用教师模型生成力数据，提升仅含能量的训练集上机器学习势函数精度。
Ensemble Knowledge Distillation for Machine Learning Interatomic Potentials
- 多教师模型先拟合量子化学能量，再推导原子受力。
- 学生模型同时学习真实能量与教师集成的平均力，准确率超当前最优。
- 特别适合缺乏梯度数据的高精度量子化学数据集，提升分子动力学稳定性。
机器学习原子间势函数（MLIPs）的性能高度依赖于训练数据量及量子化学（QC）理论级别。高保真度的QC方法生成的数据通常仅限于小分子，且可能缺少能量梯度，难以训练高精度的MLIP。本文提出集成知识蒸馏（EKD）方法，用于在仅含能量数据的训练集上提升MLIP精度。首先，多个教师模型在量子化学能量上训练，并为数据集中所有构型生成原子力。随后，学生模型同时学习真实量子化学能量和教师模型集成平均力。该方法应用于ANI-1ccx数据集，其中能量由耦合簇理论计算。结果表明，所得学生模型在COMP6基准测试中达到新状态领先水平，并显著提升分子动力学模拟的稳定性。
原文摘要 · Abstract (English)
The quality of machine learning interatomic potentials (MLIPs) strongly depends on the quantity of training data as well as the quantum chemistry (QC) level of theory used. Datasets generated with high-fidelity QC methods are typically restricted to small molecules and may be missing energy gradients, which make it difficult to train accurate MLIPs. We present an ensemble knowledge distillation (EKD) method to improve MLIP accuracy when trained to energy-only datasets. First, multiple teacher models are trained to QC energies and then generate atomic forces for all configurations in the dataset. Next, the student MLIP is trained to both QC energies and to ensemble-averaged forces generated by the teacher models. We apply this workflow on the ANI-1ccx dataset where the configuration energies computed at the coupled cluster level of theory. The resulting student MLIPs achieve new state-of-the-art accuracy on the COMP6 benchmark and show improved stability for molecular dynamics simulations.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。