DeePAW: A universal machine learning model for orbital-free ab initio calculations

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Su, Tianhao, Hu, Shunbo, Wu, Yue, Oyang, Runhai, Wang, Xitao, Li, Musen, Reimers, Jeffrey, Zhang, Tong-Yi
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915874440478720
author Su, Tianhao
Hu, Shunbo
Wu, Yue
Oyang, Runhai
Wang, Xitao
Li, Musen
Reimers, Jeffrey
Zhang, Tong-Yi
author_facet Su, Tianhao
Hu, Shunbo
Wu, Yue
Oyang, Runhai
Wang, Xitao
Li, Musen
Reimers, Jeffrey
Zhang, Tong-Yi
contents Developing universal machine learning models for ab initio calculations is the frontier of materials cutting edge research in the new era of artificial intelligence. Here, we present the Deep Augment Way model (DeePAW) that is a universal machine learning (ML) model for orbital-free (OF) ab initio calculations, based on the density functional theory (DFT). DeePAW is currently the best OFDFT ML model according to the three criterions, 1) covering the largest number of elements, 2) having the widest application capability to diverse crystal structures, and 3) achieving the highest prediction accuracy without further fine-tuning. These scientific merits and innovations of DeePAW are stemmed from the novel SE(3)-equivariant double massage passing neuron networks. Besides predicting electron density distributions, DeePAW predicts formation energies of crystals as well and therefore paves an efficient avenue for multiscale materials modeling beyond conventional electronic structure calculation methods.
format Preprint
id arxiv_https___arxiv_org_abs_2603_18650
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle DeePAW: A universal machine learning model for orbital-free ab initio calculations
Su, Tianhao
Hu, Shunbo
Wu, Yue
Oyang, Runhai
Wang, Xitao
Li, Musen
Reimers, Jeffrey
Zhang, Tong-Yi
Materials Science
Databases
Developing universal machine learning models for ab initio calculations is the frontier of materials cutting edge research in the new era of artificial intelligence. Here, we present the Deep Augment Way model (DeePAW) that is a universal machine learning (ML) model for orbital-free (OF) ab initio calculations, based on the density functional theory (DFT). DeePAW is currently the best OFDFT ML model according to the three criterions, 1) covering the largest number of elements, 2) having the widest application capability to diverse crystal structures, and 3) achieving the highest prediction accuracy without further fine-tuning. These scientific merits and innovations of DeePAW are stemmed from the novel SE(3)-equivariant double massage passing neuron networks. Besides predicting electron density distributions, DeePAW predicts formation energies of crystals as well and therefore paves an efficient avenue for multiscale materials modeling beyond conventional electronic structure calculation methods.
title DeePAW: A universal machine learning model for orbital-free ab initio calculations
topic Materials Science
Databases
url https://arxiv.org/abs/2603.18650