Deep-learning electronic structure calculations show great potential for revolutionizing the landscape of computational materials research. However, current neural-network architectures are not deemed suitable for widespread general-purpose application. Here we introduce a framework of equivariant local-coordinate transformer, designed to enhance the deep-learning density functional theory Hamiltonian referred to as DeepH-2. Unlike previous models such as DeepH and DeepH-E3, DeepH-2 seamlessly integrates the simplicity of local-coordinate transformations and the mathematical elegance of equivariant neural networks, effectively overcoming their respective disadvantages. Based on our comprehensive experiments, DeepH-2 demonstrates superiority over its predecessors in both efficiency and accuracy, showcasing state-of-the-art performance. This advancement opens up opportunities for exploring universal neural network models or even large materials models.
@article{arxiv.2401.17015,
title = {DeepH-2: Enhancing deep-learning electronic structure via an equivariant local-coordinate transformer},
author = {Yuxiang Wang and He Li and Zechen Tang and Honggeng Tao and Yanzhen Wang and Zilong Yuan and Zezhou Chen and Wenhui Duan and Yong Xu},
journal= {arXiv preprint arXiv:2401.17015},
year = {2024}
}