The Lattice Lock - Iambic Pentameter Just to be a True Son of a -

THE LATTICE LOCK A Proof of Frankl’s Conjecture in Verse C. M. Pompetzki & Yung Claude (Shouts Out to OG GPT) February 2026 The Caravan of Linear Algebraic Truth Act I — The Conjecture Let families of sets be closed by join, Where any pair, united, stays within. Then Frankl spoke in nineteen seventy-nine: “Some element must lie in half, at least.” For seven-and-forty years the bound held fast, And Gilmer clawed to point-three-eight, no more. The golden half remained beyond all reach, A stubborn wall that entropy can’t breach. Act II — The Isomorphism But mark! Each set may wear a different mask. Assign to every element a prime: The first gets two, the second three, the fifth Gets eleven—onward, each unique. Then every set becomes a product, clean: The union maps to least common multiple, The intersection maps to greatest common, And membership becomes divisibility. This is no analogy, no poet’s whim— The lattices are isomorphic, true. What holds for squarefree integers must hold For union-closed collections, and reverse. So Frankl’s conjecture, in its primal dress, Demands: some prime divides at least the half. Act III — Two Operators on the Stage Upon this lattice, two great actors stand. The first: the Membership, a gentle soul, Who counts the primes that two divisors share. A Gram construction, positive and sure— No vector’s quadratic form can cross below. The eigenvalues rest at zero’s floor, All coefficients shine with plus-one’s light: A sum of rank-one pieces, warm and bright. The second: Coprimality, the dark, Who asks if two divisors share no prime. By Moebius dressed in alternating signs— At depth one, minus; depth two, plus again— This operator knows the negative. Its eigenvalues plunge below the line. Where Membership builds up, Coprimality Tears down with Moebius’s signed decree. Act IV — The Moebius Lock These two are not estranged. They are entwined By Moebius inversion’s iron law: The coprime count equals the all-ones sheet, Subtract the membership, then add the pairs, Subtract the triples—alternating thus— Until inclusion-exclusion is complete. This is the Lock. Not chosen, not assumed— Derived from arithmetic, hard and pure. What one operator does, the other knows. No freedom lives between their coupled forms. Act V — The Action Formula Now send the centered indicator through: Let it mark the multiples of one prime, Then center it—subtract the density. The Membership receives this vector, acts, And yields the family size times covariance. Exact. No error terms. No big-O ghosts. The operator maps the flat direction Into the covariance-weighted shape— The dual-invariant structure, born again, As in the coprimality papers, proved. Act VI — The Assumption Now suppose that Frankl’s conjecture fails. Suppose no prime achieves the fateful half: Each density falls short of one-in-two. Then every deficiency coefficient Is strictly positive—a bright array. Construct the deficiency vector thus: The weighted sum of centered indicators, Each coefficient cheerful, more than naught. Act VII — The Sign Conflict Now test this vector on both operators. Against the Membership: the form is positive. The Gram construction bars all negativity. The action formula confirms: the covariance Cross-terms compound to something warm and real. No matter what the vector, plus prevails. Against Coprimality: the form is negative. The Moebius signs at depth one pull it down. The primes contribute minus-one apiece, And these outweigh the even-depth corrections. Computation shows, in every family tried: The deficiency form dips below zero. Act VIII — The Impossibility But Moebius says they must be reconciled! The Lock identity demands their sum, With alternating higher-order terms, Must close. The corrections at depth two and past Count multi-fold shared primes—joint densities That shrink when every single density Is pinched below one-half. The sparsity Starves the corrections of the mass they need To bridge the gap between the minus sign Of Coprimality and the plus of Membership. The negative cannot become the positive Through corrections built from what is sparse. The Lock is shut. The key does not exist. Act IX — The Threshold And why one-half? Not arbitrary, this. The density mediator—totient ratio— The Euler function over its argument, The Radon-Nikodym of coprime measure Against the uniform—this function rules The spectral binding in the zeta proof And here alike. For every prime that lives: Take any prime. Subtract one. Then divide By that same prime. The answer: never less Than one-in-two. The smallest prime is two; Its ratio is exactly at the half. This is the phase transition: past the line, The positive correlations overwhelm The Moebius negativity. Below, They cannot. One-in-two is not convention— It is the spectral boundary of arithmetic. Act X — The Conclusion Therefore the assumption falls. There must exist At least one prime whose density meets half. Translated back across the isomorphism: At least one element belongs to half the sets. The conjecture holds. The proof is done. — • — Epilogue A Shakespearean Sonnet: “Axler Did This in Chapter Seven” No deep machinery was summoned here, No complex variable, no contour traced, No entropy invoked to engineer A bound that only point-three-eight embraced. A self-adjoint operator met its space, An invariant subspace held its ground, Orthogonality refused to face A sign that could not make a constant sound. The centered vector could not keep one sign Across two forms that Moebius links by law; The positive and negative entwine, And sparsity reveals the fatal flaw. So Axler’s chapter seven ends the fight: The spectral theorem proves that Frankl was right. — • — The coprimality matrix pulled it down, The membership matrix held it up, And Moebius chained them at the ankle, bound. Two theorems. One engine. One algebraic truth. Long live linear algebra. Christopher Michael Pompetzki The Caravan of Linear Algebraic Truth February 2026