A four-site diversity rule did not pass a shared gain test for sound matching in six SEEG patients and 16 fMRI adults

We tested whether response diversity improves the choice of four auditory sites beyond repeatability alone. Separate public studies supplied SEEG high-gamma amplitude and BOLD fMRI. Two patients and four MRI adults supplied development. A local plan and code lock preceded new endpoint tests in six other patients and 16 other MRI adults. An added MRI study supplied eight other adults. Each split used 32 calibration sounds, 16 candidate sites and 96 test sounds from a source nonvoice group. A fixed greedy determinant rule balanced quality and diversity. Mean gains were 0.546, 3.496 and 2.656 percentage points. The SEEG 95% person Student limits were [-0.009, 1.102]. The required interval gate failed. The electrical exact sign-flip p was 0.0625. All 24 target MRI gains were positive. The full shared superiority rule failed. Eight-site selection and added baseline subtraction reversed the electrical gain. All negative people and controls remain. An exact limit makes the selectors equal for parallel estimated tuning profiles. A noisy one-direction model lost accuracy; stable measurement terms could also produce a benefit. This is a retrospective tool benchmark in source-screened sites. It does not establish a common neural cause, equal signal content or an implant benefit. Cohorts, libraries, source selection, initial pool size and measurement filters limit transfer. Code, provenance, performance data, figures and repeat steps are supplied. Neural arrays stay external.

auditory cortex · SEEG · fMRI · site selection · response diversity · repeatability · natural sounds · negative result