All procedures were reviewed and performed in accordance with the Champalimaud Centre for the Unknown Ethics Committee guidelines. 65 male C57BL mice aged between 2 – 3 months at the start of experiments were used in the study. Animals were housed under a 12 hours light/dark cycle with experiments performed during the light cycle. 17 subjects were used in the two-step task baseline behavior dataset. 4 subjects were used in the ACC imaging. 2 subjects were used for electrophysiology controls for the optogenetics. 14 subjects (8 JAWS, 6 GFP controls) were used for the two-step task ACC manipulation only. 14 subjects (8 JAWS, 6 GFP controls) were used for the probabilistic reversal learning task ACC manipulation only. 14 subjects (8 JAWS, 6 GFP controls) were first trained and tested on the two-step ACC manipulation, then retrained for a week on the probabilistic reversal learning task and tested on the ACC manipulation in this task. 7 JAWS-GFP animals were excluded from the study due to poor or mis-located JAWS expression. In the group that was tested on both tasks, 1 Jaws
