Fig. 1: Spatial memory task.

A During each navigation–encoding period of the task, participants freely and actively navigated a virtual beach environment using a game controller and successively encountered two or three unique objects in treasure chests at random locations on the beach. Participants aimed at encoding the objects and their associated locations to recall them during the retrieval period. After the navigation–encoding period, participants were moved to an elevated recall position where they completed a distractor task to prevent the participants from actively rehearsing the objects and their associated locations. After the distractor task, participants performed two types of memory recall. B During location-cued object recall, participants were presented with a location on the beach and recalled the associated object they had found at that location during encoding. C Histogram of session-wise memory performance during object recall (left) and improvement of object-recall memory performance over time, i.e., during the first half of all trials versus the second half of all trials (right; two-sided paired t-test: t(26) = –3.402, P = 0.002; n = 27 sessions). We computed object-recall performance in each session as the ratio of the number of successfully recalled objects divided by the total number of object recalls. For instance, in session number one, the participant successfully recalled 9 of 33 objects, resulting in an object-recall performance of 0.273. D During object-cued location recall, participants viewed an object and aimed at recalling the associated location on the beach. E Histogram of trial-wise memory performance during location recall (left) and improvement of location-recall memory performance over time (right; two-sided paired t-test: t(26) = –2.786, P = 0.010; n = 27 sessions). We computed location-recall performance in each trial as the drop error (Euclidean distance) between the participant’s response location and the correct location of the object, normalized by the possible distribution of drop errors. The closer to 1, the better memory performance. Red dotted line, chance performance. Gray lines, session-wise data; blue line, average. The virtual beach environment was created using the Unity 3D graphics engine and the Hand-painted Island pack (obtained from the Unity asset store under the Standard Unity Asset Store EULA; https://assetstore.unity.com/packages/3d/environments/fantasy/hand-painted-island-pack-36959#asset_quality). Source data are provided as a Source Data file.