DYNAMIC LINK OF MEMORY - CHAOTIC MEMORY MAP IN NONEQUILIBRIUM NEURAL NETWORKS

We present a model for a dynamic link of memory in terms of a self-org anized chaotic transition in nonequilibrium neural networks. The chaot ic transition blocks pinning on parasitic(false) memory and allows a s uccessive retrieval of true memory. We introduced a new neurodynamics, which allows neural networks to be temporarily unstable, keeping stab ility due to convergent dynamics. The results obtained in this paper s uggest that the cortical chaos may serve for dynamically linking true memory as well as a memory search.