Honest status (binding — must match the live popup): Diagnostic only (signs derived; magnitudes demonstrated-fitted) — direction held. STATUS-UPGRADES:0. Frozen branch
dcc66f1b2685/ manifest metaa5b1e6f9d951is READ-ONLY. This dossier is the deep version of the 30-second gate popup; it expands, it does not upgrade. Every number is traceable to a corpus source cited in §5; nothing is back-solved to a known answer.
Headline. We caught our own threshold ledger cheating — and proved it with a machine that is forbidden from ever seeing the answer.
Grand unification's most famous near-miss is that the three Standard-Model gauge couplings, run up from the electroweak scale, almost meet at a single high energy — and whether they truly meet hinges on small "threshold" corrections from heavy particles crossing the unification scale. SG-7 is the gate where the framework's 13D K₆ program asks its own frozen Kaluza–Klein spectrum to supply exactly those corrections. The honest test is brutal and it is the same test every unification scheme in the literature faces: are the corrections computed from the geometry, or are they quietly reverse-fitted to force the lines to cross?
This dossier establishes three things and is careful about what it does not establish:
The honest grade, stated once and held throughout. SG-7 is a serious candidate / a diagnostic that recovers threshold signs — NOT a validated unification. It is OPEN. The signs are the genuine win; the magnitudes hinge on a single unfixed heat-kernel scheme object. The gate does not move on the strength of this dossier, and this dossier never says it does. Diagnostic only. [DIAGNOSTIC]
What this dossier establishes and does not. It establishes the precise structure of the threshold problem, the exact arithmetic of where the published ledger reproduces (trivially) and where it fails (the deep regeneration), the target-blind machine that demonstrates the failure, and a concrete specialist work plan for each open hole. It does not establish that the threshold magnitudes are derived, that $M_U$ is a prediction, that the spectrum is the unique unifier, or that "no row is a free parameter." Those are explicitly withdrawn or open, and §7 names every one.
In a four-dimensional grand unified theory, the three running inverse couplings $\alpha_i^{-1}(\mu)$ obey one-loop renormalization-group equations
$$
\alpha_i^{-1}(\mu) = \alpha_i^{-1}(M_Z) - \frac{b_i}{2\pi}\ln\frac{\mu}{M_Z} + (\text{two-loop}) ,
$$
with Standard-Model beta coefficients (GUT-normalized $\alpha_1 = \tfrac{5}{3}\alpha_Y$)
$$
b_1^{\rm SM} = \frac{41}{10},\qquad b_2^{\rm SM} = -\frac{19}{6},\qquad b_3^{\rm SM} = -7 ,
$$
fixed entirely by the Standard-Model field content (3 generations + 1 Higgs + the gauge sector). [Source: 01_DOSSIER.md §1.1 item 1; 00_HANDOFF_README.md §0.] In the plain Standard Model these three lines do not meet at a single point; the famous near-miss is closed in the MSSM by superpartner thresholds, and in any non-supersymmetric or higher-dimensional scheme by the finite threshold corrections of whatever heavy spectrum the model carries.
The open problem — shared across the entire field, not unique to this program — is that the closing threshold corrections are model-dependent inputs in every scheme. Nobody has a first-principles derivation that fixes them from geometry alone with nothing adjustable. Whoever's spectrum is doing the closing, the question "are these corrections computed, or fitted?" is the honest discriminator, and it is rarely answered cleanly.
SG-7 sets up the test in the cleanest possible form. The threshold correction enters as a finite vector $\delta = (\delta_1, \delta_2, \delta_3)$ added to the RG transport, and the geometry's heavy KK spectrum is supposed to produce it. With the compactification scale identified with the unification scale ($m_c = M_U$), the logarithmic KK-running term cancels and only the finite Seeley–DeWitt $a_2$ remainder survives:
$$
\delta_i = \frac{1}{2\pi}\,\Delta_i^{\rm finite}\!\left(K_6,\; S^2,\; S^1_Y/\mathbb{Z}_2,\; \text{Wilson-line}\right) .
$$
[Source: 01_DOSSIER.md §1.1 item 3.] The structural claim is that thresholds are a finite topological number, not a tuned logarithm. The witness is the 8-row heat-kernel ledger (G.3.2), whose three column sums are the printed threshold triple
$$
\delta = (+4.8424,\ -3.1112,\ -1.7313)\ \pm\ 1.6\times10^{-3} .
$$
[Source: 01_DOSSIER.md §0.1, §1.1 item 4; 07_CLOSURE_RESULT_AND_FINDING.md §6 table.]
01_DOSSIER.md §4.9, §A.1 R9.]01_DOSSIER.md top downgrade note.] That downgrade is the reason this gate is Diagnostic-only, and this dossier treats the program's own ledger with exactly the skepticism it would apply to a rival's.The honest position: the field does not have a clean first-principles threshold derivation; neither do we; but we have built and run a machine that tests whether anyone's claim is real, and we ran it against ourselves.
SG-7 is a consistency / pruning gate evaluated on the already-selected survivor — no new geometry is searched. The pipeline is one forward chain [Source: 01_DOSSIER.md §1.1]:
α_i⁻¹(M_Z) [3 measured] --two-loop MS-bar RG--> α_i⁻¹ near M_U (almost meet)
+ KK spectrum of K_gauge --heat-kernel a_2 ledger (G.3.2)--> δ = (δ_1,δ_2,δ_3)
--insert into RG--> α_1(M_U) = α_2(M_U) = α_3(M_U) at M_U ~ 1e16 GeV, residual 9.6e-11
The residual at $M_U$ is quoted at $9.6\times10^{-11}$ — far inside the propagated PDG band $\sim10^{-3}$ — but this is the residual of the printed ledger re-summed and inserted, i.e. it measures self-consistency of the fitted numbers, not derivation. [Source: 01_DOSSIER.md §0.1, W5.]
Running couplings + SM beta functions. $b = (41/10, -19/6, -7)$, fixed by SM content. RG transport frozen at two-loop $\overline{\rm MS}$ (hash f531205a9159). These are inputs (given-E): SG-7 inherits the SM content E from SG-2/SG-3 and does not derive it.
The KK spectrum on each compact factor. The gauge sector lives on $K_{\rm gauge} = K_6 \times S^2 \times S^1_Y/\mathbb{Z}_2$:
- $K_6 = SU(3)/T^2$ carries $SU(3)_c$ with Laplacian tower $m_n^2 = n(n+2)/R_{K_6}^2$ (radius hash 634438ce0776);
- $S^2$ carries $SU(2)_L$ with $m_n^2 = n(n+1)/R_{S^2}^2$ (radius hash 2381d472c62e);
- $S^1_Y/\mathbb{Z}_2$ carries $U(1)_Y$ with the orbifold-projected tower (radius hash 0e8b8dba2cf0).
The chiral families are zero modes (no KK chiral copies above $m_c$), so $n_q^{\rm KK} = 0$ for matter while the gauge KK towers contribute fully. [Source: 01_DOSSIER.md §1.1 item 2, §5.3 hash index.]
The finite heat-kernel remainder. With $m_c = M_U$ the logarithmic term vanishes and only the finite Seeley–DeWitt $a_2$ remainder survives, giving the $\delta_i = (1/2\pi)\,\Delta_i^{\rm finite}$ relation of §2.2. The identification $m_c = M_U$ "up to an $\mathcal{O}(1)$ factor" is the soft seam — that $\mathcal{O}(1)$ is exactly where an undeclared normalization could hide (this is residual R5).
The 8-row G.3.2 ledger. Each row is a (compact factor × field class) contribution to one column $(\Delta_1, \Delta_2, \Delta_3)$. The eight rows are: K₆ gauge/ghost $SU(3)$; K₆ matter $SU(3)$; $S^2$ gauge/ghost $SU(2)$; $S^2$ matter $SU(2)$; $S^1_Y$ hypercharge zero-packet; $S^1_Y$ hypercharge heavy-packet; Higgs Wilson-line at $n_H=1$; orbifold/boundary fixed points (plus an inert bin). [Source: 07_CLOSURE_RESULT_AND_FINDING.md §2.] The three column sums are the $\delta$-triple.
The discipline layer. G.10 No-Hidden-Knob audit (a 10-row knob table, every knob frozen and listed before comparison); G.5 fail-closed rules; the G.9 certificate checklist; and three structural cross-checks (gauge–ghost identity per factor; family-count load-bearing-ness; Wilson-line load-bearing-ness).
Summing the printed rows reproduces the printed $\delta$ to four decimals [Source: 01_DOSSIER.md §2.3(a)]:
- $\delta_1$ column $\{-0.84,\ +3.2140,\ +1.0470,\ +1.4214\} \to +4.8424$;
- $\delta_2$ column $\{-4.02,\ +0.92,\ -0.2110,\ +0.1998\} \to -3.1112$;
- $\delta_3$ column $\{-2.49,\ +0.79,\ -0.0313\} \to -1.7313$.
All three reproduce to four decimals. This is a tautology: it checks that the injected rows add up, not that the rows come from anywhere. It is exactly — and only — what the captured reproduce_all.py terminal log verifies.
Applying the G.3.2a coefficient formulas literally to regenerate the rows from invariants, with
$$
c^{\rm gauge}+c^{\rm ghost} = -\frac{1}{3}\,\frac{T_{\rm adj}}{24\pi}\int R\sqrt{g},\qquad
c^{\rm matter} = +\frac{1}{3}\,\frac{T(R)}{24\pi}\int R\sqrt{g}\cdot n_{\rm gen},
$$
and the geometric integrals $\int_{K_6} R\sqrt g = 12\pi^3$, $\int_{S^2} R\sqrt g = 8\pi$ (both from G.3.2a), gives [Source: 01_DOSSIER.md §2.3(b)]:
| Row | Printed value | Literal-formula value | printed / formula |
|---|---|---|---|
| 1 — $SU(3)$ gauge+ghost net ($\to\delta_3$) | $-2.4900$ | $-4.9348$ | 0.505 (~2× off) |
| 2 — quark matter on $K_6$ ($\to\delta_3$) | $+0.7900$ | $+2.4674$ | 0.320 (~3× off) |
| 3 — $SU(2)$ gauge+ghost net ($\to\delta_2$) | $-4.0200$ | $-0.2222$ | 18.090 (~18× off) |
| 4 — doublet matter on $S^2$ ($\to\delta_2$) | $+0.9200$ | $+0.1667$ | 5.520 (~5.5× off) |
This is the decisive verification. The chain invariant → row → column-sum is broken at the middle link: the column sums are right because the rows are injected to make them right, but the rows themselves do not fall out of the stated formulas. The literal formulas are under-specified — they omit at least the per-factor normalization that converts the Seeley–DeWitt $a_2$ on each factor into the running-coupling correction (the "$\mathcal{O}(1)$ factor" in $m_c = M_U$, and the projector/overlap weights $c_{a,b,i}$ named-but-not-valued in G.3.2). The ~18× on the $SU(2)/S^2$ row versus ~2× on the $SU(3)/K_6$ row is suggestive: $S^2$ and $K_6$ have very different volumes and KK degeneracies, so a volume/degeneracy normalization is a natural origin of a factor that is large for $S^2$ and small for $K_6$.
To convert the "fitted-to-target" concession from an assertion into a demonstration, the program built the AXIOM-CANONICAL-ROW-TRACE-COMPILER [Source: 07_CLOSURE_RESULT_AND_FINDING.md §2–§3]:
For the frozen admissible SG-7 spectral domain there is a unique target-blind row-label compiler $\ell:\ \Omega_{\rm adm} \to \{1,\dots,8,\ \text{inert}\}$, induced only by compact support, actor type, gauge-charge insertion, orbifold/boundary status, chirality, bundle descent, Wilson/Hosotani data, and the frozen admissibility predicate. With row projectors $P_a = \mathbb{1}_{\ell^{-1}(a)}$, the threshold rows are the finite parts $$ R_{a,i} = \mathrm{FP}_{s=0}\ \mathrm{Tr}_{H_{\rm KK}}\!\left[P_a\, Q_i^2\, \mu^{2s}\, D_{\rm KK}^{-s}\right], \qquad \delta_i = \frac{1}{2\pi}\sum_{a=1}^{8} R_{a,i}. $$
The axiom is value-free only if the row-label compiler and the finite-part scheme are frozen before any comparison with the printed ledger. That freeze-before-compare discipline is what the tool mechanically enforces:
tests/test_compiler.py).delta column is rejected with TargetLoadingError. The guard forbids any column whose name contains delta, target, printed, expected, fit, fitted, ledger_value, row_value, answer, observed_output — in both the mode table and projectors.json. So the compiler cannot be fed the answer; it can only compile rows from frozen primitives (eigenvalues, degeneracies, charge insertions, actor/factor/boundary/Wilson/admissibility labels). [Source: 07_CLOSURE_RESULT_AND_FINDING.md §3.]The executed run (2026-06-25). A real frozen mode table real_frozen_modes.csv (150 modes) was built target-blind — eigenvalues and degeneracies from the fixed K₆/S²/S¹_Y KK spectra, gauge-charge insertions from the actual SM rep Dynkin indices, scheme freedoms set by canonical, declared, un-tuned rules ($Z_X = 1$, $\mu = 1$, cutoff $(p,q) \le 12$ inherited from the frozen K₆ table). The compiler ran clean, then the result was compared to the printed triple post-hoc [Source: 07_CLOSURE_RESULT_AND_FINDING.md §6 table; 00_HANDOFF_README.md STATUS UPDATE]:
| $\delta_1$ (U(1)_Y) | $\delta_2$ (SU(2)_L) | $\delta_3$ (SU(3)_c) | |
|---|---|---|---|
| Target-blind (this run) | $-63.897$ | $+70.627$ | $+227172.5$ |
| Printed (fitted) ledger | $+4.8424$ | $-3.1112$ | $-1.7313$ |
| Match? | ✗ sign+mag (13×) | ✗ sign+mag (23×) | ✗ sign+mag ($1.3\times10^5$) |
No match on any component — signs flip on all three, and $\delta_3$ is off by five orders of magnitude.
The $\delta_3$ blow-up is structural, not a bug (independently confirmed): K₆ gauge-ghost and K₆ matter ride the identical KK tower, differing only by the fixed charge×coefficient ratio
$$
q_3 \cdot \text{coef} = (3 \cdot -1)\ \text{vs}\ (1.5 \cdot +1) = -2:1 ,
$$
which can never cancel — so the canonical insertion cannot produce the small printed $\delta_3$. [Source: 07_CLOSURE_RESULT_AND_FINDING.md §6; 00_HANDOFF_README.md STATUS UPDATE item 4.]
This is a demonstration that the obvious reconstruction fails, NOT a clean falsification of the geometry [Source: 07_CLOSURE_RESULT_AND_FINDING.md §6 caveats 1–2]:
log_mu finite-part in the compiler is a self-admitted placeholder (finite_part.py: "runnable finite-part placeholder … should replace or freeze before comparison") — not the actual ζ-regularized KK finite part. The wild magnitudes (especially $\delta_3$) are partly placeholder-scheme artifacts of the missing subtraction.spectrum_K6.csv was not co-located in the run package, so its tower was internal-consistency-checked but not byte-traced to source (a flagged UNKNOWN; it does not change the disposition — even granting the towers, the triple fails on sign).So the structural direction of the miss is real, but the precise $\delta_3$ magnitude is partly a placeholder artifact, and we say so wherever the finding appears.
Three internal cross-checks bear on the structural claims. The honest split [Source: 01_DOSSIER.md §2.1 W6/W7, §2.2]:
These are the reasoning moves that produced the progress, now shareable at working-physicist depth.
(1) Weaponize the concession, don't paper over it. The instinct under a downgrade is to defend the published numbers. The productive move was the opposite: concede the ledger was fitted, then build the apparatus that would either vindicate it cleanly or demonstrate the failure beyond dispute. A demonstrated weakness is worth more to a reader than a defended one, because it ships with a runnable proof.
(2) Target-blindness as a mechanical predicate, not a promise. The κ³/π discipline (the program's standing falsification test: an axiom or normalization counts only if it would be written without knowing the target value) is usually a methodological pledge. Here it was made mechanical: the compiler raises TargetLoadingError on any column whose name hints at the answer. A reviewer does not have to trust that we were blind; the machine cannot be fed the answer. This is the difference between "we did not look" and "we could not look."
(3) Separate the two checks that everyone conflates. The single most important diagnostic distinction in this gate is between the trivial check ("do the printed rows sum to the printed $\delta$?" — always yes, a tautology) and the deep check ("do the rows regenerate from the invariants?" — no, off ~2×–18×). The published reproduce_all.py log verifies only the trivial check. Mistaking it for a closure is the gate's signature mis-close. The whole §3.3 vs §3.4 structure exists to keep these apart.
(4) Signs survive without the geometry; magnitudes do not. The sign structure (gauge+ghost negative, matter/hypercharge positive) is generic heat-kernel / asymptotic-freedom physics — it follows from the field content E and the measured couplings, and would survive even if the specific compactification were wrong. That is why it is bankable. The magnitudes, by contrast, require the specific zeta-regularized KK-tower sum on the specific factors, which is exactly the unfixed object. Recognizing this split is what lets us bank the signs honestly while leaving the magnitudes open.
(5) The $-2:1$ ratio is a real structural reading, with a caveat. That K₆ gauge-ghost and matter ride the identical tower with a fixed $-2:1$ charge×coef ratio is a genuine non-cancellation observation — it explains why a naive canonical insertion cannot produce a small $\delta_3$. But the load-bearing physics question is whether this is a real obstruction or an artifact of the placeholder subtraction. We carry the caveat with the reading.
(6) The scheme object is shared across three gates. The missing per-factor normalization $Z_{\mathcal X}$ + overlap $c_{a,b,i}$ is the same single geometry-unfixed Seeley–DeWitt finite object that Gap-01's $a_6$ graviton coefficient and SG-6's $c_{\rm loop}$ ride. This is not a coincidence to be glossed — it is leverage: reconcile the object once and the fix propagates to all three gates (§6, §7).
Branch dcc66f1b2685 · manifest meta a5b1e6f9d951 · $R_{K_6}$ 634438ce0776 · $R_{S^2}$ 2381d472c62e · $R_Y$ 0e8b8dba2cf0 · RG transport (two-loop $\overline{\rm MS}$) f531205a9159 · $M_Z$ a6852c7a6b00 · uncertainty rule 61b0d93507e7 · $M_{\rm Pl}$ df5976a365c3 · $\alpha_i^{-1}(M_Z)$ 6a3b6ef06697 · $\mathbb{Z}_2$ on $S^1_Y$ ac4d2df3e708 · $\mathbb{Z}_6$ identification a68ee92a75be · Wilson-line cycle $\gamma$ 640e1d7f7773 · winding $n_H$ f65094fd8fd1 · $\eta_{BK}$ 84e94518d3f5. Bundle files (REFERENCED, not independently re-run in the dossier audit): reproduce_all.py 30d4d3049051 · appendix_F_threshold_outputs.csv a9b61c5f8049 · appendix_F_heat_kernel_ledger.csv 9a6c7c08dc3c. $\delta$-triple $(+4.8424,-3.1112,-1.7313)\pm1.6\times10^{-3}$: no per-row first-principles hash — injected reals. [Source: 01_DOSSIER.md §0 / §5.3.]
[Source: 01_DOSSIER.md §2.1.]
| # | Witness | Grade | Reproduces? |
|---|---|---|---|
| W1 | SM beta functions $b=(41/10,-19/6,-7)$ | hand-checkable | Yes — standard, recompute by hand from the SM spectrum |
| W2 | Two couplings almost meet under SM running | machine-lane | Yes in principle; not re-run here |
| W3 | 8-row ledger column sums = $\delta$ | hand-checkable | Yes (trivially) — summing the printed rows; a tautology |
| W4 | G.3.2a row formulas → row values | symbolic | NO — literal formulas miss by ~2×–18× (§3.4) |
| W5 | $\delta$ inserted → unification at $M_U$, residual $9.6\times10^{-11}$ | machine-lane | AUDIT — asserted from appendix_F_threshold_outputs.csv; not independently re-run |
| W6 | Family-count load-bearing (cross-check 2) | hand-checkable | Plausible but NOT re-run — row scaling asserted |
| W7 | Higgs $n_H=1$ load-bearing (cross-check 3) | hand-checkable | Yes (consistent) — row 7 = the $n_H=0$ miss $(+1.047,-0.211,0)$ |
| W8 | G.10 No-Hidden-Knob audit | symbolic/audit | Conditional — an auditable claim, not a theorem |
| W9 | reproduce_all.py captured terminal output |
machine-lane | AUDIT — a pasted log; verifies only the trivial re-sum (W3), not W4 |
| W10 | Freeze hashes + meta a5b1e6f9d951 |
machine-lane | Yes in principle; not re-run here |
The runnable package was executed independently — not a captured log. Published artifacts inside the package: real_frozen_modes.csv (150 modes), outputs/real_run.csv, the build_real_modes.py generator, FINDING.md, the compiler, and tests/test_compiler.py (3/3). On the bundled example table the compiler reproduces stored values to printed digits (e.g. Delta1_sum = −6.86021259463) — TOY numbers from the example table, NOT SG-7 thresholds, and must not be read as such. The real-table result is the §3.5 table. The finding was verified SAFE by an independent adversarial pass plus an independent re-run of the column sums. [Source: 07_CLOSURE_RESULT_AND_FINDING.md §3, §6.]
real_frozen_modes.csv and confirm the blind triple $(-63.897, +70.627, +227172.5)$ and that the anti-target-fitting guard fires on a planted delta column.reproduce_all.py log demonstrates.The whole gate reduces to two real targets — R1 (a mechanical owner artifact) and R2 (one bounded heat-kernel computation) — with R3/R4/R5/R8 dependent on or folding into those two, and R6/R7/R9 being given-E / inherited / scope statements. Each open hole below is a work-package a specialist can act on immediately. Standing falsification test for every item: a normalization or axiom reverse-engineered to hit the already-known $\delta$ is true-by-construction (the κ³/π pattern) and RELOCATES the residual; it does not close it.
(a) Precise statement. The G.3.2a per-row formulas, applied literally, miss the printed rows by ~2× to ~18× (§3.4); the chain invariant → row → column-sum is broken at the middle link. The missing object is the per-factor normalization $Z_{\mathcal X}$ and representation-overlap weight $c_{a,b,i}$ that convert the bare Seeley–DeWitt $a_2$ on each factor into the running-coupling correction. The exact quantity to compute is the zeta-regularized KK-tower finite part of the one-loop gauge-coupling correction per factor ($K_6$, $S^2$, $S^1_Y/\mathbb{Z}_2$) under the declared regulator, target-blind.
(b) Why it's hard / prior-attempt lessons. The first-pass blind run (§3.5) used a self-admitted placeholder for the finite part (log_mu in finite_part.py), so its wild magnitudes — especially the $\delta_3$ blow-up — are partly placeholder artifacts of the missing subtraction, not a clean falsification. Traps to avoid, named so they are not repeated: (i) Do not report the trivial re-sum (§3.3) as a closure — it verifies a tautology. (ii) Do not reverse-engineer $Z_{\mathcal X}$ to hit $(+4.8424,-3.1112,-1.7313)$ — that is the κ³/π pattern and relocates the fit from three numbers to one normalization. (iii) Surface the shared-scheme coupling: this normalization is the same unfixed Seeley–DeWitt object shared with Gap-01's $a_6$ and SG-6's $c_{\rm loop}$, and the 2026-06-28/29 audit found the two computation routes for that object DISAGREE (route-inconsistent, not merely uncomputed) — the specific obstruction is the SU(3) Gelfand–Tsetlin off-diagonal matrix elements that gate the 5-class Levi-Civita Lichnerowicz graviton hopping on $K_6$, plus a ghost-sector discrepancy. So R2 cannot resolve even at fixed scheme until that upstream route-inconsistency is reconciled. [Source: 01_DOSSIER.md cross-gate propagation note 2026-06-29; GATE_REGRADE_… HIDDEN HOLE line 427.]
(c) Exactly what closes it. Compute the zeta-regularized KK-tower sum per factor target-blind and check whether rows 1–4 move from the literal-formula values $\{-4.9348, +2.4674, -0.2222, +0.1667\}$ to the printed $\{-2.4900, +0.7900, -4.0200, +0.9200\}$ — i.e. whether the geometry supplies exactly the ratios $\{0.505, 0.320, 18.090, 5.520\}$ from structure. Success criterion: rows fall out within the pre-declared tolerance with no per-row fudge → magnitudes DERIVED-GIVEN-E (R3/R4 dissolve with it). A refuting result is a valid close: if the target-blind zeta-sum gives definite rows that miss the printed $\delta$ in a new direction beyond the band → the printed ledger is wrong, and the gate is definitively Diagnostic-only (a confirmed FINDING). Fallback (honest, not a closure): name AXIOM-THRESHOLD-NORMALIZATION — "the running-coupling correction from each compact factor is the bare $a_2$ times the factor's KK zeta-normalization $Z_{\mathcal X}$ (a function of its volume, curvature integral, and KK degeneracy only) and the overlap $c_{a,b,i}$, both fixed by the geometry, independent of any coupling value." This is writable with no $\delta$ in sight (κ³/π-clean as a statement), but it is only AXIOM-CLOSED if verified to yield the rows — otherwise it is AXIOM-CLOSED-pending-verification, which is still OPEN.
(d) Machinery & inputs. Start from the runnable package 08_row_trace_compiler_runnable/ (the compiler, row_trace_compiler/finite_part.py to be replaced, build_real_modes.py, real_frozen_modes.csv). Replace the log_mu placeholder with the actual ζ-regularized KK heat-kernel finite part; freeze it before any comparison. The geometric inputs are the frozen radii (634438ce0776, 2381d472c62e, 0e8b8dba2cf0), the curvature integrals $\int_{K_6}R\sqrt g = 12\pi^3$, $\int_{S^2}R\sqrt g = 8\pi$, the SM Dynkin indices, and $n_H=1$. The negative theorem SG7-R2 (scalar/separable no-go) tells you the missing object is a non-separable target-blind trace — do not attempt a scalar/separable normalization repair, it is proven to fail. [Source: 00_HANDOFF_README.md §3 R2, post-atomicity re-grade.]
(e) Leverage. This is the single highest-leverage move in the gate: closing R2 converts $\delta$ from injected to derived and dissolves R3 and R4 in one move. And because the scheme object is shared, the same GT-off-diagonal + ghost reconciliation is load-bearing for Gap-01 R1 ($a_6$ value/sign) and SG-6 R5 (the modulus / $\mu_{\rm cell}$). Reconcile the routes ONCE — fix the SU(3) GT off-diagonal matrix elements (5-class LC Lichnerowicz hopping) and the ghost discrepancy, target-blind — and the fix propagates to all three gates. Caveat: SG-7 is an $a_2$-class ledger while Gap-01 is the $a_6$ graviton coefficient; the shared object is the scheme/normalization decision and its GT/ghost machinery, not the coefficient order — reconciling the routes fixes the normalization R2 needs, but Route A's zeta-sum still must land target-blind. [Source: GATE_REGRADE_… line 1172; 01_DOSSIER.md propagation note.]
(a) Precise statement. The frozen spectrum_K6.csv was not co-located in the run package, so the KK tower in real_frozen_modes.csv was internal-consistency-checked but not byte-traced to source (a flagged UNKNOWN). Separately, the load-bearing reproduce_all.py + the two CSVs were referenced (and a terminal log pasted) but not independently re-run in the dossier audit.
(b) Why it's hard / prior-attempt lessons. The trap is the gate's signature mis-close: accepting the pasted terminal log (or the trivial re-sum) as a closure of the harness. The pasted log verifies only that the printed rows re-sum to the printed $\delta$ (W3/W9) — not that the rows regenerate from invariants (W4). The §3.4 deep-check failure predicts the rows in reproduce_all.py are hard-coded, not computed; the single most important diagnostic is to open the script and check.
(c) Exactly what closes it. Co-package the byte-anchored frozen spectra (spectrum_K6.csv / spectrum_S2.md / spectrum_S1.md) with the compiler and verify them against the eigenvalue/degeneracy towers in real_frozen_modes.csv; re-run the deep check so each row is provably regenerated from frozen primitives (eigenvalues, degeneracies, charge insertions), not hard-coded; confirm the meta-hash recomputes to a5b1e6f9d951. Success criterion: (3a) trivial check — CSV column sums equal $(+4.8424,-3.1112,-1.7313)$ to tolerance $5\times10^{-4}$ and the VERIFY row reads True,True,True,True → harness VERIFIED (sum-level); this upgrades the harness from AUDIT to machine-verified but does not close R2. (3b) deep check — rows computed from G.3.2a primitives inside the script → VERIFIED (derivation-level), which would simultaneously close R2 (unlikely given §3.4). A value failing to regenerate at the declared tolerance is a valid REFUTING close → Diagnostic-only confirmed.
(d) Machinery & inputs. reproduce_all.py (30d4d3049051), appendix_F_heat_kernel_ledger.csv (9a6c7c08dc3c), appendix_F_threshold_outputs.csv (a9b61c5f8049), run target-blind against the frozen radii / RG transport / $M_Z$ / the three PDG couplings. The 08_row_trace_compiler_runnable/ package is the cleaner modern substrate.
(e) Leverage. This is the cheapest win and it is the prerequisite enabling step for R2's deep check. It also exposes which $\sum Y^2$ convention the code uses, folding in R8. It does not by itself derive the magnitudes.
(a) Precise statement. $(+4.8424,-3.1112,-1.7313)$ are hand-entered to five sig figs with no per-row first-principles hash; the entire numerical unification claim rests on three injected numbers (equivalently, eight injected rows whose three sums are derived-from-them).
(b) Why it's hard. There is no independent closure of R3 — the $\delta$ become non-injected exactly when R2's row-derivation runs forward. The trap: naming AXIOM-DELTA-IS-COLUMN-SUM ("the threshold vector is by definition the column sum of the ledger") only relocates the injection from three numbers to eight rows. It is essentially already the manuscript's position and pays no debt until the rows are earned.
(c) Exactly what closes it. R3's attack plan is R2's (§6.1). DERIVED-GIVEN-E if R2 derives the rows; otherwise R3 stays OPEN, honestly labeled.
(d/e) Machinery & leverage. Same as R2. This is the only place a quantitative fit lives in SG-7 (the $b_i$, $M_Z$, $M_U$ being SM/declared, not tunable). Closing R2 removes all three as inputs.
(a) Precise statement. A1.10 labels $M_U = 1.0\times10^{16}$ GeV "a declared closure-target convention (not a 16-sig-fig prediction)." The live risk is the "$m_c = M_U$ up to an $\mathcal{O}(1)$ factor" seam (G.3.1 Step 3), where an undeclared normalization could hide.
(b) Why it's hard / traps. The headline "couplings unify at $M_U \sim 10^{16}$ GeV" can be misread as a prediction of $M_U$. The bright-line: do not claim couplings are PREDICTED to unify at $M_U$ — the honest verb is "couplings CROSS at $M_U$." [Source: GATE_REGRADE_… line 423; 01_DOSSIER.md §4.5.]
(c) Exactly what closes it. Trace the $m_c/M_U$ ratio in the harness and confirm the $\mathcal{O}(1)$ is a fixed geometric number (e.g. $1/2\pi$ from $R_0 = (2\pi M_U)^{-1}$), not a tuned constant. If geometric → AXIOM-CLOSED (AXIOM-MU-IS-CROSSING: "$M_U$ is the scale at which the three RG-transported inverse couplings with the frozen thresholds coincide; it is an output of the crossing condition, not an independent input"). If found free → it is a hidden knob and the gate downgrades (a valid refuting close via G.10).
(d) Machinery. Folds into the R1 deep-check — trace the $m_c = M_U$ identification in reproduce_all.py.
(e) Leverage. The $\mathcal{O}(1)$ seam is exactly where R2's missing weight may partly hide; resolving it sharpens R2.
(a) Precise statement. The KK tower masses (hence the heat-kernel contributions) are fixed by the moduli read at the SG-6 Weyl-rigid witness $\vec u = (1,1,1)$, not independently derived.
(b) Why it's hard / traps. This is inherited from SG-6 and understated if phrased only as "not independently derived": the live SG-6 status records the positive-definite Hessian is diagnostic-only and "the chamber center may be a saddle rather than a minimum." So the witness point's minimum-status is itself open at SG-6. Do not present $\vec u=(1,1,1)$ as a proven minimum. [Source: GATE_REGRADE_… HIDDEN HOLE line 427.]
(c) Exactly what closes it. Prove $\vec u=(1,1,1)$ is the unique Weyl-rigid center of the $K_6$ Cartan moduli, independent of the threshold output — a clean group-theory theorem. Success → DERIVED (symmetry-protected, not tuned); naming AXIOM-CHAMBER-CENTER-SPECTRUM is the realistic AXIOM-CLOSED endpoint. A non-unique center is a refuting result.
(d) Machinery. Group theory on the $K_6 = SU(3)/T^2$ Cartan moduli; hand to the SG-6/moduli specialist.
(e) Leverage. Shared with SG-6 — closing it there closes it here.
(a) Precise statement. G.3.2a states $\sum_f Y_f^2 = 10/3$ per generation while the G.3.2 row-6 index column reads "$\sum Y^2 = 10$" ($\times 3$ generations).
(b) Why it's hard. It isn't — it is cosmetic. A reviewer running G.10.3 step 1 trips on it and may flag a phantom hidden knob.
(c) Exactly what closes it. One-line relabel making per-gen vs all-gen consistent ($10/3$ per gen $\times 3$ gens $= 10$). DISCLOSED-CORRECTED. No physics.
(d/e) Machinery & leverage. Folds into R1 — the published CSV exposes which convention the code uses.
BATTLE_GATES/. Do not send this for plugging.| Residual | Technique | Named axiom (κ³/π-clean) | Specialist target / owner artifact | Realistic endpoint |
|---|---|---|---|---|
| R2 row-formula non-reproduction | T1 / T5-guard | AXIOM-THRESHOLD-NORMALIZATION | zeta-regularized KK-tower sum per factor, target-blind; reconcile shared $a_6$ GT/ghost routes once | OPEN → AXIOM-CLOSED if verified to yield rows; DERIVED if zeta-sum lands clean; REFUTED if it misses |
| R1 deep check / byte-trace | machine-lane | (none) | byte-anchor spectra + run DEEP check (3b), not just re-sum (3a) | BLOCKED → VERIFIED (sum-level); R2 stays open |
| R3 injected $\delta$ | T6 (dependent) | AXIOM-DELTA-IS-COLUMN-SUM | (= R2) | OPEN, dissolves when R2 closes |
| R4 forcedness declared-not-proven | T5 | AXIOM-ROW-FORCEDNESS (= R2) | state "enumeration not forcedness" at headline | DISCLOSED-CORRECTED; AXIOM-CLOSED at R2 |
| R5 $M_U$ convention | T2 + T5 | AXIOM-MU-IS-CROSSING | $\mathcal{O}(1)$-seam firewall (folds into R1) | DISCLOSED → AXIOM-CLOSED if $\mathcal{O}(1)$ geometric |
| R6 $b_i$/$M_Z$ inputs | — (given-E) | (none) | — | DISCLOSED (given-E); not closeable in SG-7 |
| R7 chamber-center spectrum | T1 | AXIOM-CHAMBER-CENTER-SPECTRUM | Weyl-center uniqueness (shared w/ SG-6) | AXIOM-CLOSED; DERIVED if uniqueness proven |
| R8 $\sum Y^2$ label | owner edit | (none) | one-line relabel (folds into R1) | DISCLOSED-CORRECTED |
| R9 spectrum uniqueness | scoped | (none) | keep scope explicit; export rival-tie | OPEN (scoped) / DISCLOSED |
REDUCE-vs-RELOCATE verdict. The plan does not turn one hard problem into three harder ones: the gate reduces to two real targets (R1 mechanical, R2 one bounded computation), with the rest dependent or scope. R2 is smaller than the original gate (one zeta-regularized sum per factor, four checkable rows) and carries an explicit κ³/π falsification test so a tuned normalization cannot be banked. The one genuine danger is misclosing R1 by accepting the pasted log; the §3.3-vs-§3.4 split is the firewall. Honest expected outcome of a full campaign: R1 VERIFIED (sum-level); R2 OPEN → AXIOM-CLOSED-pending-verification or sharper-OPEN; R3 dissolved-with-R2 or OPEN; R4/R5/R8 DISCLOSED-CORRECTED; R7 AXIOM-CLOSED (shared with SG-6); R6/R9 DISCLOSED. No DERIVED-CLOSED is promised. [Source: 01_DOSSIER.md §4.10.]
The grade, held. SG-7 is OPEN / Diagnostic-only. The live chip is Diagnostic only (signs derived; magnitudes demonstrated-fitted), direction held. STATUS-UPGRADES:0. This dossier expands the spine; it does not upgrade it.
What is explicitly NOT claimed. Three bright lines, denied by the frozen docs, are never printed as proven here [Source: 00_HANDOFF_README.md bright-line block; GATE_REGRADE_… lines 418–423]:
What IS genuine (the win, stated as strength). The sign structure of the finite thresholds — gauge+ghost negative (asymptotic-freedom sign), matter/hypercharge positive, the hypercharge-matter $\sum Y^2$ term driving $\delta_1 > 0$ — is real recovered physics, terminating on the observed SM content E and the measured $\alpha_i(M_Z)$. It is generic heat-kernel / asymptotic-freedom physics that survives without the specific geometry, which is exactly why it is bankable. The Higgs winding $n_H=1$ is verified load-bearing (row 7 equals the $n_H=0$ miss $(+1.047,-0.211,0)$). And the program shipped a target-blind machine that mechanically demonstrates its own ledger is fitted — turning a conceded weakness into a runnable, reproducible finding.
Dissolved ≠ solved; selection ≠ derivation; given-E ≠ derivation-of-E. SG-7 is a consistency/pruning pass on the selected survivor, not a determination that this spectrum is the unique unifier (R9, correctly scoped). The signs are derived given E; E itself is imported, not produced here (R6). The chamber-center spectrum is read at a witness whose minimum-status is open upstream at SG-6 (R7).
The anchors paid. SG-7 rests on one ATOMIC measured anchor — ANCHOR-SG7-MEASURED-WORLD-FACTS ($E_{\rm SM}$, $\alpha_i(M_Z)$, $M_Z$; terminal as measured invariants, given-E) — plus two AXIOM-OPEN, not-atomic legs: the spectral-domain posit (AXIOM-FROZEN-ADMISSIBLE-SPECTRAL-DOMAIN, blocked on the unproven $\vec u=(1,1,1)$ Weyl-rigid uniqueness) and the threshold-trace functional (AXIOM-CANONICAL-ADMISSIBLE-MODE-TRACE, normal-form PROVEN via Thm SG7-R2B but magnitudes OPEN / computation-debt). The threshold magnitudes terminate on no measured invariant today — only on one named heat-kernel SCHEME-anchor (the per-factor KK zeta-normalization $Z_{\mathcal X}$ + overlap $c_{a,b,i}$, the scheme decision shared with $c_{\rm loop}$ and $a_6$). AXIOM-CLOSED ≠ atomic; ANCHORED ≠ DERIVED. [Source: 00_HANDOFF_README.md post-atomicity re-grade.]
The strongest honest one-sentence endpoint. SG-7 reduces to a clean floor of exactly one ATOMIC measured anchor ($E_{\rm SM}$, $\alpha_i(M_Z)$, $M_Z$ — terminal, given-E) plus two AXIOM-OPEN legs — the spectral-domain posit (blocked on an unproven $\vec u=(1,1,1)$ Weyl-rigid-uniqueness theorem) and the threshold-trace functional (normal-form proven, magnitudes open / computation-debt) — where the genuine, non-promoting wins are the recovered threshold SIGNS, the negative theorem SG7-R2 that kills every scalar/separable normalization repair, and the executed target-blind FINDING blind δ ≠ printed demonstrating the printed δ-triple is fitted/injected; so the gate stays OPEN / Diagnostic-only at the honest ceiling "serious candidate / recovers threshold signs, NOT a validated unification," with the only path to DERIVED-GIVEN-E being a real (non-placeholder) ζ-regularized KK row-regeneration on byte-anchored spectra that lands on the printed rows target-blind; STATUS-UPGRADES:0; the printed δ-triple is a post-hoc falsifier, an input to nothing. [Source: 00_HANDOFF_README.md §3A candidate endpoint.]
Dossier built from the SG-7 completion handoff package, the executed closure result + FINDING, the attack dossier (residuals R1–R9, witness ledger W1–W10, the §2.3 deep-check arithmetic), and the 2026-06-29 gate-regrade. Frozen branch dcc66f1b2685 / a5b1e6f9d951 READ-ONLY. Common construction material is the published manuscript (Paper I, GUT.html §6.7 / §5.6 / Appendix G / A1.10–A1.11 / R0) — referenced, not duplicated. STATUS-UPGRADES:0.