On Bowen's entropy inequality and almost specification for flows

Diego Sanhueza; Maria Jose Pacifico

REVIEW 4 major objections 6 minor 19 references

On Bowen's entropy inequality and almost specification for flows

T0 review · 4 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read For a continuous flow on a compact metric space, the noncompact-set topological entropy of the set of generic points of any invariant measure is bounded above by that measure's metric entropy, with equality for ergodic measures.

desk verdict A useful flow-adaptation of Bowen's entropy inequality, but Theorem B's proof has a load-bearing unproved continuous-selection step. read the letter →

arxiv 1908.08072 v2 pith:IB7FXPZU submitted 2019-08-21 math.DS

Maria Jose Pacifico , Diego Sanhueza This is my paper · ORCID

classification math.DS MSC 37B4037A3537C4537D40

keywords Bowentopologicalentropycontinuousflowsgenericpointsmetricalmostspecificationpropertysaturatedsystemsirregulartime-onemap

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

This paper proves a flow version of a classical inequality from discrete ergodic theory: for every invariant probability measure of a continuous flow on a compact metric space, the Bowen topological entropy of the set of points whose flow averages converge to that measure is no larger than the metric entropy of the measure. When the measure is ergodic, the two entropies coincide. The proof rests on a rescaling theorem showing that the Bowen entropy of any set under a continuous flow is exactly one over |t| times the Bowen entropy of the time-t map on that same set, together with a comparison between flow-generic and time-one-generic points. The authors then use the inequality to show that flows inheriting the almost specification property from their time-one maps are saturated, that geodesic flows on negatively curved closed manifolds are saturated, and that nonempty irregular sets of such flows carry full flow entropy.

What carries the argument

The load-bearing mechanism is the time-one comparison. Theorem A rescales Bowen topological entropy across flow time, $h(\Phi,Y)=\frac1{|t|}h(\varphi^t,Y)$, proved by comparing spanning covers at the scale of a Lebesgue number. Theorem 2.4 then identifies the set of quasi-regular points of the flow with that of the time-one map; its proof constructs, for each point, the $\Phi$-invariant measure $\bar\mu_x=\$int_0^{1}$(\varphi^s)_*\mu_x\,ds$ from the time-one empirical measure $\mu_x$, and uses Riesz representation to extract invariant measures from subsequential Birkhoff limits. Lemma 3.3, proved with Jensen's inequality on the concave function $-x\log x$, gives $h_\mu(\$varphi^{1}$)\ge h_{\bar\mu}(\$varphi^{1}$)$, which converts the discrete bound $h(\$varphi^{1}$,G_{\mu_x}(\$varphi^{1}$))\le h_{\mu_x}(\$varphi^{1}$)$ into the flow bound. For the applications, the named object is the almost specification property for flows, defined through a mistake function $g(t,\epsilon)$ with $g(t,\epsilon)/t\to0$; the paper proves that a flow has this property exactly when its time-one map does, so discrete consequences transfer.

What would settle it

Construct a continuous flow, an observable $\varphi$, and a point $x$ for which the set of $t\in[0,1]$ with $C_1(\varphi\circ\varphi^t)=C_2(\varphi\circ\varphi^t)$ has no continuous branch $t(\varphi)$; if such a construction also makes $G_\mu(\Phi)\not\subseteq QR(h_\mu(\Phi))$ for some invariant measure $\mu$, then Theorem B's reduction to the discrete inequality fails. A direct computation of $h(\Phi,G_\mu(\Phi))>h_\mu(\Phi)$ for any continuous flow would refute the theorem itself.

Watch

Extended reading notes

Core claim

The central claim is Theorem B: for a continuous flow $\Phi=\{(\varphi^t)\}_{t\in\mathbb R}$ on a compact metric space $X$, every $\Phi$-invariant Borel probability measure $\mu$ satisfies $h(\Phi, G_\mu(\Phi))\le h_\mu(\Phi)$, where $h(\Phi, G_\mu(\Phi))$ is the Bowen topological entropy of the set of $\mu$-generic points, and equality holds whenever $\mu$ is ergodic. The companion Theorem A is equally structural: for any subset $Y\subseteq X$, the flow entropy is exactly $h(\Phi,Y)=\frac1{|t|}h(\varphi^t,Y)$, so flow entropy and time-$t$ entropy carry the same information up to a constant rescaling. The proof strategy is to show that every flow-generic point is quasi-regular for the time-one map, that the induced time-one empirical measures have metric entropy no larger than the flow's metric entropy of $\mu$, and then to invoke the discrete version of the inequality.

Load-bearing premise

The proof that flow quasi-regular points coincide with time-one quasi-regular points assumes one can choose, continuously in the observable, a time at which two limit averages agree; no proof of that continuous selection is supplied, and the fixed-point step in Theorem 2.4 collapses if it fails.

Editorial extensions

If this is right

Any continuous flow whose time-one map is saturated is itself saturated, so saturation is inherited by continuous time from discrete time.
Geodesic flows on closed manifolds of negative curvature are saturated: for every invariant measure, the set of generic points carries exactly the measure's entropy.
A continuous flow with the almost specification property is saturated, and if an irregular set for a continuous observable is nonempty, it carries the full topological entropy of the flow.
The equality $h(\Phi,Y)=\frac1{|t|}h(\varphi^t,Y)$ means entropy results for noncompact sets can be translated between a flow and any of its time-$t$ maps with only a constant rescaling.
For ergodic measures, the flow generic set carries full metric entropy even when the measure is not ergodic for the time-one map.

Reading between the lines

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

If the continuous-selection gap in the proof of Theorem 2.4 is repaired, the same comparison scheme would likely prove Bowen-type inequalities for observables with values in Banach spaces, since only the Riesz-representation step would need adjustment.
Theorem A suggests that any dynamically defined quantity that is monotone under taking subsets and additive over covers can be rescaled across flow times, which could simplify numerical entropy estimation: a single time-1 calculation determines entropy at every sampling rate.
The almost-specification transfer indicates that suspension flows over maps with specification-like properties are natural test beds for saturatedness, since Theorem C reduces the check for such flows to a discrete check on the base map.
A full converse of Theorem C is unlikely in general: a flow whose time-one map is non-saturated but whose continuous averages repair the missing generic points would require a delicate balance between the discrete and continuous orbit structures.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Short version: this is a useful flow-adaptation paper with a real hole in the middle. The main inequality (Theorem B) is probably true, and the authors deserve credit for seeing how to extend the fixed-point-free result of Wang–Chen–Lin–Wu to arbitrary continuous flows and for stating the corollaries on saturated systems and irregular points. But as written, the proof does not support the claim.

What's genuinely good: Theorem A is a clean and useful result: the Bowen entropy of a flow on a subset equals the Bowen entropy of the time-one map scaled by 1/|t|. It makes the flow entropy for noncompact sets tractable and matches Shen–Zhao's earlier definition. The application program is sensible, and the paper is honest about what it imports from [14] and [19]; there were no invented quantities or fitted constants.

The soft spot is Theorem 2.4. The direction Q(Φ)⊆Q(φ_1) needs the claim that from the integral equality ∫_0^1 C1(φ∘φ^t)dt = ∫_0^1 C2(φ∘φ^t)dt one can choose t(φ) with C1(φ∘φ^t)=C2(φ∘φ^t) and, crucially, that t(φ) varies continuously with φ. That is not proved, and as a general statement about continuous functions on a compact space it is false; the rotation-flow example with δ_0 and δ_{1/2} shows the selection can fail to be continuous. Since Section 4 uses Theorem 2.4 to place G_μ(Φ) inside QR(h_μ(Φ)), this is not a cosmetic gap. The theorem may be standard folklore, but the proof given doesn't establish it.

There's a second, smaller problem: Corollaries 5.4 and 5.8 rely on the equivalence 'Φ has the almost specification property iff φ_1 does.' That is asserted in one sentence with no proof, and the definitions (Lebesgue-measurable mistakes for flows vs. discrete gaps for maps) are different enough that the equivalence needs a real argument. If the authors fix Theorem 2.4 and add that argument, the paper will be solid.

Who this is for: researchers working on entropy and multifractal analysis of continuous-time systems. It deserves a serious referee—the flaws are repairable and the results are likely publishable. If I were the editor, I'd send it out and ask the referee to focus on Theorem 2.4.

Referee Report

4 major / 6 minor

Summary. The paper defines a Bowen-type topological entropy for arbitrary subsets of a compact metric space under a continuous flow and proves an Abramov-type scaling identity: h(Φ,Y)=|t|^{-1}h(φ_t,Y) for every Y⊆X and t≠0 (Theorem A). It then claims a flow analogue of Bowen's inequality: for every Φ-invariant Borel probability measure μ, h(Φ,G_μ(Φ))≤h_μ(Φ), with equality when μ is ergodic (Theorem B). The proof proceeds by comparing quasi-regular points of the flow with those of the time-one map (Theorem 2.4), then applying Bowen's discrete inequality to a set QR(h_μ(Φ)). The paper also derives consequences: if the time-one map is saturated, then the flow is saturated (Theorem C); flows with almost specification are saturated (Corollary 5.4); and under almost specification the irregular set carries full topological entropy (Theorem 5.8).

Significance. If the results are correct, this is a useful contribution: it provides a workable Bowen entropy for flows, a clean scaling identity, and a flow-level Bowen inequality that transfers saturation and irregular-set results from discrete dynamics to continuous flows, including geodesic flows in negative curvature. The overall strategy is attractive and not circular: it reduces the flow statement to Bowen's discrete theorem and Walters' isomorphism lemma. However, as the manuscript stands, the key bridge Theorem 2.4 rests on an unproved continuous-selection step, and several corollaries depend on unproved equivalences or nonemptiness assertions. The main theorem is plausible and likely repairable, but the written proof does not fully support all claims.

major comments (4)

[Section 2, proof of Theorem 2.4 (around equation (3))] The proof that Q(Φ)⊆Q(φ_1) assumes that from the identity ∫_0^1 C_1(ϕ∘φ^t)dt = ∫_0^1 C_2(ϕ∘φ^t)dt for every continuous ϕ and the continuity of t↦C_i(ϕ∘φ^t), one can choose t(ϕ) such that C_1(ϕ∘φ^{t(ϕ)})=C_2(ϕ∘φ^{t(ϕ)}) and such that the map t(ϕ) is continuous in ϕ, after which a fixed-point argument gives C_1=C_2. No proof of the existence of such a continuous selection is supplied, and the displayed integral equality does not imply it. For example, for the rotation flow R_{t/2} on S^1, the two distinct φ_1-invariant measures C_1=(δ_0+δ_{1/2})/2 and C_2=(δ_{1/4}+δ_{3/4})/2 satisfy ∫_0^1 C_1(ϕ∘φ^t)dt=∫_0^1 C_2(ϕ∘φ^t)dt for every continuous ϕ, yet C_1≠C_2. Since this step is the only bridge from flow quasi-regularity to time-one quasi-regularity, Theorem 2.4 is not proved as written, and the subsequent inclusion G_μ(Φ)⊆QR(h_μ(Φ)) in Section 4 lacks support.
[Section 4, first paragraph] The displayed line 'By (4) and Teorema A, we obtain h(Φ,QR(h_μ(Φ))) = h(φ_1,QR(h_μ(φ_1))) ≤ h_μ(φ_1) = h_μ(Φ)' uses a QR-entropy bound that is not literally Bowen's inequality (4). Inequality (4) is for the set of generic points of a single measure, whereas QR(h_μ(Φ)) is a union over the uncountable family of measures ν with h_ν(φ_1)≤h_μ(Φ). The standard estimate h(f,QR(c))≤c is true but requires a separate argument (or a precise citation); as written, the upper bound in Theorem B relies on an unproved extension of (4).
[Section 5, Corollary 5.4] The proof asserts 'Φ has the almost specification property if and only if φ_1 has this property' without proof or reference. This equivalence is not immediate: the flow definition allows omitting a set of times of small Lebesgue measure inside each interval, while the discrete definition for φ_1 only controls omission of integer times. The equivalence is load-bearing because it transfers the Mesón–Vericat discrete saturation theorem to flows, and it is also used in Theorem 5.8.
[Section 5, Theorem 5.8] The opening assertion 'Observe that I_ϕ(φ_t) is not empty, for some t≠0' is nontrivial and unproved. Non-convergence of the flow averages of ϕ along a point does not automatically imply non-convergence of the Birkhoff sums of φ_t for some fixed t. Moreover, the inclusion I_ϕ(φ_t)⊆I_ϕ(Φ) used in the display comes from Corollary 5.7, whose proof depends on the unproved equation (3) in Theorem 2.4. Thus Theorem 5.8 is not established as written.

minor comments (6)

[Section 2, proof of Theorem 2.4] The maps C_i are defined as 'C_i: C(X) → X'; they should be 'C_i: C(X) → R'. Also, the formula 'C_1(1_X)=C_2(1_X)=1_X' should use the real number 1, not the constant function 1_X.
[Section 3, proof of Theorem A] The quantities n(B_i) and n(φ_τ,B_i) are used before they are defined; the discrete Bowen-entropy analogues of N(B) should be defined explicitly.
[Section 4, first paragraph] In the displayed equation, the set 'QR(h_μ(φ_1))' on the right-hand side should be 'QR(h_μ(Φ))' (or the equality h_μ(φ_1)=h_μ(Φ) should be stated immediately before the display).
[Section 5, Definition 5.3] The phrasing 'there are T_g(ε_1),...,T_g(ε_k)>0' should clarify that T_g(ε_i) is associated to each ε_i, and the concatenation times in (6) should be written more carefully (for instance, specifying that the inequality holds for each j=1,...,k).
[Throughout] There are numerous typographical errors, including 'Corolary', 'Lema', 'Teorema', and 'Pﬁste r'; the manuscript should be proofread. The sentence in Section 2 that the results are 'probably folklore' should be replaced by a precise statement about novelty or by references.
[Introduction and references] Reference [19] is a closely related preprint; the paper should discuss explicitly how Theorem B compares with the results in [19], particularly since the abstract and introduction advertise a generalization of that work.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the flow entropy inequality is derived by reduction to Bowen's discrete theorem, not by assuming its own conclusion.

full rationale

The paper's central result, Theorem B, is an external reduction: it defines flow entropy for noncompact sets from a Bowen-type cover construction, proves Theorem A relating this entropy to the time-one map entropy, and then applies Bowen's discrete inequality h(f,G_mu(f)) <= h_mu(f) together with the inclusion G_mu(Phi) subseteq QR(h_mu(Phi)). No fitted parameter is introduced, no normalization is chosen to force the conclusion, and no target quantity is used as an input. The equality h(Phi,Y)=1/|t| h(phi_t,Y) in Theorem A is not definitional: the flow entropy is defined independently via covers and escape times, and the equality is proved from that definition. The proof of Theorem B invokes external results (Bowen's Theorem 2, Walters' isomorphism lemma) rather than citations by the present authors. The closest issue is in the proof of Theorem 2.4, where the claim that one can choose t(phi) continuously is asserted without proof; that is a potential correctness gap, not a circular reduction, since Theorem 2.4 is not assumed as an input and its conclusion is not built into any definition. Accordingly, the circularity score is 0.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No numerical fitting or invented entities. The paper imports Bowen's discrete entropy theorem and variational principle; the only new assumptions are the unproved equivalence between almost specification for flows and for the time-one map, and an unproved implication from flow-irregular to time-t irregular. The continuous-selection step in Theorem 2.4 functions as an unstated axiom in the proof.

assumptions (5)

standard math Bowen's discrete entropy inequality and the QR-set bound from [3, Theorems 1 and 2]
Used as black boxes in Section 4 to pass from flow-generic points to QR(h_μ).
standard math For each t, (X, B, μ, φ^1) and (X, B, φ^t_*μ, φ^1) are isomorphic, so their entropies are equal (Lemma 3.2)
Taken from Walters [18, Theorem 4.11] and used to prove Lemma 3.3.
ad hoc to paper A continuous selection t(φ) with the required continuity exists in the proof of Theorem 2.4
Asserted without proof in Section 2; the fixed-point argument needs it.
ad hoc to paper A flow has the almost specification property if and only if its time-one map does
Stated without proof in Corollary 5.4; neither direction is derived.
ad hoc to paper If I_ϕ(Φ) is nonempty, then I_ϕ(φ_t) is nonempty for some t ≠ 0
Asserted at the start of Theorem 5.8 with no argument.

how reviews work

0 comments

Cite this review

Pith. "Pith review of On Bowen's entropy inequality and almost specification for flows." pith.science (2026). https://pith.science/paper/IB7FXPZU

@misc{pith2026190808072,
  author       = {Pith},
  title        = {Pith review of: On Bowen's entropy inequality and almost specification for flows},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IB7FXPZU}},
  note         = {Machine review of arXiv:1908.08072}
}

read the original abstract

We study the Bowen topological entropy of generic and irregular points for certain dynamical systems. We define the topological entropy of noncompact sets for flows, analogous to Bowen's definition. We show that this entropy coincides with the Bowen topological entropy of the time-1 map on any set. We also show a Bowen's inequality for flows; namely, that the metric entropy with respect to every invariant measure for a continuous flow is an upper bound for the topological entropy of the set of generic points with respect to the same measure, and the equality is always true if the measure is ergodic. We propose a definition of almost specification property for flows and prove that a continuous flow has the almost specification property if the time-1 map satisfies this property. Using Bowen's inequality for flows, we show that every continuous flow with the almost specification property is saturated, extending a result of Meson and Vericat in [22]. Under the same hypotheses, we extend a result of Thompson on the entropy of irregular points in [30].

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 19 canonical work pages

[14]

and Zhao, Y., Entropy of a ﬂow on non-compact sets , Open Syst

Shen, J. and Zhao, Y., Entropy of a ﬂow on non-compact sets , Open Syst. Inf. Dyn. 19 (2012), no. 2, 1250015, 10. MR 2930067

work page 2012
[19]

Bowen entropy of generic point for fixed-point free flows

Wang, Y., Chen, E., Lin, Z. and Wu, T., Bowen topological of generic point for ﬁxed-point free ﬂows , arXiv:1901.02135, 2019. M. J. Paciﬁco and D. Sanhueza Instituto de Matemática, Universidade Federal do Rio de Jan eiro, C. P. 68.530 CEP 21.945-970, Rio de Janeiro, RJ, Brazil. E-mail: paciﬁco@im.ufrj.br, sanhueza.diego.a@gmail.co m

work page Pith review arXiv 1901
[1]

272, Birkhäuser Verlag, Basel, 2008

Barreira, L., Dimension and recurrence in hyperbolic dynamics , Progress in Mathematics, vol. 272, Birkhäuser Verlag, Basel, 2008. MR 2434246

work page 2008
[2]

non-typical

Barreira, L. and Schmeling, J., Sets of “non-typical” points have full topological entropy and full Hausdorﬀ dimension , Israel J. Math. 116 (2000), 29–70. MR 1759398

work page 2000
[3]

Bowen, R., Topological entropy for noncompact sets , Trans. Amer. Math. Soc. 184 (1973), 125–136. MR 0338317 (49 #3082)

work page 1973
[4]

Colebrook, C., The Hausdorﬀ dimension of certain sets of nonnormal numbers , Michigan Math. J. 17 (1970), 103–116. MR 0260697

work page 1970
[5]

Eggleston, H., The fractional dimension of a set deﬁned by decimal properti es, Quart. J. Math., Oxford Ser. 20 (1949), 31–36. MR 0031026

work page 1949
[6]

and Lin, S., Topological entropy for divergence points, Ergodic Theory Dynam

Ercai, C., Küpper, T. and Lin, S., Topological entropy for divergence points, Ergodic Theory Dynam. Systems 25 (2005), no. 4, 1173–1208. MR 2158401

work page 2005

Show all 19 references

[7]

and Peyrière, J., Generic points in systems of speciﬁcation and Banach valued Birkhoﬀ ergodic average, Discrete Contin

Fan, A.-H., Liao, L.-M. and Peyrière, J., Generic points in systems of speciﬁcation and Banach valued Birkhoﬀ ergodic average, Discrete Contin. Dyn. Syst. 21 (2008), no. 4, 1103–1128. MR 2399452

2008
[8]

Systems Theory 1 (1967), 1–49

Furstenberg, H., Disjointness in ergodic theory, minimal sets, and a problem in Diophantine approximation , Math. Systems Theory 1 (1967), 1–49. MR 0213508

1967
[9]

and Hasselblatt, B., Introduction to the modern theory of dynamical systems , Encyclopedia of Mathematics and its Applications, vol

Katok, A. and Hasselblatt, B., Introduction to the modern theory of dynamical systems , Encyclopedia of Mathematics and its Applications, vol. 54, Cambridge University Press, Cambridge, 1995, With a supplementary chapter by Katok and Leonardo Mendoza. MR 1326374 14 MARIA JOSÉ ...

1995
[10]

and Oprocha, P., A panorama of speciﬁcation-like properties and their conse quences, In Dy- namics and numbers, Contemp

Kwietniak, D., Ł¸ acka, M. and Oprocha, P., A panorama of speciﬁcation-like properties and their conse quences, In Dy- namics and numbers, Contemp. Math., vol. 669, Amer. Math. So c., Providence, RI, 2016, pp. 155–186. MR 3546668

2016
[11]

and Vericat, F., Saturatedness of dynamical systems under the almost speciﬁ cation property, J

Mesón, A. and Vericat, F., Saturatedness of dynamical systems under the almost speciﬁ cation property, J. Dyn. Syst. Geom. Theor. 14 (2016), no. 1, 1–15. MR 3532420

2016
[12]

and Pitskel’, B., Topological pressure and the variational principle for non compact sets, Functional Anal

Pesin, Ya. and Pitskel’, B., Topological pressure and the variational principle for non compact sets, Functional Anal. Appl. 18 (1984), no. 4, 307–318. MR 775933

1984
[13]

and Sullivan, W., On the topological entropy of saturated sets , Ergodic Theory Dynam

Pﬁster, C.-E. and Sullivan, W., On the topological entropy of saturated sets , Ergodic Theory Dynam. Systems 27 (2007), no. 3, 929–956. MR 2322186

2007
[15]

Thompson, D., The irregular set for maps with the speciﬁcation property ha s full topological pressure , Dyn. Syst. 25 (2010), no. 1, 25–51. MR 2765447

2010
[16]

, Irregular sets, the β -transformation and the almost speciﬁcation property , Trans. Amer. Math. Soc. 364 (2012), no. 10, 5395–5414. MR 2931333

2012
[17]

Walters, P., A variational principle for the pressure of continuous tran sformations, Amer. J. Math. 97 (1975), no. 4, 937–971. MR 0390180

1975
[18]

79, Springer-Verlag, New York, 1982

, An introduction to ergodic theory , Graduate Texts in Mathematics, vol. 79, Springer-Verlag, New York, 1982. MR 648108

1982

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [14]

and Zhao, Y., Entropy of a ﬂow on non-compact sets , Open Syst

Shen, J. and Zhao, Y., Entropy of a ﬂow on non-compact sets , Open Syst. Inf. Dyn. 19 (2012), no. 2, 1250015, 10. MR 2930067

work page 2012

[2] [19]

Bowen entropy of generic point for fixed-point free flows

Wang, Y., Chen, E., Lin, Z. and Wu, T., Bowen topological of generic point for ﬁxed-point free ﬂows , arXiv:1901.02135, 2019. M. J. Paciﬁco and D. Sanhueza Instituto de Matemática, Universidade Federal do Rio de Jan eiro, C. P. 68.530 CEP 21.945-970, Rio de Janeiro, RJ, Brazil. E-mail: paciﬁco@im.ufrj.br, sanhueza.diego.a@gmail.co m

work page Pith review arXiv 1901

[3] [1]

272, Birkhäuser Verlag, Basel, 2008

Barreira, L., Dimension and recurrence in hyperbolic dynamics , Progress in Mathematics, vol. 272, Birkhäuser Verlag, Basel, 2008. MR 2434246

work page 2008

[4] [2]

non-typical

Barreira, L. and Schmeling, J., Sets of “non-typical” points have full topological entropy and full Hausdorﬀ dimension , Israel J. Math. 116 (2000), 29–70. MR 1759398

work page 2000

[5] [3]

Bowen, R., Topological entropy for noncompact sets , Trans. Amer. Math. Soc. 184 (1973), 125–136. MR 0338317 (49 #3082)

work page 1973

[6] [4]

Colebrook, C., The Hausdorﬀ dimension of certain sets of nonnormal numbers , Michigan Math. J. 17 (1970), 103–116. MR 0260697

work page 1970

[7] [5]

Eggleston, H., The fractional dimension of a set deﬁned by decimal properti es, Quart. J. Math., Oxford Ser. 20 (1949), 31–36. MR 0031026

work page 1949

[8] [6]

and Lin, S., Topological entropy for divergence points, Ergodic Theory Dynam

Ercai, C., Küpper, T. and Lin, S., Topological entropy for divergence points, Ergodic Theory Dynam. Systems 25 (2005), no. 4, 1173–1208. MR 2158401

work page 2005