Alphabeta Math
LemmaStatement: Literature-sourcedProof: AI-adaptedPipeline-generatedprecheck passjudge 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.

Rouquier generator complexes have canonical derived graph models

Statement

For every 1≤i≤n−1 the generator complexes of The positive and negative Rouquier generator complexes are canonically isomorphic in the bounded derived category to shifts of standard graph bimodules: Fi≅Rsi(−1),Fi−1≅Rsi(1)in Db(Re-grmod), the isomorphisms being induced by the quasi-isomorphisms fi below. More generally, for a signed word σ=σi1ϵ1⋯σirϵr with product w=si1ϵ1⋯sirϵr and exponent sum e(σ)=∑kϵk, F(σ)≅Rw(−e(σ))in Db(Re-grmod), where F(σ) is the iterated signed tensor totalization Fi1ϵ1⊗R⋯⊗RFirϵr and Rw is the standard graph bimodule. All isomorphisms are obtained from the comparison system ct,u of Canonical comparisons between standard graph tensor products and do not depend on the chosen words beyond their permutation product and exponent sum, with the displayed graph models and fixed generator maps understood.

Facts & Assumptions

Given: A simple reflection si, the bimodule Bi with generators u=1⊗1 (degree −1) and w0=1⊗δi, δi=αi/2, the element ρ−=δiu−w0, and the complexes Fi,Fi−1 of The positive and negative Rouquier generator complexes.

[F1]

The two exact sequences. The multiplication εi:Bi→R(1) is a surjective degree-zero bimodule map with εi(u)=1 and εi(w0)=δi, and ker⁡εi=Rρ− with Rρ−≅Rsi(−1) a graded sub-bimodule generated in degree 1; the map ηi:R(−1)→Bi is injective and the quotient Bi/ηi(R(−1))=Bi/Rρ+ is generated by the image of u with right action u⋅g≡si(g)u, so that it is isomorphic to Rsi(1); here ρ+=δiu+w0 satisfies ρ+⋅g=gρ+ and ρ−⋅g=si(g)ρ− (Standard graph bimodules, support filtrations and characters, The positive and negative Rouquier generator complexes).

[F2]

Two canonical chain maps. The assignment 1↦1⊗αi−αi⊗1=−2ρ− defines a degree-zero bimodule map fi:Rsi(−1)→Fi, concentrated in cohomological degree 0; it is a chain map because εi(ρ−)=0. The assignment 1⊗1↦1 defines the twisted multiplication ψ:Bi→Rsi(1), ψ(r⊗r′)=r si(r′), a degree-zero bimodule map with ψ(ηi(1))=0; on the quotient Bi/ηi(R(−1)) it induces the identification with Rsi(1) and defines a chain map gi:Fi−1→Rsi(1), concentrated in cohomological degree 0. (Standard graph bimodules, support filtrations and characters, Canonical comparisons between standard graph tensor products)

[F3]

Cohomology of the generator complexes. H0(Fi)=ker⁡εi=Rρ−≅Rsi(−1), H1(Fi)=coker⁡εi=0, and H−1(Fi−1)=ker⁡ηi=0, H0(Fi−1)=Bi/ηi(R(−1))≅Rsi(1); both complexes are otherwise concentrated in the displayed degrees. [F1]

[F4]

Localization. The localization functor Kb(Re-grmod)→Db(Re-grmod) sends quasi-isomorphisms to isomorphisms, and tensor totalization with a bounded complex of bimodules flat on the tensoring side preserves quasi-isomorphisms. Here all generator complexes and graph models are flat on both sides: the former have finite-free terms and the latter are twisted rank-one regular modules. Consequently the tensor comparisons used below are compatible with localization (The localization functor sends quasi isomorphisms to isomorphisms, Derived category of an abelian category, Bounded above flat tensor complexes preserve quasi isomorphisms).

[F5]

Multiplication of graph bimodules. The balanced assignment Rx⊗RRy→Rxy, a⊗b↦a x(b), is a degree-zero isomorphism of graded bimodules with inverse 1↦1⊗1; shifts satisfy M(a)⊗RN(b)≅(M⊗RN)(a+b) (Canonical comparisons between standard graph tensor products, Standard graph bimodules, support filtrations and characters).

Proof

technique · direct
1.1F1F2

The map fi of [F2] is a chain map between the complexes Rsi(−1) concentrated in degree 0 and Fi in degrees 0,1: the only condition is that the composite of fi with the differential εi vanishes, which holds since εi(ρ−)=0 by [F1]. It is a bimodule map: for g∈R one has fi(1⋅g)=fi(si(g))=−2si(g)ρ− and fi(1)⋅g=−2ρ−⋅g=−2si(g)ρ− by [F1], and R-linearity on the left is clear.

2.1F1F2F3step 1.1

Comparing with [F3], fi induces the identity identification H0(Rsi(−1))=Rsi(−1)→Rρ− (up to the unit −2) and there are no other cohomology groups on either side; hence fi is a quasi-isomorphism, and so is gi between Fi−1 and Rsi(1).

3.1F4step 2.1

By [F4] the quasi-isomorphisms fi,gi become isomorphisms in Db, giving Fi≅Rsi(−1) and Fi−1≅Rsi(1).

4.1F4F5step 3.1

For a signed word, tensoring the isomorphisms of step 3.1 over R and using that the totalization of bimodule complexes is compatible with localization in each variable [F4], together with the shift computation M(a)⊗RN(b)≅(M⊗RN)(a+b) and the multiplication isomorphism of [F5] iterated over the word, gives F(σ)≅Rsi1ϵ1⊗R⋯⊗RRsirϵr(−∑kϵk)≅Rw(−e(σ)) in Db.

5.1F2F4F5step 4.1∎

With the generator maps fi,gi fixed, tensor their derived isomorphisms (using fi−1 for a positive letter and gi for a negative letter), then compose with the graph multiplication μt of [F5]. This specifies the word map to Rw(−e(σ)) with its full shift retained; no replacement of e(σ) by the length of a reduced permutation word is made. Associativity of graph multiplication makes this construction compatible with the canonical rebracketings. Comparisons between different word models are obtained by composing their specified isomorphisms through this common target when their permutation and exponent agree.

Remarks

This is Rouquier's §3.2.1 and §3.2.4 identification of the generators with the standard graph bimodules, in the library normalization: the unit scalar −2 and the shifts (∓1) are the exact dictionary entries, so no unit factors are dropped. The word statement is proved here because the later uniqueness and decategorification arguments use the identification F(σ)≅Rw(−e(σ)) in Db as a consequence of the comparison system. The statement is a derived-category statement; it does not assert that fi is a homotopy equivalence, and no homotopy-category identification is made here.

Depends on

Used by

Dependency tree · two levels

25 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