Alphabeta Math
TheoremStatement: AI-adaptedProof: AI-adaptedPipeline-generatedjudge pass (gpt-6.1-sol)audited 2026-10-02
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.

Termwise Hochschild homology respects bimodule chain homotopies

Statement

Let k be a field, let A be a unital associative k-algebra, and let F=(Fi,dFi) and G=(Gi,dGi) be bounded cochain complexes of k-central A-bimodules with differentials of internal degree zero (Bounded graded bimodule complexes and signed tensor totalization). Let f,g:F→G be bimodule-linear cochain maps of cochain degree zero and let h:F→G be a bimodule-linear cochain homotopy of cochain degree −1 with f−g=dGh+h dF on F∙ (A chain homotopy).

Then for every j≥0 the induced maps HHj(A,f),HHj(A,g):HHj(A,F∙)→HHj(A,G∙) on the termwise Hochschild complexes of Termwise Hochschild homology and iterated homology are cochain-homotopic; the homotopy is induced by HHj(A,h):HHj(A,Fi)→HHj(A,Gi−1), and hence HHj(A,f) and HHj(A,g) induce the same map Hi(HHj(A,F))→Hi(HHj(A,G)) for every i (Cohomology object of a cochain complex). Consequently a bimodule chain-homotopy equivalence F→G induces isomorphisms Hi(HHj(A,F))→Hi(HHj(A,G)) for all i,j. Everything here is choice-free, and no invariance of the termwise groups under arbitrary quasi-isomorphisms is asserted.

Facts & Assumptions

Given: a field k, a unital associative k-algebra A, bounded cochain complexes F,G of k-central A-bimodules with internal-degree-zero differentials, bimodule-linear cochain maps f,g:F→G of cochain degree zero, and a bimodule-linear cochain homotopy h:f≃g of cochain degree −1 with f−g=dGh+hdF.

[F1]

The Hochschild chain complex of a k-central bimodule M has Cj(A,M)=M⊗kA⊗kj with boundary bj the alternating sum of faces; the termwise complex of a bounded complex F in Hochschild degree j is HHj(A,F∙) with differentials HHj(A,dFi) induced by the bimodule maps dFi, and its cohomology is Hi(HHj(A,F∙)) (Hochschild chains and Hochschild homology with coefficients, Termwise Hochschild homology and iterated homology).

[F2]

HHj(A,Fi)=Hj(C∙(A,Fi)) is the homology of the Hochschild chain complex, and a chain map u:C∙→D∙ induces a well-defined map Hj(u) (A chain map induces a well-defined map on homology).

[F3]

A chain homotopy s between chain maps of chain complexes satisfies fn−gn=dn+1sn+sn−1dn in each degree; homotopic chain maps induce the same map on homology (A chain homotopy).

[F4]

A map of k-central A-bimodules commutes with every Hochschild face, since the faces multiply the coefficient by algebra elements on either side; hence a bimodule map u:M→N induces a chain map C∙(A,u):C∙(A,M)→C∙(A,N) natural in the bimodule (Hochschild chains and Hochschild homology with coefficients, Enveloping algebra and the bimodule–module dictionary).

[F5]

For fixed j and composable bimodule maps the assignment u↦Cj(A,u) is additive: Cj(A,u+v)=Cj(A,u)+Cj(A,v), because the tensor product of a map with an identity is linear in the map; it also preserves identities and composition, so HHj(A,−) is additive on maps (Hochschild chains and Hochschild homology with coefficients).

[F6]

Chain-homotopic maps induce the same map on homology; after reindexing cochain degree i as homological degree −i, homotopic cochain maps induce the same map on cohomology (Chain-homotopic maps induce the same map on homology).

Proof

technique · direct
1.1F1F2F4givenalgebra

Fix j≥0. For each i the bimodule map dFi:Fi→Fi+1 commutes with every Hochschild face by [F4], so it induces a chain map C∙(A,dFi):C∙(A,Fi)→C∙(A,Fi+1), and hence a map HHj(A,dFi) on homology by [F2]. The same applies to dG, f, g and h; the homotopy h has cochain degree −1, so C∙(A,h) is a degree-(−1) family of maps of Hochschild complexes.

1.2F1F2F3F5givenalgebra

Because C∙(A,−) is additive on maps by [F5], applying it to the homotopy identity f−g=dGh+hdF gives exactly C∙(A,f)−C∙(A,g)=C∙(A,dG)C∙(A,h)+C∙(A,h)C∙(A,dF). Passing to homology with [F2], this is the displayed homotopy identity HHj(A,f)−HHj(A,g)=HHj(A,dG)HHj(A,h)+HHj(A,h)HHj(A,dF) in Hochschild degree j, valid in every cochain degree; the family HHj(A,h) has cochain degree −1 and is a cochain homotopy of the termwise complexes by [F3].

2.1F2F6step 1.1step 1.2givenalgebra

Applying [F6] in each cochain degree i, the cochain-homotopic maps HHj(A,f) and HHj(A,g) induce the same map Hi(HHj(A,F))→Hi(HHj(A,G)) on the cohomology of the termwise complex, and the induced map depends only on the cochain-homotopy class of the map of coefficient complexes. Everything in the argument is a computation of maps of k-vector spaces, so no choice is used, and no statement about arbitrary quasi-isomorphisms is made. In the graded case the statement is ungraded unless f,g,h are also internal-degree-zero; under that extra condition the induced homotopy and maps preserve internal degree.

3.1F2F3step 2.1givenalgebra∎

Now let f:F→G be a bimodule chain-homotopy equivalence, with bimodule-linear homotopy inverse g:G→F of cochain degree zero and two bimodule-linear homotopies fg≃idG and gf≃idF of cochain degree −1. By 2.1 and functoriality, the induced maps on Hi(HHj(A,−)) compose to the identity in both orders, so they are inverse isomorphisms for every i,j.

Depends on

Used by

Nothing in the library uses this result yet.

Dependency tree · two levels

28 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