The human brain appears to store information in an associative
manner, a partially presented pattern will be resolved to it's fully
stored version. We model such a content-addressable system through a
dynamical system comprising many interacting “neurons”. The flow of this
system through it's phase space and it's ultimate destination represent
the retrival and value of the desired memory. We present the basic
features of the model and investigate some of it's properties under
alteration of the connection pattern.