NEPMaker: Active learning of neuroevolution machine learning potential for large cells
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914475643240448 |
|---|---|
| author | Wang, Junjie Pan, Shuning Zhang, Haoting Jia, Qiuhan Ding, Chi Fan, Zheyong Sun, Jian |
| author_facet | Wang, Junjie Pan, Shuning Zhang, Haoting Jia, Qiuhan Ding, Chi Fan, Zheyong Sun, Jian |
| contents | Machine learning potentials (MLPs) achieve near first-principles accuracy but often fail for atomic environments outside the training distribution. Active learning can mitigate this limitation; however, its application to large-scale simulations is hindered by the prohibitive cost of labeling entire configurations. Here, we develop a D-optimality-driven active learning framework for the neuroevolution potential (NEP) implemented within the GPUMD package, named NEPMaker. Extrapolative atomic environments are identified on-the-fly and embedded into locally periodic structures, where boundary atoms are optimized to remain close to the training distribution. This strategy enables large-scale simulations to directly contribute to dataset construction, significantly reducing extrapolation errors while improving model robustness and transferability. The proposed framework provides a scalable route for constructing reliable machine learning potentials in complex materials systems, including those involving defects, interfaces, and phase transitions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_13848 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | NEPMaker: Active learning of neuroevolution machine learning potential for large cells Wang, Junjie Pan, Shuning Zhang, Haoting Jia, Qiuhan Ding, Chi Fan, Zheyong Sun, Jian Computational Physics Machine learning potentials (MLPs) achieve near first-principles accuracy but often fail for atomic environments outside the training distribution. Active learning can mitigate this limitation; however, its application to large-scale simulations is hindered by the prohibitive cost of labeling entire configurations. Here, we develop a D-optimality-driven active learning framework for the neuroevolution potential (NEP) implemented within the GPUMD package, named NEPMaker. Extrapolative atomic environments are identified on-the-fly and embedded into locally periodic structures, where boundary atoms are optimized to remain close to the training distribution. This strategy enables large-scale simulations to directly contribute to dataset construction, significantly reducing extrapolation errors while improving model robustness and transferability. The proposed framework provides a scalable route for constructing reliable machine learning potentials in complex materials systems, including those involving defects, interfaces, and phase transitions. |
| title | NEPMaker: Active learning of neuroevolution machine learning potential for large cells |
| topic | Computational Physics |
| url | https://arxiv.org/abs/2604.13848 |