Quantitative stability of critical points for the nonlocal-Sobolev inequality in Heisenberg group

Jialin Wang; Jijie Xu; Shuijin Zhang

REVIEW 4 major objections 4 minor 1 cited by

Quantitative stability of critical points for the nonlocal-Sobolev inequality in Heisenberg group

T0 review · 4 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper proves that near-solutions of the nonlocal Sobolev Euler-Lagrange equation on the Heisenberg group are linearly close to the explicit optimizer family, and that the same control holds for weakly interacting bubble sums when the…

desk verdict Plausible and clearly stated extension of quantitative stability to the nonlocal Sobolev inequality on the Heisenberg group, but the supplied text is too corrupt to verify the proof, and the abstract omits the coercivity/nondegeneracy hypotheses that the linear rate depends on. read the letter →

arxiv 2508.08614 v2 pith:2D7V5WAJ submitted 2025-08-12 math.AP

Shuijin Zhang , Jijie Xu , Jialin Wang This is my paper · ORCID

classification math.AP MSC 35J6043A8035R09

keywords quantitativestabilitynonlocalSobolevinequalityHeisenberggroupHardy-Littlewood-SobolevFolland-SteinEuler-Lagrangeequationbubblesolutionsofcriticalpoints

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

The paper asks how rigid the nonlocal Sobolev inequality on the Heisenberg group is, once a function is almost a solution of its Euler-Lagrange equation. The central claim is that the gradient distance $\delta(u)=\|\nabla_H u-\nabla_H U_{\lambda,\zeta}\|_{L^2}$ from $u$ to the explicit optimizer family is bounded by a constant times $\Gamma(u)$, the dual norm of the Euler-Lagrange residual. This turns a compactness statement into a quantitative, linear one: functions with small residual cannot drift away from the optimizer manifold. The same linear control is claimed for sums of widely separated bubbles when the homogeneous dimension is $Q=4$. If true, the optimizer family is the only place where near-critical functions can concentrate, and the rate of approach is governed exactly by the equation's residual.

What carries the argument

The argument is carried by the Euler-Lagrange residual and its dual norm $\Gamma(u)$, together with the local coercivity of the linearized operator around each optimizer $U_{\lambda,\zeta}$. Nondegeneracy means that on the subspace orthogonal to the tangent directions generated by varying the scale $\lambda$ and the Heisenberg translation $\zeta$, the linearized operator has a positive spectral gap; that gap is what converts a small residual $\Gamma(u)$ into a small gradient gap $\delta(u)$. The bubble family $U_{\lambda,\zeta}$ supplies the explicit reference set relative to which the distance $\delta(u)$ is measured.

What would settle it

Take a one-parameter family $u_\varepsilon = U_{\lambda,\zeta} + \varepsilon v$ with $v$ orthogonal to the tangent directions of the optimizer family and compute the ratio $\Gamma(u_\varepsilon)/\delta(u_\varepsilon)$ as $\varepsilon\to 0$; if the ratio tends to zero for some $v$, the claimed linear bound is false. Such a $v$ would be an eigenfunction of the linearized Euler-Lagrange operator with eigenvalue approaching zero.

Watch

Extended reading notes

Core claim

The paper establishes the quantitative stability estimate $\delta(u) \le C\, \Gamma(u)$ for functions $u$ close to solving $-\Delta_H u = \left(\int_{\mathbb{H}^n} |u(\eta)|^{Q^\ast_\mu}|\eta^{-1}\xi|^{-\mu}\,d\eta\right)|u|^{Q^\ast_\mu-2}u$, where $\delta(u)$ is the $L^2$ gradient gap to the family $U_{\lambda,\zeta}$ of explicit optimizers and $\Gamma(u)$ is the norm in the dual of $S^{1,2}(\mathbb{H}^n)$ of the Euler-Lagrange expression $\Delta_H u + \left(\int_{\mathbb{H}^n} |u(\eta)|^{Q^\ast_\mu}|\eta^{-1}\xi|^{-\mu}\,d\eta\right)|u|^{Q^\ast_\mu-2}u$. For sums $\sum_{i=1}^\nu U_{\lambda_i,\zeta_i}$ of weakly interacting bubbles, the same linear bound is claimed when $Q=4$.

Load-bearing premise

The proof needs every optimizer function to be isolated up to its natural symmetries: the only nearly-free deformations should be changing the scale or the location. If some other deformation were almost free, a small residual $\Gamma(u)$ could sit alongside a large gradient gap $\delta(u)$, and the linear bound would break.

Editorial extensions

If this is right

Near-solutions of the Euler-Lagrange equation form a bounded neighborhood of the optimizer manifold, with the residual and the gradient gap equivalent up to a constant.
Minimizing sequences that approach equality in the nonlocal Sobolev inequality must converge to the optimizer family, at a rate controlled by their Euler-Lagrange residual.
At $Q=4$, multi-bubble near-solutions obey the same linear rigidity, so widely separated bubbles cannot hide a residual while the gradient profile drifts away.
The linear rate is the natural quantitative form of the Euler-Lagrange characterization of the sharp constant $C_{HL}(Q,\mu)$.

Reading between the lines

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

The special role of $Q=4$ suggests a genuine dimensional threshold: for larger homogeneous dimensions, interaction between bubbles may be weaker, so the linear multi-bubble control could fail or require a different rate.
The same coercivity mechanism should yield explicit constants by computing the spectral gap of the linearized operator on the first nontrivial variation; such a computation would also identify the sharp stability threshold.
If the nondegeneracy assumption were to fail for some $\mu\in(0,Q)$, there should be a continuous family of near-critical functions with $\Gamma(u)\to 0$ but $\delta(u)$ bounded away from zero; finding one would directly delimit the range of $\mu$ where the theorem can hold.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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 paper announces a linear stability estimate delta(u) <= C Gamma(u) for the nonlocal Sobolev inequality on the Heisenberg group, plus a multi-bubble version for Q=4. That is a natural and worthwhile extension of results known in Euclidean and fractional settings, and the statement is clean. The authors import the sharp constant and optimizer classification, so the circularity burden is low. I see no invented entities or fitted constants.\n\nThe problem is that I cannot check the proof: the full text in the provided copy is garbled beyond use. So my verdict rests on the abstract and on what the argument must contain. The stress-test note is on target: the linear rate needs a spectral gap for the linearized operator around every optimizer, i.e. coercivity on the orthogonal complement of the conformal directions and exact identification of the kernel. Without that, a spurious zero mode would degrade the rate to square-root. The abstract says nothing about this, so the reader cannot tell whether the hard part is done. The unexplained Q=4 restriction for bubble sums reinforces that dimension-specific estimates are load-bearing.\n\nThere is also a small but real defect in the displayed definition: delta(u)=||nabla_H u - nabla_H U_{lambda,zeta}|| has no infimum over lambda,zeta, so literally it is not a distance to the optimizer set. The intended invariant formulation must be delta(u)=inf_{lambda,zeta} ||nabla_H u - nabla_H U_{lambda,zeta}||. I assume this is a typo, but as written it would make the theorem meaningless for a family of optimizers.\n\nThe citation pattern appears reasonable; the abstract cites the sharp constant and classification rather than claiming them. The spelling error 'Hiesenberg' is minor.\n\nWho is this for? Analysts working on sharp functional inequalities, stability of optimizers, and blow-up analysis in sub-Riemannian geometry. If the proof delivers the coercivity estimate, the paper is a useful tool. But the burden is on the authors to state and prove the spectral-gap property; the abstract hides it.\n\nMy recommendation: if the full text is readable, this deserves a serious referee. The result is important enough and the statement is precise enough to warrant referee time, even though the abstract alone cannot certify correctness. I would not cite it until the proof is verified.

Referee Report

4 major / 4 minor

Summary. The manuscript claims a quantitative stability theorem for the nonlocal Sobolev inequality on the Heisenberg group: when a function u is close to solving the Euler-Lagrange equation, the natural L^2-gradient distance to the optimizer family U_{\lambda,\zeta} is linearly controlled by the dual norm \Gamma(u) of the Euler-Lagrange expression. The abstract further claims the analogous linear stability for weakly interacting sums of bubbles when the homogeneous dimension is Q=4. The proof is said to rest on importing the sharp constant and optimizer classification from earlier work. However, the supplied full text is severely corrupted and largely unreadable, so the proof and the precise hypotheses cannot be verified from the manuscript as provided.

Significance. If the linear stability bound is correct, it would be a meaningful quantitative refinement of the known qualitative compactness and stability theory for the nonlocal Sobolev inequality in the Heisenberg group, and the Q=4 bubble-sum statement would add a new multi-bubble stability result. The statement is explicit and does not involve fitted constants, which is a strength. The use of the previously established sharp constant and optimizer classification is external grounding rather than circular reasoning. Nevertheless, because the body of the paper is unreadable and no complete proof, lemma statement, or theorem statement can be checked, I cannot certify soundness from the present submission.

major comments (4)

[Full text (all sections after the Abstract)] The supplied body consists of corrupted text with replacement characters and broken equations; I could not read any complete theorem, lemma, or proof. Since the central claims rest entirely on these unreadable arguments, this is a blocking issue. The authors must provide a clean, readable PDF or LaTeX source before a substantive content review can be completed.
[Abstract, definition of δ(u)] The displayed definition δ(u)=||∇_H u-∇_H U_{\lambda,\zeta}||_{L^2} depends on an arbitrary choice of the parameters (\lambda,\zeta), so it is not a distance to the optimizer family. As written, the linear bound δ(u)≤C Γ(u) cannot have the stated geometric meaning. The theorem should either define δ(u)=inf_{\lambda,\zeta} ||∇_H u-∇_H U_{\lambda,\zeta}||_{L^2} or specify an equivalent gauge-fixing procedure and then prove the bound for that quantity.
[Abstract, Q=4 restriction] The restriction of the weakly interacting bubble result to Q=4 is unexplained in the abstract and in the readable fragments. If the proof relies on dimension-specific interaction estimates, the paper should either prove the result for every admissible Q or state precisely which estimate fails for Q≠4. As it stands, the restriction indicates a load-bearing hypothesis whose verification is not visible.
[Abstract, linear-rate claim] The claimed linear bound δ(u)≤C Γ(u) requires a coercivity or spectral-gap estimate for the linearized nonlocal Euler-Lagrange operator at each optimizer, on the complement of the finite-dimensional parameter manifold. Neither the abstract nor the readable text states or proves such an estimate. Without it, a spurious approximate null direction would reduce the best possible rate to at most δ(u)≲√Γ(u). This is the decisive analytical hypothesis and must be stated and proved explicitly.

minor comments (4)

[Abstract] There is a spelling error 'Hiesenberg' in the first sentence; it should be 'Heisenberg'.
[Abstract] The phrase 'the the set of optimizers' contains a duplicated article and should be corrected.
[Abstract] The equation label 'non-critical-n' appears in the abstract but no corresponding numbered display is visible in the readable text; the numbering should be made consistent.
[Abstract] The notation Q^*_μ is used without a reminder of its dependence on Q and μ beyond the first occurrence; a brief restatement near the Euler equation would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found; the claimed stability estimate relates two distinct quantities and rests on external input.

full rationale

I could not identify any step in the available text where a claimed prediction reduces to its own inputs by construction. The central claim is the stability estimate δ(u) ≤ C Γ(u), which compares the distance from u to the optimizer family with the dual norm of the Euler-Lagrange residual. These are distinct mathematical objects, and the inequality is the theorem being proved rather than a definitional identity. The sharp constant C_HL(Q,μ) and the optimizer family U_{λ,ζ} are imported from earlier external results; that is standard grounding, not circularity. No fitted parameters, no data subsets, and no self-citation chain are visible in the abstract or in the readable fragments of the full text. The body of the paper is too corrupted to quote any explicit reduction, and under the requirement to exhibit a specific circular step with quoted evidence, no such step can be asserted. Therefore the appropriate finding is no significant circularity.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No free parameters are fitted. The proof relies on the known sharp inequality and on the classification and nondegeneracy of the optimizer family, both imported from prior work. The Q=4 restriction is an additional dimension-specific assumption for the bubble result.

assumptions (3)

standard math The sharp constant C_HL(Q,mu) for the nonlocal Sobolev inequality on the Heisenberg group is known and the inequality holds for all u in S^{1,2}(H^n).
Stated as the starting inequality in the abstract; the proof uses it as an external benchmark rather than deriving it.
domain assumption The family U_{lambda,zeta} contains all optimizers of the inequality and the Euler-Lagrange equation has no other nondegenerate critical points in S^{1,2}(H^n).
The distance delta(u) is defined relative to this family; the main stability estimate is only meaningful if this set is complete. The abstract cites it without proof.
domain assumption For weakly interacting bubble sums, the interaction estimates close only in homogeneous dimension Q=4.
This dimension restriction is asserted in the abstract without explanation; the general proof of the multi-bubble result depends on it.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Quantitative stability of critical points for the nonlocal-Sobolev inequality in Heisenberg group." pith.science (2026). https://pith.science/paper/2D7V5WAJ

@misc{pith2026250808614,
  author       = {Pith},
  title        = {Pith review of: Quantitative stability of critical points for the nonlocal-Sobolev inequality in Heisenberg group},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2D7V5WAJ}},
  note         = {Machine review of arXiv:2508.08614}
}

abstract

We investigate the quantitative stability of the nonlocal Sobolev inequality in Heisenberg group \begin{equation*}\label{non-Sobolev} C_{HL}(Q,\mu) \left(\int_{\mathbb{H}^{n}}\int_{\mathbb{H}^{n}}\frac{|u(\xi)|^{Q^{\ast}_{\mu}}|u(\eta)|^{Q^{\ast}_{\mu}}}{|\eta^{-1}\xi|^{\mu}}\mathrm{d}\xi\mathrm{d}\eta\right)^{\frac{1}{Q^{\ast}_{\mu}}}\leq \int_{\mathbb{H}^{n}}|\nabla_{H}u|^{2}d\xi,\qquad\forall u\in S^{1,2}(\mathbb{H}^{n}), \end{equation*} where $Q=2n+2$ is the homogeneous dimension of the Hiesenberg group $\mathbb{H}^{n}$, $\mu\in(0,Q)$ and $Q^{\ast}_{\mu}=\frac{2Q-\mu}{Q-2}$ are two parameters corresponding to the Hardy-Littlewood-Sobolev inequality and Folland-Stein inequality on Heisenberg group, $C_{HL}(Q,\mu)$ is the sharp constant of the nonlocal-Sobolev inequality. Specifically, when $u$ is close to solving the Euler equation \begin{equation*}\label{non-critical-n} -\Delta_{H} u=\left(\int_{\mathbb{H}^{n}}\frac{|u(\eta)|^{Q^{\ast}_{\mu}}}{|\eta^{-1}\xi|^{\mu}}\mathrm{d}\eta\right)|u|^{Q^{\ast}_{\mu}-2}u,\qquad\xi,\eta\in\mathbb{H}^{n}, \end{equation*} the natural distance between $u$ and the the set of optimizers $U_{\lambda,\zeta}$, defined as $\delta(u)=||\nabla_{H}u-\nabla_{H}U_{\lambda,\zeta}||_{L^{2}}$, can be linearly bounded by the functional derivative term \begin{equation*} \Gamma(u)=\left\|\Delta_{H}u+\left(\int_{\mathbb{H}^{n}}\frac{|u(\eta)|^{Q^{\ast}_{\mu}}}{|\eta^{-1}\xi|^{\mu}}\mathrm{d}\eta\right)|u|^{Q^{\ast}_{\mu}-2}u\right\|_{(S^{1,2}(\mathbb{H}^{n}))^{-1}}. \end{equation*} And for the weakly interacting bubble solutions $\mathop{\sum}\limits_{i=1}^{\nu}U_{\lambda_{i},\zeta_{i}}$, the aforementioned quantitative stability result holds when the dimension $Q=4$.

Review history (2 revisions) →

Discussion (0). Continue with ORCID 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

Optimal Stability Bounds, Minimizers, and Critical Points for a Critical Nonlocal Sobolev Inequality on the Heisenberg Group
math.AP 2026-08 reject novelty 6.0 of 10

The paper claims optimal Bianchi-Egnell constants for a nonlocal Sobolev inequality on the Heisenberg group, but the key strict spectral bound is unproved.

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.