|
Negative-feedback processes are ubiquitous in neural systems. One common type of process is a spike-triggered adaptation current. Adaptation currents can introduce temporal correlations in the spike train, where a longer interspike interval (ISI) between spikes is more likely followed by a shorter ISI and vice versa, a pattern that is common in biological neurons. We derive and analyze a probabilistically independent decomposition of adaptation-mediated correlated spike trains. Our decomposition suggests a biologically plausible way that adaptation states of neurons can represent, communicate, and decode the high levels of information contained in their temporal spike pattern sequences. |
|