Alphabeta Math
LemmaStatement: Literature-sourcedProof: AI-adaptedPipeline-generatedjudge pass (gpt-6-luna)
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 countably generated phi-irreducible chain has a positive-measure small set

Statement

Let (X,X) be countably generated, let P be a probability kernel that is ϕ-irreducible (Phi-irreducibility for a general-state Markov kernel), and let ϕ be a nonzero σ-finite irreducibility measure. Then there are an integer m≥1, a measurable C with ϕ(C)>0, a probability ν with ν(C)>0, and ϵ>0 such that Pm(x,⋅)≥ϵν(⋅) for every x∈C.

Facts & Assumptions

Given: Countable generation, the probability kernel, and ϕ-irreducibility.

[F1]

Phi-irreducibility for a general-state Markov kernel: every positive-ϕ set is reached with positive probability at some finite positive time from every state.

[F2]

Small-set minorization and geometric drift: a small set is a common fixed-time minorization by a probability measure.

[F3]

Iterated kernel densities retain convolution lower bounds: iterated kernel densities satisfy convolution lower bounds and their sum is positive ϕ-almost everywhere from every state.

[F4]

A positive triple set has uniformly overlapping sections: a positive-measure set of composable density pairs has sections with a uniform positive overlap.

[F5]

Measurable spaces and measurable sets: countable generation supplies finite partitions and countable testing algebras.

[A1]

The Axiom of Choice: the density versions and measure representatives in [F3] are selected under AC.

[F7]

Tonelli's theorem for nonnegative measurable functions on a sigma-finite product: nonnegative product integrals and countable sums can be interchanged.

Proof

Given: The hypotheses and facts [F1]–[F5], [F7], and [A1].

1.1F3F5F7A1given

Write ϕ as a countable disjoint union of measurable finite-measure sets Ej. Set cj=2−j/(1+ϕ(Ej)) and c=∑jcjϕ(Ej). Since ϕ is nonzero and the Ej cover the space, 0<c≤1. Then ϕˉ(A)=c−1∑jcjϕ(A∩Ej) is an equivalent probability measure: it has density cj/c>0 on every Ej, and hence exactly the same null sets and positive sets as ϕ. Replace ϕ by ϕˉ for the rest of the proof. By [F3], for every x the sum ∑npn(x,y) is positive for ϕ-almost every y. Tonelli and countability imply that for ϕ⊗3-almost every (x,y,z), this sum is positive at (x,y) and the corresponding sum at (y,z) is positive. Hence the product of the two sums is positive almost everywhere; its countable expansion shows that some fixed r,s≥1 satisfy ∫pr(x,y)ps(y,z) ϕ(dx)ϕ(dy)ϕ(dz)>0. For a sufficiently small η>0, the set where both factors exceed η still has positive product measure.

2.1F3F4step 1.1

Apply [F4] to A={(x,y):pr(x,y)>η} and B={(y,z):ps(y,z)>η}. Obtain positive-ϕ sets C0,D0 and γ>0 such that their sections overlap by at least γ. The convolution bound [F3] gives pr+s(u,v)≥η2γ for every u∈C0,v∈D0.

3.1F1F5F7step 2.1

Since C0 is ϕ-positive, irreducibility gives ∑k≥1Pk(x,C0)>0 for every x. Integrating over D0 and using Tonelli, some k≥1 has ∫D0Pk(x,C0)ϕ(dx)>0. A positive-level-set subdivision yields G⊆D0 with ϕ(G)>0 and δ>0 such that Pk(x,C0)≥δ for every x∈G.

4.1F2F3step 2.1step 3.1∎

For every x∈G and z∈D0, the second convolution inequality in [F3] gives pr+s+k(x,z)≥∫C0Pk(x,dy)pr+s(y,z)≥δη2γ. Let ν(A)=ϕ(A∩D0)/ϕ(D0) and choose 0<ϵ≤min⁡{1,δη2γϕ(D0)}. Integrating the density over A∩D0 gives Pr+s+k(x,A)≥ϵν(A) for every x∈G. Since G⊆D0 and ϕ(G)>0, ν(G)>0. Thus G is the required small set.

Depends on

Used by

Dependency tree · two levels

37 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