Pretraining Strategy for Neural Potentials
This work addresses the challenge of data efficiency and performance in molecular force field fitting for computational chemistry, representing an incremental improvement over existing pretraining methods.
The paper tackles the problem of improving Graph Neural Networks (GNNs) for fitting potential energy surfaces in water systems by proposing a mask pretraining method, resulting in enhanced accuracy and convergence speed compared to training from scratch or using other pretraining techniques.
We propose a mask pretraining method for Graph Neural Networks (GNNs) to improve their performance on fitting potential energy surfaces, particularly in water systems. GNNs are pretrained by recovering spatial information related to masked-out atoms from molecules, then transferred and finetuned on atomic forcefields. Through such pretraining, GNNs learn meaningful prior about structural and underlying physical information of molecule systems that are useful for downstream tasks. From comprehensive experiments and ablation studies, we show that the proposed method improves the accuracy and convergence speed compared to GNNs trained from scratch or using other pretraining techniques such as denoising. On the other hand, our pretraining method is suitable for both energy-centric and force-centric GNNs. This approach showcases its potential to enhance the performance and data efficiency of GNNs in fitting molecular force fields.