Pith. sign in

REVIEW 5 minor 25 references

Maximal pattern complexity and structure of null systems

T0 review · 0 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A compact metrizable system is null exactly when every finite open cover has polynomially bounded maximal pattern complexity.

desk verdict Strong resolution of the Huang–Ye polynomial pattern complexity problem, with explicit nonminimal null constructions; the proof is convincing and only minor expositional issues remain. read the letter →

arxiv 2608.06103 v1 pith:22XPSWAA submitted 2026-08-06 math.DS math.CO

Jie Li , Kangbo Ouyang This is my paper · ORCID
classification math.DSmath.CO MSC 37B4037B0568Q3222A05
keywords nulldynamicalsystemmaximalpatterncomplexityfat-shatteringdimensionempiricalcoveringnumberstronglyexoticgrouptopologicalsequenceentropyequicontinuitytwo-scattering
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

A compact metrizable dynamical system is null, meaning its topological sequence entropy vanishes along every sequence of times, exactly when, for every finite open cover, its maximal pattern complexity grows at most polynomially in the number of time points. The paper proves this equivalence, resolving a problem left open in earlier work that had only a quasipolynomial bound. It also shows equicontinuity is characterized by sublinear, in fact bounded, maximal pattern complexity, and that any non-equicontinuous system has at least linear growth for some cover. Finally, it constructs transitive nonminimal null systems, one uniformly rigid with two fixed points and one two-scattering, answering open questions about how nullness behaves outside the minimal setting.

What carries the argument

The central objects are the maximal pattern complexity p*_{X,U}(n), the largest join-cover complexity over n-element sets of times, and the orbit-distance function classes F_Q={x↦(d(T^k x,q))_{k,q}}. The argument's main engine is the equivalence between nullness and finiteness of fat-shattering dimension at every positive scale for every such class, together with an empirical covering theorem, proved in an appendix from a published fixed-scale estimate, that turns finite fat-shattering into polynomial empirical covering numbers. The constructions of transitive examples use scalar orbit closures X_h defined from a continuous observable h on a quotient of a Banach space, with uniform rigidity, fixed points, and two-scattering controlled respectively by positive-gap shattering bounds, dense torsion, and strong exoticity of the quotient group.

What would settle it

Find a compact metrizable null system and a finite open cover U such that p*_{X,U}(n) is not bounded by any C n^d; that would refute Theorem 1.1 directly. Conversely, a non-null system whose maximal pattern complexity is polynomial for every finite open cover would also refute it. For Theorem 1.2, a non-equicontinuous system whose maximal pattern complexity for every finite open cover is sublinear would be a falsifier.

Watch

Extended reading notes

Core claim

The paper's central discovery is that the combinatorial growth of orbit patterns over arbitrary finite sets of times is governed precisely by the dynamical class: nullness is equivalent to polynomial growth of maximal pattern complexity for every finite open cover, while equicontinuity is equivalent to bounded, or equivalently sublinear, growth, with a linear lower bound separating the non-equicontinuous case. The proof associates to every finite set of points Q the orbit-distance class of functions f_x(k,q)=d(T^k x,q), shows that its fat-shattering dimension is finite at every positive scale exactly when the system is null, and then converts polynomial empirical covering numbers of these classes into polynomial bounds on maximal pattern complexity. The two examples are built as orbit closures of scalar observables on quotients of Banach spaces, producing a uniformly rigid null system with two fixed points and, using a strongly exotic quotient, a two-scattering null system.

Load-bearing premise

The argument relies on two unproved external facts: the standard equivalence between equicontinuity and total boundedness of the orbit sup-metric, and the known construction of a strongly exotic quotient of ℓ⁴ whose only continuous positive-definite functions are constant; if either fact fails, the corresponding half of the paper's conclusions stops being supported.

Editorial extensions

If this is right

  • Every null system, not just zero-dimensional ones, has polynomially bounded maximal pattern complexity for every finite open cover; superpolynomial but subexponential pattern growth is impossible.
  • A system is equicontinuous exactly when its maximal pattern complexity for every finite open cover is bounded, and every non-equicontinuous system has a finite open cover with p*_{X,U}(n) ≥ n+1.
  • Transitive null systems can be nonminimal: there exists a uniformly rigid transitive null system with two fixed points that is neither uniquely ergodic nor mean equicontinuous.
  • There exists a transitive nonminimal null system that is two-scattering, so scattering can coexist with nullness outside the minimal setting.
  • For a transitive system, nullness is determined by scalar factors: for every continuous [0,1]-valued observable, the family of forward names has finite fat-shattering dimension at every scale and polynomial empirical covering numbers.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The fat-shattering and empirical-covering route suggests a direct connection between pattern complexity and uniform convergence estimates from statistical learning theory; the same machinery may yield quantitative bounds on other orbit-growth quantities, such as complexity along structured time sets, for null systems.
  • The scalar-orbit-closure construction suggests that the dynamical properties of a transitive null system are already encoded in its one-dimensional factors; a natural testable extension would be to ask whether every transitive null system is a factor of a null system generated by a single scalar observable.
  • One could test whether the two-scattering construction can be carried out with a Hilbert quotient; the paper's dichotomy between its ℓ²-based uniformly rigid example and its ℓ⁴-based two-scattering example suggests the strongly exotic quotient is essential for two-scattering.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

0 major / 5 minor

Summary. The paper studies maximal pattern complexity p*_{X,U}(n) for compact metrizable dynamical systems. Its main result, Theorem 1.1, characterizes nullness by polynomial maximal pattern growth for every finite open cover, resolving a question of Huang and Ye. Theorem 1.2 characterizes equicontinuity by bounded, sublinear, or arbitrarily small polynomial growth of p*_{X,U}, and establishes a linear lower bound for non-equicontinuous systems. The paper then constructs transitive nonminimal null systems (Theorem 1.4): one uniformly rigid with two fixed points and not mean equicontinuous, and another two-scattering. The proofs use fat-shattering dimension, empirical covering numbers, orbit-distance function classes, and scalar orbit closures built from quotients of ℓ2 and ℓ4.

Significance. If correct, this is a substantial contribution. It settles the polynomial-growth problem for maximal pattern complexity of general compact metrizable null systems and answers long-standing structural questions about transitive nonminimal null systems. The paper is notable for giving complete proofs, including a self-contained appendix deriving the polynomial empirical-covering theorem from the published estimate of Alon et al. The constructions are explicit and use independent known results (Kerr–Li on IN-pairs, Banaszczyk on strongly exotic groups) without fitting parameters to the conclusions. The result would be of interest to researchers in topological dynamics, local entropy theory, and learning-theoretic approaches to complexity.

minor comments (5)
  1. [Title and author block] The displayed title and author block contain typographical artifacts ('PA TTERN', 'OUY ANG'); these should be corrected in the final version.
  2. [Section 4, Proposition 4.2] The equivalence between equicontinuity and total boundedness of the sup-metric d_+ is quoted as a standard fact without proof or reference; since it is load-bearing for Theorem 1.2, a one-sentence proof or an explicit reference would make the argument easier to verify.
  3. [Appendix A] The proof uses the notion of 'P_{ρ/4}-dimension' from [4] without defining it; because the appendix aims to be self-contained from the cited estimate, a short definition of this dimension would help the reader.
  4. [Section 6.2, proof of Theorem 1.4(ii)] The standard basis vectors e_i of ℓ4(N;R) are used without being introduced; please define this notation when the quotient G4 is constructed.
  5. [Proposition 6.5] The construction of the sequence (a_n) should state explicitly that a_n = 0 for all indices not among the selected N_j, and the inductive proof of span_R{g_1,...,g_n} = span_R{e_1,...,e_n} would be clearer if the base case g_1 = e_1 is written out.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main implications are derived from independent combinatorial and dynamical theorems, not from self-citing the target claims.

full rationale

The main chain, Theorem 1.1, is not circular: nullness is connected to finite fat-shattering of orbit-distance classes in Proposition 3.1 using the Kerr–Li IN-pair characterization, and polynomial empirical covering is proved in Appendix A from the Alon–Ben-David–Cesa-Bianchi–Haussler fixed-scale estimate; neither ingredient defines p* in terms of itself. The reverse implication in Theorem 1.1 uses only the definition of topological sequence entropy. Theorem 1.2 relies on Proposition 4.2, whose only unproved input is the classical equivalence between equicontinuity and total boundedness of d_+; this is a standard theorem, not a restatement of the boundedness of p*. The constructions in Theorem 1.4 use Banaszczyk's external strongly-exotic quotient theorem and the paper's own positive-gap shattering estimates; no fitted parameter is relabelled as a prediction. The only citations involving author J. Li ([20], [21]) are used for the standard fact that transitive mean equicontinuous systems are uniquely ergodic and for background, respectively; [20] is a published parameter-free theorem whose assumptions do not include the present constructions, so under the review rules it is independent support rather than circularity. No uniqueness claim is imported from the authors' prior work, and no ansatz is smuggled in via citation. The derivation chain is therefore self-contained at the level of the paper's main assertions.

Assumptions & free parameters 0 free parameters · 9 assumptions · 0 invented entities

No free parameters are fitted: all constants arise from existence assertions in the theorems or from fixed choices like R=1/10 in the construction. No new physical or mathematical entities are postulated; the scalar orbit closures X_h are derived objects, not additional assumptions. The axioms are standard theorems from dynamics, learning theory, and group theory, with the Banaszczyk theorem and the d_+ total-boundedness fact the least self-contained.

assumptions (9)
  • standard math Kerr-Li theorem: a system is null iff it has no non-diagonal IN-pair (Theorem 2.2).
    Used to connect nullness to finite fat-shattering in Proposition 3.1 and again in Section 7.
  • standard math Alon-Ben-David-Cesa-Bianchi-Haussler fixed-scale covering estimate ([4, Lemma 3.5]).
    Basis of Theorem 2.4, which produces the polynomial empirical covers used in Theorem 1.1; the appendix uses it but does not reprove it.
  • domain assumption Equicontinuity of (X,T) is equivalent to total boundedness of d_+(x,y)=sup_{k≥0} d(T^kx,T^ky).
    Invoked as a standard fact in Proposition 4.2 without proof; it underpins the linear lower bound in Theorem 1.2.
  • standard math Banaszczyk's theorem: the constructed quotient ℓ4/Γ is strongly exotic, so every continuous positive-definite function on it is constant ([5 Theorem 6], [6 Theorem (5.1)(d)]).
    Needed for Proposition 5.2 and Theorem 1.4(ii) to make the final contradiction in the two-scattering proof.
  • standard math Every transitive mean equicontinuous system is uniquely ergodic ([20, Corollary 3.4]).
    Used to conclude that the uniformly rigid null system with two fixed points is not mean equicontinuous.
  • standard math de Bruijn-Erdős compactness theorem for graph colouring.
    Used in Proposition 5.2 to extend finite M-colourability of the distance graph to a global colouring.
  • standard math Discrete abelian groups are amenable, so ℓ∞(Gd) admits a translation-invariant mean.
    Used to build the continuous positive-definite function φ in the proof of Proposition 5.2.
  • standard math Finite minimax theorem of von Neumann.
    Used in Proposition A.5 to pass from a worst-case distribution over data points to a mixture over hypotheses.
  • standard math Infinite Ramsey theorem for finite colourings of pairs.
    Used in Proposition 4.2 to find one reference point q0 that works for infinitely many pairs.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Maximal pattern complexity and structure of null systems." pith.science (2026). https://pith.science/paper/22XPSWAA

@misc{pith2026260806103,
  author       = {Pith},
  title        = {Pith review of: Maximal pattern complexity and structure of null systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/22XPSWAA}},
  note         = {Machine review of arXiv:2608.06103}
}
read the original abstract

A compact metrizable system is null if its topological sequence entropy vanishes along every sequence of times. We prove that nullness is equivalent to polynomial maximal pattern complexity for every finite open cover, while equicontinuity is equivalent to sublinear maximal pattern complexity. The first characterization is obtained from finite fat-shattering at every positive scale and polynomial empirical covering of orbit-distance classes. We also construct transitive nonminimal null systems with properties excluded in the minimal setting: one is uniformly rigid and has two fixed points, and another is two-scattering. These results settle several long-standing open problems from the literature on polynomial maximal pattern growth and on the structure of transitive nonminimal null systems.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 25 canonical work pages

  1. [1]

    Hoeffding,Probability inequalities for sums of bounded random variables, J

    W. Hoeffding,Probability inequalities for sums of bounded random variables, J. Amer. Statist. Assoc. 58(1963), no. 301, 13–30

    work page 1963

  2. [2]

    von Neumann,Zur Theorie der Gesellschaftsspiele, Math

    J. von Neumann,Zur Theorie der Gesellschaftsspiele, Math. Ann.100(1928), no. 1, 295–320

    work page 1928

  3. [3]

    Scale-Sensitive Shattering: Learnability and Evaluability at Optimal Scale

    S. Aiyer, Y . Mansour, S. Moran, H. Shao and T. Waknine,Scale-sensitive shattering: learnability and evaluability at optimal scale, arXiv:2605.13684v1, 2026

    work page Pith review arXiv 2026

  4. [4]

    N. Alon, S. Ben-David, N. Cesa-Bianchi and D. Haussler,Scale-sensitive dimensions, uniform convergence, and learnability, J. ACM44(1997), no. 4, 615–631

    work page 1997

  5. [5]

    Banaszczyk,On the existence of commutative Banach–Lie groups which do not admit continuous unitary representations, Colloq

    W. Banaszczyk,On the existence of commutative Banach–Lie groups which do not admit continuous unitary representations, Colloq. Math.52(1987), no. 1, 113–118

    work page 1987

  6. [6]

    Banaszczyk,Additive subgroups of topological vector spaces, Lecture Notes in Mathematics, vol

    W. Banaszczyk,Additive subgroups of topological vector spaces, Lecture Notes in Mathematics, vol. 1466, Springer-Verlag, Berlin, 1991

    work page 1991

  7. [7]

    Blanchard, B

    F. Blanchard, B. Host and A. Maass,Topological complexity, Ergodic Theory Dynam. Systems20 (2000), no. 3, 641–662

    work page 2000

  8. [8]

    N. G. de Bruijn and P. Erd˝os,A colour problem for infinite graphs and a problem in the theory of relations, Nederl. Akad. Wetensch. Proc. Ser. A54= Indag. Math.13(1951), 371–373

    work page 1951

Show all 25 references
  1. [9]

    G. Gao, J. Ma, M. Rong and T. Tran,Variants of VC dimension and their applications to dynamics, Pure Appl. Math. Q.21(2025), no. 6, 2425–2449

    2025

  2. [10]

    Glasner and X

    E. Glasner and X. Ye,Local entropy theory, Ergodic Theory Dynam. Systems29(2009), no. 2, 321–356. 34 J. LI AND K. OUY ANG

    2009

  3. [11]

    T. N. T. Goodman,Topological sequence entropy, Proc. London Math. Soc. (3)29(1974), no. 2, 331–350

    1974

  4. [12]

    Huang, S

    W. Huang, S. M. Li, S. Shao and X. Ye,Null systems and sequence entropy pairs, Ergodic Theory Dynam. Systems23(2003), no. 5, 1505–1523

    2003

  5. [13]

    Huang and X

    W. Huang and X. Ye,Combinatorial lemmas and applications to dynamics, Adv. Math.220(2009), no. 6, 1689–1716

    2009

  6. [14]

    Huang and X

    W. Huang and X. Ye,Topological complexity, return times and weak disjointness, Ergodic Theory Dynam. Systems24(2004), no. 3, 825–846

    2004

  7. [15]

    Kamae and L

    T. Kamae and L. Q. Zamboni,Sequence entropy and the maximal pattern complexity of infinite words, Ergodic Theory Dynam. Systems22(2002), no. 4, 1191–1199

    2002

  8. [16]

    Kamae and L

    T. Kamae and L. Q. Zamboni,Maximal pattern complexity for discrete systems, Ergodic Theory Dynam. Systems22(2002), no. 4, 1201–1214

    2002

  9. [17]

    Kerr and H

    D. Kerr and H. Li,Independence in topological and C∗-dynamics, Math. Ann.338(2007), no. 4, 869–926

    2007

  10. [18]

    A. G. Kushnirenko,On metric invariants of entropy type, Russian Math. Surveys22(1967), no. 5, 53–61

    1967

  11. [19]

    A. N. Le, R. Pavlov and C. Schlortt,On subshifts with low maximal pattern complexity, Trans. Amer. Math. Soc., to appear; arXiv:2508.13420v1, 2025

    2025 arXiv

  12. [20]

    J. Li, S. Tu and X. Ye,Mean equicontinuity and mean sensitivity, Ergodic Theory Dynam. Systems 35(2015), no. 8, 2587–2612

    2015

  13. [21]

    J. Li, X. Ye and T. Yu,Mean equicontinuity, complexity and applications, Discrete Contin. Dyn. Syst.41(2021), no. 1, 359–393

    2021

  14. [22]

    A. L. T. Paterson,Amenability, Mathematical Surveys and Monographs, vol. 29, American Mathe- matical Society, Providence, RI, 1988

    1988

  15. [23]

    Qiu and J

    J. Qiu and J. Zhao,Null systems in the non-minimal case, Ergodic Theory Dynam. Systems40 (2020), no. 12, 3420–3437

    2020

  16. [24]

    Schlortt,On the structure of sequences with minimal maximal pattern complexity, Ergodic Theory Dynam

    C. Schlortt,On the structure of sequences with minimal maximal pattern complexity, Ergodic Theory Dynam. Systems46(2026), no. 8, 2131–2148

    2026

  17. [25]

    Walters,An introduction to ergodic theory, Graduate Texts in Mathematics, vol

    P. Walters,An introduction to ergodic theory, Graduate Texts in Mathematics, vol. 79, Springer- Verlag, New York–Berlin, 1982. (Jie Li) SCHOOL OFMATHEMATICS ANDSTATISTICS, JIANGSUNORMALUNIVERSITY, XUZHOU, JIANGSU, 221116, P.R. CHINA Email address:jiel0516@mail.ustc.edu.cn (Kan...

    1982

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.