Research

Paper

TESTING March 24, 2026

Universal and efficient graph neural networks with dynamic attention for machine learning interatomic potentials

Authors

Shuyu Bi, Zhede Zhao, Qiangchao Sun, Tao Hu, Xionggang Lu, Hongwei Cheng

Abstract

The core of molecular dynamics simulation fundamentally lies in the interatomic potential. Traditional empirical potentials lack accuracy, while first-principles methods are computationally prohibitive. Machine learning interatomic potentials (MLIPs) promise near-quantum accuracy at linear cost, but existing models still face challenges in efficiency and stability. We presents Machine Learning Advances Neural Network (MLANet), an efficient and robust graph neural network framework. MLANet introduces a dual-path dynamic attention mechanism for geometry-aware message passing and a multi-perspective pooling strategy to construct comprehensive system representations. This design enables highly accurate modeling of atomic environments while achieving exceptional computational efficiency, making high-fidelity simulations more accessible. Tested across a wide range of datasets spanning diverse systems, including organic molecules (e.g., QM7, MD17), periodic inorganic materials (e.g., Li-containing crystals), two-dimensional materials (e.g., bilayer graphene, black phosphorus), surface catalytic reactions (e.g., formate decomposition), and charged systems, MLANet maintains competitive prediction accuracy while its computational cost is markedly lower than mainstream equivariant models, and it enables stable long-time molecular dynamics simulations. MLANet provides an efficient and practical tool for large-scale, high-accuracy atomic simulations.

Metadata

arXiv ID: 2603.22810
Provider: ARXIV
Primary Category: cs.LG
Published: 2026-03-24
Fetched: 2026-03-25 06:02

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.22810v1</id>\n    <title>Universal and efficient graph neural networks with dynamic attention for machine learning interatomic potentials</title>\n    <updated>2026-03-24T05:14:33Z</updated>\n    <link href='https://arxiv.org/abs/2603.22810v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.22810v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>The core of molecular dynamics simulation fundamentally lies in the interatomic potential. Traditional empirical potentials lack accuracy, while first-principles methods are computationally prohibitive. Machine learning interatomic potentials (MLIPs) promise near-quantum accuracy at linear cost, but existing models still face challenges in efficiency and stability. We presents Machine Learning Advances Neural Network (MLANet), an efficient and robust graph neural network framework. MLANet introduces a dual-path dynamic attention mechanism for geometry-aware message passing and a multi-perspective pooling strategy to construct comprehensive system representations. This design enables highly accurate modeling of atomic environments while achieving exceptional computational efficiency, making high-fidelity simulations more accessible. Tested across a wide range of datasets spanning diverse systems, including organic molecules (e.g., QM7, MD17), periodic inorganic materials (e.g., Li-containing crystals), two-dimensional materials (e.g., bilayer graphene, black phosphorus), surface catalytic reactions (e.g., formate decomposition), and charged systems, MLANet maintains competitive prediction accuracy while its computational cost is markedly lower than mainstream equivariant models, and it enables stable long-time molecular dynamics simulations. MLANet provides an efficient and practical tool for large-scale, high-accuracy atomic simulations.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.LG'/>\n    <published>2026-03-24T05:14:33Z</published>\n    <arxiv:comment>10 pages, 6 figures, 6 tables</arxiv:comment>\n    <arxiv:primary_category term='cs.LG'/>\n    <author>\n      <name>Shuyu Bi</name>\n    </author>\n    <author>\n      <name>Zhede Zhao</name>\n    </author>\n    <author>\n      <name>Qiangchao Sun</name>\n    </author>\n    <author>\n      <name>Tao Hu</name>\n    </author>\n    <author>\n      <name>Xionggang Lu</name>\n    </author>\n    <author>\n      <name>Hongwei Cheng</name>\n    </author>\n  </entry>"
}