He Li

Publications

Deep-learning density functional theory Hamiltonian for efficient ab initio electronic-structure calculation

H. Li*, Z. Wang*, N. Zou*, M. Ye, R. Xu, X. Gong, W. Duan, Y. Xu

Nature Computational Science 2, 367-377 (2022)

A deep neural network method (named DeepH) is developed to learn the mapping function from atomic structure to density functional theory (DFT) Hamiltonian, which helps address the accuracy-efficiency dilemma of DFT and is useful for studying large-scale materials.

[Paper] [Code] [Doc] [Research Briefing] [Sharing Link]

General framework for E(3)-equivariant neural network representation of density functional theory Hamiltonian

X. Gong*, H. Li*, N. Zou, R. Xu, W. Duan, Y. Xu

Nature Communications 14, 2848 (2023)

An E(3)-equivariant deep-learning method (named DeepH-E3) to represent density functional theory (DFT) Hamiltonian as a function of material structure, which can naturally preserve the Euclidean symmetry even in the presence of spin-orbit coupling.

[Paper] [Code]

Deep-learning electronic-structure calculation of magnetic superstructures

H. Li*, Z. Tang*, X. Gong, N. Zou, W. Duan, Y. Xu

Nature Computational Science 3, 321-327 (2023)

An extended DeepH (named xDeepH) method to represent density functional theory (DFT) Hamiltonian of magnetic materials for efficient electronic structure calculation using the E(3) and time-reversal equivariant neural network.

[Cover Article] [Paper] [Code] [Research Briefing] [Sharing Link]

Deep-Learning Density Functional Perturbation Theory

H. Li*, Z. Tang*, J. Fu, W. Dong, N. Zou, X. Gong, W. Duan, Y. Xu

Physical Review Letters 132, 096401 (2024)

A framework leveraging deep learning and automatic differentiation circumvents the computational bottleneck associated with density functional perturbation theory calculations.

[Editors' Suggestion] [Paper]

Efficient hybrid density functional calculation by deep learning

Z. Tang*, H. Li*, P. Lin*, X. Gong, G. Jin, L. He, H. Jiang, X. Ren, W. Duan, Y. Xu

arXiv:2302.08221

Use DeepH to learn the hybrid-functional Hamiltonian from self-consistent field calculations of small structures, and apply the trained neural networks for efficient electronic-structure calculation by passing the self-consistent iterations.

[Paper]

Education

Tsinghua University, Beijing, China

Ph.D. candidate in physics, Institute for Advanced Study

Advisor: Prof. Wenhui Duan

Sept. 2019 -

Peking University, Beijing, China

B.S. in electronics, School of Electronics Engineering and Computer Science

Thesis: Variational Monte Carlo using Capsule Networks as a Quantum Wave Function Ansatz

Sept. 2015 - Jul. 2019

Beijing National Day School, Beijing, China

Senior High School

Sept. 2012 - Jul. 2015

He Li (李贺)

Email: lihe.eecs pku.edu.cn

I am a Ph.D. student in the Institute for Advanced Study at Tsinghua University, advised by Wenhui Duan. Previously, I obtained my B.S. degree from the School of Electronics Engineering and Computer Science, Peking University.

Research Interests: I am interested in the application of machine learning methods in computational physics and materials science.