REVIEW 3 major objections 4 minor 32 references
Sparse Random Covers and Growth of Torsion in First Homology
T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read This paper constructs random open covers of higher-rank locally symmetric spaces and proves they bound both the minimal generator number and the logarithm of first-homology torsion by $C_X\,\mathrm{vol}(M)\,R^{(1-r)/2}(\log R)^2$, with…
desk verdict Strong new vanishing theorems for homology torsion and generator growth in higher-rank locally symmetric spaces; the manifold results are solid, but the affine-building extension relies on an unverified citation for exotic buildings. 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 central object is the scaffolded Poisson cover: a maximal $1$-separated set of points (the scaffold) with radius-$2$ balls guarantees coverage, and an independent Poisson point process of intensity $1/v(t)$ sprinkles large radius-$\rho(t)$ balls, where $\rho(t)=t+h^{-1}\log\log t+c$ and $v(s)$ is the ball volume in the symmetric space. A large ball is retained wherever it meets the scaffold, and the small scaffold ball is retained only at missed points; the threshold $\rho(t)$ is tuned by the asymptotics $v(s)\simeq e^{hs}s^{(r-1)/2}$. The resulting cover is good—every nonempty finite intersection is contractible—and its nerve models the manifold. The expected number of nerve edges is computed by three terms: Poisson–Poisson pairs via the standard expectation formula for unordered pairs of a Poisson process, scaffold–scaffold edges via a polynomial tail bound on missed centres, and mixed edges via independence between the inner ball and the surrounding annulus. The topological conversion is that a good cover with $E$ edges gives $d(\pi_1M)\leq E$ and $\log|H_1(M;\mathbb Z)_{\mathrm{tors}}|\leq E\log\sqrt3$, so the sublinear edge count is the whole mechanism. This is the only place where higher-rank geometry enters: the exponent $(r-1)/2$ in the volume growth is what makes the edge count decay.
What would settle it
Take a sequence of torsion-free cocompact lattices $\Gamma_i<G$ with injectivity radii $R_i\to\infty$ and compute the ratio $\max\{d(\Gamma_i),\log|H_1(M_i;\mathbb Z)_{\mathrm{tors}}|\}/\bigl(\mathrm{vol}(M_i)\,R_i^{(1-r)/2}(\log R_i)^2\bigr)$; if the limsup is infinite, Theorem 1.1 is false. For the normalized vanishing assertions, it suffices to exhibit any Benjamini–Schramm convergent sequence along which $\log|H_1(M_i;\mathbb Z)_{\mathrm{tors}}|/\mathrm{vol}(M_i)$ does not tend to zero.
Extended reading notes
Core claim
The central claim is that a scaffolded Poisson process yields a good cover of any torsion-free quotient $M=\Gamma\backslash X$ whose nerve has, in expectation, $O(\mathrm{vol}(M))\,R^{(1-r)/2}(\log R)^2$ edges when $R=\mathrm{InjRad}(M)$ is large. Because the nerve of a good cover is homotopy equivalent to $M$, a one-skeleton edge count $E$ bounds both invariants: $d(\Gamma)\leq E$ and $\log|H_1(M;\mathbb Z)_{\mathrm{tors}}|\leq E\log\sqrt{3}$. The paper's quantitative heart, Proposition 3.3, computes the expected edge count: Poisson–Poisson pairs contribute of order $\mathrm{vol}(M)\,t^{-(r-1)/2}(\log t)^2$, missed scaffold centres contribute polynomially small terms, and mixed pairs contribute another logarithmic factor; the only geometric input is the ball-volume asymptotics $v(s)\asymp e^{hs}s^{(r-1)/2}$ of the symmetric space. Choosing $t\approx R/4$ gives Theorem 1.1. For Benjamini–Schramm convergent sequences the same construction runs on a thick region, with the thin part controlled by imported good-cover and injectivity-radius estimates; the building analogue uses uniform ball-volume growth and a uniform lower bound on translation lengths in place of manifold geometry.
Load-bearing premise
The argument depends on imported estimates—bounded-degree good covers, exponential thin-part decay, positive injectivity-radius lower bounds, and uniform ball-volume growth—holding with constants depending only on $X$ (or on the building $B$); if any of those fails for some admissible family, the stated theorems do not follow from this proof.
Editorial extensions
If this is right
- In a normal residual tower of a fixed cocompact higher-rank lattice, logarithmic systole growth upgrades Theorem 1.1 to an explicit bound $\max\{d(\pi_1 M),\log|H_1(M;\mathbb Z)_{\mathrm{tors}}|\} \leq C\,\mathrm{vol}(M)(\log\mathrm{vol}(M))^{(1-r)/2}(\log\log\mathrm{vol}(M))^2$.
- For any torsion-free Benjamini–Schramm convergent sequence of irreducible higher-rank manifolds, the volume-normalized logarithm of first-homology torsion, the generator rank, and the first-homology dimension over every field all vanish simultaneously.
- When $G$ is simple, the Benjamini–Schramm hypothesis is automatic along any sequence of lattices whose covolumes tend to infinity, so the vanishing theorem applies to all such torsion-free sequences without further assumptions.
- The same scaffolded construction proves the analogous vanishing for compact quotients of thick affine buildings of Euclidean rank at least two, including exotic buildings and quotients arising from semisimple groups over local fields.
Reading between the lines
- Editorial inference: the mechanism should transfer to other families of quotients with ball-volume growth of the form $e^{hs}s^{(r-1)/2}$ and a bounded-degree good cover of the thin part, so the results are likely not special to symmetric spaces or affine buildings.
- Editorial inference: the logarithmic factor $(\log R)^2$ probably is an artifact of the chosen radius $\rho(t)=t+h^{-1}\log\log t+c$; a sharper tail bound for missed scaffold points could plausibly remove it, matching the conjectural shape $R^{(1-r)/2}$.
- Editorial inference: the same nerve-edge estimate should have higher-degree analogues, with degree $q$ controlled by $(q+1)$-tuple intersection counts rather than pair counts; the paper only carries out degree one, but nothing in the construction seems degree-specific.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper introduces a 'scaffolded Poisson process' construction to build random good covers of higher-rank locally symmetric spaces and affine buildings. The main theorem (Theorem 1.1) asserts that for every torsion-free lattice with global injectivity radius R large, both the generator number d(Γ) and log|H_1(M;Z)_tors| are bounded by C_X vol(M) R^{(1-r)/2} (log R)^2, where r≥2 is the real rank. The proof samples sparse Poisson balls on the manifold and fills missed regions with small scaffold balls, then uses the Mecke formula, ball-volume estimates, the nerve lemma, and Gabber's inequality to control the expected number of edges in the cover. From this, the paper derives Benjamini–Schramm vanishing of normalized first-homology torsion (Theorem 1.2), a sublinear rank theorem for general orbifolds (Theorem 1.3), and analogous statements for affine buildings (Theorem 1.5). The manifold-theoretic core is a genuine proof: the constants depend only on X and the key inequalities are derived rather than assumed.
Significance. If correct, Theorem 1.1 gives the first general quantitative sublinear bound for generator rank and first-homology torsion in higher-rank locally symmetric spaces under a large injectivity-radius assumption, and Theorem 1.2 answers the Abért–Gelander–Nikolov question on vanishing of normalized first-homology torsion along Benjamini–Schramm convergent sequences. The proof is elegant and largely self-contained: it avoids model-specific geometry and reduces everything to ball-volume growth and a Poisson-process bookkeeping. The scaffolded-cover idea is a genuine improvement over deterministic covers, since it makes the expected nerve sparse without losing coverness. The main caveat is the affine-building extension, whose proof relies on external regularity and volume-growth estimates that are not verified in full generality. The manifold theorems are not affected by this caveat.
major comments (3)
- [Section 7, Lemma 7.1] Lemma 7.1(i) asserts two-sided uniform ball-volume bounds (9) for every locally finite thick affine building admitting a compact quotient. For rank r_j ≥ 2 factors the proof simply cites [30, Theorem 2.4] for regularity and [30, Theorem 5.15] for the annulus estimate, but it does not verify that these theorems apply to all exotic cocompact affine buildings, including non-Moufang and non-strongly-transitive rank-2 cases. Since the Poisson estimates in Section 7 require (9) uniformly in the centre as both an upper and a lower bound, a single factor with different spherical growth would change the exponent t^{(1-r)/2}(log t)^2 and break the Benjamini–Schramm diagonal argument. Please either provide a direct proof of (9) under the stated hypotheses or state explicitly the additional assumptions on the building that make the cited results applicable.
- [Section 4.4, Lemma 4.6] The compact BS-convergent case depends on deep external estimates: the Dobrowolski injectivity-radius bound from [14, Proposition 2.3] and the exponential thin-part decay [14, Theorem D]. The proof uses these only through the trace-field degree d_n and constants said to depend on X, but the precise hypotheses under which these estimates hold are not stated. In particular, the text says 'By Margulis arithmeticity' without recalling the exact form of the theorem needed for irreducible lattices in semisimple groups of rank at least two. Please state the required hypotheses and verify that every compact torsion-free BS-convergent sequence of irreducible lattices in this setting satisfies them.
- [Section 4.2, proof of Theorem 1.2] The noncompact case of Theorem 1.2 uses a diagonal argument to choose t_n so that log t_n · vol((M_n)^{<3ρ(t_n)+b_X})/vol(M_n) → 0. The argument is described only in one sentence, but it is load-bearing because t_n must simultaneously exceed the threshold of Proposition 4.3 and make the thin-part ratio small. Please spell out the construction with an explicit choice of index intervals, for example t_n = j on ranges where the ratio is at most 1/(j log(j+2)), and justify that the resulting t_n can be taken to tend to infinity while preserving all earlier bounds.
minor comments (4)
- [Theorem 1.1] The statement says 'every torsion-free lattice Γ<G' but the proof in Section 3 assumes M is compact; for noncompact finite-volume quotients the global injectivity radius is zero, so the hypothesis InjRad(M) ≥ R_X is never satisfied. It would be clearer to write 'compact torsion-free lattice' explicitly.
- [References [12] and [13]] References [12] and [13] appear to be the same paper by Frączyk, 'Growth of mod-2 homology in higher-rank locally symmetric spaces', listed twice with slightly different formatting. One duplicate should be removed.
- [Section 7, notation] The notation v_-(s), v_+(s) for building ball volumes is useful, but the proof of Lemma 7.1(i) could state explicitly that for the ℓ2 product metric the volume entropy h_B = (∑ h_j^2)^{1/2} follows from a Laplace-method calculation, since the displayed 'after decomposing a product ball' formula is not immediate for the ℓ2 ball.
- [Section 6, edge types] In the random graph G_t, the choice of the particular Poisson point for edges of type (iii) is arbitrary; the text says the estimates do not depend on the choices. It would be helpful to note that the graph connectivity argument uses only that at least one such edge is present for each retained covered vertex, so any deterministic choice works.
Circularity Check
No circularity: the main bounds are derived from volume-growth estimates and Poisson-process calculations, and the author's self-citations are only contextual.
full rationale
The central derivation in Sections 2 and 3 is self-contained and non-circular. The scaffolded Poisson cover is constructed from a deterministic maximal 1-separated set and a Poisson point process; Proposition 3.3 bounds the expected number of nerve edges using the ball-volume ratios of Lemma 2.2 and the Mecke formula. The target quantities d(Gamma) and log|H_1(M;Z)_tors| are then bounded by the nerve edge count via the standard nerve lemma and Gabber's lemma. No parameter is fitted to the target quantities, and no equation defining the construction is equivalent to the claimed bound. The Benjamini-Schramm arguments in Sections 4 and 6 import external results, including results of Fraaczyk-Hurtado-Raimbault, Gelander, and Dobrowolski; importing external theorems is evidence, not circularity, and none of these inputs is the conclusion being proved. The author's self-citations, such as [19], [20], [27], and [32], appear only in background and context remarks and are not load-bearing for the main proofs. The affine-building extension in Section 7 relies on external regularity and volume-growth estimates from Parkinson [30]; even if one worries about the generality of those cited estimates for exotic buildings, that is a correctness or assumption risk, not circular reasoning, because the claimed theorem is not used as an input to its own proof. No circular step can be exhibited by comparing equations or by showing that a fitted parameter has been renamed as a prediction.
Assumptions & free parameters
assumptions (6)
- standard math Uniform ball-volume growth in higher-rank symmetric spaces: v(s) approximately e^{h s} s^{(r-1)/2} for s>=1 (Lemma 2.1, citing Leuzinger [25]).
- standard math Poisson process Mecke formula (Lemma 2.3), nerve lemma for good covers, and Gabber's estimate on H1-torsion (Lemma 2.5).
- domain assumption Gelander homotopy core for noncompact manifolds (Lemma 4.2, [16]) and Gelander core for arbitrary irreducible lattices (Theorem 5.2, [17]).
- domain assumption Fraczyk-Hurtado-Raimbault theorem: bounded-degree good covers for compact arithmetic manifolds and exponential thin-part decay with trace-field degree (Theorem 4.4 and [14, Theorem D]).
- domain assumption Margulis arithmeticity of irreducible higher-rank lattices and Dobrowolski injectivity-radius lower bounds (used in Lemma 4.6).
- domain assumption For affine buildings: regularity of irreducible rank-2 factors (Parkinson [30, Thm 2.4]) and discreteness of positive translation lengths (Bridson [10]), giving uniform ball-volume growth and injectivity radius (Lemma 7.1).
Cite this review
Pith. "Pith review of Sparse Random Covers and Growth of Torsion in First Homology." pith.science (2026). https://pith.science/paper/SNLZSQI3
@misc{pith2026260805037,
author = {Pith},
title = {Pith review of: Sparse Random Covers and Growth of Torsion in First Homology},
year = {2026},
howpublished = {\url{https://pith.science/paper/SNLZSQI3}},
note = {Machine review of arXiv:2608.05037}
}
abstract
We construct random open covers of higher-rank locally symmetric spaces using a construction we call scaffolded Poisson processes. Let $X=G/K$ be a symmetric space of noncompact type and real rank $r\ge2$. We prove that there are constants $C_X,R_X<\infty$, depending only on $X$, such that every torsion-free lattice $\Gamma<G$, with $M=\Gamma\backslash X$ and global injectivity radius $R=\operatorname{InjRad}(M)\ge R_X$, satisfies \[ \max\left\{d(\Gamma), \log\left|H_1(M;\mathbb Z)_{\mathrm{tors}}\right|\right\} \le C_X\mathrm{vol}(M)R^{(1-r)/2}(\log R)^2 \] where $d(\Gamma)$ denotes the minimal size of a generating set. We then prove the corresponding vanishing statements along Benjamini--Schramm convergent sequences. The vanishing of normalized torsion in first homology answers a question of Ab\'ert, Gelander, and Nikolov, and confirms the degree-one vanishing with trivial integral coefficients predicted by a conjecture of Bergeron and Venkatesh in the higher-rank setting. Finally, we prove the analogous statements for affine buildings.
Reference graph
Works this paper leans on
-
[30]
Buildings and hecke algebras.J
James Parkinson. Buildings and hecke algebras.J. Algebra, 297(1):1–49, 2006
work page 2006
-
[1]
On the growth ofL2-invariants for sequences of lattices in Lie groups.Ann
Miklós Abért, Nicolas Bergeron, Ian Biringer, Tsachik Gelander, Nikolay Nikolov, Jean Raimbault, and Iddo Samet. On the growth ofL2-invariants for sequences of lattices in Lie groups.Ann. of Math. (2), 185(3):711–790, 2017
work page 2017
-
[2]
Miklós Abért, Nicolas Bergeron, Mikołaj Frączyk, and Damien Gaboriau. On homology torsion growth.J. Eur. Math. Soc. (JEMS), 27(6):2293–2357, 2025
work page 2025
-
[3]
Rank, combinatorial cost, and homology torsion growth in higher rank lattices.Duke Math
Miklós Abért, Tsachik Gelander, and Nikolay Nikolov. Rank, combinatorial cost, and homology torsion growth in higher rank lattices.Duke Math. J., 166(15):2925–2964, 2017
2017
-
[4]
Point processes, cost, and the growth of rank in locally compact groups
Miklós Abért and Sam Mellick. Point processes, cost, and the growth of rank in locally compact groups. Israel J. Math., 251(1):48–155, 2022. SPARSE RANDOM COVERS AND FIRST HOMOLOGY GROWTH 19
work page 2022
-
[5]
Homology and homotopy complexity in negative curvature
Uri Bader, Tsachik Gelander, and Roman Sauer. Homology and homotopy complexity in negative curvature. J. Eur. Math. Soc. (JEMS), 22(8):2537–2571, 2020
work page 2020
-
[6]
Higher property T and below-rank phenomena of lattices, 2025
Uri Bader and Roman Sauer. Higher property T and below-rank phenomena of lattices, 2025. Preprint
work page 2025
-
[7]
W. Ballmann, M. Gromov, and V. Schroeder.Manifolds of Nonpositive Curvature, volume 61 ofProgress in Mathematics. Birkhäuser Boston, Boston, MA, 1985
work page 1985
Show all 32 references
-
[8]
Torsion homology growth in arithmetic groups
Nicolas Bergeron. Torsion homology growth in arithmetic groups. In Volker Mehrmann and Martin Skutella, editors,European Congress of Mathematics: Berlin, July 18–22, 2016, pages 263–287. European Mathemati- cal Society, Zürich, 2018
2016
-
[9]
The asymptotic growth of torsion homology for arithmetic groups
Nicolas Bergeron and Akshay Venkatesh. The asymptotic growth of torsion homology for arithmetic groups. J. Inst. Math. Jussieu, 12(2):391–447, 2013
2013
-
[10]
Martin R. Bridson. On the semisimplicity of polyhedral isometries.Proc. Amer. Math. Soc., 127(7):2143–2146, 1999
1999
-
[11]
Torsion homology of arithmetic lattices andK2 of imaginary fields.Math
Vincent Emery. Torsion homology of arithmetic lattices andK2 of imaginary fields.Math. Z., 277(3–4):1155– 1164, 2014
2014
-
[12]
Growth of mod-2homology in higher-rank locally symmetric spaces.Duke Math
Mikołaj Frączyk. Growth of mod-2homology in higher-rank locally symmetric spaces.Duke Math. J., 171(2):247–271, 2022
2022
-
[13]
Growth of mod-2 homology in higher-rank locally symmetric spaces.Duke Mathematical Journal, 171(2):247–271, 2022
Mikolaj Fraczyk. Growth of mod-2 homology in higher-rank locally symmetric spaces.Duke Mathematical Journal, 171(2):247–271, 2022
2022
-
[14]
Topological complexity of arithmetic locally symmetric spaces.Invent
Mikołaj Frączyk, Sebastian Hurtado, and Jean Raimbault. Topological complexity of arithmetic locally symmetric spaces.Invent. Math., 244(1):143–164, 2026
2026
-
[15]
Poisson–Voronoi tessellations and fixed price in higher rank.Ann
Mikołaj Frączyk, Sam Mellick, and Amanda Wilkens. Poisson–Voronoi tessellations and fixed price in higher rank.Ann. of Math. (2), to appear. Preprint arXiv:2307.01194
-
[16]
Homotopy type and volume of locally symmetric manifolds.Duke Math
Tsachik Gelander. Homotopy type and volume of locally symmetric manifolds.Duke Math. J., 124(3):459–515, 2004
2004
-
[17]
Volume versus rank of lattices.J
Tsachik Gelander. Volume versus rank of lattices.J. Reine Angew. Math., 661:237–248, 2011
2011
-
[18]
Invariant random subgroups over non-Archimedean local fields.Math
Tsachik Gelander and Arie Levit. Invariant random subgroups over non-Archimedean local fields.Math. Ann., 372(3–4):1503–1544, 2018
2018
-
[19]
On the minimal size of a generating set of lattices in Lie groups.J
Tsachik Gelander and Raz Slutsky. On the minimal size of a generating set of lattices in Lie groups.J. Lie Theory, 30(1):33–40, 2020
2020
-
[20]
A quantitative Selberg’s lemma.Groups Geom
Tsachik Gelander and Raz Slutsky. A quantitative Selberg’s lemma.Groups Geom. Dyn., 2025. Published online first
2025
-
[21]
Cambridge University Press, Cambridge, 2002
Allen Hatcher.Algebraic Topology. Cambridge University Press, Cambridge, 2002
2002
-
[22]
Betti numbers of negatively curved orbifolds with coefficients in arbitrary fields, 2026
Guy Kapon and Raz Slutsky. Betti numbers of negatively curved orbifolds with coefficients in arbitrary fields, 2026. Preprint
2026
-
[23]
Sara Lapan, Benjamin Linowitz, and Jeffrey S. Meyer. Systole inequalities up congruence towers for arithmetic locally symmetric spaces.Comm. Anal. Geom., 31(4):847–878, 2023
2023
-
[24]
Cambridge University Press, Cambridge, 2018
Günter Last and Mathew Penrose.Lectures on the Poisson Process, volume 7 ofInstitute of Mathematical Statistics Textbooks. Cambridge University Press, Cambridge, 2018
2018
-
[25]
Kazhdan’s property (T),L2-spectrum and isoperimetric inequalities for locally symmetric spaces.Comment
Enrico Leuzinger. Kazhdan’s property (T),L2-spectrum and isoperimetric inequalities for locally symmetric spaces.Comment. Math. Helv., 78:116–133, 2003
2003
-
[26]
The cheap embedding principle: Dynamical upper bounds for homology growth, 2025
Kevin Li, Clara Löh, Marco Moraschini, Roman Sauer, and Matthias Uschold. The cheap embedding principle: Dynamical upper bounds for homology growth, 2025. Preprint
2025
-
[27]
On the asymptotic number of generators of high rank arithmetic lattices.Michigan Math
Alexander Lubotzky and Raz Slutsky. On the asymptotic number of generators of high rank arithmetic lattices.Michigan Math. J., 72:465–477, 2022
2022
-
[28]
ApproximatingL2-invariants by their finite-dimensional analogues.Geom
Wolfgang Lück. ApproximatingL2-invariants by their finite-dimensional analogues.Geom. Funct. Anal., 4(4):455–481, 1994
1994
-
[29]
Survey on approximatingL2-invariants by their classical counterparts: Betti numbers, torsion invariants and homology growth.EMS Surv
Wolfgang Lück. Survey on approximatingL2-invariants by their classical counterparts: Betti numbers, torsion invariants and homology growth.EMS Surv. Math. Sci., 3(2):269–344, 2016
2016
-
[31]
On torsion in the cohomology of locally symmetric varieties.Ann
Peter Scholze. On torsion in the cohomology of locally symmetric varieties.Ann. of Math. (2), 182(3):945– 1066, 2015
2015
-
[32]
A product-neighbourhood criterion for fixed price one.arXiv preprint arXiv:2607.20273, 2026
Raz Slutsky. A product-neighbourhood criterion for fixed price one.arXiv preprint arXiv:2607.20273, 2026. Merton College, Merton Street, OX1 4JD, Oxford, United Kingdom Email address:raz.slutsky@maths.ox.ac.uk
2026 arXiv
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.