# Adversarial mathematical audit of the trunk Date: 2026-07-11 Scope: the unsigned constraint, information-metric statics, dark-state statement, and the n48 frequency knee. This audit does not address the later matter, tensor, or cosmological constructions. ## Verdict in one paragraph The Gaussian fidelity metric formula and the constrained-minimum formula are mathematically sound within the finite harmonic lattice. The stronger causal claims are not. In one dimension, the Coulomb Green function and attractive sign do not diagnose the Fubini--Study metric: they persist for the identity and for other local positive-definite costs. They are consequences of the unsigned bipartite incidence operator together with positivity. The n48 Lorentzian is likewise the transfer function of the assumed first-order relaxation ODE, and zero response at rest follows from defining the monitored quantity to be the residual driven to zero. Moreover, n48 fails three of its four original preregistered gates; its advertised 0.3% result is explicitly post hoc. These are valid conditional toy-model properties, not independently derived predictions of the substrate Hamiltonian. ## 1. What checks out For a real Gaussian ground state psi_Omega(x) proportional to det(Omega)^(1/4) exp[-x^T Omega x / 2], expanding the exact overlap gives the quadratic fidelity susceptibility G_ef = (1/8) Tr[Omega^-1 (partial_e Omega) Omega^-1 (partial_f Omega)]. The Frechet derivative used in `n46_infometric.py`, (partial_e Omega)_nm = (v_n(i)-v_n(j))(v_m(i)-v_m(j)) /(omega_n + omega_m), is the correct divided difference for the matrix square root. The script's finite-difference overlap check reports relative error 1.03e-6. For a strictly positive quadratic cost G and feasible constraint A delta-k = -c, the Lagrange-multiplier solution is delta-k* = -G^-1 A^T (A G^-1 A^T)^+ c and the minimized cost is E(c) = (1/2) c^T (A G^-1 A^T)^+ c, subject to projecting onto the feasible subspace. That algebra is correct. ## 2. The one-dimensional force is encoded by A, not selected by G On the even ring, the unsigned incidence matrix has symbol A(q) = 1 + exp(-iq), and hence |A(q)|^2 = 4 cos^2(q/2). Its zero is at the staggered point q=pi. Writing q=pi+p gives |A(pi+p)|^2 = p^2 + O(p^4). For any translation-invariant positive local cost with a finite nonzero symbol G(pi), the minimized source kernel behaves as G(q)/|A(q)|^2 = G(pi)/p^2 + regular terms. The long-range lattice Green function is therefore forced by the zero of the declared constraint operator. The information metric changes its coefficient and short-range corrections; it does not create the pole. This was checked directly at N=200 for the same pair set used by n46: | cost G | fitted slope | Green-fit fractional RMS | |---|---:|---:| | identity | 0.0900000 | 1.25e-15 | | identity plus 0.2 nearest-neighbour correlations | 0.0540000 | 1.15e-15 | Both reproduce the claimed exact Green form and attractive sign without any vacuum information geometry. Thus the paper's statements that the sign is an output of the Fubini--Study price and that "nothing is declared" are too strong. The decisive structure was declared in A. Positivity also fixes the pair-energy sign once the admissible staggered charge sector is chosen. A positive quadratic energy penalizes separation of the opposite-parity source pair, so calling the resulting slope a discovery of attraction double-counts an implication of the chosen constraint and source class. The phrase "unique remaining vacuum-owned positive functional" is not a mathematical theorem. Excluding ground-state energy because it is concave does not establish uniqueness of the fidelity metric; infinitely many positive quadratic functionals and monotone quantum metrics remain unless extra axioms are stated and a uniqueness theorem is proved. ## 3. The knee is the assumed relaxation transfer function n48 assumes d(delta-k)/dt = -gamma [delta-k - delta-k*(c(t))]. For a sinusoidal target f(t), the steady-state lag x obeys x_dot + gamma x = -f_dot delta-k*_0. Consequently |x(omega)|^2 proportional to omega^2/(gamma^2 + omega^2). This is exactly the advertised Lorentzian. No property of the oscillator Hamiltonian, Fubini--Study metric, or Lindblad spectrum selects this response; it is the transfer function of the postulated single-pole overdamped ODE. Changing the relaxation law changes the prediction. Multiple relaxation modes produce a sum of filters, inertia produces resonances, and memory kernels need not be Lorentzian. Likewise, n48 monitors v = A delta-k + c while its dynamics drives delta-k to a target satisfying A delta-k* = -c. Zero steady response is therefore an algebraic fixed-point property of the definitions. This does not by itself demonstrate that a physical spatial superposition is a dark state of a single quantum Lindblad generator. The abstract Lindblad theorem is narrower and correct: if one explicitly constructs operators L_e and both dressed branch states are common eigenstates with the same eigenvalue (zero here), their superposition lies in a decoherence-free subspace. The repository's n48 computation does not perform that construction; it evolves two classical branch amplitudes and assigns the squared residual difference as a record-overlap rate. ## 4. The registered n48 gates did not pass An independent rerun of `n48_motion_decoherence.py` gives: | gate | result | |---|---| | P1, no static floor after the registered transient | FAIL: ratio 3.69e-4 | | P2, quadratic onset in the registered frequency interval | FAIL: slope 1.8677 | | P3, high-frequency saturation | PASS: slope 0.0002 | | P3, registered knee interval | FAIL: reported nu about 0.1 gamma | | P4, low-frequency gamma scaling | PASS: exponent -1.997 | The preregistration itself labels the later 0.3% Lorentzian comparison "FLAGGED POST-HOC" after changing from cyclic frequency nu to angular frequency omega and diagnosing an insufficient transient. Therefore the proper evidential statement is not "the preregistered knee passed." It is: the original gates failed, and a post-hoc comparison confirmed the analytic transfer function of the assumed ODE. The script also confirms a correction important for dimensional statements: at low frequency its squared residual scales as gamma^-2, not gamma^-1. ## 5. Reproducibility warning On the current NumPy installation, both n46 and n48 emit repeated overflow, divide-by-zero, and invalid-value warnings inside matrix multiplications even though the final finite results reproduce the stored claims. A clean audit should make floating-point exceptions fatal, record the BLAS/NumPy versions, report condition numbers, and verify the same outputs with a second numerical backend. A validation script should not silently continue after `invalid` warnings. ## 6. Claims that survive this audit The following narrower statements survive: 1. The stated finite Gaussian lattice has a computable positive fidelity metric. 2. Minimizing a positive quadratic cost under the unsigned bipartite constraint gives a Green-function interaction in the admissible sector. 3. A violations-only monitor has a decoherence-free subspace if the relevant dressed quantum branch states are common zero modes of explicitly defined Lindblad operators. 4. A single-pole overdamped relaxation model driven sinusoidally has zero steady residual at zero frequency and a Lorentzian squared-lag response. None of these statements yet establishes that nature implements the unsigned constraint, the chosen cost, the single-pole relaxation law, or the required joint matter--substrate Lindblad operators. ## 7. Next decisive mathematical gates 1. State and prove a uniqueness theorem for the information metric, or retract uniqueness and sweep a declared family of positive local costs. 2. Repeat the force analysis with A varied independently of G. Report which conclusions follow from the constraint symbol alone. 3. Construct the joint matter--substrate Hilbert space and explicit L_e. Demonstrate that two spatially separated dressed branches share the same Lindblad eigenvalues and remain dark under the full Hamiltonian. 4. Derive the relaxation kernel from that generator rather than postulating a scalar first-order ODE. Report all poles and branch-dependent modes. 5. Reregister frequency gates using omega, sufficient transient suppression, and a comparison against multi-pole and non-Markovian alternatives. # Second pass: induced attraction, active branch, noise, and the Q bridge ## 8. The mediator sign formula is correct but its advertised scope is not For the damped auxiliary mode used in n33, zero-frequency elimination gives gamma_eff = g^2 kappa/(Delta^2 + kappa^2/4), Omega_eff = -g^2 Delta/(Delta^2 + kappa^2/4). This algebra is correct. For a sum of modes with positive weights and with every declared Delta_m positive, Omega_eff is negative term by term. The n33 source convention maps that sign to repulsion; changing Delta to negative maps it to attraction. What has not been proved is the public phrase "every passive bath repels." The calculation proves a statement about the specified rotating-wave, linearly damped oscillator family after identifying passivity with Delta_m > 0. Detuning is frame-relative, and positive laboratory-frequency modes may have either detuning relative to a drive. More general passive susceptibilities, counter-rotating couplings, spectral densities, spatial boundary conditions, and multi-quadrature source couplings are outside the proof. The reference to a universal Casimir sign is also not a mathematical consequence of this model. The script section labelled "full vs effective (transient + NESS)" does not simulate a transient. It solves two static singular linear systems with `lstsq` and compares their zero-frequency steady states. The reported exact agreement therefore validates static elimination, not dynamical agreement of the full and effective models. ## 9. The so-called active branch contains no gain or inversion In n183 the negative-Delta branch uses H_aux = Delta a^dagger a, Delta < 0, dot-rho includes kappa D[a]. The dissipator is the same vacuum-loss channel D[a] in both branches. It drives the auxiliary mode toward |0>, not toward a population-inverted state. No pump, negative-temperature reservoir, Bogoliubov gain term, population dynamics, pump depletion, or pump noise appears. Consequently: * stability under Delta -> -Delta is expected; * the magnitude of the vacuum-noise spectrum is even in Delta; * the dispersive response is odd in Delta; * no Caves-type added gain noise is required, because the model has not implemented an amplifier. The n183 computation correctly verifies an odd-response/even-noise symmetry of this linear detuning model. It does not verify the physical realizability of a pumped active vacuum. The repository's later amendment acknowledges this, but stronger public phrases such as "inverted medium" and "active vacuum is stable" remain mathematically unsupported by n183. The commutator-preservation and covariance inequalities establish that the written linear quantum stochastic equations preserve canonical algebra and admit a Gaussian state on the projected subspace. They cannot establish that an omitted pump realizing the desired negative-energy effective mode is stable or noiseless. ## 10. The n183 output boundary condition has the wrong sign The Langevin convention in the script is dot-a = ... - (kappa/2)a + sqrt(kappa) a_in. For that convention the standard passive boundary relation is a_out = a_in - sqrt(kappa) a up to a simultaneous convention change in the input coupling. The script instead uses a_out = a_in + sqrt(kappa) a. For an uncoupled one-sided passive cavity on resonance, the script's choice gives a field-amplitude transfer of 3 rather than a unit-modulus reflection. Thus its output spectrum and its statement that the reflected output is an identified idler are not valid under its own input convention. The internal matter spectra computed from the resolvent are unaffected by this particular sign error, so the evenness under Delta reversal survives. ## 11. The reported bridge coefficient 0.4834 is an averaging artifact The n128 reduced mediator permits an exact calculation. For a constant source c, |E_int| = g^2 |Delta| c1 c2/(Delta^2 + kappa^2/4). For a sinusoidal source of amplitude dc, exact cycle averaging at omega -> 0 gives Gamma = kappa g^2 dc^2/[4(Delta^2 + kappa^2/4)]. Therefore the registered ratio is R = Gamma c1 c2/(|E_int| dc^2) = kappa/(4|Delta|) = Q/4, so zeta = R/Q = 0.25. The repository reports 0.4834 because `dynamic_rate` does not integrate a complete modulation cycle. In the Q sweep it chooses omega = 1e-4 kappa, T = 2400/kappa, which covers only T omega/(2 pi) = 0.0382 cycles. The drive consequently remains near its maximum, so the code approaches the DC factor zeta = 0.5 rather than the sinusoidal cycle-average factor 0.25. The small difference from 0.5 is the partial-cycle window. Direct complex susceptibility evaluation gives, for Q = 10, 100, 1000, 10000: | averaging convention | R/Q | |---|---:| | true DC displacement | 0.5 | | sinusoid, complete-cycle mean square | 0.24999999 | The Q^1 exponent is an algebraic property of this reduced model and survives. The claimed coefficient 0.4834 does not. Any absolute rate assembled using it is high by approximately 1.93 relative to the script's stated sinusoidal protocol, before the much larger geometry and units uncertainties. ## 12. The Q relation is a shared-parameter identity, not an explanation Within the single-mode model, the static dispersive coefficient scales as |Omega_eff| proportional to g^2 |Delta|/(Delta^2+kappa^2/4), while the low-frequency record rate scales as Gamma proportional to g^2 kappa/(Delta^2+kappa^2/4). Their ratio is necessarily proportional to kappa/|Delta| because the same chosen susceptibility denominator and coupling g occur in both definitions. This is a useful consistency identity, but it does not independently explain gravity's weakness. That interpretation additionally assumes: 1. the static coefficient is Newton's interaction; 2. the monitor record rate is the physical decoherence rate; 3. the same mediator and normalization control both sectors; 4. the detuning ratio maps to the observed gravitational/electromagnetic hierarchy; 5. the anchor formula fixes the physical knee. None of those five transfers is established by the reduced-model ratio. ## 13. Updated verdict on this layer The defensible result is narrow: a damped linear mediator has an odd dispersive response, even vacuum-noise power under detuning reversal, and a dynamic/static ratio proportional to kappa/|Delta|. The repository correctly computes much of that linear-system algebra. It has not shown that attraction requires every physically passive reservoir to fail, that the negative-detuning model is a pumped active vacuum, that a real pump adds no relevant noise, or that the Q identity fixes a physical gravity/decoherence relation. The only reported numerical coefficient in the bridge, 0.4834, fails exact cycle averaging. ## 14. Next decisive gates for the monitoring layer 1. Replace the ideal negative-detuning loss model with an explicit pumped reservoir including pump and idler degrees of freedom; derive its reduced drift and noise rather than copying D[a]. 2. Correct the input-output sign convention and verify unitary reflection in the g -> 0 passive-cavity control. 3. Rerun n128 over at least ten complete periods after a separately bounded transient; compare against the exact zeta = 1/4 result. 4. State the passivity theorem with a precise Hamiltonian, state class, spectral assumptions, and coupling operators. Search for a counterexample outside the single RWA oscillator family. 5. Keep the odd-response/even-noise relation as an analog-model prediction; do not treat an analog pass as evidence for an active gravitational vacuum. # Third pass: inertia, redshift, and the PPN mapping ## 15. The adiabatic inertia formula and uniform scaling survive For a Hamiltonian H(x) whose nondegenerate ground state follows a slow coordinate x, the second-order adiabatic kinetic correction M_ad = 2 hbar^2 sum_(n != 0) ||^2/(E_n-E_0) is standard and correctly specialized to the Gaussian oscillator model. The two-site anchor in n55 agrees with the closed form to 3.0e-16. Under the script's particular uniform rescaling, the gap-weighted expression scales as k^(-5/2). Independent rerunning reproduces the reported relative deviations 8.8e-7 and 3.6e-7. This is a valid scaling result for the specified Gaussian dressing and rescaling convention. It is not yet inertial mass in a physical matter theory. The coordinate x is the externally translated center of a prescribed consumption pair; there is no dynamical matter Hamiltonian, canonical x momentum, relativistic dispersion, or demonstration that this coefficient equals passive and active gravitational mass. The calculation supplies a Born--Oppenheimer mass correction for a parameterized family of vacuum states. ## 16. The constructed clock does not oscillate in its claimed regime The proposed frequency is omega^2 = [gamma C'' + E0'']/M_ad. The monitored term gamma C'' must dominate the concave unitary term E0'' for the desired sign. The n55c rerun gives, at F=0.2, E0''/C'' = -38.9, so domination requires gamma > about 39. But underdamped motion requires omega > gamma, which in the monitored asymptotic regime implies gamma < C''/M_ad = 0.012. The two intervals are disjoint by over three orders of magnitude. The actual reported points have | gamma | domination ratio | omega/gamma | status | |---:|---:|---:|---| | 50 | 1.3 | 0.0072 | overdamped | | 100 | 2.6 | 0.0085 | overdamped | | 400 | 10.3 | 0.0052 | overdamped; above phonon-band edge | | 800 | 20.6 | 0.0037 | overdamped; above phonon-band edge | Thus the model has no operational clock in the regime used to infer the PPN parameter. The repository calls this assumption A6 and proposes an unbuilt quantum internal-level completion. Such a completion could have different scaling and damping; it cannot be used as evidence that the present model has already calculated gamma_PPN. ## 17. The formula for gamma_PPN is conditional exponent arithmetic Given all of the following definitions, C'' scales as k^-2, M_ad scales as k^-5/2, gamma_drain scales as k^a, the formal frequency scales as omega_clock scales as k^(a/2 + 1/4). Calling s = a/2 + 1/4 and separately assigning the substrate wave speed the response 1/2 gives gamma_dictionary = (1/2)/s - 1 = (1-2a)/(1+2a). The algebra is correct. The result is not a derived PPN parameter. It assumes the mapping from uniform k rescaling to a local metric potential, identifies one substrate excitation with light, identifies an overdamped curvature mode with a clock, and uses coordinate light speed without constructing rods, coordinates, null geodesics, or a four-dimensional metric. Most importantly, the local test does not reproduce the uniform result. The n55 rerun gives s_uniform = 0.25002, s_bump(w=24) = 0.4150, a 66% shift. The source-driven test has a non-linear source exponent 1.261 rather than one. No scale-separation convergence study establishes that the uniform exponent is the local long-range gravitational response. Applying Cassini's bound to a is therefore unjustified. The choice a=0 is also not derived. It is the value needed to make the dictionary return gamma=1. The monitor rate being written without k in one effective equation does not prove that its microscopic physical rate is invariant under changes of the substrate that define all local clocks and units. ## 18. The earlier registered Dicke gate actually killed the proposed map The directly registered matter-maintenance clock in n53 produced s = -1.988770, gamma_dictionary = -1.251412, and failed its redshift-sign gate. The later construction changes the clock from maintenance energy to sqrt(V''/M_ad), multiplies the information cost by an assumed monitor rate, and works in a uniform-rescaling limit. It is a new modeling route, not a repair derived from the failed clock definition. Because its operational regime is empty, it does not supersede the failed n53 result as a physical clock calculation. The strongest honest statement is that a selected ratio of uniform scaling dimensions can be made equal to the algebraic weak-field GR ratio when a is set to zero. ## 19. The quantum-redshift claim is inserted through the Hamiltonian n72 adds a multiplicative coupling whose local interaction energy is U(x) = lambda h_test epsilon_ss(x). Quantum phase evolution then gives omega = E/hbar. Potential tracking and linearity follow because U was defined to be linear in both the test energy and the local epsilon field. Universality under h_test -> h_test/2 is therefore a control on linear arithmetic, not an emergent equivalence principle. This construction does not derive the response of a transition frequency. A real clock measures an energy difference; universal redshift requires the same fractional shift for all relevant internal contributions, including binding and interaction energies. With no matter spectrum, the model cannot show that both clock levels acquire the required proportional shifts. The statement that classical clocks see only tides is also narrower than its presentation. The proof covers selected local parity-neutral substrate drift observables or the gradient-coupled n33 drift equations. It does not cover every possible classical clock coupled to a local potential. Meanwhile omega=E/hbar is ordinary quantum mechanics, so no distinctive redshift mechanism has been derived. ## 20. Two overstatements in the drainage dichotomy proof The Fourier argument correctly identifies the staggered zero of the unsigned incidence operator and the filter zero of parity-neutral observables in the declared linear class. Its one-dimensional numerical witness is strong. Two broader claims do not follow as written: 1. A positive-definite translation-invariant kernel obeys |K(d)| <= K(0), but Bochner's theorem does not imply that K(d) is strictly monotone with separation. A unique quadratic infrared zero fixes the asymptotic Coulomb form, not monotonicity at every lattice distance. 2. The claimed d>1 tidal suppression was not numerically verified. Its own T1b gate failed (18.212% versus 18.213%) and was declared void because the one-dimensional linear response vanished exactly. That cannot serve as a numerical verification of the higher-dimensional statement. The safe theorem is correspondingly narrower: in the one-dimensional unsigned model, the chosen parity-neutral observables have no linear long-range response; in the signed model, the stated positive kernel fixes the asymptotic sign under additional monotonicity conditions. ## 21. Updated verdict on the metric layer The gap-weighted Gaussian metric and its k^(-5/2) scaling are genuine finite- model results. Everything after that is a chain of conditional identifications. The chain fails inside the present toy because its clock regime is empty and its local response does not match its uniform scaling limit. Consequently gamma_PPN = (1-2a)/(1+2a) should be labeled a dictionary identity under A1--A6, not a prediction or a Cassini-tested result. The model has not calculated a post-Newtonian metric. ## 22. Next decisive gates for inertia and PPN 1. Construct an actual underdamped quantum clock within the same Hamiltonian and monitor, including linewidth, and measure a transition frequency. 2. Demonstrate convergence of a localized smooth well toward the uniform scaling response while independently taking lattice size and well width to infinity. The target is s -> 1/4 and source exponent -> 1. 3. Derive the monitor-rate exponent a from pump and bath microphysics before comparing with Cassini. 4. Construct the effective metric and compute a standard gauge-invariant PPN observable such as light deflection or Shapiro delay without defining gamma from the desired response ratio. 5. Repeat the dichotomy tests in d=2 and d=3 for explicit parity-neutral observables; do not cite the void one-dimensional T1b gate as evidence. # Fourth pass: anharmonic closure, Nordtvedt, Mercury, and the pump bill ## 23. The derivative integration-by-parts identity is correct With compatible discrete operators L = D^T D and fields satisfying L phi_A = h_A, the identities sum 2 (D phi_A).(D phi_B) = 2 phi_A.h_B, sum (D phi_A).(D phi_A) = phi_A.h_A follow immediately by discrete integration by parts. The n112c rerun reproduces the reported coefficient 2.000000000000 with residuals between 1e-14 and 1e-16. This establishes that the chosen derivative field-energy source weights the linear toy's self and cross terms with the same algebraic coefficient. It does not by itself establish the physical Nordtvedt parameter eta_N = 0. That requires a relativistic definition of total inertial mass, active mass, passive mass, binding energy, center-of-mass motion, and the external-field equations. None exists in the model. LLR and MICROSCOPE are therefore not "structural passes" of this identity; they test a much larger dynamical statement. The cubic-source divergence is also reproduced: its integrated effective charge grows from 0.148 at n=24 to 0.600 at n=48. This correctly rules out that unregulated phi^2 source as a localized 3D monopole correction in the specified periodic construction. It does not prove that the derivative vertex is unique among all local nonlinearities. ## 24. The published Mercury reproduction script is broken The current `n113_derivative_mercury.py` does not reproduce its documented result. An independent execution gives: * every nonzero-xi solve reaches the 80-iteration limit; * fields/energies grow to 1e18--1e42; * the fitted Delta-beta slope prints as zero; * the reported linearity residual is 40%. The documentation says the divergent Newton solver was replaced by exact first-order perturbation. The file still calls `solve_mediator`, which is the divergent nonlinear Newton iteration. It then continues and prints a physical sign even after convergence failure. This is a reproducibility and gate- enforcement defect. Reconstructing the actual first-order expansion gives eps = eps_0 + xi eps_1, L eps_0 = s, L eps_1 = -g(eps_0), and does reproduce Delta-beta = 2.163032581... xi. Thus the documented coefficient can be recovered analytically, but not from the cited executable as currently written. The script should fail closed when Newton does not converge and should implement the perturbative solution it claims to execute. ## 25. The Mercury-fitted xi lies outside controlled perturbation theory Setting Delta-beta = 1/2 in the first-order relation gives xi = 0.23116... At xi = 0.2312, the reconstructed first-order terms are E1 = -6.07875, E2 = +18.47904. The nonlinear correction is about 3.04 times the linear term. A first-order expansion is not controlled when its correction exceeds the leading term. The original linearity sweep only used |xi| <= 0.1, but even there the perturbative correction is already comparable to or larger than E1 at the upper endpoint. Consequently the fitted value cannot be inserted into the first-order formula without solving the full nonlinear problem and demonstrating existence, stability, branch uniqueness, and convergence. The available full Newton solver instead diverges. The value xi = 0.2312 is therefore not a valid calibrated parameter of a demonstrated nonlinear solution. There is a deeper interpretive issue: Delta-beta itself is defined through an imported toy dictionary, Delta-beta = E2 E_t0/E1^2. No effective metric, orbit equation, or perihelion precession is computed. The target 1/2 is taken from Mercury and used to set xi, so Mercury is a calibration under this dictionary, not a prediction. ## 26. The positive pump correction is a model result, not Lambda The n114 rerun reproduces a positive order-xi^2 correction for all eight reported points. It is even in xi and identical under Delta reversal. The baseline coefficients include | N | g | kappa | dP/xi^2 | |---:|---:|---:|---:| | 6 | 0.1 | 2 | 1314.77 | | 8 | 0.1 | 2 | 1955.95 | | 6 | 0.2 | 2 | 346.15 | | 6 | 0.1 | 3 | 1682.65 | The sign is stable on this small grid, while the coefficient varies by more than a factor of five. A regulator sweep from 1e-4 to 1e-7 gives 1266.4, 1310.3, 1314.8, 1315.2 at the baseline point, supporting a finite positive regulator limit for that calculation. What is computed is dP = (kappa/2) Tr Delta-Sigma_aux, an auxiliary damping-power correction in a finite Gaussian open system. It is not an energy density, pressure, stress tensor, equation of state, or generally covariant vacuum expectation value. In particular: * branch independence weakens any connection to the alleged active pump; * no volume or thermodynamic continuum limit is established; * no subtraction/renormalization prescription is supplied; * no map to T_mu_nu = -rho g_mu_nu is derived; * no FRW solution or conservation equation is present; * physical units and magnitude are absent. Positive dissipated power does not imply positive cosmological constant. The statement "Lambda > 0" is an interpretation added after the calculation, not a mathematical consequence of it. ## 27. The Mercury--Lambda sign chain does not close The advertised chain is Mercury -> xi > 0 -> dP proportional to +xi^2 -> Lambda > 0. Its first arrow uses an uncontrolled first-order fit and an imported beta dictionary. Its second arrow does not depend on the sign of xi at all. Its third arrow equates finite-system dissipation with a covariant cosmological source without a bridge. Even if the positive dP calculation is accepted exactly, Mercury contributes nothing to its sign: dP is positive for both +xi and -xi. The only shared fact is that both discussions contain the symbol xi. This is not a quantitative or sign-selecting link between perihelion precession and dark energy. ## 28. Updated verdict on the nonlinear layer Two narrow mathematical results survive: 1. the derivative source obeys an exact lattice integration-by-parts identity; 2. the finite regulated Gaussian monitor has a positive second-order auxiliary power correction on the tested parameter grid. The physical claims do not survive. beta is calibrated rather than derived, the fitted xi is outside controlled perturbation theory, the cited Mercury script is broken, eta_N is not calculated in a relativistic body problem, and the pump power is not a cosmological constant. ## 29. Next decisive gates for the nonlinear layer 1. Replace n113's Newton code with the declared first-order implementation and make nonconvergence fatal. 2. Solve the full nonlinear derivative field at xi about 0.231 with continuation from zero; report Hessian stability, branch structure, and finite-size convergence. Failure to find a stable branch kills the calibration. 3. Derive beta from an effective metric and a two-body orbit, then calculate perihelion precession rather than importing Delta-beta = E2/E1^2. 4. Build a composite-body dynamics with independently calculated inertial and gravitational masses before assigning eta_N. 5. Take a controlled volume/continuum limit of dP and derive a conserved stress tensor. Only then test whether its equation of state is w=-1. # Fifth pass: tensor constraints, helicity, propagator, and wave speed ## 30. The TT null space is correct linear algebra For the nine selected directions d_a, the sampling map delta-k_a = d_a^i d_a^j h_ij has rank six. With the post-n166 weights (axis weights 1/2 and face-diagonal weights 1), the four audit covectors span trace(h), and q_i u_j + q_j u_i for u in R^3. Their Euclidean orthogonal complement in the six-dimensional space of symmetric spatial tensors is exactly the two-dimensional traceless, transverse subspace. The n197 rerun confirms this for its nine directions to machine precision. It also confirms that the original uniform-weight claim was false off axis, with transversality violation about 0.19. This is a valid projector construction. The weights satisfy the declared fourth-moment isotropy identity to 1.1e-16 for the selected direction set. They were introduced after the original uniform-weight construction failed, so they are a load-bearing amendment of the lattice model, not a consequence of the original action. ## 31. A TT projector is not the constraint algebra of GR The n197 interpretation conflates three different objects: 1. constraint equations; 2. gauge transformations generated by first-class constraints; 3. vectors orthogonal under a chosen Euclidean inner product. In canonical linearized gravity, the Hamiltonian and momentum constraints are functions of h_ij and its conjugate momentum pi_ij. Gauge transformations are Hamiltonian flows generated by those constraints. The Hamiltonian constraint is not simply trace(h), and the momentum constraints are not simply the three longitudinal tensors q_(i u_j). Demonstrating that four row vectors span `trace plus longitudinal` does not establish Poisson brackets, first-class closure, preservation under time evolution, or gauge equivalence. The script then counts 6 - 4 = 2 in the spatial tensor space, while the documentation additionally claims 10 - 4 gauge - 4 constraints = 2. The latter count cannot follow from the former. The model contains no dynamical h_00 or h_0i, no conjugate momenta for a ten-component field, and no independent demonstration of four gauge generators plus four constraints. Using the same longitudinal row space first as "watched constraints" and then again as gauge directions risks double counting precisely what must be proved. The asserted closure is also insufficient: linear functionals commute as ordinary row vectors, but GR closure is a Poisson/Dirac constraint algebra under Hamiltonian evolution. A continuity identity does not replace that calculation. ## 32. Helicity two follows tautologically after imposing TT Once a symmetric tensor is restricted to the TT plane for momentum along z, its plus and cross basis transforms under axial rotations by 2 theta. The complex combinations therefore have helicity plus/minus 2. This is correct representation theory, but it does not add an independent dynamical result. The null-rotation part of the massless little group was not calculated. It is declared to act as the unproved gauge transformations discussed above. Consequently the construction has shown spin-two transformation under spatial rotations in the continuum-form projector, not a unitary massless representation of the Poincare group. ## 33. The propagator claim is incomplete and not executable from the record The n175 record argues that a free two-phonon bubble has nonnegative spectral weight and reports favorable Euclidean frequency and momentum derivatives. Those observations can support positivity of a particular projected bubble. No corresponding n175 executable is present in the repository, so its quoted numbers cannot be independently rerun from the cited artifact. More importantly, the calculation does not provide the inverse of a complete Fierz--Pauli kinetic operator. It omits h_00, h_0i, gauge fixing, ghosts from non-TT sectors, interactions, and the monitored pinned state. Euclidean slope signs do not by themselves establish causal support, Lorentz invariance, or a positive-residue Minkowski pole for the full field. Masslessness at n178 is imposed through u'' = -Pi_shear(0), which tunes the constant inverse-propagator term to zero. Calling this a thermostat steady state does not derive the tuning unless the proposed heating/avalanche dynamics is explicitly modeled and shown to converge to it. ## 34. The claimed universal coupling is only a spatial scalar-field identity The identity sum_a delta-k_a (d_a.grad epsilon)^2 = h^ij T_ij is algebraically correct when h^ij is defined by the same directional sampling and T_ij is the stress of that scalar substrate field. It does not establish (1/2) h_mu_nu T^mu_nu for a conserved relativistic stress tensor common to all matter. There is no matter sector, no T_00/T_0i completion, and no demonstration that binding, spin, fermions, or other fields share the same normalization. Calling this "universality derived" exceeds the identity's scope. The normalization rho_G remains uncomputed. Therefore Newton's static coupling, radiative amplitude, quadrupole power, and the double-pulsar decay rate have not been connected. ## 35. The computed gravitational-wave speed fails decisively The independent Rust calculation is reproducible. It finds, in the declared free-vacuum leading-order continuum extrapolation, | complex | c_g^2/c_scalar^2 | |---|---:| | cubic | about 0.515 | | nine-direction, isotropy weights | about 0.813 | | best scanned nine-direction weights | about 0.85 | | thirteen-direction families | about 0.34--0.40 | No scanned complex is luminal. The isotropy-weighted model actually used by n197 gives c_g/c about sqrt(0.813) = 0.902, roughly ten percent slow. This is not a small lattice correction relative to the observational requirement; the needed equality is at the approximately 1e-15 level. The proposed bare-gradient repair adds a new coefficient b and sets b = c^2 I - S_induced using the desired luminal speed. That is calibration on the observable being predicted, explicitly triggering failure condition E of the frozen Fierz--Pauli gate. Its relatively small effect on one trace stiffness does not make c_g = c derived, nor does it show that all other dispersion, preferred- frame, and static observables remain acceptable to the necessary precision. At present the honest continuum result is c_g != c. Deferring the frozen kill because unknown interactions might repair a ten-percent discrepancy turns the failure condition into an open-ended escape. A specific preregistered interaction calculation would have to exist before the result, not afterward. ## 36. Weinberg--Deser cannot complete the missing construction Soft-graviton universality and the Deser bootstrap require premises such as Lorentz invariance, a genuine massless spin-two gauge field, a consistent S-matrix or local action, and universal coupling to conserved stress energy. Those are exactly the missing parts of the present construction. Invoking the theorems cannot supply their own premises. The framework therefore cannot inherit the EIH equations, factor four, quadrupole formula, nonlinear Einstein equations, or strong-field predictions from having found a two-dimensional TT null space. ## 37. Updated verdict on the radiative sector The surviving mathematical result is: > A weighted directional-sampling model can be constructed whose four chosen > row covectors annihilate exactly the TT subspace, and a projected free-scalar > bubble induces positive temporal and spatial quadratic coefficients. It has not derived linearized diffeomorphism invariance, the Fierz--Pauli action, a complete propagator, universal matter coupling, normalization to G, or luminal propagation. Its explicit speed calculation currently disagrees with the required light cone by order ten percent. ## 38. Next decisive gates for the tensor sector 1. Write the full canonical variables, symplectic form, Hamiltonian, and four constraint functions. Compute their Poisson brackets and preservation under evolution without identifying row orthogonality with gauge symmetry. 2. Derive h_00 and h_0i dynamically and produce the complete quadratic continuum action term by term beside Fierz--Pauli. 3. Add gauge fixing and invert the full operator; report every pole and residue, not only the TT projection. 4. Treat the present nonluminal speed as a failed gate. Any proposed repair must be independently motivated and preregistered before fitting b to the observed light cone. 5. Construct one conserved model matter stress tensor and calculate both its static force and radiated power with the same normalization. # Sixth pass: topology, fermions, charge, and dark matter ## 39. The graph-topology statements are real but do not constitute matter For a graph with added handles, the first Betti number counts independent cycles, harmonic one-forms represent cohomology classes, and compact U(1) phase circulation is integer when it is well defined. Antiperiodic boundary conditions on a ring also remove its zero mode. These are valid graph and spectral facts. None of them derives the identification graph handle = elementary particle. That identification needs localized finite-energy solutions, dynamical translation, creation/annihilation operators, scattering states, a continuum limit, and a rule connecting graph surgery to unitary quantum evolution. The repository instead starts with a modified graph and interprets its static topology as a particle species. The claim of indestructibility is likewise conditional on forbidding the graph-rewiring operations that would change the class. Since the model has no dynamics for adjacency, it neither permits nor derives the absence of such operations. Proton stability and exact baryon conservation cannot follow from a topology whose time evolution has not been defined. ## 40. The compact-winding executables contradict their public summary The base `n123_compact_u1.py` rerun gives N=1 -> winding 0, energy 0, N=2 -> winding 0, energy 0, N=3 -> winding 0, energy 0. Every imposed nonzero winding unwinds. Nevertheless the script prints "U2 -> PASS: quadratic ladder with compact softening." Its U3 anchor then correctly prints FAIL. This is a gate-implementation defect: U2 has no conditional check and reports success for a nonexistent ladder. The tube repair `n123b_compact_u1_tube.py` holds only N=1: | imposed N | relaxed winding | energy | |---:|---:|---:| | 1 | 1 | 3.13719 | | 2 | 0 | 0 | | 3 | 0 | 0 | The implemented graph uses four interior bridge vertices, hence five bridge edges, plus a two-edge lattice return path: the tested cycle has length seven. It is odd, not the claimed minimal even cycle of length six, and therefore conflicts with the asserted bipartite/C-conjugation structure. The public selection note instead quotes cycle length six and energy 3.634. No n137 executable producing those values exists. The keV mass calculation therefore lacks a reproducible model matching its own stated topology. ## 41. Antiperiodic boundary conditions do not derive fermions The n121 ring result is elementary: inserting an odd number of sign cuts gives antiperiodic boundary conditions, so translation around the ring produces -1 and the zero mode is lifted. The minus sign was introduced through the twisted boundary condition. The required physical steps remain absent: 1. why spatial 2 pi rotation of a three-dimensional localized object equals translation around this internal ring; 2. why exchange of two graph handles is homotopic to that rotation in the actual configuration space; 3. why the quantum wavefunctional takes the nontrivial representation of that fundamental group; 4. why locality, microcausality, and positive energy produce the usual spin- statistics relation. The repository explicitly admits that the Finkelstein--Rubinstein homotopy was not established. That is the step that would turn a rotational sign into exchange antisymmetry. Hence Pauli exclusion, fermionic statistics, composite statistics, and the absence of massless fermions are not derived. ## 42. The Alice charge-quantization argument is incomplete An odd cycle indeed frustrates a staggered sign convention and can carry Z2 holonomy. Calling it an Alice string additionally requires a gauge group with a disconnected component acting by charge conjugation, a well-defined gauge connection, and charged representations transported around the defect. Those objects have not been constructed. The argument then declares the phase law exp(i q phi) and its normalization, which already specifies how charge couples to a compact U(1) phase. Requiring single-valuedness can quantize representations of that declared compact group, but it does not derive the existence or normalization of electric charge from the audited lattice. There is also a structural tension: the electrostatic construction relies on a bipartite lattice, whereas an odd cycle is precisely an obstruction to bipartiteness. The theory needs a consistent global definition explaining where odd cycles can exist without destroying the staggered Gauss sector. No confinement dynamics follows from saying an isolated fractional phase is inconsistent. Quark confinement requires physical fractional-charge fields, gauge flux, a color sector, and an energy law for separation, all absent. ## 43. The keV mass band is a chain of arbitrary identifications The quoted interval 1.9--13.5 keV is formed by multiplying two unrelated dimensionless numbers--a twisted ring gap and a winding energy--by the assumed electron-anchor scale hbar gamma = 3.73 keV. The calculation does not determine which number is a particle's rest energy, so their span is reported as a band. The anchor itself was not derived, and the implemented winding energy does not match the quoted value. This is dimensional anchoring, not a mass prediction. There is no dispersion relation E^2 = p^2 c^2 + m^2 c^4, inertial mass calculation for the winding, or equivalence between its passive gravitational energy and its inertia. ## 44. The dark-scattering result is conditional dimensional analysis For an assumed classical potential V = beta/r^3, the scale estimate sigma proportional to beta^(2/3) m^(-2/3) v^(-4/3) is dimensionally reasonable, and the Monte Carlo rerun reproduces exponent -1.332. But the potential's three-dimensional amplitude and angular dynamics were not derived from the topological lattice. The simulation draws a fresh random frozen orientation factor for every trajectory and imposes a hard capture cutoff, so it is an illustrative classical model rather than a quantum identical-particle transfer cross section. The later amplitude inserts beta = hbar gamma (6 l_s)^3, which is a dimensional assignment based on the assumed anchor, assumed handle size, and an external upper bound on l_s. It is not a calculated overlap coefficient. Consequently the bound 3e-12 cm^2/g and its proposed dark-matter kill condition are not predictions of a normalized scattering theory. ## 45. Kibble--Zurek scaling does not predict the relic The original Python KZ script currently fails: its slow-quench samples contain zero defects, producing log(0), NaN exponent, and failed gates. The larger Rust analog does reproduce a clean exponent about 0.568 for a standard 2D model-A complex-field quench. That is a valid result for the analog simulation. The repository correctly admits that this is not the substrate's dynamics and that the abundance is unpredicted. Missing pieces include the transition temperature, quench history, correlation-length normalization, three- dimensional defect type, annihilation/annealing history, expansion, and the mapping from defect count to present energy density. "Born at rest" does not by itself eliminate structure-formation constraints. Defect networks can acquire gradient and kinetic energy, and a nonthermal keV species still needs a phase-space distribution and transfer function before Lyman-alpha compatibility can be claimed. ## 46. Updated verdict on the topology wing The surviving results are familiar mathematical structures: * graph handles produce homology classes; * compact phases can have integer winding; * twisted boundary conditions produce a sign and a spectral gap; * generic model-A quenches produce Kibble--Zurek scaling; * an assumed inverse-cube potential has v^(-4/3) classical scaling. The mappings to electrons, fermions, baryon number, electric charge, proton stability, and dark-matter particles are not constructed. The numerical selection principle and keV mass band also lack a consistent reproducible handle geometry. ## 47. Next decisive gates for the topology wing 1. Define adjacency dynamics and exhibit a stable, localized, translatable finite-energy soliton without fixing the graph topology externally. 2. Construct the two-soliton configuration space and prove the FR exchange homotopy and its quantum representation. 3. Repair n123/n123b so gates fail closed, then produce the exact claimed even six-cycle and its converged energy in 3D. 4. Derive a relativistic dispersion and inertial mass for the defect before attaching physical units. 5. Derive a normalized quantum scattering amplitude and cosmological transfer function; only then compare with direct detection, halo scattering, and Lyman-alpha data. # Seventh pass: electrostatics, gauge claims, and the photon reading ## 48. The scalar Coulomb-like sector is reproducible Near the staggered zero of the unsigned incidence operator, the constrained kernel has quadratic dispersion. The n46c rerun gives S_3D(L=96,q1) = 493.4419, S_3D(L=96,q2) = 491.6525, with a 0.36% infrared ratio and 0.36% change from L=48. This supports a finite three-dimensional coefficient and Coulomb-like static Green function in the declared lattice normalization. The two apparent charge signs arise from sublattice parity: after removing the staggered carrier, sources on opposite parity have opposite effective sign. Positive quadratic energy then gives the familiar attractive/repulsive pair pattern in the feasible sectors. This remains a scalar constrained-static model. It has not yet identified a mobile charged excitation whose parity label is preserved during translation. A one-site translation swaps the two sublattices and therefore flips the putative charge unless an independent internal label and transport law are constructed. ## 49. Clock invisibility is not U(1) gauge invariance The dichotomy result says selected parity-neutral local observables have no linear response to the staggered long-range channel. Gauge invariance instead requires a redundant local description A_mu -> A_mu + partial_mu chi, together with transformations of charged matter, Gauss constraints, Ward identities, and invariance of all physical observables. None is supplied by a filter zero in one class of clocks. The later construction also says a charged phase acquires q lambda t/hbar. Thus charged matter is intended to see the scalar potential even though the earlier absence of clock response is called gauge invariance. Only constant potential shifts are removable; spatial potential differences and Wilson loops are physical. The model has not demonstrated the required local redundancy or its action on matter. ## 50. Charge superselection is not derived and conflicts with dark states n69 claims that fast monitoring of charge-Gauss violations decoheres superpositions of different charge configurations. The trunk assumes the monitor observes only constraint residuals. If two fully dressed static branches both satisfy A delta-k + c = 0, then every residual Lindblad operator has the same zero eigenvalue on both branches, and their superposition is dark. The monitor cannot simultaneously be blind to all satisfied static branches and rapidly distinguish their local charges. The repository later reaches the same conflict from experiment: n171 kills the proposed charge-dependent audit because it would destroy electron interference, and declares the monitor charge-blind. That retracts n69-E4's mechanism for charge superselection. Physical charge superselection concerns coherent superpositions of different total asymptotic charge sectors and follows from the gauge structure/Gauss law of QED. No joint charged-matter Hilbert space or asymptotic algebra exists here, so this result remains absent. ## 51. A static zero of W is not a massless photon pole The exact statement is that the static source-response operator W(q) vanishes quadratically at the staggered point because A(q*) = 0. This protects the Coulomb Green function against the scalar regulator mu. Photon mass is a property of the pole of a dynamical transverse vector propagator. The lattice result contains no time kinetic term, frequency- dependent inverse propagator, canonical vector potential, electric/magnetic fields, or Lorentz-covariant Maxwell action. Therefore the bound on m_gamma cannot be applied to W's static zero. The n159 polarization count is another kernel identity: one divergence row acting on three edge-envelope components has a two-dimensional transverse kernel. That is the Helmholtz decomposition of a vector-like lattice field. It does not establish two propagating photon polarizations until dynamics, norm, helicity, gauge quotient, and coupling to conserved current are built. No n159 executable is present to reproduce the quoted numerical details. ## 52. Minimal coupling is asserted through missing covariance The time-component statement q lambda is simply the definition of how the putative charge couples to the static potential. The spatial component is then introduced "by covariance": q lambda dt is declared to covariantize into q(lambda dt - A.dx). But Lorentz covariance, A_i, moving matter, and a conserved current are the missing results. Invoking covariance cannot derive them. Consequently the Lorentz force and full minimal coupling have not been obtained. Likewise, saying constant billing differences are unobservable supplies a global energy-zero freedom, not local U(1) gauge symmetry. ## 53. The fine-structure normalization fails Using the declared scalar vertex and Coulomb convention gives alpha^-1 = 4 pi S_inf about 6200, roughly 45 times the observed low-energy value. The project records this miss. Introducing an unmeasured transverse stiffness S_T small enough to repair the answer does not help until S_T is derived independently; requiring a factor about 1/2500 in stiffness is itself a large unexplained hierarchy and would conflict with the Lorentz symmetry invoked to derive the vector component. The finite-size fit for S_inf is meaningful only in lattice units. Its mapping to electric charge depends on field normalization and the matter vertex, both freely rescalable before a canonical Maxwell kinetic term and normalized charged state exist. The absence of a beta function, vacuum polarization, g-2, magnetic interactions, and charge transport further prevents comparison with QED. ## 54. The charge--mass inequality is bookkeeping, not a physical bound For a stipulated nonnegative vector c_i, |sum (-1)^i c_i| <= sum c_i is the triangle inequality. Calling the right side mass and calibrating the electron to saturate it turns the inequality into |q| <= m in chosen units. It does not predict the electron's saturation or establish a physical bound on millicharged particles; those depend on the arbitrary mass/charge conversion fixed using the electron itself. The proton and neutron discussion similarly assigns observed mass fractions to smooth and staggered bookkeeping after the fact. With no quarks, gluons, binding dynamics, or calculated masses, agreement with QCD's qualitative mass budget is analogy rather than a model result. ## 55. Updated verdict on the electromagnetic sector The surviving mathematical statement is: > A bipartite constrained scalar lattice has a staggered quadratic zero, a > Coulomb-like static Green function, and a two-dimensional transverse kernel > for a vector envelope near that zero. It has not derived local U(1) gauge symmetry, charge superselection, mobile charged matter, Maxwell dynamics, photons, magnetic fields, the Lorentz force, or the fine-structure constant. Its only parameter-free alpha normalization misses by about a factor of 45. ## 56. Next decisive gates for electromagnetism 1. Construct a mobile charged state whose charge is invariant under a one-site translation and whose current satisfies a continuity equation. 2. Derive canonical A_0 and A_i variables, their gauge transformations, and the Maxwell action from the lattice dynamics. 3. Compute the full transverse frequency-dependent propagator and show a positive massless pole with two helicities and luminal speed. 4. Prove total-charge superselection in the joint matter/gauge Hilbert space without contradicting the violations-only dark-state theorem. 5. Canonically normalize the matter vertex and gauge kinetic term, then predict alpha and at least one magnetic or radiative observable before comparing with QED precision data.