Correction to: Scientific Reports https://doi.org/10.1038/s41598-025-28083-6, published online 08 December 2025

The original version of this Article contained errors in Figure 1 where there was an extra object to the right of the original bounds. In Figure 4, the Beta characters were not viewable in the statistics in panels (a) and (b). The original Figures 1 and 4 and accompanying legend appear below.

The original Article has been corrected.

Fig. 1
figure 1

The memorability judgment task. In Experiment 1, participants viewed a sequence of face images, and for each image indicated with a button press whether they thought it was a memorable face (by pressing the J key) or a forgettable face (by pressing the F key). Immediately following their response, feedback was displayed below the image informing the participant of the image’s correct memorability category. This was repeated for 180 unique images. The design of Experiment 2a was identical to that of Experiment 1, except that stimuli consisted of scene images. Face images shown come from a subset of images in the 10k US Adult Faces database that were collected manually from Creative Commons image resources, and therefore may be shown in an online open access publication1.

Full size image
Fig. 4
figure 4

Task performance over time, and most improved stimuli. a, b Each line corresponds to a simple linear regression predicting logit-transformed average accuracy (y-axis) from time (x-axis) for a single participant. Bolded lines correspond to the average regression lines. (a) For Exp. 1 (faces), regression slopes were not significantly different from zero. (b) For Exp. 2a (scenes), slopes were significantly greater than zero. (c) Memorable and forgettable scene stimuli with the greatest average improvement in accuracy when viewed in the first quarter compared with the last quarter of Exp. 2a trials. All stimuli shown here increased in accuracy by at least 24%.

Full size image