Proceedings

Environment-Aware Multiscale Geometric Interaction for Equivariant Molecular Spectral Prediction

Haoran Li, Weiran Cui, Minghui Li

Download BibTeX Cite

Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence. Main Track. Pages 4501-4508.

https://doi.org/10.24963/ijcai.2026/501

Keywords

  • Machine Learning: Geometric learning
  • Machine Learning: Deep learning architectures
  • Machine Learning: Representation learning
  • Machine Learning: Sequence and graph learning

Abstract

Predicting molecular spectra requires modeling 3D conformation and solvent modulation. However, E(3)-equivariant networks based on local message passing exhibit limited sensitivity to long-range geometric dependencies, affecting the discrimination of globally distinct conformers. We introduce the Multiscale Geometric Interaction Layer (MGIL), which integrates global context by augmenting local features with centroid-referenced anchors, geometric moments, and virtual nodes. This design explicitly encodes global anisotropy while maintaining equivariance. Furthermore, we propose a Solvent Field Modulator (SFM) to encode solvent topology for conditional feature adaptation. Experiments demonstrate that MGIL enhances the capture of global structural variations, yielding consistent performance gains across spectral prediction benchmarks while maintaining linear computational efficiency.

Haoran Li, Weiran Cui, and Minghui Li. 2026. Environment-Aware Multiscale Geometric Interaction for Equivariant Molecular Spectral Prediction. In Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence, IJCAI-26. IJCAI, 4501–4508. https://doi.org/10.24963/ijcai.2026/501
Welcome Future Conferences AI Journal Division IJCAI Proceedings About IJCAI IJCAI Awards