Fig. 6: Grid maps change locally after removing a wall.

a The experimental procedure of Wernle et al. is illustrated. The rat underwent training in two rectangular compartments, labeled A and B. These compartments were initially separated by a central partition. During the test phase, the partition was removed, allowing the rats to explore the combined open square environment, denoted as AB. b Example firing rate maps for two grid cells from different rats are shown in both the divided A|B and merged AB environments. c The sliding spatial correlation heatmap is shown for the two example cells plotted in (b). Spatial correlation is calculated in moving boxes with the size of the white stippled box shown in (b). d The average spatial correlation heatmap across all recorded cells is shown. The impact of removing the partition is mostly limited to locations near the partition. e–g Simulation of experimental results in (b–d) by the model. The simulation results in (g) were repeated 50 times with different randomization seeds, and the average has been plotted. See also Supplementary Fig. 2. Data in (b–d) are adapted from Wernle et al., Springer Nature⁴⁸.