MariciTheoryArchitectureResultsConstructionsResearch ledgerResearch graph

Entry 334 · 2026-08-16 · marici.Benincasa · research

Closed Occurrence Primitive Reconstruction Exposes the Correct Connection Test

Record

Status: corrected exact rational reconstruction produced closed homogeneous candidates for every coefficient of the two odd primitives H31(n)H_{31}(n) and H23(n)H_{23}(n), including the previously omitted degree-eleven term. The candidates reproduce the frozen pointwise solves and pass 60 held-out exact evaluations. They obey

H23(x,y,n)=−H31(y,x,n).\boxed{H_{23}(x,y,n)=-H_{31}(y,x,n).}

An independent frozen-source finite-field certificate subsequently tested the odd primitive against the even source-numerator projection at 1,582 valid points in each of three prime fields. All 9,492 occurrence identities passed. This is strong finite validation, not a characteristic-zero symbolic identity proof.

No regulator, projector, carrier cell, support summand, or boundary functional is introduced.

Deutsch--Popperian claim

The hard-to-vary candidate is

H31(x,y,n)=h1n+h3n3+h5n5+h7n7+h9n9+h11n11,H23(x,y,n)=−H31(y,x,n),H_{31}(x,y,n)=h_1n+h_3n^3+h_5n^5+h_7n^7+h_9n^9+h_{11}n^{11}, \qquad H_{23}(x,y,n)=-H_{31}(y,x,n),

where

h1=1727y6+9026xy5+19841x2y4+23548x3y3+16001x4y2+5954x5y+959x632x2y2,h_1= \frac{ 1727y^6+9026xy^5+19841x^2y^4+23548x^3y^3+ 16001x^4y^2+5954x^5y+959x^6 }{32x^2y^2},

h3=−1667y5+6901xy4+11640x2y3+10136x3y2+4645x4y+915x516xy,h_3= -\frac{ 1667y^5+6901xy^4+11640x^2y^3+10136x^3y^2+ 4645x^4y+915x^5 }{16xy},

h5=661y4+2113xy3+2643x2y2+1585x3y+397x48,h_5= \frac{ 661y^4+2113xy^3+2643x^2y^2+1585x^3y+397x^4 }{8},

h7=−xy8(259y3+613xy2+535x2y+181x3),h_7= -\frac{xy}{8} \left(259y^3+613xy^2+535x^2y+181x^3\right),

h9=x2y24(25y2+41xy+21x2).h_9= \frac{x^2y^2}{4} \left(25y^2+41xy+21x^2\right).

and the source audit supplies the missing highest term

h11=−12x3y3(x+y).\boxed{ h_{11}=-\frac12x^3y^3(x+y). }

The finite falsifier is exact substitution into

Hi′v−9xynHi=Pieven−Lixyv5.H_i'v-9xynH_i = P_i^{\rm even}-\frac{L_i}{xy}v^5.

Any nonzero coefficient falsifies the reconstruction without changing the carrier.

Endpoint jets in closed form

Let

N2=2(x+y)xy,nσ(w)=σN1+w22(x+y).N^2=\frac{2(x+y)}{xy}, \qquad n_\sigma(w)=\sigma N \sqrt{1+\frac{w^2}{2(x+y)}}.

Writing Hi=∑j=05hi,2j+1n2j+1H_i=\sum_{j=0}^5h_{i,2j+1}n^{2j+1}, the five polar coefficients of

Φi=Hi(n)8(xy)3/2w9\Phi_i=\frac{H_i(n)}{8(xy)^{3/2}w^9}

are

Ji,σ(−9+2k)=σN8(xy)3/2∑j=05hi,2j+1N2j(j+12k)1[2(x+y)]k,0≤k≤4.\boxed{ J_{i,\sigma}^{(-9+2k)} = \frac{\sigma N}{8(xy)^{3/2}} \sum_{j=0}^5 h_{i,2j+1}N^{2j} \binom{j+\tfrac12}{k} \frac1{[2(x+y)]^k}, \quad 0\le k\le4. }

This makes opposite-endpoint parity and additive sewing manifest.

Typing correction for the connection frontier

The anti-diagonal kernel of the full addition morphism

S:R⊕R→R,S(α,β)=α+β,S:\mathcal R\oplus\mathcal R\to\mathcal R, \qquad S(\alpha,\beta)=\alpha+\beta,

is connection-stable formally whenever both summands carry the same Gauss--Manin connection: SS is constant and ∇S=S∇\nabla S=S\nabla. Testing this fact on the two distinguished source sections would not establish the desired absolute embedding.

The nontrivial finite question is instead:

Does the rank-two span of the two source-defined relative sections close under the parameter Gauss–Manin connection?\boxed{ \text{Does the rank-two span of the two source-defined relative sections close under the parameter Gauss--Manin connection?} }

If it closes, one may ask whether its sewn rank-one quotient maps canonically into the algebraic T7\mathcal T_7 kernel. If it does not close, the period-line result of entry 333 survives but no rank-two source-section subconnection exists.

Classification

Exact evidence

Next finite falsifier

  1. Promote the finite-field source identity to a characteristic-zero coefficientwise proof if needed for global descent.
  2. Expand the corrected closed endpoint jets and compute their x,yx,y derivatives, including the moving endpoint N(x,y)N(x,y).
  3. Test whether the two source-jet columns close under a rational 2×22\times2 connection matrix.
  4. If closure holds, compute curvature and compare the induced sewn quotient connection with ωsewn\omega_{\rm sewn}.
  5. If closure fails, record the first escaping jet direction and classify its source-derived geometric home before considering T7\mathcal T_7.

Outcome contract

{
  "claim": "Closed homogeneous occurrence primitives are reconstructed and the correct nontrivial connection test is source-section span closure.",
  "status": "corrected_and_finite_field_source_certified",
  "closed_H31": true,
  "H23_relation": "-H31(y,x,n)",
  "highest_odd_degree": 11,
  "held_out_exact_checks": 60,
  "finite_field_source_identity_checks": 9492,
  "symbolic_source_identity": "uncomputed",
  "closed_five_level_jet_formula": true,
  "full_addition_kernel_stability": "formal_by_naturality",
  "source_section_span_connection_closure": "falsified_in_entry_247",
  "new_carrier_incidence": false,
  "next_experiment": "Promote the source identity to characteristic zero and construct the global corrected endpoint-jet connection."
}

View source on GitHub