Alphabeta Math
ExampleConstruction: AI-generatedVerification: 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.

A polynomial PID has a nonzero Kunneth Tor class

Example

Assume AC and let k be a field. Over the PID R=k[t], take C=D=(R→tR) in degrees 1,0, with respective bases e1,e0 and f1,f0. With the direct-sum total complex and Koszul differential, Hi(C⊗RD)≅{R/(t)i=0,1,0i∉{0,1}. The class [e1⊗f0−e0⊗f1] maps to the generator [1] of Tor⁡1R(R/(t),R/(t))≅R/(t). Thus the tensor cross product alone does not exhaust degree-one homology, although the Kunneth sequence has a section.

Facts & Assumptions

Given: A field k, the two displayed complexes, and The Axiom of Choice.

[F1]

Products of nonzero polynomials over a domain are nonzero and degrees add: Over an integral domain, degrees add under multiplication of nonzero polynomials.

[F2]

Polynomial division over a field gives a zero remainder or one of degree smaller than the nonzero divisor: Division algorithm for polynomials over a field.

[F3]

Every Euclidean domain is a PID: Every Euclidean domain is a principal ideal domain.

[F4]

The AC-qualified natural Kunneth sequence uses the cycle tensor cross product: The natural PID Kunneth sequence is exact.

[F5]

Under the same hypotheses its quotient has a linear section: The PID Kunneth sequence admits a section after choices.

[F6]

Balanced Tor is computed from either supplied projective resolution under DC: The balanced Tor bifunctor.

[F7]

The cycle-boundary kernel calculation identifies the Kunneth quotient as the corestricted H(ρ⊗1) with positive sign: The cycle-boundary tensor sequence has the Kunneth kernel and cokernel.

[F8]

A self-map of a set can be iterated from any prescribed initial element: The recursion theorem.

[F9]

DC is the assertion that every entire relation on a nonempty set has such a sequence from a prescribed initial point: The axiom of dependent choice: a relation in which every element is related to something admits an N-indexed chain.

Verification

1.1F1F2F3given

Since k is a field, it is a domain. By [F1], k[t] is a domain, t≠0, and t is not a unit: tf=1 for nonzero f would give 1+deg⁡f=0. By [F2], degree on the nonzero polynomials satisfies the Euclidean division condition (the convention requires no multiplicative monotonicity). Thus [F3] applies and R is a commutative PID. Multiplication by t on R is injective, so H0C=H0D=R/(t) and all other homology groups of the two input complexes vanish.

1.2F4given

Write u=e1⊗f0, v=e0⊗f1, w=e0⊗f0, and h=e1⊗f1. The tensor total complex has terms Rh, Ru⊕Rv, Rw in degrees 2,1,0. Its Koszul formula gives d1(u)=tw, d1(v)=tw, and d2(h)=tv−tu. Thus in the ordered basis (u,v), d1(a,b)=t(a+b) and d2(c)=(−tc,tc); d1d2(c)=t(−tc+tc)=0.

2.1step 1.1step 1.2

The image of d1 is tRw, so H0=Rw/tRw≅R/(t). Since t is not a zero divisor, d1(a,b)=0 exactly when a+b=0. Hence ker⁡d1=R(u−v) and im⁡d2=tR(u−v), giving H1≅R/(t) via [a]↦[a(u−v)]. Also d2(c)=0 forces tc=0 and hence c=0, so H2=0. There are no chain terms in the other degrees.

2.2givenstep 1.1F6F8F9

First derive the DC hypothesis of [F6] from the assumed AC. Given an entire relation E on a nonempty set S and a∈S, AC chooses a successor function f:S→S with xEf(x) for every x. By [F8], iterate f from a; the resulting sequence satisfies [F9]. The length-one free resolution R→tR→R/(t) of the first module, tensored with R/(t), has differential zero and no degree-two term. Thus [F6] gives Tor⁡1R(R/(t),R/(t))=R/(t), represented by 1⊗[1]. In the cycle-boundary presentation B0C=tRe0↪Z0C=Re0, identify its degree-one free module R by a↦ate0. This is an isomorphism because t is nonzero in a domain.

3.1step 1.2step 2.1step 2.2F7

The map ρ1:C1→B0C sends e1 to te0, while ρ0 is zero. Therefore ρ⊗1 sends the degree-one cycle u−v to te0⊗f0, whose homology image is te0⊗[f0]. Under the resolution identification of step 2.2, this is exactly 1⊗[1], the positive generator. Since 1∉(t), the image and the original class are nonzero. This calculation works also in characteristic two, where minus equals plus but a+b=0 still means b=−a.

4.1F4F5step 1.1step 2.1step 3.1∎

In degree one the tensor term of [F4] is zero because only degree-zero input homology is nonzero. Hence the quotient is an isomorphism, with section explicitly [a]↦[a(u−v)], well defined by step 2.1 and inverse by step 3.1. It is a section of the kind guaranteed by [F5]. In degree zero, multiplication gives R/(t)⊗RR/(t)≅R/(t) with inverse [a]↦[a]⊗[1]: balancing gives [a]⊗[b]=[ab]⊗[1]. The cross product carries this tensor generator to [w]. These computations verify the zero and one degree endpoints and all higher vanishing.

Depends on

Used by

Nothing in the library uses this result yet.

Dependency tree · two levels

45 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