Mesoscopic Scale

Mesoscopic scale research focuses on understanding systems' behavior at an intermediate level, bridging the gap between microscopic details and macroscopic properties. Current investigations leverage machine learning, particularly convolutional neural networks and autoregressive latent variable models, to efficiently model and analyze complex mesoscopic dynamics in diverse fields like materials science, neuroscience, and collective motion. This approach allows for improved simulations, more accurate characterizations of emergent behavior, and a deeper understanding of information processing and dissipation within these systems, ultimately impacting fields ranging from materials design to the analysis of complex social and biological systems.