Figure 1

Explanatory figures. (a) Illustration of shared spike count variability over repeated trials of a condition for a neuron pair (left), for a population (centre) and exhibited as alternative single trial trajectories through low-dimensional space (right). (b) Illustration of multi-neuron first spike patterns (left) and single spike patterns (right). (c) Left: Illustrative (artificial) first spike stimulus-response distribution for two neurons and a single stimulus condition. Each point represents the first spike times of the two neurons on a trial on which both neurons spiked. A spike time noise correlation is observable within the isolated cluster. The vertical distance from a point to the line \(y=x\) represents the relative spike time difference for the first spike pair. Right: If the angle of the noise correlation is \(45^{\circ }\), the relative spike time difference remains fixed (with random noise) w.r.t. the latent process/processes underlying the correlation. If the angle is different from \(45^{\circ }\), the relative spike time difference varies w.r.t. the process/processes. (d) The latent process underlying the noise correlation may be due to adaptative (left) and/or non-adaptative (right) changes in the shared excitability-level of the two neurons from trial-to-trial. (e) Fixed relative spike time differences vs time-warped spike time differences. For the latter, the relative spike time difference is dependent on the non-adaptative single trial shared excitability-level. (f) Illustration of correlated predicted cortical state for disjoint pairs of neurons.