Learning Smooth and Expressive Interatomic Potentials for Physical Property Prediction
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913805277069312 |
|---|---|
| author | Fu, Xiang Wood, Brandon M. Barroso-Luque, Luis Levine, Daniel S. Gao, Meng Dzamba, Misko Zitnick, C. Lawrence |
| author_facet | Fu, Xiang Wood, Brandon M. Barroso-Luque, Luis Levine, Daniel S. Gao, Meng Dzamba, Misko Zitnick, C. Lawrence |
| contents | Machine learning interatomic potentials (MLIPs) have become increasingly effective at approximating quantum mechanical calculations at a fraction of the computational cost. However, lower errors on held out test sets do not always translate to improved results on downstream physical property prediction tasks. In this paper, we propose testing MLIPs on their practical ability to conserve energy during molecular dynamic simulations. If passed, improved correlations are found between test errors and their performance on physical property prediction tasks. We identify choices which may lead to models failing this test, and use these observations to improve upon highly-expressive models. The resulting model, eSEN, provides state-of-the-art results on a range of physical property prediction tasks, including materials stability prediction, thermal conductivity prediction, and phonon calculations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_12147 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Learning Smooth and Expressive Interatomic Potentials for Physical Property Prediction Fu, Xiang Wood, Brandon M. Barroso-Luque, Luis Levine, Daniel S. Gao, Meng Dzamba, Misko Zitnick, C. Lawrence Computational Physics Machine Learning Machine learning interatomic potentials (MLIPs) have become increasingly effective at approximating quantum mechanical calculations at a fraction of the computational cost. However, lower errors on held out test sets do not always translate to improved results on downstream physical property prediction tasks. In this paper, we propose testing MLIPs on their practical ability to conserve energy during molecular dynamic simulations. If passed, improved correlations are found between test errors and their performance on physical property prediction tasks. We identify choices which may lead to models failing this test, and use these observations to improve upon highly-expressive models. The resulting model, eSEN, provides state-of-the-art results on a range of physical property prediction tasks, including materials stability prediction, thermal conductivity prediction, and phonon calculations. |
| title | Learning Smooth and Expressive Interatomic Potentials for Physical Property Prediction |
| topic | Computational Physics Machine Learning |
| url | https://arxiv.org/abs/2502.12147 |