% Alpha power change across trials. Left: Baseline alpha power increases over the trials of the habituation and acquisition phases. A transient reduction in baseline power occurs after participants experience the first US in trial bin 16. Right: The by-trial evolution of cue-evoked alpha power reduction to each tone from 600 to 900 ms after onset across the experimental session. The blue boxes indicate when participants were asked to provide affective ratings of the cues. In these blue boxes, the numbers 1 through 4 correspond to ratings during the early habituation, late habituation, early acquisition, and late acquisition periods of the study, respectively. The CS+ prompted a reduction in alpha power which grew larger over the course of the study. Shaded in dark green, the all-or-nothing model became significantly from the 21st to 27th trial bin. For the 27th trial bin shaded in light green, the generalization model became significant as the alpha-reduction grew for the GS1. Notably, alpha changes continued to change after affective ratings indicated that most participants had understood the contingencies: Most participants rated the CS+ as more unpleasant and arousing than the other cues during the early acquisition phase (rating 3, blue box). These early acquisition ratings were made after the 10th acquisition trial, corresponding to the16th trial bin in the figure. This supports the perspective that selective alpha power reduction may index ongoing processes of memory formation instead of reflecting attentional engagement. This reasoning is further elaborated on in the discussion.