发表机构
Zhongguancun Academy; Zhongguancun Institute of Artificial Intelligence; Kairos Materials; Shanghai Jiao Tong University; Shanghai University; Beijing Institute of Technology; Shanghai Institute of Ceramics, Chinese Academy of Sciences（中关村学院; 中关村人工智能研究院; Kairos Materials; 上海交通大学; 上海大学; 北京理工大学; 中国科学院上海硅酸盐研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究构建多组分材料基准并评估11个预训练力场模型，发现力误差与训练参考覆盖度及局部几何异质性相关，压缩侧误差显著高于拉伸侧，为力场选择和训练数据设计提供指导。
AI 中文摘要
通用机器学习力场对由成分设计产生的多组分环境的泛化能力仍未得到充分评估。我们构建了一个包含7,599个多组分构型的基准数据集，这些构型受高熵设计、元素替代和阴离子混合的启发。我们评估了十一个预训练模型在能量、力和应力方面与密度泛函理论的一致性，并将评估扩展到弹性、振动和吸附相关性质。力误差通过训练参考覆盖度、局部几何异质性、距离方向性和元素响应进行分析。到训练参考环境的距离揭示了覆盖度差异与误差增加之间的定性关联，而在相似距离下仍存在显著变化。误差较高的组表现出更大的局部几何异质性，尽管OMat24对这些环境提供了广泛覆盖。相对于训练参考对的中位数，误差在中位数附近保持较低，在压缩侧急剧上升，在拉伸侧增加较弱。在匹配元素对和绝对距离偏差后，压缩侧力误差是拉伸侧误差的1.81至1.95倍。模型预测的成对相互作用曲线在压缩下表现出更大的曲率。独立元素系统中的拟合难度与原子位移引起的电子带能响应和费米能级移动相关，在多组分系统中也观察到类似模式。在参数匹配比较中，最大度数分别为2和4的球谐表示分别降低了43个元素中38个和40个元素的测试力误差，而元素难度差异仍然存在。这些发现为实验成分设计中的力场选择提供了信息，并确定了训练数据采样和模型表示的目标。
英文摘要
Universal machine learning force-field generalization to multicomponent environments generated by compositional design remains insufficiently assessed. We construct a benchmark of 7,599 multicomponent configurations inspired by high-entropy design, elemental substitution and anion mixing. Eleven pretrained models are evaluated against density functional theory for energies, forces and stresses, with assessment extended to elastic, vibrational and adsorption-related properties. Force errors are analysed through training-reference coverage, local geometric heterogeneity, distance directionality and elemental response. Distances to training-reference environments reveal a qualitative association between coverage differences and increasing errors, while substantial variation remains at similar distances. Higher-error groups show greater local geometric heterogeneity, although OMat24 provides broad coverage of these environments. Relative to training-reference pair medians, errors remain low near the median, rise steeply on the compression side and increase more weakly on the extension side. After matching element pairs and absolute distance deviations, compression-side force errors are 1.81-1.95 times extension-side errors. Model-predicted pairwise interaction curves show greater curvature under compression. Fitting difficulty in independent elemental systems correlates with electronic band-energy responses to atomic displacements and Fermi-level shifts, and a similar pattern is observed in multicomponent systems. In parameter-matched comparisons, spherical-harmonic representations with maximum degrees of 2 and 4 lower test force errors for 38 and 40 of 43 elements, respectively, while differences in elemental difficulty remain. These findings inform force-field selection for experimental compositional design and identify targets for training-data sampling and model representations.
Comments28 pages, 10 figures
