A fixed vertebrate contrast failed its shared test in 14 SEEG patients and five fMRI adults

We tested a fixed vertebrate versus invertebrate response relation within selected animal groups. Separate public studies supplied human SEEG amplitude and BOLD fMRI. Three patients supplied local development. A code lock preceded reserved tests in 14 other patients and five MRI adults. We used different pictures for two category means, site scales from other categories, and raw cross-half distances. A full category endpoint projection removes norm-only distance structure. One source AlexNet image-distance map was also removed. Mean partial correlations were 0.03972 and 0.19959. Their 95% person Student limits were [-0.04845,0.12788] and [0.04426,0.35492]. Seven of 14 electrical scores and all five MRI scores were positive. The MRI rule passed. The electrical rule did not. The shared positive rule failed, p=0.17419. It required person sign-flip, person Student and whole-category label tests in each modality. Removing insect reversed the clinical mean. MRI single-category deletion means stayed positive. Distinct-image group contrast agreement was uncertain in SEEG. The clinical panel contains only three invertebrate groups, all arthropods. Sparse placement, separate cohorts and images, source order assumptions, untested appearance factors and measurement filters limit inference. No positive neural generalization or causal modality difference was confirmed. The package gives a reusable benchmark, code, source pins and terms, every result, figures and repeat instructions.

human neuroscience · SEEG · fMRI · animal vision · vertebrates · representational geometry · subject dependence · reproducibility