Neural density functionals: Local learning and pair-correlation matching
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866909314190409728 |
|---|---|
| author | Sammüller, Florian Schmidt, Matthias |
| author_facet | Sammüller, Florian Schmidt, Matthias |
| contents | Recently Dijkman et al. (arXiv:2403.15007) proposed training classical neural density functionals via bulk pair-correlation matching. We show their method to be an efficient regularizer for neural functionals based on local learning of inhomogeneous one-body direct correlations [Sammüller et al., Proc. Natl. Acad. Sci. 120, e2312484120 (2023), 10.1073/pnas.2312484120]. While Dijkman et al. demonstrated pair-correlation matching of a global neural free energy functional, we argue in favor of local one-body learning for flexible neural modelling of the full Mermin-Evans density functional map. Using spatial localization gives access to accurate neural free energy functionals, including convolutional neural networks, that transcend the training box. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_03327 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Neural density functionals: Local learning and pair-correlation matching Sammüller, Florian Schmidt, Matthias Soft Condensed Matter Statistical Mechanics Recently Dijkman et al. (arXiv:2403.15007) proposed training classical neural density functionals via bulk pair-correlation matching. We show their method to be an efficient regularizer for neural functionals based on local learning of inhomogeneous one-body direct correlations [Sammüller et al., Proc. Natl. Acad. Sci. 120, e2312484120 (2023), 10.1073/pnas.2312484120]. While Dijkman et al. demonstrated pair-correlation matching of a global neural free energy functional, we argue in favor of local one-body learning for flexible neural modelling of the full Mermin-Evans density functional map. Using spatial localization gives access to accurate neural free energy functionals, including convolutional neural networks, that transcend the training box. |
| title | Neural density functionals: Local learning and pair-correlation matching |
| topic | Soft Condensed Matter Statistical Mechanics |
| url | https://arxiv.org/abs/2406.03327 |