Fig. 3: Autobiographical mental image feature ratings were reconstructed from each DMN subsystem.

Feature reconstruction/individual differences outcomes replicating Fig. 2A, B within each default mode network (DMN) subsystem are displayed (please see Fig. 2 caption for analytic details). To recap, Core-DMN and medial temporal DMN (MT-DMN) contain brain areas that are often associated with autobiographical simulation, whereas fronto-temporal DMN (FT-DMN) contains areas that are often associated with language and conceptual representation (see Supplementary Fig. 7). However, the current results suggest language semantics and self-generated mental images share some common representations in each DMN subsystem. MT-DMN appears to be the most distinct in terms of information encoded (see Fig. 4). A Feature reconstruction accuracies. Statistical significances were determined using one-sample signed ranks tests to compare reconstruction accuracy correlation coefficients to zero (n = 50 participants, one-tailed). P-values were adjusted according to the false discovery rate (FDR)³⁵. Red star notation for statistical significances is: ***FDR(p) < 0.001, **FDR(p) < 0.01, *FDR(p) < 0.05. B Individual differences analyses. Red bars indicate significant decoding accuracies where FDR(p) < 0.05 (permutation test, one-tailed). Orange bars indicate decoding accuracies where permutation tests returned p < 0.05, but values did not survive FDR correction. Gray bars indicate seemingly inconsistent outcomes where individual differences were detected for features that were not reconstructed with significant accuracy in (A). Please see main text for further details. Brain images were generated using BrainSpace v1.10⁵⁷. Source data are provided as a Source Data file.