Overview of the experimental paradigm.

<p>(a) During scanning, subjects engaged in a number comparison task, which was followed by a monetary reward in some trials if correctly solved, while other trials were not rewarded. Prior to this task, a trial-unique cue from one of two categories was presented. Subjects had to infer from their experiences whether they were in an A-type block where the category of a cue predicted reward or not or in a B-type block, where cue category did not predict anything. During the outcome phase, a face surrounded by a green frame signaled a reward, while faces with red frames signaled no reward. After completion of 10 blocks and a break of 30 min, a surprise recognition memory task for all items presented as cues and during the outcome phase followed. (b) Schematic overview of trials within predictable and unpredictable blocks. Within predictable A-type blocks, the category of cue items reliably predicted the possibility of a reward after the number comparison task. In unpredictable B-type blocks, such a prediction was not possible, and rewards were thus received unpredictedly.