Enter the chemical composition (or a list of up to 2,000 compositions separated by commas) of a candidate Li conducting solid state electrolyte to return a binary classification prediction of whether a material will have a conductivity > 10^-4 S cm^-1, and a regression prediction of the materials conductivity in log₁₀(S cm^-1). These classification predictions were found to have an accuracy of 0.71, with regression predictions having a mean absolute error of 0.99. This is based on a CrabNet architecture, trained on The Liverpool Ionics Dataset, and is a hosted version of the deep learning model introduced in A database of experimentally measured lithium solid electrolyte conductivities evaluated with machine learning and reported in The Liverpool MaterialS Discover server: A Suite of Computational Tools for the Collaborative discovery of Materials. Please consider citing these papers if you use this tool in your work.
If you would like to access this tool via API click here for more information