molecular-representation

SenCos-GEM introduces a physics-guided molecular representation learning framework that uses Squeeze-and-Excitation modules and a law-of-cosines constraint to improve 3D geometric understanding, achieving state-of-the-art results on MoleculeNet regression benchmarks.

This paper introduces a tri-branch modular fusion neural network that integrates 3D geometry, SMILES embeddings, and physicochemical descriptors for molecular property prediction, achieving a 20.6% error reduction on QM9 with fewer than one million parameters.

MOLAR proposes a noise-aware framework for learning multimodal molecular representations from noisy labels by separating clean-property inference from observed label noise, outperforming baselines on molecular benchmarks.