Fixed tests did not show shared gains from a 1 s sound history term in six iEEG patients and twenty fMRI adults

We tested if a 1 s sound history term improves later auditory response prediction beyond static sound transforms and linear histories. Separate public studies supplied human intracranial EEG and BOLD fMRI. Three patients and five adults supplied development. Fixed new endpoint tests used six other patients and twenty other adults. Mean added test R² was 0.01205 and 0.01180 percentage points. Their 95% person intervals were -0.04156 to +0.06566 and -0.13813 to +0.16173 points. Both shared rules failed: positive gain and advantage over a matched future-history term. Their two-claim Holm p values were 0.435472 and 0.428620. A planned scale check placed all person and crossed-block upper limits below 0.5 points for this learner. This does not bound biological adaptation. A boundary identity shows that past and future products have equal complete-record totals. A stated filter reduced their centered disagreement to about one fifth in two fixed sound records. That check uses no neural outcomes. The contribution is a controlled benchmark, a direction diagnostic, and a scoped negative result. Clinical sampling, separate cohorts and stories, source processing, model restrictions and fixed stimuli limit transfer. No shared adaptation mechanism or cause is established. The package gives code, permitted compact inputs, source pins and terms, plans, all results, figures and repeat steps. Fresh source rebuilds and independent score checks passed.

auditory adaptation · intracranial EEG · fMRI · natural speech · negative result · temporal averaging · response prediction