结构。我们引入了Mosaic，一种混合网格-粒子表示方法，它将空间划分为包含原子槽位的锚定区域。每个槽位编码了占用状态、原子种类以及相对于其锚点的连续位移，从而为局部条件生成、层次化生成和可扩展组装提供了持久空间地址。由于Mosaic定义了一种表示而非模型，它可以与不同的生成骨干网络配对使用。在保持分子骨干网络固定的情况下，其局部支持和坐标先验改善了分子连接性和有效性，所得模型能够高效地跨分子尺寸采样，并实现细粒度的空间控制。在更大的材料尺度上，相同的表示能够基于粗略的空间布局生成原子级多晶结构。最值得注意的是，局部性将大规模生成转化为一个有界的组装问题，从而能够生成包含超过一百万个原子的结构。总之，这确立了局部空间寻址作为跨尺度物质可操控生成的实际基础，从单个分子到具有更广泛工程意义的大型结构。
英文摘要
Generative models are opening new possibilities in molecular discovery and materials science. Realizing this potential, however, requires both local control over atomic placement and the ability to assemble large, heterogeneous structures. We introduce Mosaic, a hybrid mesh-particle representation that partitions space into anchored regions containing atom slots. Each slot encodes occupancy, atomic species, and a continuous displacement from its anchor, providing persistent spatial addresses for local conditioning, hierarchical generation, and scalable assembly. Because Mosaic defines a representation rather than a model, it can be paired with different generative backbones. With the molecular backbone held fixed, its local support and coordinate prior improve molecular connectivity and validity, the resulting model samples efficiently across molecular sizes and enables fine-grained spatial control. At larger material scales, the same representation enables generation of atomistic polycrystals conditioned on coarse spatial layouts. Most notably, locality turns large-scale generation into a bounded assembly problem, enabling structure generation with more than one million atoms. Together, this establishes local spatial addressing as a practical foundation for steerable generation of matter across scales, from individual molecules to large-scale structures of broader engineering significance.
