Abstract
Attractor networks are a type of dynamical system that evolves to stable activity. Theoretical modeling and experimental observation have both arrived at these networks as a possible mechanism for some brain functions. In this project I use firing rate model units to map out the entire attractor landscape of finite networks with varying structure and connection strength. We then analyse their distribution of attractor basins and entropy. We find that attractor basins are distributed as a power law function of their size in random networks. Furthermore, we find that the properties of this power law evolve to rapidly disfavor small basin sizes as connection strength increases. This pattern allows them to maintain high entropy even when the number of stable states falls drastically. This general power law pattern persists even when introducing small world networks, though we see a peculiar pattern of noise before it is clearly established at high connection.