REVIEW 3 major objections 4 minor
The Gromov--Ros conjecture in complex hyperbolic space
T0 review · 3 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash
Pith's one-line read This paper proves that in every complex dimension m≥2, the finite-perimeter isoperimetric regions of complex hyperbolic space are exactly the geodesic balls, up to isometry and null sets.
desk verdict Serious strategy and a major claim, but the load-bearing trace identity is inconsistent with the displayed matrix; the paper is not sound as written. 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 object is the pointwise trace identity (3.9): for the radial–angular deformations F_{t,φ_j}(ρ,θ)=(e^{tφ_j(θ)}ρ,θ), the sum of the second perimeter derivatives minus n times the radial first derivative equals -−R_{n,x}(β,γ)/(1+x)^2, where x=sinh^2 r and β,γ are the squared horizontal and Reeb components of the measure-theoretic normal. R_{n,x} is an explicit quadratic polynomial that is nonnegative for all n≥4, x≥0 and vanishes exactly when β=γ=0, i.e. when the normal is radial. The identity converts the geometric problem into an algebraic positivity statement on the direction sphere.
What would settle it
Re-run the algebraic expansion of Theorem 3.4 at a specific numerical point, e.g. n=4, x=1, β=γ=1/4, comparing the left-hand side of (3.9) computed from the explicit formulas (3.6)–(3.8) with -−R_{4,1}(1/4,1/4)/4; any mismatch refutes the identity and hence the proof. Alternatively, machine-certify the polynomial identity a^2 S + R_{n,x}=0 over the integers in n,x,β,γ using an integer-arithmetic checker.
Extended reading notes
Core claim
The central claim is that in CH^m with holomorphic sectional curvature -4 (real dimension n=2m≥4), if a finite-perimeter set E and a geodesic ball B have the same positive finite volume, then Per(E)≥Per(B), with equality only when E is a geodesic ball up to ambient isometry and a null set. The argument uses a volume geometric median and exact-volume log-partition polar stretches F_ψ(ρ,θ)=(e^{ψ(θ)}ρ,θ), which are globally bi-Lipschitz and have Jacobian e^{nψ}. The reduced-boundary area formula expresses the perimeter of the stretched set through the ambient cofactor, and summing second derivatives over the coordinate functions of the direction sphere yields the identity -−R_{n,x}(β,γ)/(1+x)^2
Load-bearing premise
The proof rests on the pointwise trace identity (3.9), which is verified by direct expansion rather than a fully certified computation; a sign or coefficient error in that identity would break the step forcing the boundary normal to be radial.
Editorial extensions
If this is right
- If the theorem is correct, the isoperimetric profile of CH^m is explicit: I_n(V)=ω_{n-1}(nV/ω_{n-1})^{(n-1)/n}√(1+(nV/ω_{n-1})^{2/n}), achieved only by geodesic balls.
- Every prescribed positive volume admits a bounded finite-perimeter minimizer, and the classification applies without any smoothness or regularity assumptions on the boundary.
- Smooth stable constant-mean-curvature domains are geodesic balls, so no non-spherical stable CMC hypersurfaces exist in complex hyperbolic space for m≥2.
- The sharp convex Bergman contraction inequality, the associated norm embeddings, and the Faber–Krahn concentration inequality become unconditional on the unit ball in every dimension m≥2.
- The SU(m,1) Lieb–Wehrl entropy inequality follows with its full equality cases; in the high-weight flat limit it recovers holomorphic Gaussian hypercontractivity.
Reading between the lines
- The trace identity is essentially a Hessian-trace positivity statement, so a natural extension would be to seek analogous spherical coordinate trace polynomials for quaternionic and Cayley hyperbolic spaces; the polynomial structure there is likely different.
- The log-partition exact-volume correction and its Gaussian covariance interpretation suggest a probabilistic route to isoperimetric inequalities in which the equilibrium measure is characterized by covariance nonnegativity.
- If the trace identity were ever found to contain a sign or coefficient error, the radiality step would break immediately, but the explicit algebraic certificate offered in the paper makes this local and checkable rather than opaque.
- The analytic consequences (Bergman contraction, Faber–Krahn, Lieb–Wehrl) now rest on a single geometric input; any future counterexample to the geometric theorem would automatically propagate to those inequalities.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims to prove the Gromov–Ros conjecture for complex hyperbolic space CH^m, m≥2, in the finite-perimeter BV setting: every isoperimetric region of positive finite volume is, up to null sets and ambient isometries, a geodesic ball, and equality in the isoperimetric inequality holds only for geodesic balls. The proof is organized around a pointwise trace identity obtained by summing second derivatives of the perimeter of exact-volume radial-angular polar stretches over the coordinate functions of the direction sphere. The resulting polynomial R_{n,x}(β,γ) is explicitly shown to be nonnegative and to vanish only for radial normals. The same trace identity is then used to prove smooth rigidity of stable constant-mean-curvature hypersurfaces. The geometric theorem is applied to remove previously conditional hypotheses in convex Bergman contraction, Faber–Krahn, and Lieb–Wehrl inequalities on the unit ball, with a Gaussian–Fock scaling limit.
Significance. If the proof is correct, this settles a long-standing conjecture and provides a clean explanation of the anisotropic phenomenon: the volume density becomes Euclidean in the radius ρ, while the trace of the second variations detects nonradiality through an explicit positive polynomial. The paper has notable strengths: the exact-volume log-partition family is elegant, the nonnegativity proof is elementary, and Remark 3.9 gives an exact polynomial certificate in Z[n,x,β,γ]. The direct reduced-boundary argument avoids a priori regularity of isoperimetric boundaries, which is a genuine advance. The analytic consequences are conditional only on the geometric theorem and are correctly identified as such. However, as written, the central computation contains displayed scaling inconsistencies that prevent a reader from verifying the decisive identity from the printed equations; these must be corrected before the proof can be accepted.
major comments (3)
- [§2.7, Eq. (2.10) and §3.1, Eqs. (3.4)–(3.5)] The displayed differential matrix (2.10) and the inversion in §3.1 are mutually inconsistent, and neither matches the actual derivative of F_{t,φ}(ρ,θ)=(e^{tφ}ρ,θ). The correct off-diagonal entries in the orthonormal polar frames are u t a_t^{-1/2} h (horizontal) and u t (a a_t)^{-1/2} q (Reeb), not u t a_t^{1/2} h and u t (a a_t)^{1/2} q as printed. Solving L^T w=ν from (2.10) gives w_H = u^{-1}(v - α t a_t h/√a) and a Reeb component involving (τ - α t a_t q), which is not the K(t) stated in (3.5). Equation (3.5) itself uses yet another scaling. Since (3.9) is derived by expanding this K(t), the decisive step is not reproducible as printed. A corrected pullback-metric computation with ξ=ρdφ does reproduce the stated R_{n,x}, so the identity appears true, but the manuscript must be repaired at this load-bearing point.
- [Appendix A, Eqs. (A.1)–(A.3)] The appendix claims an independent verification of the trace identity, but the covector used is not the pullback of the deformation. With ϑ_T=ρ√a η, one has ρ dφ = ∑ h_i ϑ_i + (q/√a) ϑ_T, not q√a ϑ_T. Correspondingly, the coefficient of ϑ_0⊙ξ in (A.2) should be 2/√a, not 2√a. As printed, the appendix does not verify (3.9). If these two scaling errors are corrected, the computation does recover R_{n,x}; the appendix is therefore fixable, but it currently does not serve its stated purpose.
- [§6.5, proof of Theorem 6.9] The statement 'complete and homogeneous, hence cocompact' is not a valid inference: homogeneity of the isometry group does not by itself make a noncompact manifold cocompact. Complex hyperbolic space does admit compact quotients by cocompact lattices, so the appeal to Ritoré [20, Thm 4.25] can be justified, but the sentence should cite the existence of a cocompact lattice or quote Ritoré's hypotheses accurately. This matters because the all-volume existence of isoperimetric regions is imported from that theorem.
minor comments (4)
- [§3.1] The notation w_N, w_H, w_T is used ambiguously: after solving L^T w=ν, w is a covector in the target frame, while ν is a source vector. A brief clarification would help.
- [§3.4] In the display before Eq. (3.8), the notation ‘E’ is used both as the set E and as the coefficient α²/a. This is typographically confusing; rename one of them.
- [Theorem 3.4] The phrase 'direct expansion' at the end of the proof is trusted, but given the scaling errors elsewhere, it would be useful to include the coefficient-by-coefficient computation promised in Remark 3.9, perhaps as an ancillary file or a longer appendix.
- [§6.5] The proof of Theorem 6.9 says equality in (6.19) implies E is itself an isoperimetric region and then cites Ritoré's Lemma 4.27 for a bounded representative. Please state the precise boundedness hypothesis of that lemma, since the region E is only assumed to have finite perimeter and positive finite volume.
Circularity Check
No significant circularity: the central trace identity is derived from the metric and the cofactor formula, and no fitted or predicted quantity reduces to an input by construction.
full rationale
The paper's central derivation is self-contained rather than circular. The pointwise trace identity (3.9) is obtained in Theorem 3.4 by direct expansion from the polar metric (2.1), the differential matrix (2.10), the cofactor identity (3.3), and the surface Jacobian expansion (3.5); the polynomial R_{n,x} is then shown nonnegative with radial-only vanishing in Section 3.5. No parameter is fitted to data, and no 'prediction' is equivalent to an input: the volume geometric median is defined from the set itself, and the exact-volume log-partition family is constructed so that each competitor has the same volume by formula (6.1) and Proposition 6.3, not by matching the desired perimeter. The finite-perimeter classification uses global minimality, the reduced-boundary transport formula, BV U(m)-invariance, and a self-contained one-dimensional endpoint comparison. The paper explicitly states that prior partial self-cited results are not used as inputs: 'None of these partial results is used as an input below' (Section 1), and Remark 7.10 similarly says several self-cited stability results are 'not used in the present proof.' The external dependence on Ritoré [20] supplies existence and bounded-representative results for isoperimetric regions in complete cocompact manifolds; this is an independent external theorem, not a self-citation, and it does not assume the target geodesic-ball classification. The analytic consequences in Section 7 are explicitly conditional theorems of Li and Su [13] and Singh [25] whose geometric hypothesis is discharged by Theorem 6.9; using the proved theorem to make earlier conditional statements unconditional is not circular. Any concern about the internal correctness of the matrix inversion or the sign of the trace identity (e.g., a possible inconsistency between (2.10), (3.5), and the Appendix A computation) is a proof-checking or correctness matter, not a circularity of the kind defined here. Under the required standard, the derivation chain does not reduce to its own inputs, so the appropriate score is 0.
Assumptions & free parameters
assumptions (4)
- domain assumption Polar metric in CH^m has the form g=dr^2 + sinh^2 r g_H + sinh^2 r cosh^2 r η^2 under HolSec=−4 (Eq. 2.1).
- domain assumption There exists an isoperimetric region for every volume, and every minimizer has a bounded representative (Ritoré [20, Thm 4.25, Lemmas 4.26–4.27]) applied to CH^m as a complete cocompact manifold.
- standard math The reduced-boundary area formula for bi-Lipschitz maps (Maggi [16] §§11.2,17.1) and the BV slicing theorem (Ambrosio–Fusco–Pallara [1] §3.11).
- domain assumption The analytic consequences in Section 7 assume that the conditional theorems of Li–Su [13] and Singh [25] are valid once the isoperimetric input is supplied.
Cite this review
Pith. "Pith review of The Gromov--Ros conjecture in complex hyperbolic space." pith.science (2026). https://pith.science/paper/3DUGV3NF
@misc{pith2026260720055,
author = {Pith},
title = {Pith review of: The Gromov--Ros conjecture in complex hyperbolic space},
year = {2026},
howpublished = {\url{https://pith.science/paper/3DUGV3NF}},
note = {Machine review of arXiv:2607.20055}
}
abstract
Let $\CH^m$, $m\ge2$, denote complex hyperbolic space with holomorphic sectional curvature $-4$. We prove that every finite-perimeter isoperimetric region in $\CH^m$ is a geodesic ball. The proof is based on exact-volume radial-angular deformations centered at a volume geometric median. A second-variation trace identity reduces the problem to the positivity of an explicit polynomial in the horizontal and Reeb components of the measure-theoretic normal. This forces the reduced boundary to have radial normal almost everywhere, from which a BV radiality argument and a one-dimensional comparison yield the classification by geodesic balls. The same trace identity also shows that every smooth bounded domain that is critical and stable for perimeter under a fixed-volume constraint is a geodesic ball. As consequences, previously conditional weighted-Bergman contraction, Faber--Krahn, and Lieb--Wehrl inequalities on $\mathbb B^m$ become unconditional in every complex dimension $m\ge2$, with their equality cases. We also describe a basis-free log-partition formulation of the volume correction and its Gaussian covariance interpretation; in the high-weight flat limit, the resulting Bergman contractivity recovers holomorphic Gaussian hypercontractivity.
Reviewed August 1, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.