Explainable Deep Learning for Uncovering Actionable Scientific Insights for Materials Discovery and Design

TL;DR

This paper introduces methods to interpret deep learning models in scientific research, enabling extraction of actionable insights for materials discovery by integrating domain knowledge into the analysis process.

Contribution

It presents a novel approach that incorporates domain-specific attributes into generative models to better understand and interpret deep neural networks in scientific applications.

Findings

Enhanced interpretability of deep models in materials science

Ability to extract actionable scientific insights

Facilitates fundamental discoveries in materials research

Abstract

The scientific community has been increasingly interested in harnessing the power of deep learning to solve various domain challenges. However, despite the effectiveness in building predictive models, fundamental challenges exist in extracting actionable knowledge from deep neural networks due to their opaque nature. In this work, we propose techniques for exploring the behavior of deep learning models by injecting domain-specific actionable attributes as tunable "knobs" in the analysis pipeline. By incorporating the domain knowledge in a generative modeling framework, we are not only able to better understand the behavior of these black-box models, but also provide scientists with actionable insights that can potentially lead to fundamental discoveries.