arXiv：cs.AI· Pavlo O. Dral, Yi-Fan Hou, Yuxinxin Chen, Hassan Nawaz, Mateusz Witkowski, Szymon \'Smiga, Ireneusz Grabowski·· 20 小时前
机器学习势（MLP）能否比其训练方法更准确？
Can Machine Learning Potentials Be More Accurate Than the Methods They Are Trained On?
AI 导读
一篇 Perspective 提出，机器学习势（MLP）可通过在量子化学方法这一"第三空间"上的泛化实现非经验性的精度超越，即比其所训练的密度泛函或波函数方法更准确。作者提出方法可迁移 MLP 的实现路径，并指出追求超越训练方法精度还能让通用 MLP 比单方法模型更准确，即便前者仍不及对应量子化学方法。该思路将影响训练数据集构建与模型评估。
正文
Abstract:Although machine learning potentials (MLPs) generalize across configurational and chemical compound spaces, their accuracy is commonly believed to be limited by their training methods, with improvements beyond this limit achieved so far only by refitting to experimental observables. In this Perspective, we argue that MLPs can also be made more accurate than their training methods non-empirically by generalizing across a third space, that of quantum-chemical methods. We propose routes to method-transferable MLPs that predict at unseen methods, with working implementations in both density-functional and wavefunction-based theories showing that they can be more accurate than any of their training methods. Aiming beyond the accuracy of the training methods could also make universal MLPs more accurate than single-method models, even when the method-transferable MLPs remain less accurate than those QM methods. Method-transferable MLPs will shape the construction of training datasets and the evaluation of models, pushing the accuracy limits of chemical simulations.
| Subjects: | Chemical Physics (physics.chem-ph); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Computational Physics (physics.comp-ph) |
| Cite as: | arXiv:2610.18027 [physics.chem-ph] |
| (or arXiv:2610.18027v1 [physics.chem-ph] for this version) | |
| https://doi.org/10.48550/arXiv.2610.18027 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Pavlo O. Dral [view email]
[v1]
Tue, 13 Oct 2026 09:19:53 UTC (1,526 KB)
来源：arXiv：cs.AI · arxiv.org