Alphabeta Math
ExampleConstruction: AI-generatedVerification: AI-generatedPipeline-generatedprecheck passjudge pass (gpt-6-luna)verified 2026-11-05 (gpt-6.1-sol)
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Modules and fusion for the one-dimensional vertex algebra

Example

Take V=C1, Y(a1,x)(b1)=ab1, T=0, and conformal vector ω=0, with central charge zero. Its grading-restricted generalized modules are exactly finite-dimensional complex vector spaces E with YE(a1,x)=a1E and weight zero. Use the trivial auxiliary groups A=A~={0} throughout: V(0)(0)=V, E[0](0)=E for every module E, and all other components vanish. The category has all finite-dimensional complex vector spaces and all linear maps. For finite-dimensional E,F,G, every logarithmic intertwining operator is constant in the variable and has the form YB(e,x)f=B(e,f) for a unique bilinear map B:E×F→G. Thus dim⁡I(GE F)=(dim⁡E)(dim⁡F)(dim⁡G). For every z≠0, the P(z)-tensor product is E⊗CF, with universal map e⊗f↦e⊗f; the completion equals the module. Contragredients are the usual finite-dimensional duals, and all multipoint matrix coefficients are constant, so their convergence and continuation hold directly. This example uses no generic HLZ existence or convergence theorem.

Facts & Assumptions

Given: Finite-dimensional complex vector spaces, including the zero space, and the displayed scalar state-field map.

[F1]

The vertex-algebra and module axioms are those of Vertex algebra in the state–field convention and Module for an ordinary vertex algebra. For the conformal grading and finite-dimensional generalized weight spaces use Conformal vertex algebra of CFT type and Grading-restricted generalized module.

[F2]

The formal logarithmic series, Jacobi, derivative and Möbius-bracket conditions are those of Logarithmic intertwining operator.

[F3]

The P(z)-map and representing property are those of P(z)-intertwining map and P(z)-tensor product in a module category.

[F4]

Contragredient module of a generalized module, C1-cofinite generalized module in the ungraded HLZ convention and Quasi-finite-dimensional generalized module specify respectively the opposite-field restricted dual, the positive-weight (-1)-mode quotient and the finite weight-cutoff condition.

Verification

1.1givenF1algebra

Only the (−1)st mode is nonzero. Truncation, creation and vacuum hold immediately, translation has both sides zero, and scalar fields commute, so locality holds with exponent zero. The zero conformal field has L(n)=0 for every n and satisfies Virasoro with c=0; L(−1)=T, V0=C1 and all other graded pieces vanish. Thus V has exactly the stipulated conformal structure.

2.1F1step 1.1algebra

On any module, linearity and vacuum force YE(a1,x)=a1E, hence YE(ω,x)=0 and LE(n)=0. Conversely this action satisfies Jacobi because each term is a scalar multiple of the basic formal identity x0−1δ(x1−x2x0)−x0−1δ(x2−x1−x0)=x2−1δ(x1−x0x2). This identity follows coefficientwise by expanding each integer power with the binomial convention. The only possible generalized eigenvalue of the zero LE(0) is zero. Thus grading restriction is exactly finite-dimensionality of E; its real lower weight bound is zero. Every linear map between these spaces commutes with all fields.

3.1F2step 2.1algebra

Since LE(−1)=0, the derivative identity forces ∂xY(e,x)f=0. Write each coefficient as xaPa(log⁡x) with Pa a finite polynomial. Its derivative is xa−1(aPa+Pa′). For a≠0, its top coefficient forces Pa=0; for a=0 one has Pa′=0. Distinct exponents remain distinct after differentiation. The only remaining term is therefore a constant B(e,f), bilinear in e,f. Conversely any such B has the required truncation, zero derivative and zero Möbius brackets. Its Jacobi identity is the scalar delta identity in step 2.1 multiplied by aB(e,f) for the state a1. Thus the assignment B↦YB is a linear bijection, and bases of E,F,G show its dimension is the displayed product.

4.1F3step 2.1step 3.1algebra

The completion of G is G, since it has only weight zero. With the specified trivial auxiliary grading, every linear map preserves auxiliary degree and its two input degrees add to zero. A P(z)-map is consequently just a linear map E⊗F→G: weight compatibility is automatic and lower truncation is automatic outside that single weight. Its Jacobi identity reduces to x0−1δ(x1−zx0)−z−1δ(x1−x0z)=x0−1δ(z−x1−x0), multiplied by its value and the scalar state; this is the identity in step 2.1 with x2=z, with the prescribed formal expansions. All three brackets vanish. Every linear map is therefore admitted. Composition with the identity of E⊗F gives precisely the required natural bijection of Hom spaces. This proves the P(z)-universal property directly for every z≠0.

4.2F4step 2.1step 3.1algebra

Because V+=0, C1(E)=0 and its quotient has dimension dim⁡E. Each strict weight cutoff is either zero or E, hence finite dimensional. The opposite transform fixes 1, so its transposed scalar field on E∗ is again scalar multiplication, and evaluation E→E∗∗ is a module isomorphism by finite-dimensional linear algebra. Every finite product or iterate of the operators from step 3.1 is an actual finite composition of bilinear maps, independent of all insertion variables. There is a single nonzero weight projection, so the projected series are finite. Matrix coefficients converge absolutely, extend as single-valued constant functions, and have a regular expansion with exponent zero and logarithmic degree zero at every boundary divisor.

5.1step 2.1step 4.1step 4.2algebra∎

The associator sends (e⊗f)⊗g to e⊗(f⊗g), unit maps are scalar multiplication, and the braiding sends e⊗f to f⊗e. Their inverses are the reverse rebracketing, insertion of 1, and the reverse flip. They are linear module maps and natural because their formulas commute with arbitrary linear maps on the factors. Both routes through the associativity pentagon send ((e⊗f)⊗g)⊗h to e⊗(f⊗(g⊗h)); both triangle routes send (e⊗a)⊗f to e⊗af. In the first hexagon both routes send (e⊗f)⊗g to f⊗(g⊗e), and in the second both send e⊗(f⊗g) to (g⊗e)⊗f. Pure tensors span, so these verify all coherence equations. Choosing a basis ei of E with dual ei∗, evaluation ei∗⊗ej↦δij and coevaluation 1↦∑iei⊗ei∗ have triangle composites e↦∑ieiei∗(e)=e and λ↦∑iλ(ei)ei∗=λ. These formulas also cover dimension zero by the empty sum. Hence this particular example has duality by direct verification.

Depends on

Dependency tree · 2 levels