Memory capacity and heterogeneity in swelled particle systems*

Systems of particles that swell and repel store memories of the sizes to which the particles are swelled.  In simulations, we investigate how system parameters such as the amount of noise, number and density of particles, interaction strength, and degree of anisotropy affect memory capacity.  Additionally, we have developed a method of reading memories which relies on the distribution of nearest neighbor distances, allowing the system to remain unperturbed during reading.  These statistics give us insight into which portions of these heterogeneous systems are responsible for memories.  We show that individual clusters chosen from a system can be reliably used to read memories.