# n55b — the Pauli review and its amendments (2026-07-09)

A second, fully hostile referee (Pauli persona, no authorship stake)
reviewed n55. His full report is preserved in the session record. This
file records his findings, our verifications, and what the theory looks
like after the cut. Verdict he issued: "publishable after amendments" —
the amendments were run the same day.

## Finding 1 (FATAL as stated) — CONFIRMED, then bounded
The hybrid dynamics discarded the unitary sector's Born-Oppenheimer
potential. Verified with our machinery: E0'' / (gamma C'') = -684.7 at
the n55 parameters — the discarded term is 685x the kept one, opposite
sign; at those parameters the clock does not oscillate. "Both halves
forced" is STRUCK from the standing claims. Then the amendment
measurement: the ratio scales as 1/F_test (685, 140.4, 38.9 at F =
0.01, 0.05, 0.2 — the product gamma_min * F = 6.9, 7.0, 7.8). The
domination window is NOT empty; it is a REGIME:
    gamma_drain * F  >~  7   (toy units),
and at F = 0.2, gamma ~ 40 puts omega ~ 0.6 BELOW the phonon band top
(2.04), escaping the anti-adiabatic ceiling. AMENDED CLAIM: the formula
gamma_PPN = (1-2a)/(1+2a) holds in the strong-monitoring / heavy-mass
regime; general relativity is this theory's strong-monitoring limit.
Outside the regime the closed sector's concave repulsion (n45)
dominates and there is no clock at all. Still owed: the s_op
measurement INSIDE the regime with the three-term dynamics (E0 kept),
and Pauli's radiation-linewidth and overdamping conditions
(omega/gamma, omega vs w_min) as explicit gates — n55c.

## Finding 5 (pre-registration violation) — OWNED, gate run late
G4 was registered in n55 and never run. That is a process failure on
the headline sentence ("EP as an identity"), recorded here at full
prominence. Run late and labeled late: uniform-limit s_op for dipole
delta = 1 vs 3 agree to 3.1e-6 (gate 1e-3). The EP-identity sentence
survives in the uniform limit; the violation stands in the record.

## Finding 3 (optical dictionary) — partially discharged by measurement
A wave was propagated (first time in the program): group transit
through a flat-top slab gives delta-c/c = 0.0219 vs the dictionary's
0.0200 — the optical response HALF is confirmed at the ~10% level with
known edge/finite-size systematics. Precision verification owed at
larger N. Pauli's deeper cut STANDS UNDISCHARGED and is now a named
assumption: nothing in the program selects which excitation is
"light"; if electromagnetism is a dressing-sector excitation, gamma =
0 for every a. Assumption A4, on the books.

## Finding 4 (LDA / bump convergence) — owed, named
The bump s_op (0.415 vs 0.25) has never been shown to converge to the
uniform value with scale separation. Convergence study owed (larger N).
Assumption A3.

## Finding 6 (the handcuff) — conceded and quantified
Knee-rigidity test of a at solar Delta-k/k = 4e-8 requires 2.3e-13
fractional knee metrology: not a laboratory constraint, ever. Cassini
binds a; the knee's role is discovery (knee vs floor vs nothing), not
precision. "Across materials" is withdrawn (k is the vacuum's, not the
material's). The knee-rate vs read-rate conflation (n48 has both) is
a real ambiguity, now assumption A5: gamma_drain in the formula is the
rate that prices maintenance, and its identification with the n48 knee
is a modeling statement, not a theorem.

## Finding 7 (the embedding family) — accepted as the honest posture
The registration's posture ("gamma_drain is an input") replaces the
memorandum's ("the theory has no choice"). The one-parameter
multiplicative family V = R(k) * C is assumption A1; non-multiplicative
embeddings are unexplored territory, named.

## Findings 8 (code hygiene) — accepted
The RuntimeWarnings (benign Accelerate subnormal noise, proven benign
by the exact anchors) are now documented here; embedding B was defined
only in the uniform limit (correct where used); the G1 anchor validates
the kernel, not the position-coordinate route (coordinate-route anchor
owed in n55c); the G6 exponent drift (1.101 -> 1.261) is unexplained
and flagged.

## What survives both referees, independently re-derived by the second
- The k^{-5/2} law for the gap-weighted fidelity metric (Pauli:
  "this survives me", his own re-derivation).
- The n45 concavity theorem.
- The conditional identity gamma = (1-2a)/(1+2a), now with its four
  assumptions STATED: A1 multiplicative embedding; A2 the
  strong-monitoring regime gamma*F >~ 7 (with underdamping/adiabaticity
  gates owed); A3 LDA/bump convergence; A4 light couples as substrate
  acoustics; A5 the pricing rate is the n48 knee rate.
- The prior-art position: claim 1 VIRGIN (n55_prior_art.md), window
  closing.

## The standing sentence, amended
General relativity's gamma = 1 is this theory's strong-monitoring
limit, reached when the vacuum audits its ledger in ledger units
(a = 0) and audits it fast (gamma * F >~ 7) — under assumptions
A1-A5, each named, each attackable, none hidden.