Developed through the collaboration of former group member Dr. Chen Shen and current PhD student Yixuan Zhang, we are excited to introduce SuperSalt—a machine learning–based framework for modelling molten salts. SuperSalt enables accurate simulations and predictions of molten salt properties, providing valuable insights into structure–property relationships. This capability will help researchers design and optimize salts for advanced energy storage technologies and the demanding conditions of next-generation nuclear reactors. The work has just been published in Nature Communications: https://www.nature.com/articles/s41467-025-62450-1