Alphabeta Math
LemmaStatement: Literature-sourcedProof: AI-adaptedPipeline-generatedprecheck passjudge pass (gpt-6-sol)audited 2026-09-30
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.

Relative projective-line cohomology and apolarity

Statement

Assume the Axiom of Choice. Let π:E→S be a Zariski locally trivial P1-bundle of complex schemes and let L be an invertible sheaf whose degree on every geometric fibre is the same integer n≥−1. Write Kπ=ωE/S. Then Rqπ∗L=0(q>0),Rqπ∗(L⊗Kπ⊗(n+1))=0(q≠1). With the nonzero apolarity normalization determined by the ordered residue x0−1x1−1↦1, there is a natural isomorphism aL:π∗L→ ∼ R1π∗(L⊗Kπ⊗(n+1)). It commutes with restriction on S, bundle isomorphisms and changes of local projective coordinates. When n=−1, both sheaves in this isomorphism are zero.

Facts & Assumptions

Given: π,E,S,L,Kπ,n as in the statement.

[F1]

On a sufficiently fine Zariski cover U of S with EU≅PU1, there is an invertible MU on U such that L∣EU≅O(n)⊗π∗MU. The evaluation construction gives MU=π∗(L(−n))∣U. (Local normal form for a line bundle on a projective-line bundle)

[F2]

For any ring A, the ordered two-chart Čech calculation gives H0(PA1,O(d))=Sym⁡d(A2)∨ for d≥0 and zero for d<0, while H1 vanishes for d≥−1 and for d≤−2 is free on the Laurent classes x0−ax1−b with a,b≥1 and a+b=−d; higher cohomology vanishes. These formulas commute with ring maps through the same Čech complex. (Cohomology of O(d) on projective space)

[F3]

Over a field, the Laurent coefficient functional sends x0−1x1−1∈H1(Pk1,O(−2)) to 1 and pairs H0(O(n)) perfectly with H1(O(−n−2)). (Residue pairing between H^0 and top cohomology of projective space)

[F4]

A higher direct image is the cohomology sheaf of the direct image of an injective resolution; for a quasi-compact separated morphism and a quasi-coherent sheaf, on each affine base open U its restriction is the associated sheaf of Hq(EU,−), compatibly with restriction to smaller affine opens. (Higher direct image of a sheaf, Higher direct images localize over an affine base)

[F5]

The Axiom of Choice is The Axiom of Choice and the Axiom of Dependent Choice is The axiom of dependent choice: a relation in which every element is related to something admits an N-indexed chain. AC implies DC: choose, for each element of the domain of a serial relation, one successor, then iterate this choice function from a given starting element. Thus the DC-qualified affine-localization supplier [F4] is available under the statement's AC hypothesis.

Proof

1.1F1construct

Work over a sufficiently small affine U=Spec⁡A that trivializes E and MU in [F1]. Put V=A2, use the line convention P(V), and write L∣EU=O(n)⊗π∗MU. The relative Euler sequence, or its direct two-chart differential calculation, gives Kπ∣EU=O(−2)⊗π∗(det⁡V)∨. Indeed ΩP(V)/U1 is the determinant of the kernel of V∨⊗O(−1)→O, and the displayed formula has central λIV-weight zero, as a relative canonical bundle must. Hence L⊗Kπ⊗(n+1)=O(−n−2)⊗π∗(MU⊗(det⁡V)−n−1).

2.1F1F2F4step 1.1

Apply [F2] to these two twists and [F4] to identify the resulting modules with higher direct images on U. For n≥0 this gives π∗L∣U≅Sym⁡n(V∨)⊗MU,Rqπ∗L∣U=0 (q>0), and Rqπ∗(LKπn+1)∣U=0 for q≠1. For n=−1 the two twists are both O(−1)⊗MU and [F2] makes every direct image zero. The calculations hold after every affine restriction because their Čech matrices are defined over A and tensor with the new base ring.

3.1F2F3step 1.1step 2.1algebra

For n≥0, the monomial bases of [F2] give a perfect pairing over A: a degree-n monomial x0ax1n−a pairs to 1 with the Laurent class x0−a−1x1−(n−a)−1 and to 0 with the other basis classes. This is the same ordered residue normalization as [F3] after every field specialization. It identifies H1(O(−n−2)) with Sym⁡n(V∨)∨⊗det⁡V=Sym⁡nV⊗det⁡V as a GL(V)-representation: the determinant factor records how the ordered Laurent residue changes under a coordinate matrix. After the determinant twist of step 1.1, R1π∗(LKπn+1)∣U≅Sym⁡nV⊗(det⁡V)−n⊗MU. The canonical wedge pairing in rank two gives a GL(V)-equivariant isomorphism V∨≅V⊗(det⁡V)−1; its nth symmetric power identifies the right side with Sym⁡n(V∨)⊗MU=π∗L∣U. Choose the scalar of this map so the dual pair of ordered monomials has residue 1; the wedge and residue formulas then determine the same nonzero invariant normalization on every chart.

4.1F1F2F3F4F5step 2.1step 3.1∎

On an overlap two projective trivializations differ by a PGL2-matrix. After an affine refinement lift it to g∈GL2; replacing g by λg changes the action on Sym⁡nV∨ by λ−n and the action on Sym⁡nV⊗(det⁡V)−n by λnλ−2n=λ−n. The transition of MU is the same on both sides because both functors are linear in L. Thus the GL2-equivariance of step 3.1 and equal central weights show that the local maps agree on overlaps independently of the lift, and they glue to aL. The Čech constructions, wedge map and descent use only restriction and tensor operations, so aL is compatible with base restriction and coordinate changes. The explicit projective-line Čech calculation of [F2] supplies the required cohomology comparison, and the completed local normal form [F1] is used at steps 1.1–2.1. AC is inherited through [F1], [F2] and [F4].

Depends on

Used by

Dependency tree · two levels

72 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.

Sources