Fig. 1: Overview of task curriculum, trial structure, and rule design.

A Curriculum overview. Rats were trained on a sequence of odor-guided discrimination tasks governed by two orthogonal rules: ‘Non-match’ and ‘cue-id’. For each rule, two different odor sets, comprised of 8 different odors (problem A and B), were used for at least 5 consecutive sessions. Control group was trained solely on ‘cue-id’ rule. B Schematic of trial structure. C Schematic of the task design in 6 consecutive trials and the correct response according to the different rules. In each trial, one odor was presented (1–8), and the rat had to decide whether it was associated with a reward.