Fig. 1. Task and behavioral results.

a Schematic of the free-recall task. Left: Task timeline for a single recall trial, highlighting the list structure and timing during the encoding period. Subjects completed 25 lists per experimental session. Right: Category representations of stimuli. Each point indicates a word’s semantic representation along with the first three principal components of the word2vec²⁸ space. Larger points denote the midpoint of all stimuli within a category. Words with similar meaning are spatially proximal. b Recall performance by serial position. Compared to average recall (dashed line), subjects (n = 69, indicated by individual points) recalled significantly more items from early serial positions. c Recall performance by category. Plotting convention follows panel (b). d Factor scores for example recall sequences. Factor scores above 0.5 indicate above chance levels of temporal or categorical organization of recall sequences. e Recall organization. Subjects (n = 69, indicated by individual points) clustered recalls along temporal and semantic dimensions, as indicated by above chance (dashed line) factor scores. Statistical inference was performed with two-sided, one-sample t tests versus chance (e) or average recall performance, with FDR correction for multiple comparisons (b, c). ***p < 0.001, **p < 0.01, *p < 0.05. Data are presented as mean values ± SEM. Source data are provided as a Source Data file.