Fig. 1: Learning behavior of mice in a two-alternative tone-texture discrimination task.

a Schematic of Experimental setup (adapted from reference 23). b Task design. Each trial consisted of a tone followed by a paired texture presented to the whisker pad of the mice. Then, mice chose between two lick ports to receive a sugar water drop as reward. c Learning curves of all mice (n = 16), subdivided into sessions of 80–120 trials. Naive is defined as performance <55%, learning as 55–75% and expert as ≥75%. Line colors indicate mice. d Lick pattern in correct trials, represented as percentage of trials with a lick at each time point, in naïve (gray line) and expert (orange line) mice. e Quantification of lick rates during tone window (left), early texture window (middle), and late texture window (right), in correct trials (black line represents mean). f Response time in correct trials. g Lick pattern in incorrect trials, represented as percentage of trials with lick at each time point, in naive and expert mice. h Quantification of lick rates during tone window (left), early texture window (middle), and late texture window (right), in incorrect trials. i Response time in incorrect trials. j Pupil diameter (z-scored) over trial time (left) and quantification over learning (right). Gray and orange in the left panel indicate naive and expert conditions; colors in the right panel indicate task windows. k Face movement (z-scored) over trial time (left) and quantification over learning (right). Colors as in (j). (15 mice, d–i: n = 138 [naive], 192 [learning], 171 [expert] sessions; j, k: n = 134, 191, 166 sessions due to missing/broken behavior videos; two-sided Wilcoxon rank-sum test; *p < 0.05, **p < 0.01, ***p < 0.001; p-values were corrected for multiple comparisons using FDR method in j, k; mean ± SEM). Source data are provided as a Source Data file.