Fig. 9: Schematic diagram for steps to calculate transfer entropy (TE) from input to output spikes of a single pyramidal neuron.

a Input signal is represented by red arrows (darker red for Past, lighter red for Future). Output signal is represented by green arrows (darker green for Past, lighter green for Future). Spikes in both input and output signals were binned, and the spike counts were used to generate probability distributions to calculate entropy. b The entropies of the signals are overlapped, represented by areas of overlap of the sets in the Venn diagram. The overlap of the entropies is partly because of the flow of information between the different signals, and partly because of chance. c The blue region, representing the overlap between H(input^Past) and \(H({{\rm{output}}}^{{\rm{Future}}})\), is the mutual information (MI) between the Past of the input signal and the Future of the output signal (\(I(H({{\rm{input}}}^{{\rm{Past}}});H({{\rm{output}}}^{{\rm{Future}}}))\)). The striped portion of the blue region represents TE from the Past of the input signal to the Future of the output signal. TE was calculated as the MI between H(input^Past) and \(H({{\rm{output}}}^{{\rm{Future}}})\), given the Past of the output signal (\(I(H({{\rm{input}}}^{{\rm{Past}}});H({{\rm{output}}}^{{\rm{Future}}})| H({{\rm{output}}}^{{\rm{Past}}}))\)).