npj Computational Materials (Dec 2024)
DPA-2: a large atomic model as a multi-task learner
- Duo Zhang,
- Xinzijian Liu,
- Xiangyu Zhang,
- Chengqian Zhang,
- Chun Cai,
- Hangrui Bi,
- Yiming Du,
- Xuejian Qin,
- Anyang Peng,
- Jiameng Huang,
- Bowen Li,
- Yifan Shan,
- Jinzhe Zeng,
- Yuzhi Zhang,
- Siyuan Liu,
- Yifan Li,
- Junhan Chang,
- Xinyan Wang,
- Shuo Zhou,
- Jianchuan Liu,
- Xiaoshan Luo,
- Zhenyu Wang,
- Wanrun Jiang,
- Jing Wu,
- Yudi Yang,
- Jiyuan Yang,
- Manyi Yang,
- Fu-Qiang Gong,
- Linshuang Zhang,
- Mengchao Shi,
- Fu-Zhi Dai,
- Darrin M. York,
- Shi Liu,
- Tong Zhu,
- Zhicheng Zhong,
- Jian Lv,
- Jun Cheng,
- Weile Jia,
- Mohan Chen,
- Guolin Ke,
- Weinan E,
- Linfeng Zhang,
- Han Wang
Affiliations
- Duo Zhang
- AI for Science Institute
- Xinzijian Liu
- AI for Science Institute
- Xiangyu Zhang
- State Key Lab of Processors, Institute of Computing Technology, Chinese Academy of Sciences
- Chengqian Zhang
- DP Technology
- Chun Cai
- AI for Science Institute
- Hangrui Bi
- AI for Science Institute
- Yiming Du
- State Key Lab of Processors, Institute of Computing Technology, Chinese Academy of Sciences
- Xuejian Qin
- Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences
- Anyang Peng
- AI for Science Institute
- Jiameng Huang
- DP Technology
- Bowen Li
- Shanghai Engineering Research Center of Molecular Therapeutics & New Drug Development, School of Chemistry and Molecular Engineering, East China Normal University
- Yifan Shan
- Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences
- Jinzhe Zeng
- Laboratory for Biomolecular Simulation Research, Institute for Quantitative Biomedicine and Department of Chemistry and Chemical Biology, Rutgers University
- Yuzhi Zhang
- DP Technology
- Siyuan Liu
- DP Technology
- Yifan Li
- Department of Chemistry, Princeton University
- Junhan Chang
- DP Technology
- Xinyan Wang
- DP Technology
- Shuo Zhou
- DP Technology
- Jianchuan Liu
- School of Electrical Engineering and Electronic Information, Xihua University
- Xiaoshan Luo
- State Key Laboratory of Superhard Materials, College of Physics, Jilin University
- Zhenyu Wang
- Key Laboratory of Material Simulation Methods & Software of Ministry of Education, College of Physics, Jilin University
- Wanrun Jiang
- AI for Science Institute
- Jing Wu
- Key Laboratory for Quantum Materials of Zhejiang Province, Department of Physics, School of Science, Westlake University
- Yudi Yang
- Key Laboratory for Quantum Materials of Zhejiang Province, Department of Physics, School of Science, Westlake University
- Jiyuan Yang
- Key Laboratory for Quantum Materials of Zhejiang Province, Department of Physics, School of Science, Westlake University
- Manyi Yang
- Atomistic Simulations, Italian Institute of Technology
- Fu-Qiang Gong
- State Key Laboratory of Physical Chemistry of Solid Surface, iChEM, College of Chemistry and Chemical Engineering, Xiamen University
- Linshuang Zhang
- DP Technology
- Mengchao Shi
- DP Technology
- Fu-Zhi Dai
- AI for Science Institute
- Darrin M. York
- Laboratory for Biomolecular Simulation Research, Institute for Quantitative Biomedicine and Department of Chemistry and Chemical Biology, Rutgers University
- Shi Liu
- Key Laboratory for Quantum Materials of Zhejiang Province, Department of Physics, School of Science, Westlake University
- Tong Zhu
- Shanghai Engineering Research Center of Molecular Therapeutics & New Drug Development, School of Chemistry and Molecular Engineering, East China Normal University
- Zhicheng Zhong
- Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences
- Jian Lv
- Key Laboratory of Material Simulation Methods & Software of Ministry of Education, College of Physics, Jilin University
- Jun Cheng
- State Key Laboratory of Physical Chemistry of Solid Surface, iChEM, College of Chemistry and Chemical Engineering, Xiamen University
- Weile Jia
- State Key Lab of Processors, Institute of Computing Technology, Chinese Academy of Sciences
- Mohan Chen
- AI for Science Institute
- Guolin Ke
- DP Technology
- Weinan E
- AI for Science Institute
- Linfeng Zhang
- AI for Science Institute
- Han Wang
- HEDPS, CAPT, College of Engineering, Peking University
- DOI
- https://doi.org/10.1038/s41524-024-01493-2
- Journal volume & issue
-
Vol. 10,
no. 1
pp. 1 – 15
Abstract
Abstract The rapid advancements in artificial intelligence (AI) are catalyzing transformative changes in atomic modeling, simulation, and design. AI-driven potential energy models have demonstrated the capability to conduct large-scale, long-duration simulations with the accuracy of ab initio electronic structure methods. However, the model generation process remains a bottleneck for large-scale applications. We propose a shift towards a model-centric ecosystem, wherein a large atomic model (LAM), pre-trained across multiple disciplines, can be efficiently fine-tuned and distilled for various downstream tasks, thereby establishing a new framework for molecular modeling. In this study, we introduce the DPA-2 architecture as a prototype for LAMs. Pre-trained on a diverse array of chemical and materials systems using a multi-task approach, DPA-2 demonstrates superior generalization capabilities across multiple downstream tasks compared to the traditional single-task pre-training and fine-tuning methodologies. Our approach sets the stage for the development and broad application of LAMs in molecular and materials simulation research.
