Alphabeta Math
LemmaStatement: Literature-sourcedProof: AI-adaptedPipeline-generatedjudge pass (gpt-5.6-terra)audited 2026-09-10
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.

Enveloping quotient kernels and augmentation intersections

Statement

For countably based complex Lie algebras with supplied compatible bases, a surjection θ:L→L/R with ideal kernel R induces ker⁡U(θ)=RU(L). For any subalgebra R⊂L with such a compatible basis, R∩RU0(L)=[R,R]. Here U0(L) is the kernel of the augmentation U(L)→C. These hypotheses hold for the homogeneous subalgebras used in this page by finite-degree elimination.

Facts & Assumptions

Given: The indicated bases; R is an ideal for the first claim and only a subalgebra for the second.

[F1]

PBW gives the compatible ordered monomial bases. (PBW for countably presented Kac Moody Lie algebras).

[F2]

The tensor quotient realizes Lie homomorphisms as associative homomorphisms. (The universal enveloping algebra as a tensor quotient).

Proof

1.1F2given

If R is an ideal, RU(L) is two-sided: xr=rx+[x,r] with [x,r]∈R lets every left generator pass it. It is killed by U(θ). The class of x in U(L)/RU(L) depends only on x+R, giving a Lie map L/R→U(L)/RU(L). F2 extends it to an associative inverse of the map U(L)/RU(L)→U(L/R): both composites fix all Lie generators. This proves the kernel equality.

1.2F1F2given

For the subalgebra R, order its basis before a complement and let W span the nonempty ordered complement monomials. Multiplication and F1 identify U(L)=U(R)⊕(U(R)⊗W) as left U(R)-modules. Augmentation then gives U0(L)=U0(R)⊕(U(R)⊗W). Nonempty words yield RU(R)=U0(R) and RU0(R)=U0(R)2: in a product of two nonempty words the first letter lies in R, and the remaining word is nonempty, and conversely. Left multiplication by R therefore yields RU0(L)=U0(R)2⊕(U0(R)⊗W).

2.1F1F2step 1.1

For any algebra K with these bases, [K,K]⊂K∩U0(K)2 because [x,y]=xy−yx. Mapping to U(K/[K,K]) sends U0(K)2 into the square of its augmentation ideal. This enveloping algebra is the polynomial algebra on a basis of the abelian quotient by F1: ordered words commute and have independent monomials. Its degree-one subspace has zero intersection with the ideal of polynomials of degree at least two. Thus an element of K∩U0(K)2 maps to zero in K/[K,K], proving equality.

3.1F1step 2.1step 1.2∎

The subspace R lies in the first summand of step 1.2. Intersecting gives R∩RU0(L)=R∩U0(R)2=[R,R] by step 2.1. This calculation retains commutators that can have PBW length one; it makes no false assertion that the augmentation square has only ordered monomials of length at least two. In the homogeneous applications, finite-degree echelon bases of F1 supply all compatible bases used above.

Sources

Source comparison: Kleshchev, Lemmas 9.3.1–9.3.3, pp.122–124; corrected left U(R)-module proof for Lemma 9.3.3.

Depends on

Used by

Dependency tree · one level

2 results within one dependency step 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