Alphabeta Math
LemmaStatement: AI-adaptedProof: AI-generatedPipeline-generatedprecheck passverified 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.

Admissible triplet action and the Jordan top of the cyclic logarithmic module

Statement

Use the rank-one lattice-screening data for p≥2, and write W(p) for the vertex subalgebra generated by ω,F,H,E in VL. Its conformal vector is ω with central charge 1−6(p−1)2/p. The deformed fields on M are an actual ordinary W(p)-module action satisfying the full module Jacobi identity. Put h=(2−p)/4. The cyclic space P+p−1=W(p)eα/2 is an admissible grading-restricted generalized logarithmic module. Its lowest generalized conformal-weight space has basis u=eα/2 and w=eα/2−α/p, with L~(0)u=hu+w,L~(0)w=hw. Moreover M1=W(p)e−α/2 is an admissible submodule of P+p−1, and L~(0) is semisimple on M1. This statement does not identify its submodules or quotient endpoints and does not assert the nonsplit extension.

Facts & Assumptions

Given: The explicit lattice-screening data and integer p≥2. For oscillator degree d and charge t write wt⁡(fetα)=d+pt2−(p−1)t. The notation W(p) here uses the four-generator definition; equality with the full screening kernel is not required.

[F1]

Rank-one lattice and square-zero screening data for the triplet construction gives the explicit fields and sector-zero/sector-one spaces.

[F2]

Jacobi identities for the rank-one square-zero lattice extension proves that A is a vertex algebra, I is square-zero, and M is an actual A-module with the full Jacobi identity.

[F3]

Full module Jacobi identity for a square-zero screening deformation proves the full module Jacobi identity for the deformed action on every vertex subalgebra of ker⁡v(0).

[F4]

Admissible module for a vertex operator algebra, Grading-restricted generalized module and Logarithmic generalized module specify the grading, finiteness, lower bound and nonsemisimplicity requirements.

Proof

1.1givenF1F2algebra

Write q=(p−1)/(2p). The field of ω is T(z)=:α(z)2:/(4p)+q∂zα(z). The contraction α(z)α(w)∼2p/(z−w)2, obtained directly from the oscillator commutator, gives the singular product T(z)T(w)∼(1/2−12pq2)/(z−w)4+2T(w)/(z−w)2+∂wT(w)/(z−w). Here the double contraction of the quadratic terms is 1/2 in fourth order; the two improvement terms give q2∂z∂w(2p/(z−w)2)=−12pq2/(z−w)4; the mixed cubic poles cancel, leaving 2q∂α(w)/(z−w)2+q∂2α(w)/(z−w). The remaining quadratic single contractions give 2:α(w)2:/(4p) and its derivative. Taking mode residues proves the Virasoro relations with central charge 1−24pq2=1−6(p−1)2/p. In particular L(n)=14p∑j∈Z:αjαn−j:−q(n+1)αn. The formula for n=−1 acts on fetα as ∑k≥1kxk+1∂xkf+tx1f, the translation in [F2]. The n=0 formula is L(0)=∑k≥1kxk∂xk+pt2−(p−1)t. Thus it is diagonal with the displayed weights on both spaces. In VL these weights are nonnegative integers: pn2−(p−1)n≥0 for every integer n, with equality only at n=0. Its weight spaces are finite, since the quadratic bounds the possible charges and the polynomial space in each oscillator degree is finite.

2.1F1F2step 1.1algebra

The same contraction calculation gives T(z)Et(w)∼(pt2−(p−1)t)Et(w)/(z−w)2+∂wEt(w)/(z−w). Indeed the quadratic double contraction contributes pt2, the improvement contributes −(p−1)t, and the quadratic single contraction contributes t:αEt:=∂Et, as follows by differentiating its explicit exponential formula. In particular v=e−α/p is primary of weight one. To check the screening kernel on the generators directly, apply its exponential to ω: the annihilation exponential substitutes x1+2z−1 and x2+2z−2 into x12/(4p)+qx2. The result before its creation exponential is ω+(x1/p)z−1+z−2; the creation exponential begins 1−(x1/p)z+⋯. Hence v(1)ω=v, v(0)ω=0 and v(n)ω=0 for n≥2. Also Y(v,z)F starts with z2, so v(0)F=0. Finally Qv=x1e(1−1/p)α=pp−1Te(1−1/p)α. The zero-mode commutator formula from the adjoint Jacobi identity gives [Q,v(0)]=(Qv)(0)=0, since the zero mode of a derivative is zero. Therefore v(0)H=v(0)E=0. The same commutator formula makes v(0) a derivation of every state product, so its kernel is a vertex subalgebra containing ω,F,H,E; it contains W(p).

3.1F1F2F3F4step 1.1step 2.1algebra

Apply [F3] to A,I,v and S=W(p). It supplies the full module Jacobi identity on M, not merely the mode commutation relations. Step 2.1 gives Δ(z)ω=ω+z−1v, whence L~(n)=L(n)+v(n) and L~(0)=L(0)+v(0). The screening v(0) has weight zero and square zero on M, since it sends sector zero to sector one and annihilates sector one. For the two module charge strings, subtracting h from the weight of the charge vacuum gives respectively pn2+n and pn2−n, for n∈Z. These are nonnegative integers, vanishing only at n=0. Thus M=⨁d≥0M(d), with each degree finite and L(0)=(h+d)I on it. Since v(0) preserves degree, these are exactly the generalized L~(0)-weight spaces. Counting weights in the explicit fields gives a(k)M(d)⊂M(d+wt⁡a−k−1) for homogeneous a∈VL: the charge shift, oscillator multiplications and derivatives have exactly these weight changes. In a correction v(n)a has weight wt⁡a−n and the extra factor z−n increases the mode index by n, leaving the same shift. This is the admissible grading for the deformed fields.

4.1F1F4step 3.1algebra

Put u=eα/2 and w=eα/2−α/p. The screening exponential on u starts with z−1w, so v(0)u=w; it annihilates w by the sector rule. Both have original weight h. Therefore L~(0)u=hu+w and L~(0)w=hw. Since the cyclic submodule containing u is stable under L~(0), it also contains w. The degree-zero space of M is precisely Cu⊕Cw by step 3.1, so this is exactly the degree-zero space of P+p−1. Every invariant subspace of the direct sum of finite generalized eigenspaces of L~(0) decomposes as its intersections with them: for each vector, projection onto any one of its finitely many generalized eigencomponents is a polynomial in L~(0), obtained by the coprime factors (X−(h+d))2. Thus the cyclic submodule inherits the admissible grading, finite generalized weight spaces and lower bound. The nonzero nilpotent action on its degree-zero space proves it logarithmic in [F4].

5.1F1F2F3step 2.1step 3.1step 4.1algebra∎

The correction to F vanishes because v(n)F=0 for all n≥0. The leading term of YM(F,z)u is z−pe−α/2, so F~(p−1)u=e−α/2 and M1⊂P+p−1. It inherits the same grading by the polynomial-projection argument in step 4.1. To check semisimplicity on M1, the screening exponential on its cyclic vector has leading power z1, giving v(0)e−α/2=0. For a∈W(p) the ordinary module zero-mode commutator gives [v(0),YM(a,z)]=YM(v(0)a,z)=0. Every deformation correction has state in I, which is also annihilated by the adjoint screening zero mode; hence this commutation holds for the deformed fields too. Therefore v(0) vanishes on the entire cyclic space M1. On it L~(0)=L(0), and the explicit oscillator-weight operator is diagonal. This proves the claimed semisimplicity and locates the Jordan block in P+p−1 rather than M1.

Depends on

Dependency tree · 2 levels

Sources