Fig. 6: Subpopulations of RSC neurons recorded in individual experiments exhibit different stability of task-related representations.

a Encoding of context (top row), motor choice (top-middle row), post-decision outcome (bottom-middle row), and post-trial outcome (bottom row) by populations of neurons recorded in individual experiments. Solid line, mean; shaded area, 95% CIs. Note the gradient in population encoding, where post-trial outcome and context are more stable than motor choice and post-decision outcome. b Heat maps of context (top), motor choice (top-middle), post-decision outcome (bottom-middle), and post-trial outcome (bottom) coding performance along trial duration for the different experiments in this study. For clarity, only comparisons between day 1 and days 2 and 4 are shown. Note the higher stability of context (top row) and post-trial outcome (bottom row) encoding in days 2 and 4 between trial start and trial end, and after trial end, respectively. By contrast, motor choice and post-decision outcome exhibit lower stability. c Normalized decoding performance (averaged across experiments) integrated over the indicated 1.5 s windows relevant for each task variable (mean ± s.e.m.; n = 15, 12, 12, 11, 13 experiments in days 1–5 [context]; n = 16, 12, 14, 11, 14 experiments in days 1–5 [motor]; n = 17, 13, 14, 12, 14 experiments in days 1–5 [post-decision outcome]; n = 17, 12, 14, 12, 14 experiments in days 1–5 [post-trial outcome]). Exponential decay functions were fit to estimate the decay in encoding stability across days (black lines). Note the faster decay in motor choice and post-decision outcome encoding, and the slower decay in context and post-trial outcome encoding. d Time constants \(({{\rm{\tau }}})\) calculated after fitting exponential decay functions to the decoding performance of each task variable in each individual experiment (see Supplementary Fig. 14a, b; n = 8 experiments [context]; n = 11 experiments [motor]; n = 11 experiments [post-decision outcome]; n = 11 experiments [post-trial outcome]). Dots, individual sessions; overlaid lines, median ± 75% CIs. For clarity, one outlier τ value for post-trial outcome (33.5 days) is shown at saturation (25 days). Asterisks denote significantly different decay constants (*p < 0.05, **p < 0.01, ***p < 0.001; two-sided F-test calculated by LME models; see ”Methods” section for details). For all panels: s trial start, d decision point, e trial end. Source data for C and D are provided as a Source Data file.