RELEVANT PARAMETERS FOR A CLASS OF SEQUENCE-RETRIEVING NEURAL NETWORKS

Reconsidering a recently introduced model of sequence-retrieving neura l network, we introduce appropriate analogues of the well-known stabil ities and show how these, together with two coupling parameters lambda and theta, entirely control the dynamics in the case of strong diluti on. The model is exactly solved and phase diagrams are drawn for two d ifferent choices of the synaptic matrices ; they reveal a rich structu re. We then briefly speculate as to the role of these parameters withi n a more general framework.