Fig. 1: Task stimuli, experimental design and main results of Experiment 1.

a Stimuli presented in the predictive learning task: S + = a learning stimulus that predicts the appearance of the outcome; S− = a learning stimulus that predicts the absence of the outcome. S = generalization stimuli. The numbers reflect size, from smallest to largest. Note that the hypothesis focuses on S1-S4. b An example of a trial’s sequence: A single ring is presented on the computer screen accompanied by an 11-point rating scale for participants to indicate their prediction of whether it is pathogenic, ranging from 0 (certainly not pathogenic) to 10 (certainly pathogenic). After participants respond, a feedback screen appears, and then a screen announcing the next trial. c Mean outcome predictions (on a 0 to 10 scale) during the learning phase by expected variability, trial number, and stimulus type. Error bars depict ±1 standard errors. d Mean outcome predictions for all rings (1-10) during the first generalization block by expected variability. Error bars depict ±1 standard errors. e Mean outcome predictions for all rings (1-10) during all generalization blocks by expected variability. Error bars depict standard errors ±1.