Figure 5. Distribution of NAc firing rates across time surrounding cue-onset.

Each panel shows normalized (z-score) peak firing rates for all recorded NAc units (each row corresponds to one unit) as a function of time (time 0 indicates cue-onset), averaged across all trials for a specific cue type, indicated by text labels. (A) left: Heat plot showing smoothed normalized firing activity of all recorded NAc units ordered according to the time of their peak firing rate during the light block. Each row is a unit’s average activity across time to the light block. Dashed line indicates cue-onset. Notice the yellow band across time, indicating all aspects of visualized task space were captured by the peak firing rates of various units. (A) middle: Same units ordered according to the time of the peak firing rate during the sound block. Note that for both blocks, units tile time approximately uniformly with a clear diagonal of elevated firing rates. (A) right: Unit firing rates taken from the sound block, ordered according to peak firing rate taken from the light block. Note that a weaker but still discernible diagonal persists, indicating partial similarity between firing rates in the two blocks. Colour bar displays relationship between z-score and colour. (B) Same layout as in A, except that the panels now compare two different locations on the track instead of two cue modalities. As for the different cue modalities, NAc units clearly discriminate between locations, but also maintain some similarity across locations, as evident from the visible diagonal in the right panel. Two example locations were used for display purposes; other location pairs showed a similar pattern. (C) Same layout as in A, except that panels now compare reward-available and reward-unavailable trials. Overall, NAc units coded experience on the task, as opposed to being confined to specific task events only. Units from all sessions and animals were pooled for this analysis.

Figure 5—figure supplement 1. Distribution of NAc firing rates across time surrounding cue-onset.

Each panel shows normalized (z-score) minimum firing rates for all recorded NAc units (each row corresponds to one unit) as a function of time (time 0 indicates cue-onset), averaged across all trials for a specific cue type, indicated by text labels. (A) Responses during different stimulus blocks as in Figure 5A, but with units ordered according to the time of their minimum firing rate. (B) Responses during trials on different arms as in Figure 5B, but with units ordered by their minimum firing rate. (C) Responses during cues signalling different outcomes as in Figure 5C, but with units ordered by their minimum firing rate. Overall, NAc units coded experience on the task, as opposed to being confined to specific task events only. Units from all sessions and animals were pooled for this analysis.