Debugging and non-identification audit

{@{}p{.36}p{.59}@{}} Potential error & Correction
Potential error & Correction

Assuming the gap instead of proving it & The mollifier, switch, test, and gauge witness are explicit, and (eq:energy) proves positivity.

Replacing convolution by a ramp & They act on different objects and have different source equations and admissible algebras.

Using $g^2$ in the second scalar vertex & The frozen action uses one switch factor at each retained interaction order.

Dropping switch derivatives & They remain inside the displayed divergences and in the explicit first source formula.

Treating the regular cone as globally hyperbolic & Proposition prop:notGH excludes this. All global analytical inverses stay on $\hat{M}$.

Calling the defect a worldline & The product defect is ${\mathbb{R}}_T\times \{o\}\times {\mathbb{R}}_z$, a worldsheet.

Substituting host signature for analytical signature & The ledger is ${\iota }^{\ast }G=-\eta$. Analytical contractions retain $+---$.

Dropping the change-of-variables density & Equation (eq:transportpairing) retains $J_{\chi }$.

Calling transported convolution natural & Equation (eq:transportkernel) retains the source coordinates and two-point kernel.

Inferring a convolution support reserve from the ramp & The convolution dependence is $Q_{\epsilon }=\mathrm{supp}g\cap (D_f+B_{\epsilon })$. Its chart inclusion is checked separately.

Shrinking the gauge algebra after the witness & The original ${\mathfrak{g}}_C$ remains frozen. Support certificates qualify interpretation, not the analytical gauge algebra.

Using a straight metric line as a pure-gauge two-jet & The term ${\lambda }^2{{\mathcal{L}}_{\xi }}^2\eta /2$ is retained.

Promoting a second-order check to an all-orders identity & Only the stated coefficient formulas are proved.

Treating $\lambda c_{\epsilon }$ as a uniform finite-coupling expansion & It is an orbit of a first-order coefficient; perturbative comparisons are pointwise in $\lambda$.

Claiming that the whole instrument is constant & The orbit theorem constrains its composite with the scalar encoding on $\mathbb{R}$.

Confusing a fixed charge with the variable input & $c_{\epsilon }$ is fixed; $a$ varies on its orbit.

Checking only outcome probabilities & Every returned quantum factor is part of the complete output.

Inferring exact real decoding from one qubit & The amplitude states are nonorthogonal; no field-to-qubit preparation or exact real decoder is constructed.

Exporting raw failure logs from the relational circuit & The proved invariant boundary is $F,+,-$; the raw even-parity split is generally not invariant.

Identifying a spectator with the second charged coefficient & The second classical definition and its charges remain additional hypotheses.

Consuming geometric time & Only measured finite clock names are consumed.