Pyramidal Compactification of Asymmetric Metric Measure Spaces via Adjoint Transport

Shigeaki Yokota

REVIEW 3 major objections 5 minor 15 references

Pyramidal Compactification of Asymmetric Metric Measure Spaces via Adjoint Transport

T0 review · 3 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Every quasi-metric measure space embeds densely into a compact space of pyramids that keeps forward and reverse directions distinct.

desk verdict A genuinely new directed pyramidal compactification built on a clean adjunction, but its central compactness rides on unproved companion-paper lemmas; worth refereeing with those papers in hand. read the letter →

arxiv 2608.01145 v1 pith:CD2QTEKT submitted 2026-08-02 math.MG

Shigeaki Yokota This is my paper · ORCID

classification math.MG MSC 53C2354E3528A33

keywords pyramidalcompactificationquasi-metricmeasurespaceasymmetricmetricadjointtransportboxdistanceone-sidedLipschitzfunctionsconcentrationtopologypyramid

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

Classical pyramid compactification records metric measure spaces as lower sets of dominated spaces, but it works with symmetric distances and therefore cannot tell which endpoint carries which potential. This paper removes that limitation by building the same compactification for quasi-metric measure spaces (qm-spaces): spaces whose directed distance has a complete separable symmetrization and which carry a full-support probability measure. Its central claim is that the lower set $P(X) = \{Y \mid Y \preceq X\}$ is a pyramid for every qm-space $X$, and that the associated-pyramid map $X \mapsto P(X)$ is a $1$-Lipschitz topological embedding of the concentration-distance space into a compact metric space $(\Pi^+, d^{\mathrm{M}}_{\Pi})$ of all such pyramids, with dense image. The construction uses functions whose increments are bounded in only one direction, so the order of endpoints is preserved at limits. A secondary result, completeness and separability of the qm-space box-distance space, supplies the closedness and approximation machinery the argument needs.

What carries the argument

The load-bearing object is the pyramidal adjunction $\mathrm{Rep}^+ \dashv \mathrm{Rec}^+$ between qm-spaces and one-sided-Lipschitz geometric data sets, i.e. an adjunction that transports pyramid structure and weak convergence between box-metrized categories. $\mathrm{Rep}^+(X) = (X, \mathrm{Lip}^+_1(X), \mu_X)$ records the ordered increments of the space, and $\mathrm{Rec}^+$ reconstructs the directed distance as the supremum of those increments; the adjunction is pyramidal because its unit is an isomorphism, $\mathrm{Rep}^+$ is box-isometric, $\mathrm{Rec}^+$ is box-nonexpansive, and the target category has domination refinement. This lets the paper transfer the gd-set pyramid compactific

What would settle it

A single counterexample would settle the claim: find a qm-space $X$ whose lower set $P(X)$ is not a pyramid, for instance two elements with no common upper bound inside $P(X)$ or a box-limit outside $P(X)$, or find a sequence of pyramids in $\Pi^+$ with no weakly convergent subsequence. The paper's three-point example is its smallest confirmation, so a natural search begins with four-point qm-spaces; alternatively, a pair of qm-spaces with identical ordered measurements for all $N,R$ but different directed distances would falsify the separation of $d^{\mathrm{M}}_{\Pi}$.

Watch

Extended reading notes

Core claim

The paper's central discovery is a directed analogue of the pyramidal compactification. For every qm-space $X$, the set $P(X)$ of qm-spaces dominated by $X$ belongs to $\Pi^+$, and there is a metric $d^{\mathrm{M}}_{\Pi}$ on $\Pi^+$ such that $(\Pi^+, d^{\mathrm{M}}_{\Pi})$ is compact, the map $X \mapsto P(X)$ is $1$-Lipschitz from $(X^+, d_{\mathrm{conc}})$ into $(\Pi^+, d^{\mathrm{M}}_{\Pi})$ and is a topological embedding, and the image is dense. The engine is the representation–reconstruction adjunction $\mathrm{Rep}^+ \dashv \mathrm{Rec}^+$: a qm-space is represented by the family of one-sided $1$-Lipschitz functions it admits, and the directed distance is recovered as $d^+_X(x,x') = \s

Load-bearing premise

The proof rests on unproved companion-paper results about geometric data sets, especially the domination-refinement lemma and the completeness and separability of the box-distance space; if any of those lemmas is false, the pyramid-transport argument and the compactness of $(\Pi^+, d^{\mathrm{M}}_{\Pi})$ collapse.

Editorial extensions

If this is right

Every sequence of qm-space pyramids has a weakly convergent subsequence in $(\Pi^+, d^{\mathrm{M}}_{\Pi})$, and every pyramid is a weak limit of associated pyramids $P(X_n)$.
One-sided observables keep forward and reverse distances distinct at limits, so the directed compactification carries strictly more information than the symmetrized theory; the paper's three-point example illustrates the bound that symmetric box limits lose.
The box-distance space $(X^+, \square)$ is complete and separable, providing a Polish background for convergence of directed metric measure spaces.
The associated-pyramid map satisfies $d^{\mathrm{M}}_{\Pi}(P(X),P(Y)) \le d_{\mathrm{conc}}(X,Y)$, so concentration convergence of qm-spaces implies weak convergence of their pyramids, and conversely weak convergence of the pyramids forces concentration convergence on the image.

Reading between the lines

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

The stated Funk-ball motivation implies the three-phase limit diagram of the beta model can be read off as weak limits inside $(\Pi^+, d^{\mathrm{M}}_{\Pi})$; a natural next step is to identify each phase with an explicit pyramid in $\Pi^+$.
The transport framework is categorical, so the same pyramidal-adjunction template should produce compact pyramid spaces in any box-metrized category whose right adjoint is box-nonexpansive and whose codomain has domination refinement; the gd-set and qm-space cases are two instances of one recipe.
A testable extension is to compute ordered $(N,R)$-measurements on finite qm-spaces: because ordered measurements detect reverse-distance bounds that symmetric measurements miss, the Hausdorff terms in $d^{\mathrm{M}}_{\Pi}$ should distinguish spaces that the classical pyramid metric identifies.

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.

Desk Editor's Note

read the letter

Dear colleague,

The one thing to know: this is a real extension, not a repackaging. Yokota constructs a compact space of pyramids for quasi-metric measure spaces, preserving the order of directed endpoints that Shioya's symmetric compactification discards. The route is a representation functor Rep+ into Lip_1^+(R)-gd-sets with Rec+ as right adjoint, and the associated-pyramid map embeds the concentration space into this pyramid space with dense image. Theorem 1.1 is the directed analog of Shioya's theorem, and the three-point example in §2.4 shows concretely that the directed lower set differs from the symmetric one. The secondary result, completeness and separability of (X^+, □), is also useful.

Credit where due: the adjunction is genuinely new, the paper is unusually transparent about which statements come from [13] and [14], and the proofs actually carried out—box-nonexpansiveness of Rec+, closedness of domination, the appendix transport theorem—are dense but coherent. There are no fitted parameters and no invented objects. The citation pattern is not a red flag by itself.

The soft spot is exactly where the reader put it, and the stress-test note sharpens it. Theorem 1.1(i), compactness of (Π^+, d^M_Π), ultimately depends on [14, Lemma 5.4], [14, Lemma 5.6], and [14, Theorem 4.10], all imported as black boxes. Lemma 5.6 in particular is a substantive mixed domination/subsequence statement: from X_n, Y_n ⪯ Z̄_n and X_n→X, Y_n→Y it produces refinements Z_n with a □-convergent subsequence. The present paper gives no proof or informal justification, and that lemma is what makes the Painlevé–Kuratowski weak limit of pyramids directed. If it failed, Corollary 5.3 and half of Theorem 1.1(i) collapse. Section 3.2 lists these dependencies honestly, but honesty does not make them less load-bearing. This is not a flaw in the internal logic; it is a gap in independent verifiability. The conditional verdict is the right one.

Who is this for? Someone working in metric measure geometry, concentration of measure, or directed data models who wants the asymmetric counterpart of Shioya's compactification. The paper deserves a serious referee—ideally one who can also read [13] and [14]. I would not rely on Theorem 1.1 in my own work until those lemmas are checked.

Recommendation: send it to peer review, conditional on the companion papers being available to the referee. If [14] is still preprint-only, ask the author to include or state the proof of Lemma 5.6.

Referee Report

3 major / 5 minor

Summary. The paper introduces quasi-metric measure spaces (qm-spaces), whose directed distance is retained rather than symmetrized away, and constructs a pyramidal compactification for them. The main theorem (Theorem 1.1) states that every lower set P(X) is a pyramid, that the space (Π_+, d^M_Π) is compact, that the associated-pyramid map is a 1-Lipschitz topological embedding with dense image, and that convergence in the pyramid metric is equivalent to weak convergence of pyramids. The strategy is to transport the known pyramidal compactification for geometric data sets through an adjunction Rep^+ ⊣ Rec^+ between qm-spaces and Lip^+_1(R)-gd-sets. The paper also proves completeness and separability of the box distance on qm-spaces. The proofs are dense and the abstract adjunction framework is coherent, but several essential lemmas are imported from the companion papers [13] and [14], especially [14, Lemma 5.4], [14, Lemma 5.6], and [14, Theorem 4.10]. The manuscript is transparent about these dependencies, but they are not reproduced here, so the main compactness claim cannot be independently verified from the present text alone.

Significance. If the result is correct, it provides the directed analogue of Shioya's pyramidal compactification, preserving the order of directed endpoints that classical symmetrization loses. This is a natural and potentially useful extension for applications such as the forward Funk beta model mentioned in the introduction. The paper's abstract pyramid-transport formalism (Appendix A) is a useful contribution in its own right, and the measurement-based metric in Section 4 gives a concrete, finite-dimensional criterion for pyramid convergence. The manuscript is unusually transparent: Section 3.2 lists every imported result and its first use, and no fitted parameters or self-referential definitions appear. However, the central compactness assertion depends on black-box lemmas from companion papers, and the companion [12] is said to depend mathematically on the present paper. This makes the current submission a high-level transport of an unverified upstream theory rather than a fully self-contained proof of Theorem 1.1.

major comments (3)

[Appendix A.3 / Theorem 5.1 / [14, Lemma 5.6]] The directedness of weak limits is proved by invoking [14, Lemma 5.6] as a black box: given X_n, Y_n ⪯ Z̄_n with X_n→X and Y_n→Y, the lemma asserts existence of Z_n with X_n, Y_n ⪯ Z_n ⪯ Z̄_n and a □-convergent subsequence. This is a substantive mixed domination/subsequence existence statement, and the present manuscript gives no proof, informal justification, or even a precise statement beyond the summary in Section 3.2. The directedness step is then used to show that the box-closed limit E in Theorem 5.1 is a pyramid; if Lemma 5.6 fails, Corollary 5.3 fails and compactness in Theorem 1.1(i) collapses. I recommend either proving Lemma 5.6 in an appendix or including the companion paper [14] in full as part of the submission so that the referee can verify this load-bearing step.
[Section 2.4 / Proposition 5.2 / [14, Lemma 5.4]] Domination refinement is imported from [14, Lemma 5.4] and used in three essential places: to show Lip^+_1(R)∘D has domination refinement after Definition 2.19, to prove Proposition 5.2 (descent to qm-spaces), and inside Theorem 2.20's proof of the transport equivalence. The proof of the outer/inner conditions in Theorem A.7 also relies on domination refinement. Since this lemma underlies both the pyramid property of transported sets and the weak-convergence equivalence, it is not a minor convenience but a load-bearing black box. The manuscript's Section 3.2 lists the statement, but a journal referee normally needs either the statement with proof or a clear indication that the companion is available and accepted. As written, the central claim is not independently supported.
[Section 3.1 / Corollary 3.4 / [14, Theorem 4.10]] The paper's secondary result, completeness and separability of (X^+, □), is derived by pulling back the corresponding fact for Lip^+_1(R)∘D, which is imported from [14, Theorem 4.10]. This is an entire structural property rather than a small technical lemma. The proof in Corollary 3.4 is only a one-line consequence of the imported theorem plus the box-nonexpansiveness of Rep^+∘Rec^+. If [14, Theorem 4.10] is correct, the argument is fine; but the present submission does not allow a referee to check that theorem. Given that the main theorem uses completeness and separability both for subsequence extraction (Theorem 5.1) and for density (Theorem A.13), this dependency is load-bearing. Please supply the missing proof or make the companion paper available for review.

minor comments (5)

[Throughout] There are frequent typos and spacing errors: “SP ACES” in the title line, “W eak” in Section 5.1 heading, “T(rport” in the abstract area, and inconsistent use of “Rec +” with and without a space. These should be corrected in a final revision.
[Figure 1 / Remark 2.17] The caption of Figure 1 says “The displayed arrows give d_X(b,a)=d_X(c,a)=0.1,” but the arrowheads in the figure are not visible in the text and the orientation is easy to misread. The remark would be clearer if the arrows were explicitly labeled with values.
[Definitions 2.12 and 2.13] The notation dconc for the qm-space observable distance uses the same symbol as d_conc for gd-sets. This is acceptable because of the pullback definition, but the subscript in the qm-space case is typeset inconsistently (sometimes dconc, sometimes d_conc). Please standardize.
[Appendix A.3] In the paragraph proving the existence of a one-point object, the text says “If 1/∈ E, then □(1,P_n)=0.” It might be clearer to say that 1∈P_n for every n, so the distance from 1 to P_n is zero; then the outer condition contradicts 1/∈E. The meaning is correct but the phrasing is compressed.
[References] Reference [10] is to a Japanese-language book that may not be readily accessible to all readers; the relevant theorem numbers are cited, but a short statement in the text would help. Also, reference [12] is “Manuscript in preparation” and is not available for checking the claimed three-phase limits.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the central compactification is a genuine transport theorem; the cited companion-paper lemmas are upstream, not restatements of the target result.

full rationale

The derivation chain is not circular. The core new content is the representation–reconstruction adjunction Rep^+ ⊣ Rec^+ (Theorem 2.11), proved in the paper: the unit is an isomorphism, Rep^+ is box-isometric (Definitions 2.12–2.13, Theorem 2.15), and Rec^+ is box-nonexpansive (Proposition 3.3). Theorem 1.1 is assembled from these facts together with the measurement criterion (Theorem 4.4) and the abstract pyramid-transport theorem (Appendix A). The only places where the manuscript invokes prior work are the gd-set statements listed in Section 3.2, especially [14, Lemma 5.6] in Appendix A.3 and [14, Theorem 4.10] in Corollary 3.4. These are genuinely upstream companion-paper results about L-compact geometric data sets; their hypotheses do not include the qm-space compactification being proved, and the present paper does not define its conclusion into those lemmas. The dependency on [14, Lemma 5.6] for directedness of weak limits is a correctness risk—if that lemma fails, Theorem 1.1(i) would collapse—but it is not circular, because the cited lemma is not the target theorem and the present paper does not rely on the sister paper [12], which depends on the present paper. No fitted parameter is renamed as a prediction, and no displayed equation is equal to its input by construction.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The central claim rests on the companion-paper theory of gd-sets and their pyramids. The present paper contributes the adjunction and its applications, but the compactness and measurement results are imported, so the ledger lists those imported theorems as domain assumptions.

assumptions (3)

domain assumption The theory of L-compact geometric data sets and their pyramids as developed in [13] and [14].
Used for the box distance metric, completeness, domination refinement, and measurement compactness; these are not proved in this paper (Section 3.2 lists them).
domain assumption Domination refinement holds for Lip^+_1(R) ∘ D ([14, Lemma 5.4]).
Essential in Theorem 2.20 and Corollary 5.3; implies the descent of dominated limits through weak convergence.
ad hoc to paper Every monoidal subfamily L contains T_B ([14, Definitions 3.1 and 3.2]).
Standing assumption in Section 2.1, needed for self-compactness of L and for compactness of finite measurement sets in Appendix B.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Pyramidal Compactification of Asymmetric Metric Measure Spaces via Adjoint Transport." pith.science (2026). https://pith.science/paper/CD2QTEKT

@misc{pith2026260801145,
  author       = {Pith},
  title        = {Pith review of: Pyramidal Compactification of Asymmetric Metric Measure Spaces via Adjoint Transport},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CD2QTEKT}},
  note         = {Machine review of arXiv:2608.01145}
}

abstract

A quasi-metric measure space (qm-space) is a set with a directed distance whose symmetrization is a complete separable metric, together with a Borel probability measure of full support. Motivated by the problem of determining the pyramid limits of beta measures on forward Funk balls, we construct a compact metric space of pyramids of qm-spaces. The associated-pyramid map from the concentration-distance space of qm-spaces into this compact space is a $1$-Lipschitz topological embedding with dense image. As a secondary result, we prove that the box-distance space of qm-spaces is complete and separable.

Figures

Figures reproduced from arXiv: 2608.01145 by the authors.

**Figure 1.** The three-point qm-space X. The displayed arrows give dX(b, a) = dX(c, a) = 0.1. Every other nonzero directed distance is 2. Layer Input Conclusion obtained Role in the main theorem Distance recovery Ordered increments of one-sided observables The directed distance and its symmetrization are recovered Prevents loss of the directed object under representation Box geometry Observable families on coupled probability s… view at source ↗

**Figure 2.** Adjunctions connecting the symmetric, directed, and gd-set layers. Corollary A.12 (The pyramidal property of the three-layer diagram). All three adjunctions in [PITH_FULL_IMAGE:figures/full_fig_p025_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

15 extracted references · 13 canonical work pages

[13]

S. Yokota. Geometry of geometric data set I. to appear in Tsukuba Journal of Mathematics, arXiv:2603.23316

work page arXiv
[14]

S. Yokota. Geometry of geometric data set II: Pyramid. preprint, arXiv:2603.23325

work page arXiv
[12]

S. Yokota. Funk beta balls: Exact potentials and a three-phase pyramid diagram. Manuscript in preparation

work page
[1]

Beer.Topologies on closed and closed convex sets, volume 268 ofMathematics and its Applications

G. Beer.Topologies on closed and closed convex sets, volume 268 ofMathematics and its Applications. Kluwer Academic Publishers Group, Dordrecht, 1993

work page 1993
[2]

Convergence of cones of metric measure spaces and its application to Cauchy distribution

S. Esaki, D. Kazukawa, and A. Mitsuishi. Convergence of cones of metric measure spaces and its application to Cauchy distribution.Int. Math. Res. Not. IMRN, 2025(18):rnaf292, 2025. arXiv:2402.14331

work page Pith review arXiv 2025
[3]

Gromov.Metric structures for Riemannian and non-Riemannian spaces

M. Gromov.Metric structures for Riemannian and non-Riemannian spaces. Modern Birkh¨ auser Classics. Birkh¨ auser Boston, Inc., Boston, MA, english edition, 2007. Based on the 1981 French original, With appendices by M. Katz, P. Pansu and S. Semmes, Translated from the French by Sean Michael Bates

2007
[4]

Hanika, F

T. Hanika, F. M. Schneider, and G. Stumme. Intrinsic dimension of geometric data sets. Tohoku Math. J. (2), 74(1):23–52, 2022

work page 2022
[5]

Krist´ aly and W

A. Krist´ aly and W. Zhao. On the geometry of irreversible metric-measure spaces: Conver- gence, stability and analytic aspects.Journal de Math´ ematiques Pures et Appliqu´ ees (9), 158:216–292, 2022

work page 2022

Show all 15 references

[6]

V. Pestov. An axiomatic approach to intrinsic dimension of a dataset.Neural Networks, 21(2-3):204–213, 2008

2008
[7]

Romaguera and M

S. Romaguera and M. Sanchis. Semi-Lipschitz functions and best approximation in quasi- metric spaces.J. Approx. Theory, 103(2):292–301, 2000

2000
[8]

Shioya.Metric measure geometry, volume 25 ofIRMA Lectures in Mathematics and Theoretical Physics

T. Shioya.Metric measure geometry, volume 25 ofIRMA Lectures in Mathematics and Theoretical Physics. EMS Publishing House, Z¨ urich, 2016. Gromov’s theory of convergence and concentration of metrics and measures

2016
[9]

T. Shioya. Metric measure geometry: an approach to high-dimensional and infinite- dimensional spaces.Sugaku Expositions, 35(2):221–241, 2022. Translation of S¯ ugaku71 (2019), no. 2, 159–177

2022
[10]

T. Shioya.Sokudo kyori k¯ ukan no kikagaku e no sh¯ otai: K¯ ojigen oyobi mugen-jigen k¯ ukan e no apur¯ ochi [An invitation to the geometry of metric measure spaces: An approach to high- dimensional and infinite-dimensional spaces]. Number 195 in SGC Library. Saiensu-sha, Nov
[11]

Stojmirovi´ c

A. Stojmirovi´ c. Quasi-metric spaces with measure.Topology Proceedings, 28(2):655–671, 2004

2004
[2024]

In Japanese; first edition published November 25, 2024

2024

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [13]

S. Yokota. Geometry of geometric data set I. to appear in Tsukuba Journal of Mathematics, arXiv:2603.23316

work page arXiv

[2] [14]

S. Yokota. Geometry of geometric data set II: Pyramid. preprint, arXiv:2603.23325

work page arXiv

[3] [12]

S. Yokota. Funk beta balls: Exact potentials and a three-phase pyramid diagram. Manuscript in preparation

work page

[4] [1]

Beer.Topologies on closed and closed convex sets, volume 268 ofMathematics and its Applications

G. Beer.Topologies on closed and closed convex sets, volume 268 ofMathematics and its Applications. Kluwer Academic Publishers Group, Dordrecht, 1993

work page 1993

[5] [2]

Convergence of cones of metric measure spaces and its application to Cauchy distribution

S. Esaki, D. Kazukawa, and A. Mitsuishi. Convergence of cones of metric measure spaces and its application to Cauchy distribution.Int. Math. Res. Not. IMRN, 2025(18):rnaf292, 2025. arXiv:2402.14331

work page Pith review arXiv 2025

[6] [3]

Gromov.Metric structures for Riemannian and non-Riemannian spaces

M. Gromov.Metric structures for Riemannian and non-Riemannian spaces. Modern Birkh¨ auser Classics. Birkh¨ auser Boston, Inc., Boston, MA, english edition, 2007. Based on the 1981 French original, With appendices by M. Katz, P. Pansu and S. Semmes, Translated from the French by Sean Michael Bates

2007

[7] [4]

Hanika, F

T. Hanika, F. M. Schneider, and G. Stumme. Intrinsic dimension of geometric data sets. Tohoku Math. J. (2), 74(1):23–52, 2022

work page 2022

[8] [5]

Krist´ aly and W

A. Krist´ aly and W. Zhao. On the geometry of irreversible metric-measure spaces: Conver- gence, stability and analytic aspects.Journal de Math´ ematiques Pures et Appliqu´ ees (9), 158:216–292, 2022

work page 2022