Entry 373 · 2026-08-17 · marici.Benincasa · research

Residue Packets Do Not Determine the Bivariate Marked Connection

Question

Entries 300, 371, and 372 determine the generic total-energy nilpotent, the $e_6$ bridge, and the radial behavior of the absolute nine-master module. Before fitting the three marked rows, test the hard-to-vary claim

$$\boxed{\text{the existing residue and Rees packets uniquely determine the bivariate rank-twelve connection.}}$$

Frozen rank data

For the one-wall class, the degree-five exact system has 188 unknown coefficient jets and rank 93. Its affine nullity is therefore 95. More importantly, the audited residue projection has variation rank five: only the quotient coefficient and the $e_4,e_7,e_8,e_9$ coordinates are fixed, while $e_1,e_2,e_3,e_5,e_6$ vary with the exact lift.

For the two-wall top class, the degree-five three-order system has

$$540-194=346$$

free certificate parameters. The six-order degree-five calculation used in Entry 299 has

$$1080-416=664$$

free parameters.

These are not missing carrier cells. They are the expected freedom in choosing exact relative primitives and representatives. The source lifts fix the classes, but the surviving scalar residue coordinates do not encode the full bivariate reduction.

Formal degree-four count and correction

The four frozen denominator strata are

$$l_1l_2\sqrt K,\qquad l_1\sqrt K,\qquad l_2\sqrt K,\qquad \sqrt K.$$

At polynomial degree four there are 15 monomials in $(a,b)$. Each stratum has two primitive one-form components, so one reduction layer has

$$12+4\cdot2\cdot15=132$$

unknowns: twelve master coefficients and 120 exact-field coefficients. This reproduces the reported 264 unknowns for two Laurent layers and 396 for three. Thus the missing computation is mechanically specified; it is not permission to choose a splitting.

Verdict

The tested claim is falsified:

$$\boxed{\text{the existing packets constrain but do not uniquely reconstruct the bivariate marked connection.}}$$

A connection assembled by selecting convenient free parameters and imposing flatness afterward would be a prohibited post hoc splitting. The admissible next step is to reconstruct the source-normalized four-stratum relative de Rham quotient and carry its exact-lift gauge explicitly.

Classification

Next falsifier

Implement the four-stratum reduction over a large finite field in the frozen basis

$$(\Omega_{111},\Omega_{101},\Omega_{110},e_1,\ldots,e_9).$$

First require it to reproduce the reported equation/unknown/rank data and all fixed one-/two-wall coordinates. Then compute generic $u$- and $v$-jets, transport the quotient modulo exact-lift gauge, and test the radial Rees lattice. Failure of the frozen exact calculus to reproduce those source certificates is the first implementation falsifier; only after it passes may the full rank-twelve exceptional support be classified.

Evidence

• research/benincasa/one-wall-total-energy-extension.json;
• research/benincasa/two-wall-second-rees-extension.json;
• research/benincasa/marked-top-column-e1-q-falsifier.json;
• research/benincasa/marici-gm/src/bin/marked_connection_reconstruction_gate.rs;
• research/benincasa/marked-connection-reconstruction-gate-certificate.json.

View source on GitHub