Fig. 5: Preserved trajectory-specific tuning across pauses in task execution and across contexts.

a Schematic of continuous (i.e., daily execution of the task) and pause blocks (i.e., no exposure to the task apparatus over 7 days). Gray arrows show the days on which the mice experienced the task. Black arrows show the compared days for continuous (gray) and pause conditions (orange), separated by 7 days each. b Trajectory-specific tuning functions sorted by the first (left) and last day (right) of both continuous and pause pairs. 958 and 945 neurons, n = 8 mice. c Task pause does not affect trajectory-specific map stability over 7 days (t = −2.01, p = 0.084, 2-sided paired t-test). d Similar linearized position decoding (left, training: day 1, testing: day 7, t = 0.91, p = 0.395, 2-sided paired t-test) and prediction of trial outcome based on calcium signals in the center arm (T = 8, p = 0.600, 2-sided Wilcoxon signed-rank test) across continuous and pause blocks. e Schematic of familiar-familiar and familiar-novel pairs of recording days. Trajectory-specific tuning of all cells that were active on both days of the pair were analyzed. f Trajectory-specific tuning functions during familiar-familiar and familiar-novel sequence sorted by the first (top) or second day (bottom). 1261 and 1267 neurons. g Similar trajectory-specific spatial correlation (t = 0.94, p = 0.379, 2-sided paired t-test) for familiar-familiar (gray) and familiar-novel conditions (blue). h Accuracy of position decoding (T = 3, p = 0.039) and prediction of trial outcome based on calcium signals in the center arm (T = 2, p = 0.138, 2-sided Wilcoxon signed-rank tests) for familiar-familiar and familiar-novel conditions. Boxes show median and upper/lower quartiles. Circles are individual mice (n = 8 mice for all comparisons). Source data are provided as a Source Data file.