Quantifying Hidden Order Out of Equilibrium

Systems with more order take less information to describe. Computers take advantage of this fact to reduce the size of files such as photos and video without sacrificing quality. The Computable Information Density, or CID, is a way of quantifying the information in a system. The CID is the ratio of the length of a losslessly compressed data file to the uncompressed length of the original file. It also provides a new and essential way of studying systems, such as identifying critical points, critical exponents and the order of dynamical phase transitions.

In this lecture, Paul Chaikin will describe how lossless data compression enables the quantification of order in non-equilibrium and equilibrium systems of many interacting components, even when the underlying form of order is unknown. He will consider absorbing state models on- and off-lattice and active Brownian particles undergoing motility-induced phase separation. The technique identifies non-equilibrium phase transitions, determines their character, quantitatively predicts exponents for correlation lengths and critical slowing down, without prior knowledge of the order parameters and reveals previously unknown ordering phenomena. This technique should provide a quantitative measure of organization in condensed matter and other systems that exhibit collective phase transitions.