Information processing by the brain relies on the functions of neuronal networks. Therefore, understanding the structure and computational mechanisms of the brain circuitry is crucial for clarifying how cognitive functions emerge and how they can be best modeled for engineering applications. The manner in which neocortical and hippocampal circuits represent and process information has not been understood in detail. However, if sloppy processors like neurons can process enormous amounts of information efficiently, a large assembly of neurons is likely to operate in a parallel manner. In this report, we discuss the structure of cortical circuits that undergo self-organization through spike-timing-dependent plasticity, under the influence of two-state membrane potential fluctuations. Furthermore, we propose a stochastic rule for the generation of synapses, i.e., neuronal wiring, in a large population of cortical neurons.