Evaluating the diversity and utility of materials proposed by generative models

Alexander New, Michael Pekala, Elizabeth A Pogue, Nam Q Le, Janna Domenico, Christine D. Piatko, Christopher D Stiles

Keywords: generative adversarial networks, inverse design, materials optimization, graph neural networks

TL;DR: We analyze a generative model for inorganic materials in terms of the suitability for its use in the inverse design process.

Abstract: Generative machine learning models can use data generated by scientific modeling to create large quantities of novel material structures. Here, we assess how one state-of-the-art generative model, the physics-guided crystal generation model (PGCGM), can be used as part of the inverse design process. We show that the default PGCGM's input space is not smooth with respect to parameter variation, making material optimization difficult and limited. We also demonstrate that most generated structures are predicted to be thermodynamically unstable by a separate property-prediction model, partially due to out-of-domain data challenges. Our findings suggest how generative models might be improved to enable better inverse design.

Submission Number: 39