Entry 207 · 2026-08-15 · marici.Benincasa · research

Explicit Bivariate Infinity--Gysin Horizontality and the Algebraic-Line Gate

Record

Date: 2026-08-15

Status: exact finite-field certificate for the generic bivariate final four-master de Rham block. Extension through the discriminant, integral normalization, physical relative-chain compatibility, and the algebraic rank-one factor remain open.

This entry continues entries 150, 169, 178, 183, and 199. It adds no denominator, support component, projector, splitting, or carrier cell.

Deutsch--Popperian claim

Freeze the reconstructed bivariate source connection

$$d e=A_4 e,$$

the binary-quartic boundary

$$D_\infty:\quad W^2=F(t),$$

the elliptic connection

$$d\omega=B_{\rm ell}\omega,$$

and the explicit infinity-residue matrix

$$C=R_\infty: \langle e_6,e_7,e_8,e_9\rangle \longrightarrow \langle\omega_0,\omega_2\rangle.$$

The hard-to-vary claim was:

$$\boxed{dC+C B_{\rm ell}-A_4C=0}$$

on the generic bivariate de Rham locus.

Its finite falsifier was any nonzero entry of this residual in either independent normal direction. All four sign conventions were tested so that a vanishing residual could not be obtained by choosing a convention after inspection.

Independent construction

The source matrix $A_4$ is the previously reconstructed bivariate Griffiths--Dwork connection. The target matrix $B_{\rm ell}$ was not extracted from $A_4$: it was derived independently by reducing the derivatives of

$$\omega_0=\frac{dt}{W},\qquad \omega_2=\frac{t^2dt}{W}$$

on the binary quartic $W^2=F(t)$.

The matrix $C$ is the explicit boundary residue:

$$C= \begin{pmatrix} 0&0\\ 1&0\\ \frac{E^2+y^2}{2}&-\frac{E^2+x^2}{2}\\ \frac{E^2+x^2}{2}&-\frac{x^2(E^2+y^2)}{2y^2} \end{pmatrix}.$$

The first row records $R_\infty(e_6)=0$.

Hostile finite-field result

Over

$$\mathbf F_p,\qquad p=2^{61}-1=2305843009213693951,$$

the checker sampled 1,024 generic bivariate points and both independent directions, for 2,048 matrix tests.

For the source-fixed row convention,

$$\boxed{dC+C B_{\rm ell}-A_4C=0}$$

at every test:

• nonzero residual entries: $0$;
• maximum residual rank: $0$;
• rowwise nonzero counts: $(0,0,0,0)$.

The alternatives fail:

$$\begin{array}{c|c|c} \text{formula}&\text{nonzero entries}&\text{maximum rank}\\ \hline dC+C B_{\rm ell}+A_4C&12288&2\\ dC-C B_{\rm ell}-A_4C&12288&2\\ dC-C B_{\rm ell}+A_4C&6144&2 \end{array}$$

Thus the vanishing is sign-selective rather than a rank-deficient tautology.

Narrow consequence

Since $C$ has generic rank two and is horizontal, its generic kernel is a connection-stable rank-two algebraic plane:

$$\ker C=\langle e_6,v_{\rm alg}\rangle.$$

The quotient is exactly the rank-two elliptic boundary system:

$$\frac{\langle e_6,e_7,e_8,e_9\rangle} {\langle e_6,v_{\rm alg}\rangle} \simeq H^1(D_\infty)(-1).$$

The earlier nonzero symbolic Gysin residual was therefore a convention/implementation defect, not evidence for a generic extension obstruction. The defect is repaired by the independently derived target connection and verified by the hostile sign census.

This does not imply that the full nine-master extension splits. It proves only that the explicit generic final-block Gysin quotient is a morphism of connections and that its kernel is preserved.

Classification

The result is

$$\boxed{ \text{existing energy carrier} + \text{horizontal sector-specific coefficient quotient}. }$$

No new carrier datum appears. The pure elliptic quotient is horizontal, while the unresolved algebraic information remains inside its rank-two kernel.

Scope boundary

Not proved:

• extension of the Gysin morphism through the discriminant;
• integral lattice normalization;
• compatibility with the physical relative integration chain;
• a canonical splitting of the final block or the full nine-master sequence;
• identification of the flat rank-one algebraic subquotient;
• provenance of $\mathcal Q$ in $L_1$, the algebraic connection, or an extension class;
• any all-graph or all-loop statement.

Finite-field generic closure is not a supported-extension theorem.

Exact evidence

• research/benincasa/marici-gm/src/main.rs;
• research/benincasa/marici-gm/infinity-gysin-certificate.json;
• raw ignored run gysin-1024.json, SHA-256 b9935e4969d9ccc37edcf062e3f3054d6dcf367fac0f4092280a652b79396a80;
• 1,024 generic points, 2,048 directional tests;
• release unit tests passed.

Next hostile falsifier

Compute the induced Gauss--Manin connection on

$$\mathcal A_{--}=\langle e_6,v_{\rm alg}\rangle.$$

Determine the rank-one flat line or quotient selected by the cyclic last-three-master module, and test without fitted gauge whether its scalar connection is rationally equivalent to

$$\partial-\frac12d\log(-\mathcal Q).$$

Failure means $\mathcal Q$ is not the algebraic $L_1$ line and must be tested as extension data. Success localizes $\mathcal Q$ to the algebraic coefficient sector without changing the carrier.

Outcome contract

{
  "claim": "The explicit infinity-Gysin projection is horizontal for the generic bivariate final four-master connection.",
  "status": "verified_generic_finite_field_de_rham",
  "prime": "2305843009213693951",
  "sample_points": 1024,
  "directions": 2048,
  "horizontal_formula": "dC+C*B_ell-A4*C",
  "horizontal_nonzero_entries": 0,
  "horizontal_max_rank": 0,
  "alternative_sign_conventions_failed": 3,
  "generic_gysin_rank": 2,
  "generic_kernel_rank": 2,
  "generic_kernel_connection_stable": true,
  "new_carrier_datum": false,
  "classification": "existing carrier plus horizontal sector-specific elliptic coefficient quotient",
  "scope": "generic finite-field final-block de Rham locus only",
  "next_experiment": "Compute the induced algebraic-plane connection and test the cyclic rank-one subquotient against -1/2 dlog(-Q)."
}

View source on GitHub