2211.03448
STABLE CLT FOR DETERMINISTIC SYSTEMS
Zemer Kosloff, Dalibor Volný
correctmedium confidence
- Category
- math.DS
- Journal tier
- Strong Field
- Processed
- Sep 28, 2025, 12:56 AM
- arXiv Links
- Abstract ↗PDF ↗
Audit review
The paper proves that for every ergodic aperiodic probability-preserving system and every α in (0,2), there exists an observable f with n^{-1/α} S_n(f) converging in distribution to a symmetric α-stable law with any prescribed dispersion σ; see Theorem 1 and its proof, including the final rescaling to arbitrary σ . The candidate solution correctly invokes exactly this result and then (redundantly but harmlessly) rescales to match a target σ, and notes that convergence in distribution implies cdf convergence at all t because the limit law is non-atomic, as also noted in the paper’s preliminaries . The model’s write-up does not reproduce the paper’s constructive proof (triangular array, discretization, and embedding via the Alpern-lemma–independent partition), but it is a valid solution by citation.
Referee report (LaTeX)
\textbf{Recommendation:} minor revisions \textbf{Journal Tier:} strong field \textbf{Justification:} This work resolves a compelling existence problem for stable limit theorems in deterministic dynamical systems, achieving the iid-type normalization n\^{1/α} across the full α∈(0,2) range. The construction is careful and technically sound, leveraging a novel combination of truncated stable arrays and an embedding independent of a partition. While the exposition is strong, a few clarifications (e.g., emphasizing why no slowly varying factor appears) would further help readers. Overall, I recommend minor revisions.