A deep equivariant neural network approach for efficient hybrid density functional calculations
Nat. Commun. 15, 8815 (2024)
First author (co-first). Deep equivariant neural network for efficient hybrid DFT calculations.
Publications
You can also find my articles on my Google Scholar profile.
Journal Articles
Deep-learning electronic structure calculations
Nat. Comput. Sci. 5, 1133 (2025) [Review]
First author. Comprehensive review of deep-learning approaches to electronic structure calculations.
Physics and Deep Learning: An Introduction to the 2024 Nobel Prize in Physics
PHYSICS (Wuli) 54, 1 (2025)
First author. Popular science article bridging AI-physics research and the public (in Chinese).
Deep-learning electronic-structure calculation of magnetic superstructures
Nat. Comput. Sci. 3, 321-327 (2023) [Cover story]
Co-first author. Extending DeepH to magnetic materials with DFT accuracy.
Deep learning density functional theory Hamiltonian in real space
Phys. Rev. Lett. 137, 046401 (2026)
Co-first author. Learning DFT Hamiltonians in real space for improved accuracy and foundation models.
Deep-learning density functional perturbation theory
Phys. Rev. Lett. 132, 096401 (2024) [Editors' suggestion]
Co-first author. Deep-learning DFPT for accelerating electron-phonon coupling calculations.
Neural-network density functional theory based on variational energy minimization
Phys. Rev. Lett. 133, 076401 (2024) [Editors' suggestion]
Co-first author. Neural-network DFT incorporating physical principles via variational energy minimization.
Universal materials model of deep-learning density functional theory Hamiltonian
Sci. Bull. 69, 2514 (2024) [Cover story]
Co-first author. Universal materials model for deep-learning DFT Hamiltonians across diverse systems.
DeepH-pack: A general-purpose neural network package for deep-learning electronic structure calculations
npj Comput. Mater. (2026)
Contributing author. A unified software package integrating first-principles calculations with deep learning.
Equivariant Neural Network Force Fields for Magnetic Materials
Quantum Front. 3, 8 (2024)
Contributing author. Equivariant neural network force fields for magnetic materials simulations.
Roadmap on Advancements of the FHI-aims Software Package
Electron. Struct. accepted (2026)
Contributing author (200+ authors). Comprehensive roadmap for the FHI-aims all-electron electronic structure package.
ABACUS: An electronic structure analysis package for the AI era
J. Chem. Phys. 163, 192501 (2025)
Contributing author (58 authors). Software advancement paper for the ABACUS electronic structure package.
Preprints
AI-accelerated metallized σ-bonding screening for superconductor discovery
arXiv:2606.21251
First author. AI-accelerated high-throughput screening of superconductors using deep-learning DFT Hamiltonians.
Improving density matrix electronic structure method by deep learning
arXiv:2406.17561
First author (co-first). Deep-learning enhanced density matrix-based electronic structure methods.
DeepH-2: Enhancing deep-learning electronic structure via an equivariant local-coordinate transformer
arXiv:2401.17015
Co-first author. New-generation deep-learning electronic Hamiltonian framework with equivariant transformer.
Deep-Learning Database of Density Functional Theory Hamiltonians for Twisted Materials
arXiv:2404.06449
Contributing author. Deep-learning DFT Hamiltonian database for twisted moire materials.
AI4X Roadmap: Artificial Intelligence for the advancement of scientific pursuit and its future directions
arXiv:2511.20976
Contributing author (61 authors). Collaborative roadmap for AI-driven scientific advancement across disciplines.