Bounded cohomology of diffeomorphism groups of higher dimensional spheres

Zixiang Zhou

REVIEW 2 major objections 6 minor 6 references

Bounded cohomology of diffeomorphism groups of higher dimensional spheres

T0 review · 2 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read For n≥4, the orientation-preserving diffeomorphism groups of the n-sphere have zero bounded cohomology in every positive degree.

desk verdict Significant theorem, but a load-bearing gap in Lemma 2.7's orbit-transitivity argument makes the proof incomplete; still worth refereeing. read the letter →

arxiv 2411.14059 v1 pith:ONHVFMPH submitted 2024-11-21 math.GT

Zixiang Zhou This is my paper · ORCID

classification math.GT MSC 57S0520J06

keywords boundedcohomologydiffeomorphismgroupshigher-dimensionalspheresacyclicitysemi-simplicialcomplexesgenericrelationscoamenablesubgroupsdifferentialtopology

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proves that for every n≥4 and every 1≤r≤∞, the group Diff^r_+(S^n) of orientation-preserving C^r diffeomorphisms of the n-sphere is boundedly acyclic, meaning H^k_b(Diff^r_+(S^n))=0 for every k>0. This answers a question raised in recent work on transformation groups of Euclidean spaces and discs, extending their vanishing results to higher-dimensional spheres. If correct, the theorem implies that the comparison map from the bounded cohomology of the discrete diffeomorphism group to its ordinary cohomology is zero, so every nontrivial positive-degree ordinary class is unbounded. The proof works by showing that punctured-sphere diffeomorphism groups are boundedly acyclic, using a semi-simplicial complex generated by a generic relation on germs of arcs and differential-topology lemmas that exploit the freedom available in dimensions at least 4.

What carries the argument

The argument is carried by four devices. (1) A generic relation ⊥ on germs of arcs, which produces a boundedly acyclic semi-simplicial complex (Lemma 2.3). (2) A spectral-sequence criterion (Lemma 1.5) that computes the bounded cohomology of a group acting on such a complex from the bounded acyclicity of the quotient and of all stabilizers. (3) A coamenability lemma (Lemma 1.6) that lets one replace punctures by tubular neighborhoods without changing bounded cohomology. (4) Lemma 1.16, the main differential-topology lemma, which asserts that for n≥4 a compactly supported diffeomorphism of $S^{{n-1}}$×I can be altered to agree with a prescribed germ on the segment 0×I while becoming trivial on any finite union of disjoint disks and arcs avoiding that segment and its image. Lemma 1.16 supplies the independent control of germs along arcs that makes the stabilizer groups tractable, and it is proved using approximation by embeddings, transversality, and isotopy extension in high dimensions.

What would settle it

Exhibiting a concrete pair (f,K) in dimension 4 with f∈Diff^r_c($S^{3}$×I) and K a finite union of disks and arcs disjoint from 0×I and f(0×I) for which no such φ exists would break the stabilizer argument; alternatively, computing a single nonzero class in H^k_b(Diff^r_+($S^{4}$)) for some k>0 would refute the theorem itself.

Watch

Extended reading notes

Core claim

The central result is Theorem 2.1: for n≥4 and 1≤r≤∞, H^k_b(Diff^r_+(S^n))=0 for all k>0. The proof reduces the sphere group to the groups Diff^r_c(S^n−F_k) for finite sets F_k, which are coamenable to subgroups preserving tubular neighborhoods of F_k. Bounded acyclicity of these subgroups is established by constructing, for each k, a semi-simplicial complex Y_• whose vertices are germs of diffeomorphisms along k disjoint arcs, with a generic relation defined by disjointness of the arcs. The quotient complex is shown to be the boundedly acyclic complex of pairwise-disjoint tuples, and every stabilizer is shown to be boundedly acyclic via a chain of differential-topology reductions culminating in Lemma 1.16. A spectral sequence argument then transfers these acyclicities to Diff^r_+(S^n).

Load-bearing premise

The proof rests on the main differential-topology lemma (Lemma 1.16): in dimension at least 4, any compactly supported diffeomorphism of $S^{{n-1}}$×I can be modified to have a prescribed germ on the segment 0×I while becoming trivial on any finite union of disks and arcs disjoint from that segment and its image; if this lemma fails, the bounded acyclicity of the stabilizers, and hence of Diff^r_+(S^n), collapses.

Editorial extensions

If this is right

H^k_b(Diff^r_+(S^n))=0 for every k>0, n≥4, and 1≤r≤∞.
The full group Diff^r(S^n) is boundedly acyclic as well, because the orientation-preserving subgroup is coamenable in it.
The comparison map H^•_b(BDiff^δ_r(S^n)) → H^•(BDiff^δ_r(S^n)) is zero, so every nontrivial positive-degree class in the ordinary cohomology of the discrete group is unbounded.
For a simply connected manifold V of dimension at least 3 with nonempty boundary, deleting finitely many disks from V×I yields a manifold V_k with Diff^r_c(V_k) boundedly acyclic.

Reading between the lines

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

Editorial: the same generic-relation construction would likely prove bounded acyclicity for diffeomorphism groups of other high-dimensional manifolds with boundary, not only spheres, if the main germ lemma can be adapted.
Editorial: the theorem sharpens the known contrast with surface diffeomorphism groups, which admit infinite-dimensional spaces of quasimorphisms; the paper's introduction cites this contrast but does not explore it.
Editorial: the paper notes its methods 'seem to be hopeful' for the homeomorphism group Homeo_+(S^n) in higher dimensions, so a concrete next step is to carry the argument over to homeomorphisms.
Editorial: Lemma 1.16 is a strong localization statement in its own right and may be reusable in other problems about high-dimensional diffeomorphism groups, such as studying the bounded cohomology of groups of diffeomorphisms of manifolds with boundary.

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

The theorem is exactly the kind of result worth proving: full bounded acyclicity for Diff^r_+(S^n) when n≥4, answering a question from FFMNK24. If correct this is a real step forward, since prior work only had low-degree vanishing on spheres and the Euclidean/disk cases. The strategy is sensible—reduce to coamenable G_k, use a generic-relation semisimplicial complex, identify the quotient with a configuration complex, and apply Monod–Nariman's spectral-sequence lemma. The adaptation to higher spheres is nontrivial, and the main differential-topology lemma (Lemma 1.16) is a plausible and interesting tool.

The problems are in the execution. The stress-test note is right: Lemma 2.7, the engine for both the quotient identification and the stabilizer computation, prescribes the conjugating diffeomorphism on the model arc α_q^{-1}(y_q^0×I) rather than on the image arc g_0(α_q^{-1}(y_q^0×I)). To arrange h g_0 = g'_0 on that arc, you need h to have the specified germ on g_0(A), not on A. Since g_0 is not assumed to fix A, the stated condition does not do the job. This is not cosmetic—it breaks the induction that identifies Y_•/G_k with Z_• and feeds the stabilizer acyclicity. The proposed repair (apply Lemma 1.16 with 0×I replaced by g_0(A), or first move g_0(A) to a standard arc via Lemma 1.17) looks plausible, but it is not written.

Lemma 1.16 itself has issues. In Step 1, ξ is 1 on a neighborhood of 0×I, so h equals f there; it does not become a smooth approximation of f on that germ. And the assertion that h and f agree on K is unsupported, since on K (where ξ=0) h equals g, and g is not constrained to agree with f on K. These may be repairable with a swap of ξ and 1−ξ and a careful choice of g, but as written the lemma is not proven. The same pattern appears again in Section 3 (Lemma 3.5), which follows the same third-step argument.

To be clear, I do not see circularity or fitting; the citation pattern is appropriate, and the author engages the literature honestly. This is a promising preprint, not a completed proof. A serious referee should be sent in, but they should be asked to focus on Lemma 2.7 and the smoothing/agreement claims in Lemma 1.16. If those can be fixed, the paper would be a strong contribution.

Referee Report

2 major / 6 minor

Summary. The paper claims that Diff^r_+(S^n) is boundedly acyclic for n≥4 and 1≤r≤∞, answering a question of Fournier-Facio, Monod, Nariman, and Kupers for spheres of dimension at least four. The proof reduces the problem to the bounded acyclicity of subgroups G_k that are trivial in a neighborhood of k+1 disjoint disks, constructs a semi-simplicial complex Y_• whose vertices are germs of diffeomorphisms along model arcs, and applies a spectral-sequence criterion (Lemma 1.5) to conclude that G_k is boundedly acyclic. A generalization to Diff^r_c(V_k) for simply connected V of dimension ≥3 is stated in Section 3.

Significance. If the main theorem is correct, it is a substantial step in the bounded cohomology of diffeomorphism groups: it yields the vanishing of H^k_b(Diff^r_+(S^n)) for all k>0 and implies that every nontrivial positive-degree class in H^•(BDiff^δ_r(S^n)) is unbounded (Corollary 0.2). The strategy is coherent and builds on established tools (Monod–Nariman's Lemma 1.5, Monod's coamenability results, and the algebraic criterion of Campagnolo–Fournier-Facio–Lodha–Moraschini). The paper is also transparent about the role of each technical lemma. However, the proof as written contains two localized but load-bearing gaps: an interpolation error in Step 1 of the main differential-topology lemma (Lemma 1.16) and a mismatch between the prescribed and required germs in the orbit-transitivity argument (Lemma 2.7). Both appear repairable, but they are not merely cosmetic.

major comments (2)

[Lemma 1.16, Step 1] The interpolation h(x)=π(ξ(x)f(x)+(1−ξ(x))g(x)) is defined with ξ≡1 near 0×I and ξ≡0 outside U. As written, h equals f near 0×I, so it does not acquire a smooth germ there, and h equals g outside U, so it does not agree with f on K (since K can and should be chosen disjoint from U). Both assertions "h has a smooth germ at 0×I" and "h and f have the same germ at K" are therefore unsupported. Swapping the roles of ξ and 1−ξ and choosing U disjoint from K would give h=g near 0×I and h=f on K, repairing the step; as it stands, the proof of the main differential-topology lemma contains a genuine error at a load-bearing point.
[Lemma 2.7] In the induction step the proof applies Lemma 1.16 to obtain φ_q such that h=α^{-1}_q φ_q α_q has the same germ as g'_0 g^{-1}_0 on A:=α^{-1}_q(y_q^0×I). To arrange that h g0 and g'_0 have the same germ on A, one needs h to be prescribed on the image arc g0(A), namely h|_{g0(A)} = g'_0∘g^{-1}_0|_{g0(A)}. The stated condition prescribes h only on A, and since g0 need not fix A, it gives no control over h g0|_A. The set K in the same sentence also omits g0(A), so the argument does not establish the desired reduction. This gap is structural because Lemma 2.7 is used in Lemma 2.8 to identify stabilizers and in Lemma 2.9 to identify Y_•/G_k with Z_•, and hence feeds directly into the proof of Theorem 2.1. A repair would require a version of Lemma 1.16 applied to the arc g0(A), or an additional isotopy step moving g0(A) to a standard arc before applying the current argument; neither is present.

minor comments (6)

[Abstract and Introduction] The abstract cites [FNS24] while the introduction cites [FFMNK24] for the same question; the reference label should be consistent.
[Lemma 1.10 proof] The convex combination is missing a parenthesis: it should read π_N(ξ(x)f'(x)+(1−ξ(x))f(x)).
[Lemma 1.16, Step 2] The expression "∞< im(c)" should read "∞∉im(c)" or "∞ is not in im(c)".
[Lemma 1.16, Step 3] In the displayed formula for D(H^{-1}_3 f)(0,t), the subscript s in B_s(t) is not defined; it should presumably be B_1(t).
[Definition 2.4] The union in the definition of Λ_k is written as S_{0≤i≤k} B_k; it should be ⋃_{0≤i≤k} B_i.
[Lemma 2.8] "p−stabilizers" should be "p-stabilizers" for consistency with the rest of the paper.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the bounded acyclicity theorem is derived from external differential-topology and bounded-cohomology results, with no fitted inputs or self-citation chain.

full rationale

The paper's derivation is a standard chain of external results: it imports the semi-simplicial bounded-cohomology criterion from Monod–Nariman (Lemma 1.5), coamenability criteria from Monod (Propositions 1.2 and 1.3), the algebraic vanishing criterion from Campagnolo–Fournier-Facio–Lodha–Moraschini (Lemma 1.1), and classical differential-topology facts from Hirsch (Propositions 1.7, 1.8, 1.9, 1.10, 1.14, 1.15). The core original input, Lemma 1.16, is a differential-topology statement about constructing a diffeomorphism with prescribed germs; its proof is an explicit construction using approximation, isotopy extension, and transversality. No parameter is fitted to data, no target cohomology group is assumed in the hypothesis of any lemma, and the conclusion that Diff^r_+(S^n) is boundedly acyclic is not used to prove any of the ingredients. The author has no overlapping prior work cited as a load-bearing uniqueness theorem. The possible weaknesses identified by the reader, such as the interpolation in Lemma 1.16 Step 1 and the conjugation condition in Lemma 2.7, are internal proof gaps or correctness risks, not circularity: even if those arguments were repaired, the structure of the proof would remain a derivation from external theorems rather than a reduction of the theorem to its own assumptions. Overall circularity score is 0.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

No free parameters or invented entities. The proof relies on standard theorems from differential topology, the bounded cohomology spectral sequence and generic relation results of Monod-Nariman, coamenability results of Monod, and the algebraic criterion of Campagnolo et al. These are treated as axioms from prior literature.

assumptions (7)

standard math Hirsch's approximation, embedding, transversality and isotopy extension theorems (Propositions 1.7-1.10, Lemma 1.15)
Used throughout Lemma 1.16 and Lemma 1.17 to perturb maps, approximate by smooth maps, and extend isotopies with controlled support.
standard math [MN23, Theorem 3.3] spectral sequence for group actions on boundedly acyclic complexes (Lemma 1.5)
Main tool to compute H^•_b(G) from the quotient complex and the stabilizers of the action.
standard math [MN23, Proposition 3.2] the complex associated to a generic relation is boundedly acyclic (Lemma 2.3)
Used to prove the quotient complex Z_• and hence Y_•/G_k is boundedly acyclic.
standard math [CFFLM23, Theorem 1.3] existence of Z-commuting conjugates implies bounded acyclicity (used in Lemma 1.1)
Gives bounded acyclicity of Diff^r_c(M×R), which is the model for stabilizer groups.
standard math [Mon22, Propositions 10 and 11] coamenability and injectivity of bounded cohomology (Propositions 1.2 and 1.3)
Used to transfer bounded acyclicity from subgroups to overgroups and to relate fat-point stabilizers to the groups G_k.
domain assumption The domain assumption n≥4 and 1≤r≤∞
Transversality of arcs requires dimension at least 4; the author notes n=2,3 are not covered.
domain assumption For Theorem 3.1, V is a smooth simply connected manifold of dimension at least 3 with nonempty boundary
The generalization to Diff^r_c(V_k) assumes simple connectivity and boundary conditions to control the frame bundle and arc isotopies.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Bounded cohomology of diffeomorphism groups of higher dimensional spheres." pith.science (2026). https://pith.science/paper/ONHVFMPH

@misc{pith2026241114059,
  author       = {Pith},
  title        = {Pith review of: Bounded cohomology of diffeomorphism groups of higher dimensional spheres},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ONHVFMPH}},
  note         = {Machine review of arXiv:2411.14059}
}

abstract

In this paper we prove the vanishing of the bounded cohomology of $\text{Diff}^r_+(S^n)$ with real coefficients when $n\geq 4$ and $1\leq r\leq \infty$. This answers the question raised in \cite{FNS24} for $\geq 4$ dimensional spheres.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

6 extracted references · 2 canonical work pages

[1]

Quasi-morphisms on surface diffeomorphism groups

Jonathan Bowden, Sebastian Wolfgang Hensel, and Richard Webb. Quasi-morphisms on surface diffeomorphism groups. J. Amer. Math. Soc. , 35(1):211--231, 2022

2022
[2]

An algebraic criterion for the vanishing of bounded cohomology

Caterina Campagnolo, Francesco Fournier-Facio, Yash Lodha, and Marco Moraschini. An algebraic criterion for the vanishing of bounded cohomology. arxiv:2311.16259 , 2023

arXiv 2023
[3]

The bounded cohomology of transformation groups of euclidean spaces and discs

Francesco Fournier-Facio, Nicolas Monod, Sam Nariman, and Alexander Kupers. The bounded cohomology of transformation groups of euclidean spaces and discs. arxiv:2405.20395 , 2024

arXiv 2024
[4]

Morris W. Hirsch. Differential topology , volume No. 33 of Graduate Texts in Mathematics . Springer-Verlag, New York-Heidelberg, 1976

1976
[5]

Bounded and unbounded cohomology of homeomorphism and diffeomorphism groups

Nicolas Monod and Sam Nariman. Bounded and unbounded cohomology of homeomorphism and diffeomorphism groups. Invent. Math. , 232(3):1439--1475, 2023

work page 2023
[6]

Lamplighters and the bounded cohomology of T hompson's group

Nicolas Monod. Lamplighters and the bounded cohomology of T hompson's group. Geom. Funct. Anal. , 32(3):662--675, 2022

work page 2022

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [1]

Quasi-morphisms on surface diffeomorphism groups

Jonathan Bowden, Sebastian Wolfgang Hensel, and Richard Webb. Quasi-morphisms on surface diffeomorphism groups. J. Amer. Math. Soc. , 35(1):211--231, 2022

2022

[2] [2]

An algebraic criterion for the vanishing of bounded cohomology

Caterina Campagnolo, Francesco Fournier-Facio, Yash Lodha, and Marco Moraschini. An algebraic criterion for the vanishing of bounded cohomology. arxiv:2311.16259 , 2023

arXiv 2023

[3] [3]

The bounded cohomology of transformation groups of euclidean spaces and discs

Francesco Fournier-Facio, Nicolas Monod, Sam Nariman, and Alexander Kupers. The bounded cohomology of transformation groups of euclidean spaces and discs. arxiv:2405.20395 , 2024

arXiv 2024

[4] [4]

Morris W. Hirsch. Differential topology , volume No. 33 of Graduate Texts in Mathematics . Springer-Verlag, New York-Heidelberg, 1976

1976

[5] [5]

Bounded and unbounded cohomology of homeomorphism and diffeomorphism groups

Nicolas Monod and Sam Nariman. Bounded and unbounded cohomology of homeomorphism and diffeomorphism groups. Invent. Math. , 232(3):1439--1475, 2023

work page 2023

[6] [6]

Lamplighters and the bounded cohomology of T hompson's group

Nicolas Monod. Lamplighters and the bounded cohomology of T hompson's group. Geom. Funct. Anal. , 32(3):662--675, 2022

work page 2022