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.

Groups of multiplicative type are linearly reductive

Statement

Assume the Axiom of Choice. Let k be a field and let G be a finite-type group scheme of multiplicative type over k (Groups of multiplicative type and tori). Then every rational representation of G (Rational representations and comodules of an affine group scheme) is a direct sum of simple representations, and the simple representations are classified by the Γk-orbits of the character group X∗(G): G is linearly reductive.

Facts & Assumptions

Given: The Axiom of Choice, a field k, and a finite-type group G of multiplicative type over k.

[F1]

Assume AC. Every finite-type group of multiplicative type splits over a finite Galois extension K/k: there is a finite Galois K/k with GK diagonalizable, O(GK)=K[M], where M=X∗(GK) is a finitely generated abelian group with continuous Γk-action, and G↦X∗(G) is a contravariant equivalence between finite-type groups of multiplicative type over k and finitely generated abelian groups with continuous Γk-action. (Multiplicative type groups split over a finite Galois extension, Multiplicative type groups and Galois character modules)

[F2]

Assume AC. For a finite Galois extension E/F with group Γ, the functor V↦E⊗FV is an equivalence between finite-dimensional F-vector spaces and finite-dimensional semilinear Γ-spaces, with inverse W↦WΓ. (Galois fixed points recover finite-dimensional scalar extensions, Semilinear Galois actions, twists, and split central idempotents)

[F3]

Over a field K in which GK is diagonalizable with character group M, every rational representation W of GK decomposes as W=⨁χ∈MWχ into eigenspaces for the distinct characters. (Representations of diagonalizable groups split into character eigenspaces, Rational representations and comodules of an affine group scheme)

Proof

Given: The Axiom of Choice, a field k, a finite-type group G of multiplicative type over k, and a rational representation V of G that is finite-dimensional over k.

1.1F1F2

By [F1] choose a finite Galois extension K/k with group Γ=Gal⁡(K/k) such that GK is diagonalizable with O(GK)=K[M], M=X∗(GK) finitely generated, and the Γ-action permutes the characters with σ(eχ)=eσχ. The base change VK=V⊗kK is a representation of GK equipped with a semilinear Γ-action compatible with the comodule structure, and by [F2] the passage V↦VK is an equivalence with the corresponding category of finite-dimensional semilinear Γ-equivariant GK-representations, with inverse W↦WΓ.

2.1F3step 1.1

By [F3] the representation VK decomposes as a direct sum VK=⨁χ∈M(VK)χ of eigenspaces for the distinct characters, and the Γ-equivariance of the comodule structure gives σ((VK)χ)=(VK)σχ for σ∈Γ, so the decomposition is permuted by Γ. For each Γ-orbit O⊆M put WO=⨁χ∈O(VK)χ; then WO is a Γ-stable GK-subrepresentation and VK=⨁OWO is a sum over the finitely many orbits.

3.1F2F3step 1.1step 2.1

Fix an orbit O, choose χ∈O, and put Γχ={σ:σχ=χ} and Kχ=KΓχ. The weight space (VK)χ has a semilinear Γχ-action. Galois descent [F2] identifies it with K⊗KχE for the Kχ-space E=((VK)χ)Γχ. Choose a basis of E. Each basis vector defines a Γχ-stable K-line in (VK)χ. Translate that line by representatives of Γ/Γχ and take their direct sum across the weights in O; the result is a Γ-stable GK-subrepresentation with one-dimensional weight spaces, and it is independent of the choice of representatives because the initial line is Γχ-stable. These orbit subrepresentations, one for each basis vector, decompose WO. Each descends by [F2] to a simple G-module: a submodule after scalar extension is a sum of some of its distinct one-dimensional weight spaces, and Γ-stability and transitivity on O force either none or all. Thus the whole isotypic block need not be simple, but is a direct sum of copies of one simple module indexed by O.

4.1F1F2F3step 3.1∎

Every finite-dimensional representation is therefore a direct sum of simple modules. For each orbit O the construction with a one-dimensional Kχ-space gives a simple module, and any simple module must have a single orbit and multiplicity one by step 3.1; different orbits have different scalar-extended weights. This gives the asserted classification. For an arbitrary rational V, project its coaction onto the direct sum of weight-coalgebra blocks for each Galois orbit; these finite-dimensional blocks descend from the spans of eχ, χ∈O, and counit and coassociativity give a direct decomposition V=⨁OVO. Finite Galois descent also holds for arbitrary semilinear spaces: for each vector, its finite orbit under the Galois group spans a finite-dimensional stable subspace, to which [F2] applies. This proves surjectivity of the canonical map from the scalar extension of invariants; a finite relation among invariant vectors lies in such a finite stable subspace, and finite-dimensional descent proves injectivity. Apply this argument to each weight space and its stabilizer subgroup. In each block choose a basis of the descended possibly infinite-dimensional space E under AC; the construction of step3.1 then decomposes VO into copies of its orbit simple. Hence every rational representation is a direct sum of simples and G is linearly reductive.

Depends on

Used by

Dependency tree · two levels

41 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