Fig. 5: Cue activity of mTh_LCN→M2 axons and pull latency are both greater in the post-error trials than in the post-success trials in the non-expert sessions.

‌a Pull latency and success rate of post-success (pS) and post-error (pE) trials in H (n = 14), M (n = 14), and L (n = 14) sessions of mTh_LCN→M2 axon imaging. Gray lines indicate individual sessions and black lines indicate the mean ± SEM. *P = 0.0295 for pull latency, 0.0494 for success rate, n = 14, Wilcoxon signed-rank test, two-sided. b Activity of cue-preferring axons in post-success (blue) and post-error (orange) trials in H, M, and L sessions (n = 12, 12, and 10, respectively) aligned to the cue onset. The corresponding lever trajectories are also shown. Shading indicates ± SEM. The black lines and dots indicate the period in which the difference in activity between post-success and post-error trials was significant (P  <  0.05, Wilcoxon signed-rank test, false discovery rate [FDR]-adjusted). c Activity of pull-preferring axons in post-success (blue) and post-error (orange) trials in H, M, and L sessions (n = 13, 14, and 14, respectively) aligned to the pull onset. The corresponding lever trajectories are also shown. Shading indicates ± SEM. d Cue, pull_–0.3s, and pull activities in post-success (pS; blue) and post-error (pE; orange) trials in H, M, and L sessions. The mean ± SEM are also shown. **P = 0.00244, *P = 0.0161 for the comparison between post-success and post-error trials, *P = 0.00920 for the comparison between H, M, and L sessions, Wilcoxon signed-rank tests with Bonferroni correction (α = 0.0167), two-sided. Cue activity, n = 12, 12, and 10 for H, M, and L sessions, respectively. Pull_−0.3s and pull activities: n = 13, 14, and 14 for H, M, and L sessions, respectively. e Schematic of the time course of the combined activity of mTh_LCN→M2 cue-preferring and pull-preferring axons in H sessions. Blue traces indicate post-success trials. Orange traces indicate post-error trials. Source data are provided as a Source Data file.