Analysis of Atomistic Representations Using Weighted Skip-Connections
arXiv:1810.09751
Abstract
In this work, we extend the SchNet architecture by using weighted skip connections to assemble the final representation. This enables us to study the relative importance of each interaction block for property prediction. We demonstrate on both the QM9 and MD17 dataset that their relative weighting depends strongly on the chemical composition and configurational degrees of freedom of the molecules which opens the path towards a more detailed understanding of machine learning models for molecules.
arXiv commentNIPS 2018 Workshop: Machine Learning for Molecules and Materials
References in corpus (12)
- Neural Message Passing for Quantum Chemistry
- Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
- SchNet - a deep learning architecture for molecules and materials
- SGDR: Stochastic Gradient Descent with Warm Restarts
- Quantum-Chemical Insights from Deep Tensor Neural Networks
- Active learning of linearly parametrized interatomic potentials
- Machine learning prediction errors better than DFT accuracy
- Visualizing the Loss Landscape of Neural Nets
- SchNetPack: A Deep Learning Toolbox For Atomistic Systems
- Achieving DFT accuracy with a machine-learning interatomic potential: thermomechanics and defects in bcc ferromagnetic iron
- Learning how to explain neural networks: PatternNet and PatternAttribution
- Neural Message Passing with Edge Updates for Predicting Properties of Molecules and Materials