Coherent responses in neural networks with the Hopfield-type memory

We have set up a signal-driven scheme of a chaotic neural network, with the coupling constants corresponding to certain information, and we investigate the stochastic resonance-like effects under deterministic dynamics, comparing it with the conventional case of a Hopfield network with stochastic noise. It is shown that the chaotic neural network can enhance weak sub-threshold signals and can have higher coherence abilities between the stimulus and the response than those attained by the conventional stochastic model.