Driven by analogies with state vectors of quantum informatics and noise-based logic, we propose a general scheme and elements of neural circuitry for processing and addressing information in the brain. Specifically, we consider random (e.g., Poissonian) trains of finite-duration spikes, and, using the idealized concepts of excitatory and inhibitory synapses, offer a procedure for generating 2^N-1 orthogonal vectors out of N partially overlapping trains ("neuro-bits"). We then show that these vectors can be used to construct 2^(2^N-1)-1 different superpositions which represent the same number of logic values when carrying or routing information. In quantum informatics the above numbers are the same, however, the present logic scheme is more advantageous because it is deterministic in the sense that the presence of a vector in the spike train is detected by an appropriate coincidence circuit. For this reason it does not require time averaging or repeated measurements of the kind used in standard cross-correlation analysis or in quantum computing.