这给深度学习电子结构计算方法带来了更高的精度和更好的泛化能力,并打通了其利用电子结构大数据作深度学习的通道。 相关研究以「Generalizing deep learning electronic structure calculation to the plane-wave basis」为题,于 10 月 3 日发布在《Nature Computational Science》上。 论文链接:https://www.nature.com/...
In this context, generalizing deep learning electronic structure calculation to the PW basis would be of critical importance to future development of the field. Fig. 1: Idea of the deep learning DFT Hamiltonian under the PW basis and its applications to twisted bilayer graphene. a, The PW ...
该成果以“Deep-learningelectronic-structure calculation of magnetic superstructures ”为题发表在4月26日的《自然·计算科学》(NatureComputational Science),并入选为期刊封面文章。同期,该杂志还发表了以“Adeep-learning method for studying magnetic superstructures ”为题的Research Briefing、以“Computationally probing...
DeepH框架的方法介绍和示例研究结果以“Deep-learning density functional theory Hamiltonian for efficient ab initio electronic-structure calculation”为题发表在6月23日的《自然·计算科学》(Nature Computational Science)。在同一期中,该杂志以研究简报(Research Briefing)形式介绍了这一成果,附有专家和编辑评论,以及Be...
et al. Deep-learning density functional theory Hamiltonian for efficient ab initio electronic-structure calculation. Nat. Comput. Sci. https://doi.org/10.1038/s43588-022-00265-6 (2022). Rights and permissions Reprints and permissions About this article Cite this article Improving the efficiency of...
DeepH-pack is the official implementation of the DeepH (DeepHamiltonian) method described in the paperDeep-learning density functional theory Hamiltonian for efficient ab initio electronic-structure calculationand in theResearch Briefing. DeepH-pack supports DFT results made byABACUS,OpenMX,FHI-aimsor...
To achieve simultaneous restoration and enhancement, we study deep low-light image enhancement from a perspective of texture-structure decomposition, that is, learning image smoothing operator. Specifically, we design a low-light restoration and enhancement framework, in which a Deep Texture-Structure ...
Structure Learning in Motor Control: A Deep Reinforcement Learning Model [arXiv] Programmable Agents [arXiv] Grounded Language Learning in a Simulated 3D World [arXiv] Schema Networks: Zero-shot Transfer with a Generative Causal Model of Intuitive Physics [arXiv] SVCCA: Singular Vector Canonical ...
Anatole von Lilienfeld, Fast machine learning models of electronic and energetic properties consistently reach approximation errors better than DFT accuracy, arXiv preprint arXiv:1702.05532v1, 2017. pdf; Machine learning prediction errors better than DFT accuracy, arXiv preprint arXiv:1702.05532v2, 2017...
Despite recent advances of data acquisition and algorithms development, machine learning (ML) faces tremendous challenges to being adopted in practical catalyst design, largely due to its limited generalizability and poor explainability. Herein, we develop a theory-infused neural network (TinNet) approach...