Pith. sign in

REVIEW 2 major objections 4 minor 1 cited by

Estimating the Euclidean distortion of an orbit space

T0 review · 2 major / 4 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 The quotient of $\mathbb{C}^n$ by the $r$-th roots of unity embeds into Euclidean space with distortion exactly $r\sin(\pi/(2r))$.

desk verdict Good tools and genuine applications, but Theorem 16 as stated is not well-defined; the fix is clear and the paper deserves review after correction. read the letter →

arxiv 2506.04425 v1 pith:WAPQOS36 submitted 2025-06-04 math.MG cs.ITmath.FAmath.ITmath.RT

Ben Blum-Smith , Harm Derksen , Dustin G. Mixon , Yousef Qaddura , Brantley Vose This is my paper
classification math.MGcs.ITmath.FAmath.ITmath.RT MSC 46B8557S1554C25
keywords Euclideandistortionorbitspacequotientmetricbilipschitzembeddingcyclicgroupactionwallpapergroupsinvariantmachinelearningcontortion
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

Metric spaces that arise as quotients of Euclidean space by a group of symmetries, called orbit spaces, appear whenever data carries a symmetry ambiguity, and machine-learning pipelines often need to map such spaces into Euclidean space while distorting distances as little as possible. This paper develops a toolbox for computing the optimal distortion, and uses it to settle the exact value for several fundamental families. The headline result computes the Euclidean distortion of $\mathbb{C}^n/\mathcal{C}_r$, the quotient of complex $n$-space by the $r$-th roots of unity, showing it equals $r\sin(\pi/(2r))$; the same machinery gives exact distortions for seven wallpaper-group quotients of the plane and two-sided bounds for others. A recurring message is that low-distortion embeddings can be built from the classical polynomial invariants of the group action, suitably rescaled.

What carries the argument

The central mechanism is the quotient–orbit embedding (Theorem 11): if a map $\phi$ embeds the quotient $X/G$, and a $G$-equivariant map $\psi$ is orbit-expanding and alignment-preserving, then $x \mapsto (\phi([x]), c\,\alpha_\phi\, \psi(x))$ is a bilipschitz embedding of $X$ with distortion controlled by $\kappa(\phi)$ and the Lipschitz constant of $\psi$. For Theorem 16 the two ingredients are the degree-two tensor map $\phi(u)=u\otimes u/\|u\|$, which is optimal for the full circle action on $\mathbb{C}^n$, and the degree-$r$ tensor map $\psi(u)=u^{\otimes r}/\|u\|^{r-1}$, which separates the $\mathcal{C}_r$-orbits inside each circle-orbit; the scaled mixture $F=[\cos(\pi/(2r))\,\phi,\; \sin(\pi/(2r))\, \psi]$ gives the exact distortion $r\sin(\pi/(2r))$. Supporting the general theory are the finite-determinacy principle ($c_2$ equals the supremum over finite subspaces), the contortion $\Upsilon(G)$ (the largest distortion any quotient by $G$ can have), and the local-to-global inequality $c_2(T_pM/G_p) \le c_2(M/G)$ for wandering isometric actions.

What would settle it

Run the semidefinite-programming lower bound for a finite subset of $\mathbb{C}^2/\mathcal{C}_3$ that includes points with several relative phases on the unit sphere; if the computed distortion exceeds $3/2$, Theorem 16's exact value is false.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the Euclidean distortion of an orbit space can often be computed exactly by combining the optimal embedding of a larger quotient with an invariant that separates the remaining orbits. Theorem 16 states that for the action of the cyclic group $\mathcal{C}_r$ by scalar multiplication on $\mathbb{C}^n$, the distortion satisfies $c_2(\mathbb{C}^n/\mathcal{C}_r) = r\sin(\pi/(2r))$, and the map $F([u]) = (\cos(\pi/(2r))\, u\otimes u/\|u\|,\; \sin(\pi/(2r))\, u^{\otimes r}/\|u\|^{r-1})$ achieves this distortion. The proof reduces to the unit sphere and to verifying two trigonometric inequalities involving $|u^*v|$ and $\operatorname{Re}((u^*v)^r)$. The same local-to-global and quotient-orbit tools yield exact distortions for seven wallpaper-group types, for instance $\sqrt{2}$ for type $2{*}22$ and $2\sqrt{2-\sqrt{2}}$ for type $4{*}2$, and imply that permutation-symmetric quotients of graphs and databases have unbounded distortion as the size grows.

Load-bearing premise

The general machinery relies on the finite-determinacy principle that a quotient's distortion equals the supremum of distortions of its finite subspaces, a nonconstructive ultraproduct assertion that the paper itself flags as using the axiom of choice.

Editorial extensions

If this is right

  • The exact value $c_2(\mathbb{C}^n/\mathcal{C}_r)=r\sin(\pi/(2r))$ supplies the scalar-cyclic case in all dimensions, including $c_2(\mathbb{C}/\mathcal{C}_3)=3/2$ used in the contortion computation for groups of order three.
  • Seven wallpaper-group types now have exact distortions: reflection-wall groups of types $*333$, $*442$, $*632$, and $*2222$ embed isometrically with distortion $1$, while types $2{*}22$ and $4{*}2$ embed optimally with $\sqrt{2}$ and $2\sqrt{2-\sqrt{2}}$, respectively.
  • For $\mathrm{SO}(r)$ acting on $\mathbb{R}^{r\times n}$, the distortion lies between $\sqrt{2}$ and $2\sqrt{2}$ whenever $n \ge r \ge 2$.
  • For landmark spaces $(\mathbb{R}^r)^n$ modulo rotations and translations, embedding reduces to the centered configuration space: $c_2((\mathbb{R}^r)^n/(K \ltimes \mathbb{R}^r)) = c_2((\mathbb{R}^r)^{n-1}/K)$.
  • Quotients of weighted graphs and databases by row and column permutations have Euclidean distortion tending to infinity as the number of vertices or columns grows.

Reading between the lines

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

  • A natural generalization suggested by the quotient–orbit recipe is that for any subgroup $H$ of a compact group $G$, the optimal embedding of $V/H$ might be a scaled concatenation of the optimal $V/G$ embedding with a suitably normalized $H$-orbit-separating invariant; this could be tested computationally for small finite groups before seeking a proof.
  • The exact wallpaper values came from lower bounds at rotation centers and upper bounds by gluing; the same two ingredients could yield exact values for the remaining wallpaper types once the flat-torus distortion of their translation lattices is known.
  • The unboundedness results imply that any invariant feature map for graphs or point clouds that aims for bounded metric distortion must have target dimension growing with $n$, a concrete constraint for geometric deep learning architectures.
  • Since the proof of the finite-determinacy principle uses an ultraproduct, a constructive proof for the specific spaces treated here would be a worthwhile test of whether the exact values depend on the axiom of choice.
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

2 major / 4 minor

Summary. The paper develops tools for estimating the Euclidean distortion c2(V/G) of orbit spaces under isometric group actions, introducing equivariant embedding lemmas, a new notion of Euclidean contortion of a finite group, quotient–orbit embeddings, and a local-to-global lower bound via tangent-space isotropy quotients. These tools are then applied to compute or bound c2 exactly for several families: scalar cyclic actions on C^n, orthogonal/special-orthogonal matrix actions, alternating subgroups of reflection groups, wallpaper groups, landmark quotients by Euclidean groups, and permutation actions on graphs and databases. The headline result is Theorem 16, claiming c2(C^n/C_r)=r sin(pi/(2r)), achieved by an explicit embedding mixing u⊗u/|u| and u^{⊗r}/|u|^{r-1}; Theorem 26 claims exact values for seven wallpaper group quotients. The paper is largely self-contained, with appendices supplying proofs of auxiliary results such as Proposition 21 and background on Bochner spaces.

Significance. If the main results are correct, the paper resolves the scalar cyclic case in all dimensions, supplies exact Euclidean distortions for several wallpaper-group quotients, introduces the transferable invariant Υ(G) with exact values for |G|≤3, and gives a new general local-to-global mechanism for lower bounds. The explicit nature of the embeddings and the inclusion of complete proofs for several delicate steps (e.g., Propositions 21, 34, 35, and the slice-theorem argument in Appendix B) are notable strengths. However, the central Theorem 16 as printed is not well-defined on the quotient, and since it feeds Lemma 8 and Theorem 26(c),(d), the significance is conditional on the straightforward repair described in the major comments.

major comments (2)
  1. [§3.1, Theorem 16] The map F([u]) = (cos(π/2r)φ(u), sin(π/2r)ψ(u)) is not well-defined on C^n/C_r as printed, because φ(u)=u⊗u/‖u‖ is not C_r-invariant for r>2: for ω=e^{2πi/r}, φ(ωu)=ω²φ(u). Concretely, for r=3, n=1, u=1, v=ω, one has [u]=[v] but ‖F([u])−F([v])‖=cos(π/6)|1−ω²|>0, contradicting well-definedness. The proof itself uses the identity ‖φ(u)−φ(v)‖²=2−2|z|² with z=u^*v, which holds for φ(u)=u⊗`bar u` (equivalently the rank-one projection uu*/‖u‖), not for u⊗u. Since Theorem 16 is used to obtain Υ(C₃)=3/2 in Lemma 8 and the exact values in Theorem 26(c),(d), this is a load-bearing defect. The argument is repairable by replacing every occurrence of u⊗u with u⊗`bar u` in §3.1 (and correspondingly in §1.2.1 and §4.1) and adjusting the codomain to (C^n)⊗`overline{C^n}`, after which the displayed distance computations and bounds are consistent with the corrected map. The submitted text requires this correction before the central claim is valid.
  2. [§3.4, Theorem 26(c),(d)] The exact lower bounds for wallpaper groups of types 2*22 and 4*2 are derived from Lemma 25 (via Theorem 13) and the upper bounds are obtained by isometrically embedding R²/G into R²/±Id or R²/C₄, invoking Theorem 16 for the distortion of the latter quotients. Since Theorem 16 as stated is invalid without the correction to φ described above, the exact values √2 and 2√(2−√2) in Theorem 26(c) and (d) currently rest on an ill-defined embedding. After the repair, the same proof goes through verbatim, so this is a consequence of the first major comment rather than an independent defect.
minor comments (4)
  1. [§1.2.1 and §4.1] The text repeatedly writes the optimal U(1)-invariant map as z ↦ z⊗z/‖z‖; this should be z⊗`bar z`/‖z‖, in line with the correction required in Theorem 16, otherwise the notation is actively misleading.
  2. [§3.1, proof of Theorem 16] The proof cites 'Theorem 13 in [14]' to reduce to unit vectors; it would be helpful to state explicitly that this is a homogeneity argument, since the homogeneity of F is immediate but the cited theorem is not standard in the distortion literature.
  3. [§3.1, Figure 3 and surrounding text] In the proof of (ii), the functions f₁ and f₂ are used in the text before being defined; they are introduced only later via the equation g' = f₁−f₂. Please define them at first use.
  4. [§2.3.2 and §3.2] There are small typos that should be cleaned up: 'arbirary' in the proof of Corollary 12, 'mulivariate Bernoulli' in the proof of Lemma 19, and 'the the' in Section 1.2.1.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the new theorems are derived from explicit embeddings and from external published lemmas; the self-citations to [14] and [39] are prior published results, not unverified assumptions, and no new claim reduces to its own input by construction.

full rationale

I found no circular step in the claimed derivation chain. Theorem 16 is proved by an explicit quotient–orbit embedding and uses [14, Cor. 38] only for the already-known n=1 base case; the general-n upper and lower Lipschitz estimates are direct analytic arguments. Lemma 8's use of Theorem 16 for the order-3 contortion is a forward reference to an independently proved result, not a restatement of Lemma 8's input. The local-to-global Theorem 13 does rely on the finite-determinacy principle (Proposition 34, from [14, Prop. 31]) and on the equivariant embedding lemmas, but Proposition 34 is a published theorem with its own ultraproduct proof, and the paper explicitly notes the nonconstructive step; it is external support rather than a self-referential reduction. Proposition 35 collects known distortions from [14] and [27], and Proposition 21 (from [37]) is re-proved in Appendix E; these are stated inputs, not disguised conclusions. The wallpaper and special-orthogonal bounds combine these external inputs with the paper's own tools (Corollary 12, Theorem 13, Theorem 16) without fitting any parameter to the target quantity and then renaming the fit a prediction. Although D. G. Mixon coauthored [14] and [39], the cited results are published, proof-carrying theorems, so the self-citation is not circular. I also note, separately, a correctness issue that is not a circularity: as printed, F([u]) in Theorem 16 is not well-defined on C^n/C_r for r>2, since φ(u)=u⊗u/∥u∥ is multiplied by ω^2 when u is multiplied by ω, and the proof's identity ∥φ(u)-φ(v)∥^2=2-2|z|^2 corresponds to φ(u)=u⊗bar u rather than u⊗u. This appears repairable but is a defect in the stated definition, not a circular derivation.

Assumptions & free parameters 0 free parameters · 4 assumptions · 1 invented entities

The central claims rest on published background lemmas and on one explicitly flagged nonconstructive step, the ultraproduct finite-determinacy proof of Proposition 34. No free parameters are fitted; the only new quantity, the Euclidean contortion Υ(G), is a definition with checkable exact values in small cases rather than an unexplained postulate. Self-citations to [14] and [39] are disclosed and concern published results with independent proofs.

assumptions (4)
  • standard math Finite determinacy of Euclidean distortion (Proposition 34, from [14, Prop. 31]): c2(X) = sup over finite B of c2(B), proven via an ultraproduct and hence the axiom of choice.
    Load-bearing for Theorem 7 (contortion bound) and for claim (i) in the proof of Theorem 13 (local-to-global), which powers the wallpaper lower bounds and Theorems 31-32. The paper itself flags the nonconstructive axiom in Appendix A.
  • domain assumption Known distortion values for base quotients (Proposition 35): c2(C/C_r) = r sin(pi/(2r)), c2(C^n/T) = sqrt(2), c2(V/{+-Id}) = sqrt(2), c2(R/cZ) = pi/2, quoted from [14, Cor. 36-38] and [27, Thm. 6.1].
    Used as building blocks throughout Section 3; one coauthor of [14] is also an author here, so these are self-citations with published proofs.
  • domain assumption Wandering isometric action and small geodesically convex balls for the local-to-global theorem (Theorem 13): the stabilizer at p must be finite and the exponential map must localize the quotient metric near [p].
    Required for the lower bounds in Section 3.4 and Section 3.6; satisfied by finite group actions and wallpaper group actions in the applications.
  • standard math Coxeter and reflection-group classification facts used in Theorem 22, including the largest Coxeter entry n, the order of the rotation from two adjacent reflections, and the fundamental chamber structure.
    Background from [24]; used to reduce the alternating reflection group quotient to the planar cyclic case via Corollary 14.
invented entities (1)
  • Euclidean contortion Υ(G), the worst-case Euclidean distortion over all orthogonal representation quotients of a finite group G independent evidence
    purpose: Gives a universal bound c2(X/G) <= Υ(G)*c2(X) for any isometric action of G (Theorem 7) and a subnormal-series decomposition (Theorem 9).
    A mathematical quantity, not a physical postulate. It has falsifiable handles: exact values for groups of order at most 3, two-sided bounds for C4 and C2 x C2, and divergence for S_n, all computed in Section 2.2, so independent evidence is genuinely present.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Estimating the Euclidean distortion of an orbit space." pith.science (2026). https://pith.science/paper/WAPQOS36

@misc{pith2026250604425,
  author       = {Pith},
  title        = {Pith review of: Estimating the Euclidean distortion of an orbit space},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WAPQOS36}},
  note         = {Machine review of arXiv:2506.04425}
}
abstract

Given a finite-dimensional inner product space $V$ and a group $G$ of isometries, we consider the problem of embedding the orbit space $V/G$ into a Hilbert space in a way that preserves the quotient metric as well as possible. This inquiry is motivated by applications to invariant machine learning. We introduce several new theoretical tools before using them to tackle various fundamental instances of this problem.

Figures

Figures reproduced from arXiv: 2506.04425 by the authors.

Figure 1
Figure 1. An illustration of a glued space and a corresponding embedding of the type [PITH_FULL_IMAGE:figures/full_fig_p017_1.png] view at source ↗
Figure 2
Figure 2. A depiction of the tangent space analysis described in Section 2.4. [PITH_FULL_IMAGE:figures/full_fig_p020_2.png] view at source ↗
Figure 3
Figure 3. Plot of the functions f1, f2, and g ′ = f1 − f2 analyzed in the proof of Theorem 16. for θ ∈ [0, π r ]. Notably, g(0) = g( π r ) = 0 and g( π 2r ) = −2 sin2 ( π 4r ) cos( π 2r ) < 0. By the mean value theorem, there necessarily exist θ− ∈ (0, π 2r ) and θ+ ∈ ( π 2r , π r ) for which g ′ (θ−) < 0 and g ′ (θ+) > 0. We claim that g ′ has a unique root θ0 ∈ (0, π r ), meaning g ′ is negative over (0, θ0) and positive ov… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: (left) A Weyl chamber C for an action of D4, the dihedral group with 8 elements, on R 2 . We highlight the boundary of the Weyl chamber in red. (right) An embedding of the quotient R 2/D+ 4 in R 3 . This embedding is obtained by identifying R 2/D+ 4 ∼= C ⊔∂C C and appl…
Figure 5
Figure 5. Figure 5: (left) Wallpapers with symmetry groups of type ∗∗, 2∗22, and 4∗2, respectively. (right) An optimal bilipschitz embedding of each wallpaper pattern quotiented by its sym￾metry group. The image of each embedding forms a portion of a circular cylinder or cone. 38 [PITH_F…

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Coarea Reduction, Sparse Transfer, and Geometric Recomposition for Synchronized Singular Forms

    math.FA 2026-03 accept novelty 7.0 of 10

    For planar rotations, real phase retrieval, and finite reflection groups, linear transforms of max filter banks achieve distortion arbitrarily close to the Euclidean distortion of the orbit space.

Reference graph

Works this paper leans on

46 extracted references · 44 canonical work pages · cited by 1 Pith paper

  1. [7]

    Balan, C

    R. Balan, C. B. Dock, Lipschitz analysis of generalized phase retrievable matrix frames, SIAM J. Mat. Anal. Appl. 43(3) (2022) 1518–1571. 43

    work page 2022

  2. [1]

    Nearly Optimal Embeddings of Flat Tori

    I. Agarwal, O. Regev, Y. Tang, Nearly optimal embeddings of flat tori, (2020) arXiv:2005.00098

    work page Pith review arXiv 2020

  3. [2]

    M. M. Alexandrino, R. G. Bettiol, Lie groups and geometric aspects of isometric actions, Springer, 2015

    work page 2015

  4. [3]

    Alharbi, S

    W. Alharbi, S. Alshabhi, D. Freeman, D. Ghoreishi, Locality and stability for phase retrieval, Sampl. Theo. Sig. Proc. Dat. Anal. 22(1) (2024) 10

    work page 2024

  5. [4]

    T. Amir, T. Bendory, N. Dym, D. Edidin, The stability of generalized phase retrieval problem over compact groups, (2025) arXiv:2505.04190

    work page arXiv 2025

  6. [5]

    Andoni, A

    A. Andoni, A. Naor, O. Neiman, Impossibility of sketching of the 3D transportation metric with quadratic cost, ICALP 2016

    work page 2016

  7. [6]

    Balan, P

    R. Balan, P. Casazza, D. Edidin, On signal reconstruction without phase, Appl. Comput. Harmon. Anal. 20 (2006) 345–356

    work page 2006

  8. [8]

    Balan, N

    R. Balan, N. Haghani, M. Singh, Permutation invariant representations with appli- cations to graph deep learning, arXiv:2203.07546

    arXiv

Show all 46 references
  1. [9]

    Balan, E

    R. Balan, E. Tsoukanis, G-invariant representations using coorbits: Bi-lipschitz prop- erties, arXiv:2308.11784

    arXiv

  2. [10]

    Balan, Y

    R. Balan, Y. Wang, Invertibility and robustness of phaseless reconstruction, Appl. Comput. Harmon. Anal. 38 (2015) 469–488

    2015

  3. [11]

    A. S. Bandeira, J. Cahill, D. G. Mixon, A. A. Nelson, Saving phase: Injectivity and stability for phase retrieval, Appl. Comput. Harmon. Anal. 37 (2014) 106–125

    2014

  4. [12]

    Burnside, On groups of order pαqβ, Proc

    W. Burnside, On groups of order pαqβ, Proc. London Math. Soc. 2 (1904) 388-392

    1904

  5. [13]

    Cahill, P

    J. Cahill, P. Casazza, I. Daubechies, Phase retrieval in infinite-dimensional Hilbert spaces, Trans. Amer. Math. Soc., Ser. B 3 (2016) 63–76

    2016

  6. [14]

    Cahill, J

    J. Cahill, J. W. Iverson, D. G. Mixon, Towards a bilipschitz invariant theory, Appl. Comp. Harm. Anal. 72 (2024) 101669

    2024

  7. [15]

    Cahill, J

    J. Cahill, J. W. Iverson, D. G. Mixon, D. Packer, Group-invariant max filtering, Found. Comp. Math. (2024) 1–38

    2024

  8. [16]

    Conca, D

    A. Conca, D. Edidin, M. Hering, C. Vinzant, An algebraic characterization of injec- tivity in phase retrieval, Appl. Comput. Harmon. Anal. 38 (2015) 346–356

    2015

  9. [17]

    J. H. Conway, D. H. Huson, The Orbifold Notation for Two-Dimensional Groups, Structural Chemistry 13, 247–257 (2002)

    2002

  10. [18]

    J. H. Conway, H. Burgiel, C. Goodman-Strauss, The Symmetries of Things, CRC Press (2016)

    2016

  11. [19]

    Dadok, Polar Coordinates Induced by Actions of Compact Lie Groups, Trans

    J. Dadok, Polar Coordinates Induced by Actions of Compact Lie Groups, Trans. Amer. Math. Soc. 288, 1 (1985) 125–137

    1985

  12. [20]

    Eriksson-Bique, Quantitative bi-Lipschitz embeddings of bounded-curvature man- ifolds and orbifolds, Geom

    S. Eriksson-Bique, Quantitative bi-Lipschitz embeddings of bounded-curvature man- ifolds and orbifolds, Geom. and Top. 22 (2018) 1961–2026

    2018

  13. [21]

    Federer, Geometric measure theory, Springer (2014)

    H. Federer, Geometric measure theory, Springer (2014)

    2014

  14. [22]

    G. B. Folland, Real analysis: modern techniques and their applications, John Wiley & Sons (1999)

    1999

  15. [23]

    D. H. Fremlin, Measure theory, Vol 4 & 5 Torres Fremlin (2000)

    2000

  16. [24]

    L. C. Grove, C. T. Benson, Finite reflection groups, Springer Sci. & Busi. Med. (1996) Vol. 99

    1996

  17. [25]

    Grove, Geometry of, and via, symmetries, Uni

    K. Grove, Geometry of, and via, symmetries, Uni. Lect. Series-AMS, 27, 31–51 (2002). 44

    2002

  18. [26]

    Haviv, O

    I. Haviv, O. Regev, The Euclidean Distortion of Flat Tori, Approx. Rand. Comb. Opt. Alg. & Tech. (2010) 232–245

    2010

  19. [27]

    Heimendahl, M

    A. Heimendahl, M. L¨ ucke, F. Vallentin, M. C. Zimmermann, A semidefinite program for least distortion embeddings of flat tori into Hilbert spaces, arXiv:2210.11952

  20. [28]

    Hu, Cohomology theory in topological groups, Mich

    S. Hu, Cohomology theory in topological groups, Mich. Math. 1 (1952) 11–59

    1952

  21. [29]

    Hyt¨ onen, J

    T. Hyt¨ onen, J. V. Neerven, M. Veraar, L. Weis, Analysis in Banach spaces, Vol 12 Berlin: Springer (2016)

    2016

  22. [30]

    Jones, A

    P.W. Jones, A. Osipov, V. Rokhlin, Randomized approximate nearest neighbors al- gorithm. ONAS 108.38 (2011): 15679-15686

    2011

  23. [31]

    Kapovich, A note on properly discontinuous actions, S˜ ao Paulo J

    M. Kapovich, A note on properly discontinuous actions, S˜ ao Paulo J. Math. Sci. 18, 2 (2024) 807–836

    2024

  24. [32]

    S. Khot, A. Naor, Nonembeddability theorems via Fourier analysis, Mathematische Annalen 334 (2006) 821–852

    2006

  25. [33]

    Kramer, Some remarks on proper actions, proper metric spaces, and buildings, Adv

    L. Kramer, Some remarks on proper actions, proper metric spaces, and buildings, Adv. Geom. 22, 4 (2022) 541–559

    2022

  26. [34]

    J. M. Lee, Introduction to Riemannian manifolds, Springer, 2018

    2018

  27. [35]

    MacBeath, The classification of non-euclidean plane crystallographic groups, Canad

    M. MacBeath, The classification of non-euclidean plane crystallographic groups, Canad. J. Math. 19 (1967) 1192–1205

    1967

  28. [36]

    Makarychev, Y

    K. Makarychev, Y. Makarychev, A union of Euclidean metric spaces is Euclidean, Discrete Analysis Journal (2016)

    2016

  29. [37]

    Miranda, R

    H. Miranda, R. C. Thompson, A trace inequality with a subtracted term, Lin. Alg. and its Appl. 185 (1993) 165–172

    1993

  30. [38]

    D. G. Mixon, D. Packer, Max filtering with reflection groups, Adv. Comp. Math. 49(6) (2023) 82

    2023

  31. [39]

    D. G. Mixon, Y. Qaddura, Injectivity, stability, and positive definiteness of max filtering, Constructive Approximation (2025)

    2025

  32. [40]

    Qaddura, A max filtering local stability theorem with application to weighted phase retrieval and cryo-EM, arxiv:2403.14042

    Y. Qaddura, A max filtering local stability theorem with application to weighted phase retrieval and cryo-EM, arxiv:2403.14042

  33. [41]

    Salamon, Measure and integration, London EMS (2016)

    D. Salamon, Measure and integration, London EMS (2016)

    2016

  34. [42]

    Steenrod, The topology of fibre bundles, Princeton university press, vol 14 (1999)

    N. Steenrod, The topology of fibre bundles, Princeton university press, vol 14 (1999)

    1999

  35. [43]

    Vallentin, P

    F. Vallentin, P. Moustrou, Least distortion Euclidean embeddings of flat tori, ISSAC (2023) 13–23. 45

    2023

  36. [44]

    De Vries, The local weight of an effective locally compact transformation group and the dimension of L2(G), Coll

    J. De Vries, The local weight of an effective locally compact transformation group and the dimension of L2(G), Coll. Math. Vol. 39 IMPAN (1978) 319–323

    1978

  37. [45]

    Y. Xia, Z. Xu, Z. Xu, Stability in phase retrieval: Characterizing condition numbers and the optimal vector set, Math. Comp. (2024)

    2024

  38. [46]

    Zolotov, Bi-lipschitz embeddings of SRA-free spaces into Euclidean spaces, arXiv:1906.02477 (2019)

    V. Zolotov, Bi-lipschitz embeddings of SRA-free spaces into Euclidean spaces, arXiv:1906.02477 (2019). A Previous results in bilipschitz invariant theory We first collect two existing general tools for computing Euclidean distortion. The first regards the Euclidean distortion ...

    2019 arXiv

Pith tools

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