Skip to main content
. 2020 Mar 10;9:e51458. doi: 10.7554/eLife.51458

Figure 6. Neural activity in naive and expert animals.

Figure 6.

(A) Example field of views in the naïve (left) and expert (right) mouse with neurons that are shown in (B). Naïve mice were exposed to the virtual track for the first time after being previously habituated to being head restrained and running on a treadmill as well as receiving rewards at pseudo-random intervals. (B) Two example neurons that modified their activity as a function of training. Neuron 1 (top) showed no discernable receptive field in the naïve animal. However, in the expert animal, it showed a clear receptive field after the landmark. Neuron 2 (bottom) in contrast, showed some landmark-anchored activity that was strongly amplified in the expert condition. (C) Activity of task-active neurons in expert animals shown in both naïve and expert sessions (n = 81 short track, 80 long track). (D) Population vector cross-correlation matrix of activity in naïve and expert sessions. (E) Cross correlation value of population vectors at the actual location of the animal in naïve and expert sessions calculated from task active neurons in the respective sessions. (F) Reconstruction error in naïve and expert conditions (mean ± SEM reconstruction error short/long: 2.07 ± 0.37/2.67 ± 0.46 (expert), 5.1 ± 0.63/5.7 ± 0.62 (naive); two tailed t-test, short and long: p<0.001).