Fig. 4: I Partial Information Decomposition (PID) of the joint mutual information between pairs of brain areas encoding prediction errors.
From: Predictive learning shapes the representational geometry of the human brain
a analysis framework for computing PID from the pairwise joint mutual information (centre), where the predictors (\({X}_{1}\) and \({X}_{2}\)) are the brain signal across trials at the time window where the prediction error was mostly encoded (90–120 ms, left), and the predicted variable is the prediction error trajectory (\(Y\), right). b average redundancy and synergy of prediction error encoding across all pairs of brain areas. Dots represent participants, and asterisks indicate statistical significance (p < 0.05). Significance was determined by paired two-tailed t tests. c Redundancy and synergy t-statistics against the pre-tone baseline. Top views show significant cortical interactions. The bottom views show the betweenness centrality of each brain region. Transparency indicates statistical significance (p < 0.05 corrected), which was determined by component-based paired two-tailed permutation tests using network-based statistics (NBS). d correlation analysis between the representational shift and betweenness centrality of the synergy component for the brain regions indicated on the top. Dots represent individual participants. Shaded areas indicate 95% confidence intervals. The dotted lines indicate non-significant regression fits. Solid lines indicate significant regression fits. Significance (p < 0.05) was determined using the Pearson correlation coefficient and Bonferroni correction. Source data are provided as a Source Data file. e correlation analysis between the representational shift and betweenness centrality of the redundancy component. Dots represent individual participants. Shaded areas indicate 95% confidence intervals. The dotted lines indicate non-significant regression fits. Solid lines indicate significant regression fits. Significance (p < 0.05) was determined using the Pearson correlation coefficient and Bonferroni correction. Source data are provided as a Source Data file.
