After regional mean regression, hippocampal boundary patterns did not pass a shared test in single-neuron records from 18 patients and fMRI from 15 adults

We tested whether hippocampal response patterns at film event boundaries repeat from early to late film sections. Separate public human studies supplied intracranial single-neuron spikes and fMRI. Seven patients supplied development. We then tested 18 other patients and 15 fMRI adults. Regional mean regression and film controls preceded each spatial comparison. Mean cosines above shared time-shift references were +0.2279 and +0.0865. Their 95% person intervals were +0.0473 to +0.4077 and +0.0152 to +0.1623. When the bootstrap also varied common film blocks, the intervals crossed zero: -0.1008 to +0.3859 and -0.0185 to +0.1363. Scores were positive in 13/18 and 11/15 people. The fixed shared rule failed; its intersection p was 0.059235, and both block interval gates failed. A later spike window gave a negative mean. Omitting regional mean regression changed the study means in opposite directions. One common source with measurement noise also gave repeatable patterns after regional mean regression. Thus this test would not identify several neural sources. This was a test of a new outcome in reused cohorts and two fixed films. It was not confirmation in new cohorts. Cohort differences, sparse electrical sampling, vascular effects, source cleaning and stationarity assumptions limit scope. No positive cross-modal neural or causal claim is made. The package preserves code, compact inputs, all outcomes, source pins and terms, plans, figures and repeat steps. All 33 reserved person scores passed separate arithmetic checks. A primary replay matched exactly.

human hippocampus · event boundaries · single neurons · fMRI · cross-modal testing · regional mean regression · film sampling · negative result