NpjComput.Mater.：高效建立機器學習原子間勢的美好途徑：元學習

海歸學者發起的公益學術平臺

分享資訊，整合資源

交流學術，偶爾風月

針對各種應用的機器學習勢的開發涉及大量資料集的建立，其中包含不同理論水平的量子力學計算資料集。問題是，為每一個研究專門建立資料集並從頭開始訓練模型十分耗時，嚴重阻礙了機器學習勢的進一步發展。為建立機器學習勢的預訓練模型，雖然可從現有文獻中有效收集這些資料，但這些資料集出自不同的計算方式和具有不同的精度，甚至可能相互矛盾。元學習所尋求建立的模型雖然不是專門針對任何特定的任務，但可以快速地重新訓練到許多新的相似任務，其中每一個任務都是一個特定的學習問題。即使新資料的量相對較少，這種再訓練也可能是有效的。如何將元學習應用於建立龐大的機器學習原子間勢資料集，自然成了急需打通的路徑。

Fig. 1 | The Reptile algorithm.

來自美國洛斯阿拉莫斯國家實驗室的Alice E. A. Allen等，透過建立各種系統的機器學習原子間勢，從單個阿司匹林分子到ANI-1ccx資料集，展示了元學習的廣泛適用性。透過對多個大型有機分子資料集進行模型預訓練，研究者表明這些資料集可以組合在一起，並對模型進行了預訓練。使用預訓練模型的好處體現在了3BPA分子原子間勢模型訓練上，產生了更準確和更平滑的勢函式。元學習極大地擴充套件了機器學習原子間勢可用的擬合數據的多樣性，並建立了為機器學習原子間勢建立現成預訓練的基礎模型的可能性。

Fig. 2 | Meta-learning used for an
aspirin interatomic potential.

元學習將多種不同資料集同時擬合，建立的預訓練模型，從而限制了引數更新的空間，使模型可以被快速調整到具體任務的訓練上。這一進步改變了對現有資料集的利用方式，併為機器學習原子間勢的擬合開闢了新的途徑。該文近期發表於npj Computational Materials10: 154 (2024)，英文標題與摘要如下，點選左下角“閱讀原文”可以自由獲取論文PDF。

Learning together: Towards foundation models for machine learning interatomic potentials with meta-learning

Alice E. A. Allen, Nicholas Lubbers, Sakib Matin, Justin Smith, Richard Messerly, Sergei Tretiak & Kipton Barros

The development of machine learning models has led to an abundance of datasets containing quantum mechanical (QM) calculations for molecular and material systems. However, traditional training methods for machine learning models are unable to leverage the plethora of data available as they require that each dataset be generated using the same QM method. Taking machine learning interatomic potentials (MLIPs) as an example, we show that meta-learning techniques, a recent advancement from the machine learning community, can be used to fit multiple levels of QM theory in the same training process. Meta-learning changes the training procedure to learn a representation that can be easily re-trained to new tasks with small amounts of data. We then demonstrate that meta-learning enables simultaneously training to multiple large organic molecule datasets. As a proof of concept, we examine the performance of a MLIP refit to a small drug-like molecule and show that pre-training potentials to multiple levels of theory with meta-learning improves performance. This difference in performance can be seen both in the reduced error and in the improved smoothness of the potential energy surface produced. We therefore show that meta-learning can utilize existing datasets with inconsistent QM levels of theory to produce models that are better at specializing to new datasets. This opens new routes for creating pre-trained, foundation models for interatomic potentials.