The Freeze Test — the breakthrough number, asked to be a coefficient, and the answer

Paper 93's release curve landed the model's surface charge at 20.4 MeV — within 15% of the ledger's fitted 17.8 — and reviewers immediately proposed the obvious prize: freeze the surface at the model's number, drop one fitted knob, and re-score the whole ledger with three. This paper runs that proposal on the full 2,938-nuclide ledger and the 514 Q-value sign test. The answer is a decisive null: the model's number is not the ledger's coefficient, and forcing it in triples the error and zeroes the sign score. Rule D8 — knobs must be earned, model numbers may not be transplanted as coefficients — is vindicated here not by doctrine but by measurement. This record holds one paper on its own and is refreshed every couple of months. The complete deposit, 10.5281/zenodo.21788037, holds every paper with its code and data and is updated first; for the newest version of this paper, go there.