Unveiling the predictive power of static structure in glassy systems

Despite decades of theoretical studies, the nature of the glass transition remains elusive and debated, while the existence of structural predictors of the dynamics is a major open question. Recent approaches propose inferring predictors from a variety of human-defined features using machine learning. We learn the long time evolution of a glassy system solely from the initial particle positions and without any hand-crafted features, using a powerful model: graph neural networks. We show that this method strongly outperforms state-of-the-art methods, generalizing over a wide range of temperatures, pressures, and densities. In shear experiments, it predicts the location of rearranging particles. The structural predictors learned by our network exhibit a correlation length which increases with larger timescales to reach the size of our system. Beyond glasses, our method could apply to many other physical systems that map to a graph of local interactions.