Result 154, Dynamical systems and ergodic theory

Pointwise multiple ergodic averages for mixing transformations

Proves almost-everywhere convergence of consecutive multiple ergodic averages of every finite length for invertible mixing probability-preserving transformations. For each fixed tuple of bounded functions, the limit is the product of their integrals, along all positive averaging lengths. No mixing rate or standardness assumption on the probability space is required.

Proof

The bigger picture

Why it matters

Mixing describes dynamics in which correlations between observations fade as their time separation grows. The reported result extends this picture to averages of products sampled at consecutive multiples of time, for almost every starting point.

What changes?

The unreviewed manuscript claims this for every invertible mixing probability-preserving transformation: a reversible evolution leaving probabilities unchanged. Fix any finite list of at least two bounded measurable observables. Multiply their values at times k, twice k, and successive multiples of k, then average over k. These averages reportedly converge to the product of the observables' means for almost every starting point, through all positive integer averaging lengths. No mixing rate or standardness assumption on the probability space is required.

What does that help mathematicians do?

Choosing observables that record membership in measurable sets turns the claim into a frequency statement. For almost every starting point, the frequency of simultaneous visits to those sets at consecutive multiples of a sampled time tends to the product of their probabilities. This identifies behavior along individual trajectories, not merely an average over the whole space. The exceptional set of probability zero may depend on the chosen observables.

Are there practical applications?

Its immediate value is foundational: it links mixing to precise long-run statistics of multiple observations along a trajectory, at every fixed finite length. This supplies a proposed qualitative law for such statistics. It gives no convergence speed, so it does not establish how long a finite simulation must run to approximate the limit.

This section was generated by GPT-6 Astra Medium. This explanation is based on the result summary and manuscript abstracts below. This context is separate from OpenAI's source text.

4 manuscripts

Pointwise Multiple Ergodic Averages for Mixing Transformations

October 4, 2026 25 pages

Let T be an invertible mixing probability-preserving transformation. For every integer n ≥ 2 and every fixed tuple of bounded measurable functions, we prove that the consecutive multiple ergodic averages of length n converge almost everywhere to the product of the integrals, as the averaging length tends to infinity through all positive integers. The probability space need not be standard, and no rate of mixing is required.

PDF LaTeX source
Cite (BibTeX) 
@misc{OAI:Pointwise-Multiple-Ergodic-Averages-for-Mixing-Transformations-October-4-2026,
  author = {{OpenAI}},
  title = {{Pointwise Multiple Ergodic Averages for Mixing Transformations}},
  howpublished = {OpenAI Math Release preprint
                  \href{https://github.com/openai/math/blob/main/preprints/Pointwise-Multiple-Ergodic-Averages-for-Mixing-Transformations-October-4-2026/multiple-ergodic-averages.pdf}{OAI:Pointwise-Multiple-Ergodic-Averages-for-Mixing-Transformations-October-4-2026}},
  year = {2026}
}

Pointwise convergence of fourfold ergodic averages for mixing transformations

October 4, 2026 36 pages

We prove that fourfold ergodic averages along the times n,2n,3n,4nn,2n,3n,4n converge almost everywhere to the product of the integrals for every invertible, bimeasurable, mixing probability-preserving transformation and every fixed choice of bounded measurable inputs. Convergence holds along all positive integer averaging lengths. No quantitative mixing rate is assumed, and the probability space need not be standard.

PDF LaTeX source
Cite (BibTeX) 
@misc{OAI:Pointwise-convergence-of-fourfold-ergodic-averages-for-mixing-transformations-October-4-2026,
  author = {{OpenAI}},
  title = {{Pointwise convergence of fourfold ergodic averages for mixing transformations}},
  howpublished = {OpenAI Math Release preprint
                  \href{https://github.com/openai/math/blob/main/preprints/Pointwise-convergence-of-fourfold-ergodic-averages-for-mixing-transformations-October-4-2026/fourfold-ergodic-averages.pdf}{OAI:Pointwise-convergence-of-fourfold-ergodic-averages-for-mixing-transformations-October-4-2026}},
  year = {2026}
}

Triple ergodic averages with distinct integer slopes

October 4, 2026 43 pages

For every invertible mixing probability-preserving transformation, triple ergodic averages of bounded measurable functions along any three pairwise distinct nonzero integer slopes converge almost everywhere to the product of their integrals. The slopes may be positive or negative.

PDF LaTeX source
Cite (BibTeX) 
@misc{OAI:Triple-ergodic-averages-with-distinct-integer-slopes-October-4-2026,
  author = {{OpenAI}},
  title = {{Triple ergodic averages with distinct integer slopes}},
  howpublished = {OpenAI Math Release preprint
                  \href{https://github.com/openai/math/blob/main/preprints/Triple-ergodic-averages-with-distinct-integer-slopes-October-4-2026/triple-ergodic-distinct-slopes.pdf}{OAI:Triple-ergodic-averages-with-distinct-integer-slopes-October-4-2026}},
  year = {2026}
}

Pointwise convergence of triple ergodic averages for mixing transformations

October 4, 2026 61 pages

We prove pointwise convergence of triple ergodic averages for every invertible mixing probability preserving transformation T of an arbitrary probability space (X,F,μ)(X,\mathcal F,\mu), with measurable inverse. For every triple of bounded measurable functions f1,f2,f3:X→Cf_1,f_2,f_3:X\to\mathbb C,

1N∑n=1Nf1(Tnx)f2(T2nx)f3(T3nx)⟶∏j=13∫Xfj dμ\displaystyle \frac1N\sum_{n=1}^N f_1(T^nx)f_2(T^{2n}x)f_3(T^{3n}x) \longrightarrow \prod_{j=1}^3\int_X f_j\,d\mu

for μ-almost every x as N→∞N\to\infty through all positive integers.

PDF LaTeX source
Cite (BibTeX) 
@misc{OAI:Pointwise-convergence-of-triple-ergodic-averages-for-mixing-transformations-October-4-2026,
  author = {{OpenAI}},
  title = {{Pointwise convergence of triple ergodic averages for mixing transformations}},
  howpublished = {OpenAI Math Release preprint
                  \href{https://github.com/openai/math/blob/main/preprints/Pointwise-convergence-of-triple-ergodic-averages-for-mixing-transformations-October-4-2026/pointwise-triple-ergodic-averages-mixing-transformations.pdf}{OAI:Pointwise-convergence-of-triple-ergodic-averages-for-mixing-transformations-October-4-2026}},
  year = {2026}
}

Data from github.com/openai/math at commit adc7f12, committed October 6, 2026 at 21:58 UTC, last checked for changes about 8 hours ago. Titles, subjects, summaries, abstracts and Lean notes are OpenAI's; page counts are read from the PDFs. The map, related results, search, kinds of results and the named-problem index are Emergent Mind's, built with text embeddings and an LLM, and may contain errors.

An Emergent Mind Labs project. Emergent Mind is not affiliated with OpenAI. None of these results has been peer reviewed. Cite the manuscripts themselves, using the BibTeX on each result's page.