Schematic drawings showing the trials following positive (left) and negative (right) experience (a), and the “inference” trial following a single experience of the reversal of stimulus-reward associations (b) for the NOVEL and FAMILIAR tasks, respectively. Averaged optimal choice rate for trials following positive (left) and negative (right) outcomes for OFC11/13 silencing (c) (two-way ANOVA, positive: treatment, F(1,24) = 14.8, p = 7.9 × 10−4; subject, F(1,24) = 0.001, p = 0.97; interaction, F(1,24) = 2.2, p = 0.15; negative: treatment, F(1,24) = 14.3, p = 9.1 × 10−4; subject, F(1,24) = 2.7, p = 0.11; interaction, F(1,24) = 0.94, p = 0.34), OFC11/13-rmCD silencing (d) (positive: treatment, F(1,16) = 4.8, p = 4.3 × 10−2; subject, F(1,16) = 1.8, p = 0.20; interaction, F(1,16) = 0.001, p = 0.97; negative: treatment, F(1,16) = 6.0, p = 2.6 × 10−2; subject, F(1,16) = 2.4, p = 0.14; interaction, F(1,16) = 0.65, p = 0.43), and OFC11/13-MDm silencing (e) (positive: treatment, F(1,16) = 1.5, p = 0.23; subject, F(1,16) = 0.51, p = 0.48; interaction, F(1,16) = 0.90, p = 0.36; negative: treatment, F(1,16) = 5.2, p = 3.7 × 10−2; subject, F(1,16) = 2.4, p = 0.14; interaction, F(1,16) = 0.56, p = 0.46). Averaged optimal choice rate for inference trials after OFC11/13 silencing (f) (two-way ANOVA, treatment, F(1,20) = 33.1, p = 1.3 × 10−5; subject, F(1,20) = 2.5, p = 0.13; interaction, F(1,20) = 1.7, p = 0.21), OFC11/13-rmCD silencing (g) (treatment, F(1,16) = 2.7, p = 0.12; subject, F(1,16) = 19.6, p = 4.2 × 10−4; interaction, F(1,16) = 2.9, p = 0.11), and OFC11/13-MDm silencing (h) (treatment, F(1,16) = 14.2, p = 1.7 × 10−3; subject, F(1,16) = 43.7, p = 6.0 × 10−6; interaction, F(1,16) = 0.021, p = 0.88). For the OFC11/13 silencing (c, f), data were obtained from N = 7 and 6 sessions for each treatment in each monkey for the NOVEL (c) and FAMILIAR (f) tasks, respectively. For silencing of each pathway (d, e, g, h), data were obtained from N = 5 sessions for each treatment, each task, and each monkey.