EXP-038NEGATIVE
Advantage Decomposition / SHA-256

Within-family affine audit

If the Majority contrast is structural, it should remain directionally coherent after conditioning on exact affine truth-table families and c5 branches.

Claim statusNO END-TO-END MINING ADVANTAGE DEMONSTRATED

Diagnostic work, exact algebra and local capabilities are not treated as end-to-end mining advantage.

Date2026-08-14
HardwareClassical exact algebraic verifier
Run scopeThree canonical variants
ReproducibilityCanonicalize every truth mask, freeze family labels before outcome comparison, publish all cell sizes and repeat the same within-family audit on an independent cohort.
01 / Question & hypothesis

What was tested?

If the Majority contrast is structural, it should remain directionally coherent after conditioning on exact affine truth-table families and c5 branches.

02 / Scientific basis

Why the test is meaningful

Exact truth-mask families control for Boolean function shape. A contrast that disappears or changes across families is distribution-sensitive and cannot support a universal mechanism claim.

family(f)=canonical 16-lane truth maskΔ_{family,c5}=E[N|survivor,family,c5]−E[N|non-survivor,family,c5]universal mechanism requires coherent within-family replication
Within-family affine auditVisual reading of the published metrics and gates for EXP-038; it summarizes the registered result, not a mining advantage.EXP-038 / WITHIN-FAMILY AFFINE AUDITORIGINAL FAMILIES3FRESH FAMILIES2BRANCH VERDICTCLOSED
FIGURE / RESULT READINGVisual reading of the published metrics and gates for EXP-038; it summarizes the registered result, not a mining advantage.
03 / Method

How it was tested

Group records by exact affine family and c5, calculate product, shell and combined nonlinear support for all cells, then repeat the complete audit on the fresh cohort without removing sparse or unfavorable cells.

04 / Observed result

What happened

Original families3
Fresh families2
Branch verdictCLOSED

All masks were semantically exact. Three families appeared in the original cohort and two in the fresh cohort. Within-family differences were heterogeneous and distribution-sensitive; the canonical synthesis closed the Majority branch rather than promoting a selector.

05 / Validation

Exactness and statistical controls

The audit reported every family and cell, including zero-count and low-count cells. Fresh-family support differed from original support, which is itself evidence against a universal pooled conclusion.

06 / Interpretation

What the result means

The earlier Majority signal is not stable enough to guide candidate work. Closing this branch is a negative scientific result that narrows the search without claiming that all Majority structure is useless.

Limitations

  • Some exact families are absent from the fresh cohort.
  • Cell means are unstable when counts are small.
  • Closure applies to this registered mechanism and interface only.
07 / Reproduction

Evidence trail

Canonicalize every truth mask, freeze family labels before outcome comparison, publish all cell sizes and repeat the same within-family audit on an independent cohort.

Canonical variants

SUBENGINE-V38ASUBENGINE-V38BSUBENGINE-V38C

Source: internally audited canonical reports. Local filesystem structure, private headers and operational identifiers are excluded from publication.