Uncertainty-biased molecular dynamics for learning uniformly accurate interatomic potentials
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2023
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915002050412544 |
|---|---|
| author | Zaverkin, Viktor Holzmüller, David Christiansen, Henrik Errica, Federico Alesiani, Francesco Takamoto, Makoto Niepert, Mathias Kästner, Johannes |
| author_facet | Zaverkin, Viktor Holzmüller, David Christiansen, Henrik Errica, Federico Alesiani, Francesco Takamoto, Makoto Niepert, Mathias Kästner, Johannes |
| contents | Efficiently creating a concise but comprehensive data set for training machine-learned interatomic potentials (MLIPs) is an under-explored problem. Active learning, which uses biased or unbiased molecular dynamics (MD) to generate candidate pools, aims to address this objective. Existing biased and unbiased MD-simulation methods, however, are prone to miss either rare events or extrapolative regions -- areas of the configurational space where unreliable predictions are made. This work demonstrates that MD, when biased by the MLIP's energy uncertainty, simultaneously captures extrapolative regions and rare events, which is crucial for developing uniformly accurate MLIPs. Furthermore, exploiting automatic differentiation, we enhance bias-forces-driven MD with the concept of bias stress. We employ calibrated gradient-based uncertainties to yield MLIPs with similar or, sometimes, better accuracy than ensemble-based methods at a lower computational cost. Finally, we apply uncertainty-biased MD to alanine dipeptide and MIL-53(Al), generating MLIPs that represent both configurational spaces more accurately than models trained with conventional MD. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_01416 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Uncertainty-biased molecular dynamics for learning uniformly accurate interatomic potentials Zaverkin, Viktor Holzmüller, David Christiansen, Henrik Errica, Federico Alesiani, Francesco Takamoto, Makoto Niepert, Mathias Kästner, Johannes Computational Physics Materials Science Machine Learning Efficiently creating a concise but comprehensive data set for training machine-learned interatomic potentials (MLIPs) is an under-explored problem. Active learning, which uses biased or unbiased molecular dynamics (MD) to generate candidate pools, aims to address this objective. Existing biased and unbiased MD-simulation methods, however, are prone to miss either rare events or extrapolative regions -- areas of the configurational space where unreliable predictions are made. This work demonstrates that MD, when biased by the MLIP's energy uncertainty, simultaneously captures extrapolative regions and rare events, which is crucial for developing uniformly accurate MLIPs. Furthermore, exploiting automatic differentiation, we enhance bias-forces-driven MD with the concept of bias stress. We employ calibrated gradient-based uncertainties to yield MLIPs with similar or, sometimes, better accuracy than ensemble-based methods at a lower computational cost. Finally, we apply uncertainty-biased MD to alanine dipeptide and MIL-53(Al), generating MLIPs that represent both configurational spaces more accurately than models trained with conventional MD. |
| title | Uncertainty-biased molecular dynamics for learning uniformly accurate interatomic potentials |
| topic | Computational Physics Materials Science Machine Learning |
| url | https://arxiv.org/abs/2312.01416 |