Alphabeta Math
LemmaStatement: Literature-sourcedProof: AI-adaptedPipeline-generatedjudge pass (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.

A character of a normal subgroup admits an inverse multiple in a group representation

Statement

Assume AC. Let k be algebraically closed, G an affine finite-type k-group scheme and H⊂G a closed normal subgroup scheme. If a character χ:H→Gm occurs in a finite-dimensional representation V of G (there is a nonzero vector spanning an H-stable line of character χ), then χ−m occurs in a finite-dimensional representation of G for some integer m>0. Both G and H may be nonsmooth.

Facts & Assumptions

[F1]

High relative Frobenius has smooth scheme-theoretic image. (High relative Frobenius has smooth scheme-theoretic image)

[F2]

A nonempty reduced finite-type scheme over a perfect field has a nonempty regular locus, and regularity equals smoothness. The regular locus of any finite-type scheme over a perfect field is open. A Noetherian local ring is regular when its dimension equals the dimension of its maximal ideal modulo its square, and regular local rings are domains. (Dense regular loci on every component, Regular equals smooth over a perfect field, embedding dimension and regular local ring, Openness of the regular locus over a perfect field, regular local rings are domains and cohen macaulay)

[F3]

Closed points of finite-type schemes over an algebraically closed field are rational; a function on a reduced such affine scheme vanishing at all closed points is zero. The second assertion follows from the first by applying it to the nonempty principal open where a proposed nonzero function is invertible. (Over an algebraically closed field, every maximal ideal is an evaluation ideal)

Proof

Given: AC, G,H,V,χ and a line L⊂V of character χ.

1.1F2F3givenalgebra

First record the characteristic-zero reducedness needed below. Put A=k[G], m=ker⁡ϵ. The reduction of G is smooth: by [F2] it has a regular rational point, and translations by rational points preserve the reduction and carry that point to the identity and then to every closed point. The open regular locus therefore contains all closed points and is the whole reduction, by [F3]. For any nilpotent a∈A vanishing in Am, localization of A/m2 at its unique maximal ideal is an isomorphism, so a∈m2. Otherwise choose the least n≥2 with an=0 in Am. Multiplying a by some s∉m arranges an=0 already in A and an−1≠0 in Am. This replacement does not change whether a∈m2, since s is invertible modulo m2. The counit identities give Δ(a)=a⊗1+1⊗a+y with y∈m⊗m. Expanding Δ(a)n=0 modulo A⊗m2 gives nan−1⊗aˉ∈an−1m⊗(A/m2). Here an−1∉an−1m: otherwise (1−t)an−1=0 for some t∈m, contradicting its nonzero localization. Since n≠0 in characteristic zero, projecting the first tensor factor modulo an−1m proves aˉ=0. Thus all nilpotents belong to m2. The cotangent dimension of G at the identity equals that of its smooth reduction, and its local dimension is also unchanged by reduction. By [F2] its local ring at the identity is regular. Translations and the openness of the regular locus supplied by [F2] make G regular everywhere; [F2] makes it smooth, hence reduced.

1.2givenalgebra

For any affine group scheme, distinct characters are linearly independent in its coordinate ring. Indeed their coordinate functions are group-like elements b with Δ(b)=b⊗b and ϵ(b)=1. If one were a linear combination b=∑icibi of independent other group-like elements, comparison of Δ(b) would give ci2=ci, cicj=0 for i≠j, and ∑ici=1. Over a field precisely one coefficient is one, contradicting distinctness. Consequently the character eigenspaces in any comodule have direct sum: apply its coaction to a finite relation among weight vectors, then project the coordinate factor onto each independent group-like function. This applies to nonsmooth H.

2.1F3step 1.1step 1.2constructalgebra

Suppose G is reduced. Let W⊂V be the sum of all H-character eigenspaces. Normality says every g∈G(k) sends an H-weight vector to another H-weight vector, since h(gv)=g(g−1hg)v holds after every algebra extension. Thus G(k) preserves W. In a basis extending one of W, the matrix coefficients for the induced map W→V/W vanish at all rational points and hence vanish by [F3]; W is a G-subrepresentation. By step 1.2 it is the direct sum of its H-weight spaces. Choose a complement to L in its finite-dimensional weight space and add the other weight spaces. This makes L an H-module direct summand of W, so the embedded dual line in W∗ has character χ−1. Step 1.1 proves that this case always applies in characteristic zero.

3.1F1F3step 1.2step 2.1constructalgebra∎

In characteristic p>0, choose q=pr with the Frobenius image I⊂G(r) smooth by [F1]. Inside Sym⁡qV take the span U of the pure powers vq. If ei is a basis of V, the eiq form a basis of U, and its representation matrix is (aijq) when that on V is (aij). In particular it factors through I: these coefficients are pullbacks of the twisted matrix coefficients on G(r), restricted to its scheme-theoretic image. The Hopf identities hold on I because k[I]→k[G] and its tensor square are injective. The line Lq⊂U has H-character χq. Let W be the sum of the H-character eigenspaces in U. Normality makes W stable under G(k) as in step 2.1. Frobenius is a universal homeomorphism onto I, and its closed points over algebraically closed k are rational, so G(k)→I(k) is onto. Therefore I(k) preserves W; since I is reduced, [F3] makes W stable under I as a scheme, hence under G. Step 1.2 again makes Lq an H-module direct summand. Its dual occurs in the G-representation W∗ with character χ−q. This proves the assertion with m=q. The pure-power subspace is essential: the entire tensor power need not be killed by the Frobenius kernel. AC enters through [F1]–[F3].

Depends on

Used by

Dependency tree · two levels

59 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