Alphabeta Math
CounterexampleConstruction: AI-generatedVerification: AI-generatedPipeline-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.

Equal Euler classes do not by themselves prove homotopy-equivalent complexes

Statement refuted

The following statement is refuted: two bounded complexes with terms in SBimn and the same alternating class in K0split(SBimn) are homotopy equivalent (equivalently, the Euler or Hecke class determines the homotopy type).

Facts & Assumptions

Given: A simple reflection si, the bimodule Bi=R⊗RsiR(1) with generators u=1⊗1 of degree −1 and w0=1⊗δi, δi=αi/2, the standard graph bimodule Rsi of Standard graph bimodules, support filtrations and characters, and the two-term complexes Zi=[Bi→ 0 R(1)] and Fi=[Bi→εiR(1)] of The positive and negative Rouquier generator complexes.

[F1]

Equal classes. χ(Fi)=[Bi]−[R(1)] in K0split(SBimn); the alternating class is the alternating sum of the term classes, so the zero-differential complex with the same two terms in the same cohomological degrees has the same class χ(Zi)=[Bi]−[R(1)], and under the identification Φ of the split Grothendieck ring with the Hecke algebra this class corresponds to Hi−v. (Euler classes of Rouquier complexes are homotopy invariant and multiplicative, Decategorification of a Rouquier complex is the Hecke braid generator)

[F2]

Cohomology of the generator complex. H0(Fi)=ker⁡εi≅Rsi(−1) and H1(Fi)=coker⁡εi=0, all other cohomology of Fi being zero. (Rouquier generator complexes have canonical derived graph models)

[F3]

Cohomology of the zero-differential complex. For the bounded complex Zi with zero differential, H0(Zi)=Bi, H1(Zi)=R(1) and all other cohomology is zero. (Cohomology object of a cochain complex)

[F4]

Left R-ranks. Bi is finite free of rank two as a left R-module, while Rsi and its shifts are free of rank one as left R-modules; an isomorphism of graded bimodules restricts to an isomorphism of left R-modules, and isomorphic free modules have equal rank. (The Soergel bimodule Bi of a simple reflection, Standard graph bimodules, support filtrations and characters)

[F5]

Invariance of cohomology. Homotopy equivalent complexes have isomorphic cohomology objects, because homology factors through the homotopy category. (Homology factors uniquely through the homotopy category, Complexes, homotopies and contractibility in an additive category)

Counterexample

As a counterexample take the zero-differential complex Zi:=[  Bi→ 0 R(1)  ] with Bi in cohomological degree 0 and R(1) in degree 1, and the Rouquier generator complex Fi=[Bi→εiR(1)]. Both have the same alternating class χ(Zi)=χ(Fi)=[Bi]−[R(1)]∈K0split(SBimn),Φ(χ(Zi))=Φ(χ(Fi))=Hi−v∈HSn, but H0(Zi)=Bi and H1(Zi)=R(1), whereas H0(Fi)=Rsi(−1) and H1(Fi)=0. Since Bi is free of rank two as a left R-module while Rsi(−1) is free of rank one, the two complexes are not quasi-isomorphic and hence, by homotopy invariance of cohomology, not homotopy equivalent. Thus the decategorification map loses the differential data already at the level of the generators.

Proof technique: direct.

1.1F1

The classes agree. By [F1] the generator complex has class χ(Fi)=[Bi]−[R(1)], and Zi has the same two terms in the same cohomological degrees with zero differential, so its alternating class is the same alternating sum, χ(Zi)=[Bi]−[R(1)]=χ(Fi); under Φ both correspond to Hi−v.

1.2F2F3F4

The cohomology differs. By [F3] H0(Zi)=Bi and H1(Zi)=R(1), whereas by [F2] H0(Fi)≅Rsi(−1) and H1(Fi)=0; in particular H1(Zi) is nonzero while H1(Fi)=0, and the two left R-modules H0(Zi) and H0(Fi) have different ranks.

2.1F4F5step 1.1step 1.2∎

They are not homotopy equivalent. By [F4] the left R-module Bi has rank two while Rsi(−1) has rank one, so H0(Zi)≇H0(Fi); independently H1(Zi)≇H1(Fi). If Zi and Fi were homotopy equivalent, [F5] would make their cohomology objects isomorphic, a contradiction; hence Zi≄Fi although their classes agree, which refutes the statement.

Remarks

The counterexample uses no choice. The same phenomenon is why the derived comparisons and the homotopy-category comparisons of this page must not be conflated: in the derived category the Rouquier complexes become isomorphic to shifted graph models, while in the homotopy category the differentials carry information that the alternating class forgets already for the generators. The statement refuted is the general claim; the example above does not refute the weaker statement that two complexes with equal class and equal cohomology are homotopy equivalent, which is not claimed here in either direction.

Depends on

Used by

Nothing in the library uses this result yet.

Dependency tree · two levels

50 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