合成晶体数据管线降低材料预测扩散预训练对 DFT 弛豫的依赖
Synthetic Crystal Data Generation for Diffusion Pretraining in Material Property Prediction
做材料性质预测的同学可以看看:作者把 MatterGen 生成、protostructure 枚举和 ML 势弛豫拼成一条合成数据管线,7 个 MatBench 任务上全部跑赢随机初始化,165k 语料还追平了 MP-20k。不用 DFT 弛豫也能扩预训练数据,写论文或建材料模型都用得上。
一项 arXiv 研究提出将扩散预训练与合成晶体结构生成结合,摆脱对 DFT 弛豫数据的依赖。管线包含三种策略:perturb-and-relax、MatterGen 生成采样和 protostructure 枚举,需要弛豫时用机器学习原子间势替代 DFT。在随机初始化的 Orb-v2 主干上,作者比较了坐标去噪控制与连续时间 VP 和 sub-VP 扩散目标。
实验覆盖 MatBench 七个性质预测任务(protostructure 扩展研究另含 mp_gap 任务):所有合成预训练条件在七个任务上数值上均优于随机初始化。在固定 400 epoch 调度下,protostructure 语料从 20k 扩到 165k,多数任务上数值持平或超过 MP-20k 预训练;由于数据多样性和优化步数同时增加,该结果反映的是数据与算力的联合扩展。
这项工作表明,材料性质预测的预训练不必依赖昂贵的量子力学标签或 DFT 弛豫结构,用机器学习势弛豫加上生成式与枚举式合成数据即可扩规模。对做图神经网络材料模型的研究者,这提供了一条用 MatterGen 等工具自主扩充预训练语料的可行路径。
Synthetic Crystal Data Generation for Diffusion Pretraining in Material Property Prediction
Diffusion-based pretraining has emerged as a powerful strategy for learning transferable representations in graph neural networks for materials property prediction without requiring quantum-mechanical labels such as energies, forces, and stresses for every pretraining structure. However, the method retains a critical dependency: it requires structures residing near local minima of a potential energy surface, which have traditionally been obtained through costly DFT relaxations of known or physically synthesized compounds. This reliance constrains both the scale and the structural diversity of available pretraining data, leaving large regions of crystal structure space unexplored. In this work, we connect diffusion-based pretraining with synthetic structure generation to reduce this bottleneck. We introduce a synthetic data generation pipeline combining three complementary strategies: perturb-and-relax, generative sampling via MatterGen, and protostructure enumeration. Where structural relaxation is required, we use fast machine-learned interatomic potentials instead of performing a new DFT relaxation for each generated structure. We further compare a coordinate-denoising control with continuous-time variance-preserving (VP) and sub-VP diffusion objectives using a randomly initialized Orb-v2 backbone. The source comparison covers seven MatBench property-prediction tasks, and an eighth task, \texttt{mp\_gap}, is included in the protostructure scaling study. Every synthetic-pretraining condition numerically outperforms random initialization on all seven source-comparison tasks. Scaling the protostructure corpus from 20k to 165k under a fixed 400-epoch schedule numerically matches or exceeds MP-20k pretraining on most tasks; because this protocol increases both data diversity and optimization updates, the result reflects joint data-and-compute scaling.
This report is categorized under AITOP's "Papers" desk. Key entities and technical topics under continuous tracking include MatterGen, Orb-v2, MatBench.
Source attribution: curated from arXiv cs.LG (@Aamod Varma, Shuyi Jia, Victor Fung).