G Model Lorentz Tests | Complete Current Audit

Complete Lorentz Tests audit

G MODEL Complete Lorentz-Related Test Audit and Current Re-Test Website Test-Section Research Report | Gurbuz Dolu | 12 September 2026 Scope: all retained G Model records located in the Library, with duplicate copies and repeated write-ups de-duplicated. Historical failures are preserved; later developments are used only where they genuinely change the scope or status of an earlier test.

Distinct Lorentz-related checks 39

PASS / structural or conditional PASS 26

FAIL / rejected / no-go test routes 12

Overarching all-scale status 1 OPEN

Current controlled verdict The G Model's local four-null geometry, inverse-metric reconstruction, six-bivector Lorentz/Hodge structure, inter-cell covariant transport, co-transforming six-channel operator, and accepted finite-cell reorientation observables pass their Lorentz-covariance tests. The rigid permanently oriented lattice implementation is rejected. Later gauge/action audits structurally remove the direction of the local centroid U from accepted spinless local observables. However, complete all-scale microscopic no-preferred-frame dynamics remains OPEN because a fully derived physical graph/reconnect law and exact high-k relational propagation law are still missing.

Important counting rule A FAIL below means that a particular construction, criterion, or implementation failed its stated requirement. It does not mean that the whole G Model failed. Several failures directly produced the later corrected construction.

Re-test policy used for this report Algebraically reproducible failures were re-tested with the current four-null geometry. Definition-level historical failures were dependency-audited against later replacement structures. Newer results narrow an old OPEN result only where they address the same requirement without changing the test target.

1. Complete De-duplicated Test Register The following 39 entries count mathematically distinct checks once. Repeated manuscript copies, repeated boosts of the same identity, and later restatements are not counted as new tests.

# Test Core mathematics Status Current interpretation / re-test

1 Four-null Gram signature spec(G4) = {-4, 4/3, 4/3, 4/3} Lc^2 PASS Retest unchanged. Exact 1+3 Lorentzian signature; independent of Tc magnitude.

2 Inverse Gram reconstructs K G4^-1 K^T = eta PASS Retest max numerical residual Minkowski metric 1.67e-16.

3 Directional-derivative (DM)^T G4^-1 (DM) = eta^(mu nu) PASS Signed inverse Gram gives the Lorentz scalar d_mu M d_nu M correct Lorentz contraction.

4 Equal-positive branch wave (1/4) sum_a (k_a.d)^2 = d0^2 + (1/3) FAIL Retest confirms it is not Box = operator nabla^2 -d0^2+nabla^2. Static averaging role only.

5 Common / explicit Lorentz k'_a = Lambda k_a; k'_a.k'_b = PASS Retest beta=0.6: null residual boost invariants k_a.k_b <=3.31e-16; Gram change <=2.22e-16.

6 Six-bivector Lorentz metric signature(Gamma)=(3,3) PASS Native six-bivector geometry remains Lorentzian.

7 Physical Lorentz Hodge J_*^2=-I6; J_*^T Gamma J_*=-Gamma PASS Proper Lorentz compatibility star retained.

# Test Core mathematics Status Current interpretation / re-test

8 Centroid-projected 3:1 E_U^2=8/3; B_U^2=8/9; PASS Covariant only when the split is bivector weighting E_U^2:B_U^2=3:1 defined relative to local U; raw E/B split is frame-dependent.

9 Rigid tetrahedron exact sum n_x n_y n_z = 4/(3 sqrt(3)); 4th FAIL Retest unchanged. This rejects a rotational isotropy moments != sphere permanently aligned microscopic crystal, not the dynamic-orientation model.

10 Orientation-balanced Abar[M]=M+(Lc^2/6)nabla^2M+(Lc^ PASS Leading correction is isotropic. long-wave isotropy 4/120)nabla^4M+... Uniform orientation averaging still needs physical realization.

11 Old Euclidean parent R as R^T R=I, R^3=I but R^T Gamma R FAIL Historical route still fails as Lorentz physical Lorentz transport != Gamma transport; superseded by Omega/connection transport.

12 Inter-cell Lorentz Omega_yx=K_y Pi_yx K_x^-1; PASS Exact metric compatibility once transporter Omega^T eta Omega=eta branch correspondence is supplied.

13 Induced six-bivector T_yx=wedge^2(Omega_yx); T^T PASS Exact for proper transporter Gamma T=Gamma; T J_*=J_* T orientation-preserving matched links.

14 Branch matching on Pi_Celal in argmin ||gamma_x-Pi^T PASS UPDATED: existing centered-link existing links gamma_y Pi||_F^2 correspondence is exact modulo V4; physical link birth/death remains OPEN.

# Test Core mathematics Status Current interpretation / re-test

15 Pure relative-frame Pi=I => Omega_loop=I FAIL Old route remains flat/pure gauge. connection as generic Later area-holonomy curvature is a curvature different branch, so this failure is scoped.

16 Reciprocal M deformation A(M)=diag(e^M,e^-M,e^-M,e^-M); FAIL Retest unchanged. M belongs to as Lorentz transform A^T eta A != eta for M!=0 coframe/metric deformation, not Lorentz-frame transformation.

# Test Core mathematics Status Current interpretation / re-test

17 Fixed-matrix commutator [W,P]!=0; rank=2 for beta=3/5 FAIL Exact calculation; the criterion itself criterion was too strong. P must co-transform, not commute in every basis.

18 Co-transforming physical P P_phys -> D P_phys D^-1 PASS Local six-channel update transforms covariantly.

19 Inter-cell covariant Delta'=D_y Delta; Delta'^T Gamma PASS Cell-by-cell local Lorentz gauge difference / Gamma norm Delta'=Delta^T Gamma Delta consistency.

20 Preferred timelike centroid gamma=K^T eta K; 16=10+6 PASS UPDATED structural pass locally: U U in accepted spinless local Lorentz-fiber decomposition direction is gauge-fiber data for observables accepted spinless action/tensor observables. Full interacting network still not closed.

21 Complete all-scale All physical observables independent OPEN CONTROLLING boundary: high-k microscopic of absolute orientation / motion propagation plus physical no-preferred-frame relative to U graph/reconnect evolution are not dynamics yet derived unconditionally.

# Test Core mathematics Status Current interpretation / re-test

22 Explicit beta=0.6 local-cell Lambda^T eta Lambda=eta; 4 nulls, PASS Independent numerical falsification stress test 6 pair products, 6 areas preserved challenge reproduced with current cell geometry.

23 Continuous 360-degree W_f(theta)=Lambda_x W_f(0) PASS Exact finite-cell all-angle gauge local-frame / holonomy test Lambda_x^-1 invariance for conjugacy-class observables.

24 All 24 tetrahedral U_pi P = P U_pi for all pi in S4 PASS Exact structural pass. relabelings vs P

25 120-degree edge holonomy H^3=I; [H,P]=0 PASS Exact; Lorentz bivector lift preserves vs P Gamma and J_*.

26 Fixed-frame Hessian L_fixed: diagonal coefficient 3/2, FAIL Retest 20,000 random H,R: response rotational off-diagonal coefficient 1 covariance error remains order 1; covariance max 2.50 in current rerun.

27 Dynamic principal-axis L_dyn(H)=(3/2)[H-(tr H/3)I] PASS Exact SO(3) covariance; retest max Hessian response discrepancy 3.66e-15. Physical alignment law remains CONDITIONAL.

28 Covariant dynamic tidal / E_mn=h_m^a h_n^b nabla_a nabla_b PASS Formal Lorentz-tensor construction; shape tensor M-(1/3)h_mn h^ab nabla_a nabla_b M source-to-shape amplitude remains model-dependent.

# Test Core mathematics Status Current interpretation / re-test

29 Infinite connected +/-axis <n_i n_j>=(1/3)delta_ij PASS Exact long-wave isotropy of the network: second moment six-direction stencil.

30 Infinite connected +/-axis <n_x^4>=1/3, <n_x^2 n_y^2>=0 vs FAIL High-k lattice anisotropy remains for network: fourth moment 1/5,1/15 frozen connectivity.

31 Naive finite-cell dispersion Tc=sqrt(3)/(c k) arccos(v_g/c) without FAIL Coordinate expression is not an in arbitrary observer frame transforming cell frame arbitrary-frame scalar.

32 Covariant cell-frame k_t^2=h^mn q_m q_n; Tc=sqrt(3)/(c PASS Coordinate Lorentz covariance PASS dispersion extraction k_t) arccos(v_g,t/c) when t^mu co-transforms.

33 Exact microscopic Lorentz q^2=k_t^2-(4/a^2)sin^2(a k_t/2), FAIL k_t^2=q^2+(t.q)^2 remains. invariance of sinusoidal a=2cTc/sqrt(3) Changing Tc does not remove finite-cell dispersion preferred-frame dependence for nonzero a.

34 Continuum limit of a k_t -> 0 => q^2 -> 0 PASS Ordinary massless Lorentz-invariant finite-cell dispersion propagation recovered at leading order.

# Test Core mathematics Status Current interpretation / re-test

35 Naive 3-space Tc^2=(5 pi hbar/(12 c^3))(nabla^2 FAIL Do not use unchanged outside the source-response in M/rho) source rest frame. arbitrary frame

# Test Core mathematics Status Current interpretation / re-test

36 Covariant source-rest Q_u=(g^mn+u^m u^n)nabla_m PASS Observer-independence PASS in extraction nabla_n M; Tc^2=(5 pi hbar/(12 pressureless weak-source domain; c^3))(Q_u/rho0) full stress-energy extension remains OPEN.

37 Native area-holonomy S_conn/hbar=zeta sum PASS Structural branch: 20,000 proper Lorentz-scalar action stress <B/Lc^2,(J_*+beta I)K_f>_Gamma boosts to beta<=0.95 preserved test scalar contractions to <9e-14. Action uniqueness/normalization remain open.

38 Local Lorentz fiber / K->Lambda K leaves gamma; source PASS 20k one-cell + 20k loops + 20k tensor-source stress audit reconstruction and loop conjugacy symmetric sources + 5k P-similarity unchanged audits in record; current rerun preserved Gram to ~1.53e-14.

39 Can six Lorentz gauge 16=10+6; Lorentz fiber has 6 gauge FAIL NO-GO for deriving four H_ab directions force four directions but no independent conservation identities from Lorentz source-conservation 4-translation gauge gauge alone. This is a equations? conservation/symmetry limitation, not a Lorentz-covariance failure.

2. Core Mathematical Lorentz Results 2.1 Four-null Gram and metric reconstruction For four future-directed tetrahedral null branches, the normalized branch Gram matrix has zero diagonal and -4/3 off diagonal. Its spectrum is exactly one negative and three positive eigenvalues: G4_aa = 0, G4_ab = -4/3 (a != b) spec(G4) = {-4, 4/3, 4/3, 4/3} G4^-1 = (3/4) I4 - (1/4) 11^T K G4^-1 K^T = eta

Current numerical re-test of the reconstruction gave a maximum residual of 1.67 x 10^-16. The result is scale-free: multiplying every branch by Lc=cTc does not change the normalized signature or reconstructed metric.

2.2 Why the old equal-positive wave operator still fails (1/4) sum_a (k_a . partial)^2 = partial_0^2 + (1/3) nabla^2 but Box = -partial_0^2 + nabla^2

No single overall scalar can repair both the time sign and the 1/3 spatial coefficient. The correct Lorentz contraction uses G4^-1; equal-positive averaging remains useful only in its static averaging role.

2.3 Bivector metric and Hodge structure signature(Gamma) = (3,3) J_*^2 = -I6 J_*^T Gamma J_* = -Gamma

The local 3:1 E/B-like ratio survives only when the split is defined covariantly relative to the local timelike direction; the raw coordinate split is frame dependent.

2.4 Inter-cell transport and local gauge covariance Omega_yx = K_y Pi_yx K_x^-1 Omega_yx^T eta Omega_yx = eta Omega_yx -> Lambda_y Omega_yx Lambda_x^-1 T_yx = wedge^2(Omega_yx) T_yx^T Gamma T_yx = Gamma, T_yx J_* = J_* T_yx

The fixed-matrix commutator test is retained as an exact calculation but is no longer interpreted as a Lorentz failure. The correct physical rule is conjugation covariance: P_phys -> D P_phys D^-1 Delta_yx = F_y - T_yx F_x -> D_y Delta_yx Delta_yx^T Gamma Delta_yx = invariant

3. Orientation, Preferred-Frame, and Reconnect Re-Tests 3.1 Rigid tetrahedron: failure preserved sum_a n_ax n_ay n_az = 4/(3 sqrt(3)) != 0 T_xxxx = 1/9, T_xxyy = 1/9 isotropic sphere: 1/5 and 1/15

The exact third- and fourth-moment failures remain. Therefore a permanently aligned tetrahedral crystal is not accepted as the physical microscopic vacuum implementation.

3.2 Dynamic 360-degree local orientation: later pass K_x(theta)=Lambda_x(theta)K_x(0) W_f(theta)=Lambda_x(theta) W_f(0) Lambda_x(theta)^-1 spec W_f, tr(W_f^n), det W_f = invariant

This is an exact all-angle finite-cell gauge pass for accepted local/holonomy observables.

3.3 Fixed-frame Hessian failure and dynamic-principal-axis repair L_dyn(H) = (3/2)[H - (tr H/3) I] L_dyn(R H R^T) = R L_dyn(H) R^T

Current 20,000-sample rerun: dynamic covariance max discrepancy 3.66 x 10^-15; fixed-frame discrepancy remains order unity (max 2.50). The physical rule that a cell aligns with instantaneous principal axes remains conditional.

3.4 Preferred timelike centroid U - refined current status gamma_x = K_x^T eta K_x 16 components of K_x = 10 metric/Gram + 6 local Lorentz-fiber components K_x -> Lambda_x K_x leaves gamma_x unchanged U_x=(1/4)K_x 1 -> Lambda_x U_x

Later spinless gauge/action audits upgrade the old blanket U-direction concern: within accepted local spinless gauge-invariant observables, the direction of U is frame/gauge data. This is a STRUCTURAL PASS locally. The stronger all-scale network claim remains OPEN.

3.5 Celal matching narrows an old OPEN result Pi_Celal in argmin_{Pi in S4} ||gamma_x - Pi^T gamma_y Pi||_F^2

Existing centered-link correspondence is exact modulo V4. The unresolved problem is physical link birth/death and graph evolution, not ordinary branch correspondence on an already existing link.

4. High-k Propagation and the Remaining Lorentz Problem 4.1 Frozen finite-step networks still retain high-k anisotropy <n_i n_j> = (1/3) delta_ij [second moment PASS] <n_x^4> = 1/3, <n_x^2 n_y^2> = 0 [not isotropic 1/5, 1/15]

Long-wavelength isotropy can coexist with finite-step anisotropy. The current model rejects a frozen permanent stencil as final fabric, but the full relational graph dynamics remains open.

4.2 Sinusoidal finite-cell dispersion a = 2 c Tc / sqrt(3) chi^2 = (4/a^2) sin^2(a k_t/2) q^2 = k_t^2 - (4/a^2) sin^2(a k_t/2) k_t^2 = q^2 + (t.q)^2

When the local cell direction t^mu co-transforms, the extraction is coordinate-covariant. But the finite-cell correction still depends on k_t, so the current sinusoidal law is not an exact no-preferred-frame microscopic law. q^2 = (a^2 k_t^4)/12 - (a^4 k_t^6)/360 + ... -> 0 as a k_t -> 0

Changing the numerical value of Tc changes the physical wavelength at which finite-cell corrections become important. It does not remove the structural t^mu dependence. Therefore this failure cannot be repaired merely by selecting another Tc.

4.3 Source-response route has a cleaner Lorentz form h_u^(mu nu) = g^(mu nu) + u^mu u^nu Q_u = h_u^(mu nu) nabla_mu nabla_nu M Tc^2 = (5 pi hbar/(12 c^3)) (Q_u/rho_0) = (5 pi hbar/(12 c)) (Q_u/epsilon_0)

Here u^mu is the physical matter four-velocity, not a universal vacuum vector. In the pressureless weak-source domain this formulation passes observer-independence. The full stress-energy extension remains open.

5. Numerical Stress Audits with the Current Architecture Re-test Current result Verdict

Gram eigenvalues {-4, 1.333333333, 1.333333333, 1.333333333} PASS

Metric reconstruction max |K G4^-1 K^T - eta| = 1.67e-16 PASS

Explicit beta=0.6 boost max null residual 3.31e-16; max Gram shift 2.22e-16 PASS

M deformation as Lorentz norm(A^T eta A-eta) = 0.384 at M=0.1; 6.56 at M=1 FAIL map

Rigid tetrahedron 3rd sum n_x n_y n_z = 0.7698003589 != 0 FAIL moment

Rigid tetrahedron 4th <n_x^4>=<n_x^2 n_y^2>=1/9, not isotropic FAIL moment

Dynamic Hessian 20,000 samples; max error 3.66e-15 PASS covariance

Fixed Hessian covariance 20,000 samples; order-one error, max 2.50 FAIL

Random local boosts 10,000 boosts to beta below 0.95; Gram max drift 1.53e-14 PASS

Source tensor frame 5,000 samples; max numerical discrepancy 2.80e-13 PASS reconstruction

These are internal numerical consistency checks, not independent experimental verification.

Area-holonomy / local-action stress record The retained working record also reports 10,000 nondegenerate coframe rank tests, 20,000 random proper Lorentz transformations up to beta=0.95 for action contractions, 20,000 one-cell and 20,000 loop frame tests, 20,000 symmetric-source reconstructions, and 5,000 P-similarity checks. These support a structural local-Lorentz-gauge closure in the present spinless action branch, but not the complete physical graph dynamics.

6. Deep Note: Use of T_c in Lorentz Tests The historical audit found six distinct numerical routes that actually produced a T_c value. They are not six independent confirmations. Four are current reconstruction routes and two are historical diagnostics. None is yet a fully independent Big-G-free experimental determination of the absolute scale.

# Route T_c (s) L_c (m) Difference Status

1 Measured Big-G 2.18657023792000 6.555172662157e-3 +0.0000 ppm CALIBRATED calibration 0e-43 5

2 K_M -> T_c inversion 2.18657023792000 6.555172662157e-3 +0.0000 ppm EXACT inversion; current 0e-43 5 numerical K_M provenance not independent

3 Mercury perihelion 2.18656511942973 6.555157317309e-3 -2.3409 ppm OBSERVATIONAL back-inversion 0e-43 5 inversion / calibrated compatibility

4 Moved-source lab 2.18657023752900 6.555172660985e-3 -0.0002 ppm EXPERIMENTAL form / reconstruction demo 0e-43 5 numerical demonstration only

5 Historical K_GD 2.20046945400000 6.596841463686e-3 +0.6357% HISTORICAL / provisional bare-SI closure 0e-43 5 / superseded

6 Historical Earth 2.47000000000000 7.404873712600e-3 +12.9623% HISTORICAL / illustrative source-curvature 0e-43 5 / not independent diagnostic

Is it scientifically acceptable to use T_c here? Yes, if it is labeled correctly. For dimensional examples, cell lengths, numerical wavelengths, or plots, use the retained calibrated reference T_c,ref = 2.186570237920 x 10^-43 s and L_c = 6.555172662157 x 10^-35 m, explicitly labeled CALIBRATED / REFERENCE.

For the principal Lorentz pass/fail identities, the absolute number is generally irrelevant. The normalized Gram signature, inverse-metric reconstruction, Lorentz/Hodge algebra, transporter identities, and local gauge conjugation laws do not change when Tc is replaced by another positive reference scale because Lc factors cancel or scale homogeneously. Therefore it would be misleading to run the same algebraic Lorentz test six times and count six validations. The six Tc values are a scale-sensitivity note. The historical Earth diagnostic differs by about 12.96%, yet it does not change the normalized Lorentz verdict; it only shifts Lc and the wavelength assigned to finite-cell effects. For the sinusoidal finite-cell law, the Lorentz issue is structural for every nonzero a=2cTc/sqrt(3): changing Tc does not remove the k_t / t^mu dependence.

7. Website-Ready Scientific Summary Recommended wording for the Tests section Across the retained G Model research record, 39 distinct Lorentz-related mathematical and numerical checks were identified after de-duplicating repeated write-ups. Twenty-six pass in their stated exact, structural, or conditional domains; twelve specific constructions or criteria fail and are retained as falsified or superseded routes; one overarching question remains open: complete all-scale microscopic no-preferred-frame dynamics. The strongest established result is Lorentz covariance of the local four-null/six-bivector architecture and its accepted inter-cell gauge transport. The model exactly reconstructs the Minkowski metric from the four-null Gram matrix, preserves the Lorentzian six-bivector metric and Hodge structure, and passes explicit boost, local-frame, holonomy, and tensor-reconstruction stress tests. A rigid permanently oriented tetrahedral lattice fails higher-order isotropy and is not the accepted physical implementation. Later dynamic-orientation and local-gauge analyses substantially narrow the historical preferred-frame concern: in the accepted spinless local action/tensor sector, absolute U direction is gauge-fiber data and does not appear in accepted gauge-invariant observables. Nevertheless, the model still withholds a claim of full microscopic Lorentz invariance until the physical graph/reconnect law and exact finite-cell/high-k propagation are derived so that no all-scale observable can retain a preferred direction.

Most Important PASS Results What matters most is not the raw number of PASS results, but whether the foundational Lorentz consistency gates pass. The most important successful results are: (1) the four-null Gram has the exact 1+3 Lorentzian signature; (2) its inverse reconstructs the Minkowski metric exactly; (3) the six-bivector sector carries the proper (3,3) Lorentz metric and Hodge structure; (4) inter-cell transport and the co-transforming six-channel operator obey the correct local Lorentz transformation laws; and (5) continuous 360-degree local-frame/holonomy observables are exactly invariant under the accepted gauge interpretation. Later spinless action and source-tensor audits also show that the direction of U is gauge-fiber data in the accepted local observable sector.

These passes are the central reason the Lorentz sector remains viable after the rejected branches. They establish that the retained local mathematical architecture can be formulated Lorentz-covariantly without fitted Lorentz coefficients. They do not, by themselves, prove that the complete microscopic fabric exists in nature.

The most important PASS still required The decisive remaining target is all-scale microscopic no-preferred-frame dynamics. To close it, the native physical reconnect/graph-evolution law and finite-cell high-k propagation must be derived so that every measurable gauge-invariant observable is independent of absolute tetrahedral orientation or motion relative to a vacuum frame. A clean PASS of that test would materially strengthen the claim that the G Model can serve as a fundamental spacetime theory; an unavoidable preferred-frame observable would instead be a fundamental challenge.

How the FAIL results should be interpreted Failed implementation does not equal failed G Model. The recorded FAIL results do not constitute a falsification of the G Model as a whole. They falsify particular implementations, operators, assumptions, or test criteria explored during development. Several failures directly motivated corrected Lorentz-covariant constructions and remain in the record so that superseded routes are not hidden.

Examples include the rejection of equal-positive branch averaging as the relativistic wave operator, rejection of a permanently aligned rigid tetrahedral crystal as exactly isotropic, rejection of the old Euclidean parent R as physical Lorentz transport, and reclassification of the fixed-matrix commutator requirement after the co-transforming P rule was established. These are scientifically useful negative results because they narrow the admissible theory rather than being silently discarded.

The principal unresolved Lorentz question is whether the complete interacting finite-cell network can be shown to have no observable preferred frame at all scales. If future native dynamics proves that an observable preferred-frame signal is unavoidable, that would constitute a fundamental challenge to the G Model. Likewise, a contradiction in an indispensable core identity, an inability of any completed source dynamics to satisfy required conservation, or a genuine quantitative prediction of the completed model that conflicts with experiment would be potentially falsifying. Those stronger outcomes have not been established by the twelve recorded FAIL routes in this audit. Scientific-status boundary: these are internal mathematical/covariance tests and AI-assisted numerical audits. They are not experimental confirmation of the G Model. PASS means the declared mathematical test closes; it does not mean nature has been shown to use the proposed microscopic causal fabric.

8. Source Record Map ID Record Use

S1 G_Model_Lorentz_Covariance_and_Microscopic_Invariance Master Lorentz register and dynamic-fabric _Audit_v2_DYNAMIC_FABRIC_2026-08-28.pdf addendum.

S2 G_Model_Dynamic_360_Lorentz_Test_31AUG2026.md Continuous 360-degree local-frame and reconnect audit.

S3 G_Model_Lorentz_Covariance_Stress_Test_Landscape_08S Explicit beta=0.6 boost stress test. EP2026.pdf

S4 G_Model_Post_Manuscript_Scale_Research_Report_03SEP Lorentz audit of T_c extraction routes and 2026.pdf preferred-frame stress test.

S5 G_Model_Author_Contribution_Working_Record_2026-08-2 Later local Lorentz gauge/action stress audits 8_v7_HISTORY_READY.md and U-fiber closure.

S6 G_Model_Einstein_Native_Audit_Checkpoint_31AUG2026. 16=10+6 Lorentz gauge inventory and md conservation no-go.

S7 G_Model_CURRENT_MANUSCRIPT_55P_MMM_AAA_XY Current controlling manuscript status. Z_LAW_UPDATE_12SEP2026.pdf

S8 G_Model_Tc_Numerical_Reconstruction_Historical_Audit_1 Six numerical T_c routes and independence 2SEP2026.pdf audit.

Authorship and verification boundary G Model conceptual framework, research direction, interpretation, and final scientific responsibility: Gurbuz Dolu. AI assistance was used for retrieval, mathematical formalization, algebraic/numerical re-testing, status reconciliation, and document preparation. This is an internal research audit for transparent public presentation and independent falsification; it is not peer review or experimental validation.