Fig. 4: Brain activation profiles and connectivity changes time-locked to entorhinal grid-like codes during observation, and association with performance.

a Grid orientations and magnitudes were estimated and tested on independent data sets (GLM1 and GLM2), focusing on data within the entorhinal cortex (yellow). This yielded participant- and path segment-specific grid magnitudes for each of the 36 path segments per condition and task run. Observation periods are indicated in black (navigation periods in red, 1 example task run). Segment-wise grid magnitudes were then used to parametrically modulate observation/navigation events in a third, independent analysis (GLM3) that tested for voxel-wise brain activation/connectivity changes time-locked to fluctuations in entorhinal grid magnitude during observation. b Increased brain activation during observation periods (compared to the implicit baseline), positively scaling with entorhinal grid magnitudes (one-sample t-test, N = 47). c Same contrast associated with behavioral performance (linear regression, N = 47). d Entorhinal connectivity changes during observation periods, time-locked to increased entorhinal grid magnitudes and associated with behavioral performance (linear regression, N = 47; Supplementary Table S4). c, d The scatter plots show the relationships between the changes in parameter estimates (arbitrary units, arb. units) extracted from the respective clusters (black outlines in glass brains) and the cumulative distance error (vm). Given the clear inferential circularity, we would like to highlight that these plots serve visualization purposes only, solely illustrating the direction of the brain-behavior relationship. Connectivity is schematically indicated through dashed line. Results are shown at p < 0.05 FWE-corrected at cluster level [cluster-defining threshold of (a, d) p < 0.001 or (c) p < 0.005]. L, Left; R, Right. Source data are provided as a Source Data file.