Realizing large materials models has emerged as a critical endeavor for materials research in the new era of artificial intelligence,but how to achieve this fantastic and challenging objective remains elusive.Here,we ...Realizing large materials models has emerged as a critical endeavor for materials research in the new era of artificial intelligence,but how to achieve this fantastic and challenging objective remains elusive.Here,we propose a feasible pathway to address this paramount pursuit by developing universal materials models of deep-learning density functional theory Hamiltonian(Deep H),enabling computational modeling of the complicated structure-property relationship of materials in general.By constructing a large materials database and substantially improving the Deep H method,we obtain a universal materials model of Deep H capable of handling diverse elemental compositions and material structures,achieving remarkable accuracy in predicting material properties.We further showcase a promising application of fine-tuning universal materials models for enhancing specific materials models.This work not only demonstrates the concept of Deep H's universal materials model but also lays the groundwork for developing large materials models,opening up significant opportunities for advancing artificial intelligencedriven materials discovery.展开更多
基金supported by the Basic Science Center Project of National Natural Science Foundation of China(52388201)the National Natural Science Foundation of China(12334003)+4 种基金the National Science Fund for Distinguished Young Scholars(12025405)the National Key Basic Research and Development Program of China(2023YFA1406400)the Beijing Advanced Innovation Center for Future Chip(ICFC)the Beijing Advanced Innovation Center for Materials Genome Engineeringfunded by the Shuimu Tsinghua Scholar program。
文摘Realizing large materials models has emerged as a critical endeavor for materials research in the new era of artificial intelligence,but how to achieve this fantastic and challenging objective remains elusive.Here,we propose a feasible pathway to address this paramount pursuit by developing universal materials models of deep-learning density functional theory Hamiltonian(Deep H),enabling computational modeling of the complicated structure-property relationship of materials in general.By constructing a large materials database and substantially improving the Deep H method,we obtain a universal materials model of Deep H capable of handling diverse elemental compositions and material structures,achieving remarkable accuracy in predicting material properties.We further showcase a promising application of fine-tuning universal materials models for enhancing specific materials models.This work not only demonstrates the concept of Deep H's universal materials model but also lays the groundwork for developing large materials models,opening up significant opportunities for advancing artificial intelligencedriven materials discovery.
