Live neuron behavior on an electronic chip was recorded by Nippon Telephone & Telegraph (NTT) with a time-lapsed video under the microscope. The dissociated chick embryonic brain neurons (telencephalic neurons) were cultured, and individual neurons were placed on silicon glass plates deposited with metal oxide strips about 10 μm width for possible electronegativity neurite guidance.

We have attempted to map biological neural networks (BNN) to artificial neural networks (ANN) to determine whether the connectivity patterns dictate the information processing efficiency, or vice versa. By an image processing technique, we discovered the smallest size of intelligent BNN from those NTT video tapes that happened to fail their neurite guidance experiments. The neurite growth connecting other neurons was accomplished intentionally through selective paring in time rather than mechanically following the external guidance.

Such a dynamic interconnection has inspired a model of energy landscaping, of which a general convergence theorem that has unified Hopfield-Cohen-Grossberg and Werbos-Rumelhart theorems is given, and a nonlinear Hebbian learning is derived capable for self-determination of ANN architectures.