Multi cutoff method improves prediction accuracy in atom modeling

MCSA: Multi-Cutoff Spectral Aggregation in SO(3) Equivariant GNNs for Interatomic Potentials

Machine LearningArtificial Intelligence

Summary

Predicting how atoms interact is important for designing new materials, but it can be tricky because atoms influence each other at different distances. The authors found that changing a key setting, called the cutoff radius, affects how well the model captures these interactions. They created a way to combine information from multiple cutoff distances, improving the model’s ability to understand interactions at different scales. This leads to better predictions of energy, forces, and other material properties.

What this means in practice

  • For materials simulation engineers: Generate more accurate atomistic simulations by combining multiple spatial scale interactions to improve predictions of forces, energies, and material properties.
  • For computational chemists: Model layered materials more precisely by selecting multiple physical cutoff distances to capture varied atomic interactions effectively.

Authors

Junfu Tan, Mengfan Wu, Yu Zhu, Yu Shi, Yifang Qin, Chang Liu, Bonan Zhu, Ziheng Lu

Abstract

Machine Learning Interatomic Potentials (MLIPs) based on Equivariant Graph Neural Networks (EGNNs) have achieved remarkable progress in atomistic modeling. However, accurately representing atomic interactions across different spatial scales remains challenging. In this work, we investigate the radial spectral properties of SO(3)-equivariant message passing and establish that the cutoff radius controls the characteristic spatial-frequency scale of radial interactions. Based on this analysis, we propose Multi-Cutoff Spectral Aggregation (MCSA), a plug-and-play module that aggregates radial kernels with different cutoff radii to enrich multi-scale geometric representations. To achieve efficient implementation, we exploit the compact support of radial kernels to fuse cutoff-dependent radial weights before angular tensor-product operations, avoiding redundant angular computations while preserving rotational equivariance. Extensive experiments across representative EGNN backbones demonstrate consistent improvements in energy, force, and stress prediction, as well as enhanced accuracy in higher-order properties, including phonon dispersion and lattice thermal conductivity. Further evaluations on layered materials with distinct interaction scales demonstrate the effectiveness of physically motivated cutoff selection. Code is available at: https://github.com/JF-Tan/MCSA.