arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.23515cond-mat.mtrl-sci
收藏

通用机器学习势在全局优化中的性能

Performance of universal machine learning potentials in global optimization of inorganic crystal structures

Binghamton University-SUNY(宾汉姆顿大学(纽约州立大学))

Edan T. Marcial, Laxman Chaudhary, Olesya Gorbunova, Aleksey N. Kolmogorov

AI总结:

研究评估了多种通用机器学习势在无约束进化搜索中的性能,发现其在预测复杂晶体结构基态方面的能力差异显著,部分模型能准确捕捉细微电子结构特征的能量差异。

摘要:

通用机器学习势(uMLPs)的快速发展和训练数据集的扩展正在重塑原子模拟的最前沿,突显了同时系统性评估其能力的必要性。全局优化是大多数 demanding 的 uMLP 应用之一,因为无约束探索包括探测参考集中没有的基元。我们检查了最新一代的 uMLPs 在无约束进化搜索中,以评估这些模型是否能够一致地预测多种无机系统中的复杂晶体结构基态。我们的发现表明,所考虑的 M3GNet、MACE、SevenNet、EquiformerV2、MatterSim、GRACE、eSEN、Orb-v3 和 PET-MAD 模型在解决低能盆地内竞争相的能力方面跨度很大,从接近从头计算到基本上无预测能力。此外,对 hcp-Zn、MB₄(M = Cr、Mn 和 Fe)和 LiB_y(y≈0.9)基态的额外测试表明,几种 uMLPs 能够捕捉到来自微妙电子结构特征的细微能量差异。

英文摘要:

Rapid development of universal machine learning potentials (uMLPs) and expansion of training data sets are reshaping the state of the art in atomistic simulation, highlighting the need for concurrent systematic benchmarking of their capabilities. Global optimization is among the most demanding uMLP applications because unconstrained exploration includes probing motifs not present in reference sets. We examined twelve pretrained uMLPs in unconstrained evolutionary searches to assess whether these models can consistently predict complex nonmagnetic crystal structure ground states under ambient pressure across diverse inorganic systems. Our findings demonstrate that the considered M3GNet, MACE-MATPES, MACE-mh-1, SevenNet, EquiformerV2, EquiformerV3, MatterSim, GRACE, eSEN, Orb-v3, PET-MAD, and PET-OAM models span a wide performance range, from near {\it ab initio} to essentially non-predictive, in their ability to resolve competing phases within low-energy basins. Additional tests on hcp-Zn, MB$_4$ (M = Cr, Mn, and Fe), and LiB$_{y}$ ($y\approx 0.9$) ground states reveal that several uMLPs capture fine energy differences arising from subtle electronic structure features.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑