Each plot is taken from a specific electrode contact in a patient (three patients are represented) that correspond to the raw traces in Fig. 1. Each condition’s power spectral density differs from the contrasted condition by at least P<0.01. (a) Real-world search (invisible sensor task) elicits more power in high-theta than when the patient is stopped (patient #2). (b) Real-world search (invisible sensor task) elicits more power in low-theta from patient #4 than when the patient is stopped. (c) Recall of real-world spatial locations elicited greater low-frequency oscillations (2–8 Hz) than the stop condition in patient #5. (d) Real-world random walk without cognitive demands elicits more power in high-theta than during stopping (hippocampal contact from patient #2). (e) In patient #5, navigation in a VR task (4-on-8 virtual maze) elicits greater low-theta oscillations than stopping in that session. (f) Comparisons across all experimental conditions within a single patient and electrode (patient #2). All conditions elicit low-frequency oscillations, real-world random walking eliciting the highest. Grey bar indicates a significant difference at P<0.01.