REVIEW 3 major objections 6 minor 2 cited by
$\ell^p$-coarse Baum-Connes conjecture for $\ell^{q}$-coarse embeddable spaces
T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash
Pith's one-line read The ℓ^p-coarse Baum–Connes conjecture holds for every bounded-geometry space that coarsely embeds into an ℓ^q-space, for all p,q≥1.
desk verdict A real advance is claimed, but the proof's load-bearing product decomposition in Corollary 2.26 is unjustified; the paper deserves a careful referee, not a desk reject. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing mechanism is the homogeneous Mazur map Ψ:(E,‖·‖_q)→(E,‖·‖_2), extended from the Mazur map on spheres, whose Lipschitz constant L=1+q·2^q is independent of the dimension of E (Lemma 2.6). This dimension independence is what makes the scale s_n=2L(dim E_n)^2 in equation (2.4) produce uniform norm bounds on the rescaled Bott–Dirac operators B_{s,v}=$s^{{-1}}$D+C_v and their functional-calculus images Φ_{s,v}; the commutator and propagation estimates of Proposition 2.17 all depend on it. The other essential ingredient is Theorem 3.7, a vector-valued Marcinkiewicz–Zygmund inequality proved via the Grothendieck inequality, which controls operators on the mixed ℓ^p-$ℓ^{2}$ module ℓ^p(Z,ℓ^p(N,H)) so that the $ℓ^{2}$-theoretic estimates survive in the ℓ^p setting.
What would settle it
A concrete check would be to compute the Lipschitz constant of the homogeneous Mazur map on a sequence of finite-dimensional subspaces E_n of ℓ^q with dim E_n→∞. If these constants grow without bound, the uniform estimates in Corollary 2.14 and Proposition 2.17(5)–(6) fail, and the Mayer–Vietoris proof of the twisted assembly isomorphism cannot proceed; alternatively, finding a bounded-geometry space that coarsely embeds into some ℓ^q but whose ℓ^p-coarse Baum–Connes evaluation map is not an isomorphism would directly refute the theorem.
Extended reading notes
Core claim
The central claim is Theorem 1.2: if (X,d) is a discrete metric space with bounded geometry and X coarsely embeds into ℓ^q for some q∈[1,∞), then for every p∈[1,∞) the evaluation map ev : C^p_L(P_r(X)) → C^p(X) induces an isomorphism ev_* : lim_{r→∞} K_*(C^p_L(P_r(X))) → K_*(C^p(X)), where C^p(X) is the paper's modified ℓ^p-Roe algebra (Definition 3.2). The proof constructs a twisted ℓ^p-Roe algebra A^p(P,E) by tensorially attaching finite-dimensional Euclidean spaces through the coarse embedding, and uses a Bott–Dirac operator built from the homogeneous Mazur map Ψ:(E,‖·‖_q)→(E,‖·‖_2) to twist K-theory classes. It then shows that the twisted assembly map is an isomorphism by a Mayer–Vietoris argument over ℓ^q-balls, using the vector-valued Marcinkiewicz–Zygmund inequality to pass from $ℓ^{2}$ estimates. For p=2 the argument recovers the classical coarse Baum–Connes conjecture for ℓ^q-embeddable spaces, which is Theorem 1.1.
Load-bearing premise
The proof rests on the assumption that the homogeneous Mazur map from ℓ^q to $ℓ^{2}$ has a Lipschitz constant that does not grow with the dimension of the finite-dimensional subspaces; if that constant grew with dimension, the uniform norm and propagation estimates that make the twisted assembly map an isomorphism would collapse.
Editorial extensions
If this is right
- For every bounded-geometry space coarsely embeddable into ℓ^q, the ℓ^p-coarse Baum–Connes conjecture holds for all p≥1, so in particular the coarse assembly map is an isomorphism on K-theory.
- The K-theory K_*(C^p(X)) of the modified ℓ^p-Roe algebra is independent of p for ℓ^q-embeddable spaces, since it is isomorphic to the K-homology of X through the assembly map.
- By the descent principle, the Novikov conjecture holds for any discrete group whose Cayley graph coarsely embeds into some ℓ^q, recovering and extending the known Banach-space property (H) result.
- The theorem covers spaces that do not coarsely embed into Hilbert space, such as those coarsely embeddable into ℓ^q with q>2, which was the obstruction in earlier Hilbert-space approaches.
- The modified definition of ℓ^p-Roe algebra (with coefficients in ℓ^p(N,H)) is the right one for this argument; the question of whether its K-theory agrees with the earlier Zhang–Zhou ℓ^p-Roe algebra is left open.
Reading between the lines
- The same Bott–Dirac twist ought to work for any Banach space whose unit sphere admits a dimension-independent Lipschitz (or Hölder) map to a Hilbert space; the Mazur map is one instance, so the proof may generalize to other uniformly convex targets.
- The p-independence of K_*(C^p(X)) might hold for a wider class of spaces: if one ℓ^p-coarse Baum–Connes isomorphism holds for a space, the others likely follow by the same twisted-algebra comparison.
- The vector-valued Marcinkiewicz–Zygmund inequality of Theorem 3.7 could be useful for other ℓ^p-index problems where mixed norms appear, beyond the specific coarse-geometric setting.
- A natural test case is a coarsely ℓ^q-embeddable box space of a group that does not have finite asymptotic dimension; verifying the conjecture there would give a concrete non-Hilbert example.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proves (Theorem 1.2) that if a bounded-geometry discrete metric space X coarsely embeds into ℓ^q for some q ∈ [1,∞), then for every p ∈ [1,∞) the ℓ^p-coarse Baum–Connes conjecture holds for X, where the ℓ^p-Roe algebra is the modified one of Definition 3.2 acting on ℓ^p(Z, ℓ^p(N, H)). Section 2 treats the C*-case (p = 2): a Bott–Dirac operator is built from the homogeneous Mazur map Ψ, twisted Roe algebras A(P,E) and their localization algebras are defined, an index map Ind : K_*(C^*(P)) → K_*(A(P,E)) is constructed (Propositions 2.19–2.20), and the twisted assembly map is claimed to be an isomorphism via a localization/Mayer–Vietoris argument (Proposition 2.21, Corollaries 2.26–2.27). Theorem 2.2 (the C*-case) and Theorem 1.2 follow by diagram chasing. Section 3 introduces ℓ^p-Roe algebras with an extra Hilbert-space factor H, proves a vector-valued Marcinkiewicz–Zygmund inequality (Theorem 3.7) using the Grothendieck inequality, and asserts that the same twisted-algebra argument works verbatim for all p.
Significance. If the missing technical steps are supplied, this is a significant result: it extends Yu's coarse Baum–Connes theorem from Hilbert-space-embeddable spaces to the strictly larger class of ℓ^q-embeddable spaces (for q > 2, by Johnson–Randrianarivony), gives a uniform family of ℓ^p-versions, and yields the coarse Novikov conjecture for ℓ^q-embeddable groups. The paper is a fresh derivation with no circularity: the constants L = 1 + q·2^q (Lemma 2.6) and s_n = 2L(dim E_n)^2 in (2.4) are derived explicitly rather than fitted, and the Grothendieck constant K_G in Theorem 3.7 is explicit. The authors honestly flag Question 3.6, which limits comparison with the Zhang–Zhou ℓ^p-Roe algebras. The overall architecture follows [18, 22] faithfully, which makes the unproved product/Mayer–Vietoris steps the main risk rather than the overall strategy.
major comments (3)
- [§2.4, Corollary 2.26] The proof consists of the single assertion K_*(A(P_1(X),E)^R_U) = ∏_{i≥1} K_*(A(P_1(X),E)^R_{U_i}), with no argument. Definition 2.24 requires the E-support of each matrix entry T_s(x,y) to lie eventually in the ε-neighborhood of U and, by Definition 2.23, within the (R+ε)-neighborhood of {f(x), f(y)}; it does not require the support to lie in the neighborhood of the component U_i associated with x or y. When the separation δ satisfies δ < 2(R+ε), a nonzero entry with f(x) near U_i and f(y) near U_j may have E-support in the corridor between the two components, so the Banach algebra A(P_1(X),E)^R_U need not be the product algebra ∏_i A(P_1(X),E)^R_{U_i}. Even when the algebra does decompose, K-theory of Banach algebras does not commute with countable products in general, and a Milnor lim^1 argument would be required; none is given. Since Corollary 2.27, Proposition 2.21, and hence Theorem 2.2 rest on this step, this is load-bearing and must be repaired, for instance by proving the decomposition under a hypothesis relating δ to R (e.g., δ > 2(R+1)) and controlling the leftover corridor terms, together with a lim^1 argument.
- [§2.4, Corollary 2.27] The 'standard Mayer–Vietoris argument' is not written out. The finite-multiplicity cover U_R = ⋃_{x∈X} B_q(f(x), R+1) appearing in the proof of Proposition 2.21 is the only route from the bounded-set isomorphism of Lemma 2.25 to the full twisted assembly map, and it must pass through the unproved product decomposition of Corollary 2.26 and through Mayer–Vietoris sequences for the twisted Banach algebras A(P,E)^R_U and their localization algebras. The intersection terms and the interchange of the limits (r → ∞, R → ∞, and the tail behavior in s) require explicit verification in the ℓ^q-metric setting; as written, Corollary 2.27 is a restatement of the required result rather than a proof.
- [§3.3, proof of Theorem 1.2] The general-p case is dispatched with 'the proof is completely similar' and 'we shall be brief', and Proposition 3.11 is said to follow from the 'same computation' as Proposition 2.19. This omits: the ℓ^p analogue of the Eilenberg swindle (Proposition 2.20), which requires the homotopy and almost-idempotent estimates to hold uniformly with the Marcinkiewicz–Zygmund constant K_G from Theorem 3.7 in the mixed-norm module M_{P,E}; the ℓ^p analogues of Lemma 2.25 and Corollaries 2.26–2.27, which inherit the product-decomposition gap of Corollary 2.26; and the ℓ^p versions of the uniform-in-s estimates of Proposition 2.17(5)–(7) for the Banach-module action. These steps are load-bearing for Theorem 1.2 and should be stated and proved rather than referenced by analogy.
minor comments (6)
- [Lemma 2.7] The displayed estimate ‖B_s(ρ_R φ)‖² ≥ (1/4)‖R²ρ_R φ‖² has the wrong dimension in R (it scales as R⁴), and it is inconsistent with the subsequent bound ‖ρ_R(1+B_s²)^{-1/2}σ‖ ≤ (2/R)(1+s^{-1})‖σ‖, which requires ‖B_s ψ‖ ≳ R‖ψ‖ outside B_q(R). The proof should be corrected to match the cleaner estimate derived in Lemma 2.11.
- [Lemma 2.6] The dimension-free constant is correct: the computation uses the unit-sphere Lipschitz bound ‖ψ(u)−ψ(v)‖₂ ≤ q2^{q−1}‖u−v‖_q together with ‖v_s−u‖_q ≤ 2δ, yielding L = 1 + q·2^q, and I found no hidden dependence on dim E.
- [Abstract and Introduction] The phrase 'coarsely embedds' should read 'coarsely embeds' (it also appears in the statements of Theorem 1.1 and Definition 2.1), and 'Atiya h–Singer' contains a stray space.
- [Proposition 2.21 and proof of Theorem 2.2] The codomain of ev_* in Proposition 2.21 should be K_*(A(P_1(X),E)) rather than K_*(A(P_r(X),E)) to match its use in the diagram, and the limit 'lim_{d→∞}' in the diagram should be 'lim_{r→∞}'.
- [Definition 2.18(4)] The E-support condition on T_s(x,y) is stated only with respect to the ball centered at f(y); since Definition 2.23 later uses the symmetric condition with {f(x), f(y)}, the intended convention in Definition 2.18 should be clarified or symmetrized.
- [Question 3.6] Question 3.6 is honestly posed but leaves open whether K_*(C^p(X)) agrees with the K-theory of the Zhang–Zhou ℓ^p-Roe algebra; the introduction should state explicitly that Theorem 1.2 concerns the modified algebra, so that the comparison with [24] is not overstated.
Circularity Check
No material circularity: the ℓq-embeddable ℓp-coarse Baum–Connes isomorphism is derived from explicit non-fitted estimates (Mazur map, Grothendieck inequality) plus an Eilenberg–swindle index map; the flagged Corollary 2.26 product-decomposition step is an omitted proof (correctness risk), not a self-referential reduction.
full rationale
The derivation chain is not circular. Theorem 1.2 (the ℓp-coarse Baum–Connes conjecture for bounded-geometry spaces coarsely embedding into ℓq) is never assumed as an input. All constants are explicit and either derived in the paper or taken from external theorems: the Mazur-map Lipschitz constant L = 1 + q·2^q is proved in Lemma 2.6 from the external Lipschitz bound for the Mazur map ψ ([16], Weston); the scale s_n = 2L(dim E_n)^2 in line (2.4) is built directly from that L; and the Grothendieck constant K_G enters through the externally stated Grothendieck inequality (Theorem 3.8) and bounds T⊗1 in Theorem 3.7. No parameter is fitted to any subset of the target data; the claimed ev_*-isomorphism is attacked head-on by constructing the twisted index map Ind (Propositions 2.19 and 3.11), the Eilenberg–swindle trivialization of ı_*∘Ind (Proposition 2.20), and the localization/Mayer–Vietoris decomposition of Section 2.4 (Lemmas 2.25, Corollaries 2.26–2.27, Proposition 2.21). Self-citations ([2], [9], [18], [20], [22]) support only strictly weaker or standard statements — coarse Novikov for spaces with property (H), the Hilbert-space-embeddable case, localization-algebra K-theory, and the coarse-disjoint-union reduction; none of them asserts the target ℓq-embeddable ℓp-Baum–Connes isomorphism, so no self-citation chain forces the conclusion. Per the reviewing rule, I flag the following missing support as correctness risks, not circularity: Corollary 2.26 asserts 'K_*(A(P_1(X),E)^R_U) = ∏_{i≥1} K_*(A(P_1(X),E)^R_{U_i})' with no proof that the twisted algebra splits over δ-disjoint components and no Milnor lim^1 check that K-theory commutes with the countable product; Corollary 2.27 invokes a 'standard Mayer–Vietoris argument' without writing it out; and the proof of Theorem 1.2 says the ℓp twisted ev_*-isomorphism 'follows from the same construction and the Mayer–Vietoris argument as the ℓ2-case,' inheriting those gaps. None of these equates the theorem's conclusion to its own hypotheses by construction, so they do not constitute circularity under the stated standards. Score 1: no load-bearing circularity; minor, non-load-bearing self-citations are present.
Assumptions & free parameters
assumptions (6)
- domain assumption Nowak's theorem: for 1 ≤ q ≤ 2, coarse embeddability into ℓ^q is equivalent to coarse embeddability into Hilbert space.
- standard math Yu's localization algebra theorem: K_*(C^*_L(P)) is naturally isomorphic to the K-homology of P.
- standard math Grothendieck inequality (Theorem 3.8).
- standard math Mazur map Lipschitz and Hölder estimates from Weston [16].
- domain assumption The reduction to coarse disjoint unions of finite metric spaces from [18, Section 12.5].
- standard math Roe algebra K-theory is invariant under coarse equivalence, and Rips complexes are coarsely equivalent to X.
invented entities (2)
-
Modified ℓ^p-Roe algebra C^p(X) (Definition 3.2), acting on ℓ^p(Z, ℓ^p(N,H)) instead of ℓ^p(Z, ℓ^p(N)).
-
Twisted Roe algebras A(P,E) and A^p(P,E), and the Bott-Dirac twisted index map.
Cite this review
Pith. "Pith review of $\ell^p$-coarse Baum-Connes conjecture for $\ell^{q}$-coarse embeddable spaces." pith.science (2026). https://pith.science/paper/JIY4ZH4H
@misc{pith2026241115070,
author = {Pith},
title = {Pith review of: $\ell^p$-coarse Baum-Connes conjecture for $\ell^q$-coarse embeddable spaces},
year = {2026},
howpublished = {\url{https://pith.science/paper/JIY4ZH4H}},
note = {Machine review of arXiv:2411.15070}
}
abstract
We prove an $\ell^p$-version of the coarse Baum-Connes conjecture for spaces that coarsely embedds into $\ell^q$-spaces for any $p$ and $q$ in $[1,\infty)$.
Forward citations
Cited by 2 Pith papers
-
$K$-theory of ghostly ideals for $\ell^p$-coarsely embeddable spaces
Bounded-geometry metric spaces coarsely embeddable into ℓ^p have the property that geometric ideals in their Roe algebras have the same K-theory as the corresponding ghostly ideals.
-
Hilbert-Hadamard spaces and the equivariant coarse Novikov conjecture
If a proper Γ-space with equivariant bounded geometry equivariantly and coarsely embeds into an admissible Hilbert-Hadamard space, then the rational analytic equivariant coarse Novikov conjecture holds.
Reference graph
Works this paper leans on
-
[1]
Classifying space for p roper actions and K- theory of group C∗ -algebras
Paul Baum, Alain Connes, and Nigel Higson. Classifying space for p roper actions and K- theory of group C∗ -algebras. In C∗ -algebras: 1943–1993 (San Antonio, TX, 1993) , volume 167 of Contemp. Math. , pages 240–291. Amer. Math. Soc., Providence, RI, 1994
work page 1943
-
[2]
The coarse Novikov conj ecture and Banach spaces with Property (H)
Xiaoman Chen, Qin Wang, and Guoliang Yu. The coarse Novikov conj ecture and Banach spaces with Property (H). J. Funct. Anal. , 268(9):2754–2786, 2015
work page 2015
-
[3]
Yeong Chyuan Chung and Piotr W. Nowak. Expanders are counte rexamples to the ℓp coarse Baum-Connes conjecture. J. Noncommut. Geom. , 17(1):305–331, 2023
work page 2023
-
[4]
The coarse Baum-Connes conjecture for certain relative expanders
Jintao Deng, Qin Wang, and Guoliang Yu. The coarse Baum-Connes conjecture for certain relative expanders. Adv. Math., 424:Paper No. 109047, 52, 2023
work page 2023
-
[5]
M. Gromov. Asymptotic invariants of infinite groups. In Geometric group theory, Vol. 2 (Sussex, 1991) , volume 182 of London Math. Soc. Lecture Note Ser. , pages 1–295. Cam- bridge Univ. Press, Cambridge, 1993
work page 1991
-
[6]
Misha Gromov. Spaces and questions. pages 118–161. 2000. GA F A 2000 (Tel Aviv, 1999). 34 JINMIN W ANG, ZHIZHANG XIE, GUOLIANG YU, AND BO ZHU
work page 2000
-
[7]
On the coarse Baum-Connes conject ure
Nigel Higson and John Roe. On the coarse Baum-Connes conject ure. In Novikov conjec- tures, index theorems and rigidity, Vol. 2 (Oberwolfach, 19 93), volume 227 of London Math. Soc. Lecture Note Ser. , pages 227–254. Cambridge Univ. Press, Cambridge, 1995
work page 1995
-
[8]
William B. Johnson and N. Lovasoa Randrianarivony. lp (p > 2) does not coarsely embed into a Hilbert space. Proc. Amer. Math. Soc. , 134(4):1045–1050, 2006
work page 2006
Show all 24 references
-
[9]
The Novikov conjecture and geometry of Banach spaces
Gennadi Kasparov and Guoliang Yu. The Novikov conjecture and geometry of Banach spaces. Geom. Topol., 16(3):1859–1880, 2012
2012
-
[10]
K-th´ eorie bivariante pour les alg` ebres de Banach et conjecture d e Baum-Connes
Vincent Lafforgue. K-th´ eorie bivariante pour les alg` ebres de Banach et conjecture d e Baum-Connes. Invent. Math. , 149(1):1–95, 2002
2002
-
[11]
Piotr W. Nowak. On coarse embeddability into lp-spaces and a conjecture of Dranishnikov. Fund. Math., 189(2):111–116, 2006
2006
-
[12]
Factorization of linear operators and geometry of Banach sp aces, volume 60 of CBMS Regional Conference Series in Mathematics
Gilles Pisier. Factorization of linear operators and geometry of Banach sp aces, volume 60 of CBMS Regional Conference Series in Mathematics . Conference Board of the Mathematical Sciences, Washington, DC; by the American Mathematical Society, Providence, RI, 1986
1986
-
[13]
Coarse cohomology and index theory on complete Riem annian manifolds
John Roe. Coarse cohomology and index theory on complete Riem annian manifolds. Mem. Amer. Math. Soc. , 104(497):x+90, 1993
1993
-
[14]
Index theory, coarse geometry, and topology of manifolds , volume 90 of CBMS Regional Conference Series in Mathematics
John Roe. Index theory, coarse geometry, and topology of manifolds , volume 90 of CBMS Regional Conference Series in Mathematics . Published for the Conference Board of the Mathematical Sciences, Washington, DC; by the American Mathema tical Society, Provi- dence, RI, 1996
1996
-
[15]
The coarse geometric ℓp-Novikov conjecture for subspaces of nonpositively curved manifolds
Lin Shan and Qin Wang. The coarse geometric ℓp-Novikov conjecture for subspaces of nonpositively curved manifolds. J. Noncommut. Geom. , 15(4):1323–1354, 2021
2021
-
[16]
On the uniform classification of Lp(µ) spaces
Anthony Weston. On the uniform classification of Lp(µ) spaces. In Miniconference on probability and analysis (Sydney, 1991) , volume 29 of Proc. Centre Math. Appl. Austral. Nat. Univ. , pages 231–237. Austral. Nat. Univ., Canberra, 1992
1991
-
[17]
Higher index theory for certain exp anders and Gromov monster groups, I
Rufus Willett and Guoliang Yu. Higher index theory for certain exp anders and Gromov monster groups, I. Adv. Math., 229(3):1380–1416, 2012
2012
-
[18]
Higher index theory , volume 189
Rufus Willett and Guoliang Yu. Higher index theory , volume 189. Cambridge University Press, 2020
2020
-
[19]
Coarse Baum-Connes conjecture
Guo Liang Yu. Coarse Baum-Connes conjecture. K-Theory, 9(3):199–221, 1995
1995
-
[20]
Localization algebras and the coarse Baum-Connes conjecture
Guoliang Yu. Localization algebras and the coarse Baum-Connes conjecture. K-Theory, 11(4):307–318, 1997
1997
-
[21]
The Novikov conjecture for groups with finite asym ptotic dimension
Guoliang Yu. The Novikov conjecture for groups with finite asym ptotic dimension. Ann. of Math. (2) , 147(2):325–355, 1998
1998
-
[22]
The coarse Baum-Connes conjecture for spaces which admit a uniform em- bedding into Hilbert space
Guoliang Yu. The coarse Baum-Connes conjecture for spaces which admit a uniform em- bedding into Hilbert space. Invent. Math. , 139(1):201–240, 2000
2000
-
[23]
Open cones and K-theory for ℓp Roe algebras
Jianguo Zhang. Open cones and K-theory for ℓp Roe algebras. Chinese Ann. Math. Ser. B, 45(6):943–956, 2024
2024
-
[24]
Lp coarse Baum-Connes conjecture and K-theory for Lp Roe algebras
Jianguo Zhang and Dapeng Zhou. Lp coarse Baum-Connes conjecture and K-theory for Lp Roe algebras. J. Noncommut. Geom. , 15(4):1285–1322, 2021. (Jinmin Wang) Institute of Mathematics, Chinese Academy of Sciences Email address : jinmin@amss.ac.cn (Zhizhang Xie) Department of Mat...
2021
Reviewed August 12, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.