{"id":"94187582-b491-480b-a2f9-a1bc3df05d4a","arxiv_id":"2506.01217","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Two stochastic analogs of Q-curvature flow on even-dimensional manifolds are shown to have weak solutions, with volume governed by square Bessel and CIR processes.","lead":"This paper constructs two random versions of Q-curvature flow, a geometric evolution equation on even-dimensional spaces, and proves that weak solutions exist under standard geometric conditions. It extends stochastic Ricci flow from surfaces to all even dimensions and connects the construction to higher-dimensional Liouville measures.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 2.7's gluing of local CGMC inverse maps is asserted rather than proved; if it fails, the symmetrizing measures cannot be defined on M and Theorem 1.7 does not construct weak solutions.","rationale":"I read the paper in good faith and checked the main computational scaffolding: the gamma-subcriticality condition reduces to sigma^2 < 4pi when n=2 and matches [DS22]; the sphere computation Q_r(1)=(4pi)^{n/2}(n/2-1)! is correct; the NQF volume drift cancels because P_ref(1)=0, leaving a BESQ(0) dynamic; the LQF volume is a CIR process and the Feller-type condition follows from sigma^2 <= -2Q_ref(1)/n since rho>=1; and the tuning rho = 1 + a_n n sigma^2/4 reproduces theta = a_n(n/gamma + gamma/2) from [DSHKS24]. These checks give me confidence that the paper is not papering over the easy parts. The genuinely load-bearing step is Lemma 2.7: the measurable inversion of the CGMC map and its extension to the non-Lebesgue ground measures that appear on locally conformally flat charts. The proof in Appendix A is intricate and the gluing of local inverse maps over a finite cover is compressed to a single assertion. Because the Dirichlet form construction needs the measures nu_NQF and nu_LQF to live on M, and the generator expressions L_NQF and L_LQF involve the field psi in the term <g_i, psi>_E, a failure of the inverse map would prevent the Hunt processes from being realized on M as weak solutions. I do not see a concrete internal inconsistency in Lemma A.1; the near-scale-invariance computation for the non-Lebesgue case is plausible and the covariance algebra checks out once the definition of K_{delta,epsilon} is read correctly. But the compatibility of the local inverse maps is underproved. This is exactly the reader's weakest assumption, and my read does not move the verdict: CONDITIONAL remains the right call, pending a complete proof or independent verification of Lemma 2.7.","tokens_in":36913,"tokens_out":42607,"duration_ms":449893,"concrete_test":"Verify the compatibility claim in Lemma 2.7 directly: take a locally conformally flat closed manifold (for example S^4 with the round metric) and two overlapping coordinate charts. Write out the two local inverse maps produced by Lemma A.1 and prove, or disprove by a counterexample, that for every test function compactly supported in the overlap the two reconstructed fields agree almost surely. Equivalently, establish the equivariance of Vihko's reconstruction map under smooth coordinate changes and under multiplication of the ground measure by a smooth density factor close to 1; if the equivariance holds, the gluing is valid, and if not, Lemma 2.7 fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim depends on Lemma 2.7, which supplies a measurable inverse X_gamma for the CGMC map M^gamma_ref, allowing the symmetrizing measures nu_NQF and nu_LQF to be transported from the field space H^{0-}_ref to the measure space M where the Dirichlet forms live. The proof builds local inverse maps on a finite locally conformally flat cover using Lemma A.1, a non-Lebesgue extension of Vihko's reconstruction theorem, and then asserts compatibility on overlaps: 'the outputs of the two maps agree as fields on U1... This is clear from the construction in Lemma A.1.' No argument is given. The concrete risk is that the local reconstruction maps in different charts use different coordinate representations, ground densities, and smooth remainders; each recovers the restricted field almost surely, but a single global measurable map with the required almost-sure inversion still needs a proof of equivariance under coordinate changes and under multiplication of the ground measure by a smooth density. That equivariance is neither stated nor proved. If this gluing step is false, nu_NQF and nu_LQF cannot be viewed as measures on M, and the Hunt processes constructed in Section 4 are not processes on M solving (1.24) or (1.25). The paper itself flags the sensitivity of this step in Section 5.3, where it concedes that local conformal flatness is used only for Lemma 2.7. This is a load-bearing weakness, not a demonstrated contradiction: the central computations I checked -- the subcriticality threshold, the sphere Q-curvature identity, the cancellation of the NQF volume drift, the CIR form of the LQF volume, and the parameter match in Section 5.2 -- are internally consistent.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper defines two stochastic analogs of the even-dimensional Q-curvature flow, the normalized Q flow (NQF) and the Liouville Q flow (LQF), and proves existence of weak solutions for noise intensity below a dimension-dependent threshold. The construction follows the strategy of Dubédat–Shen for surfaces: the author first builds symmetrizing measures from co-polyharmonic Gaussian multiplicative chaos (CGMC), proves integration-by-parts formulas against a Cameron-Martin space, then uses Dirichlet form theory to obtain Hunt processes on the space of positive finite measures. The main theorem, Theorem 1.7, asserts existence of weak solutions to NQF for positive prescribing functions f under conditions (A1) and (A2), and to LQF for non-positive f under (A2'), on closed locally conformally flat even-dimensional manifolds. The paper also derives the volume dynamics, identifies a CIR-type equation for the LQF volume in a special case, and interprets one LQF invariant measure as a stochastic quantization of the Polyakov-Liouville measures of Dello Schiavo–Herry–Kopfer–Sturm.","tokens_in":37012,"tokens_out":25105,"duration_ms":247216,"significance":"If correct, Theorem 1.7 is the first existence theorem for stochastic Q-curvature flows in dimensions n ≥ 4, a natural and nontrivial extension of the two-dimensional result of Dubédat–Shen. Several key computations are genuinely verified in the text: the subcriticality threshold in n=2 reduces to σ² < 4π, matching DS22; the sphere computation Q_r(1) = (4π)^(n/2)(n/2−1)! is correct; the NQF volume drift cancels exactly via P_ref(1)=0, giving dV_t = nσ√V_t dB_t; and the Feller-type condition in Lemma 5.1 follows from σ² ≤ −2Q_ref(1)/n. The symmetrizing measures are constructed from CGMC theory rather than assumed, and the Markov processes are then shown to be symmetric with respect to them, so the argument is not circular. The appendix contains a substantial extension of Vihko's GMC reconstruction theorem to non-Lebesgue ground measures, which is necessary for the main theorem.","major_comments":[{"comment":"The gluing of the local inverse CGMC maps is asserted rather than proved. After constructing local inverse maps on a finite locally conformally flat cover, the proof states that on overlaps 'the outputs of the two maps agree as fields ... This is clear from the construction in Lemma A.1.' This is the only step that promotes the local reconstruction theorem into the global measurable map Xγ : M → D'(M), and the rest of the paper depends on it: ν_NQF and ν_LQF are transported to the measure space M through Xγ, and the Dirichlet forms of Section 4 are defined on L²(M, ν). The author should supply an explicit compatibility argument: for μ_ref-a.e. ψ, the local reconstructions obtained from Mγ_ref(ψ) restricted to two overlapping charts coincide on the overlap, or equivalently the local maps are equivariant under chart transitions and under multiplication of the ground measure by a smooth density. That equivariance is not stated or proved. This issue is load-bearing, and the paper itself flags in Section 5.3 that local conformal flatness is used only for Lemma 2.7.","section":"Section 2.2, Lemma 2.7"},{"comment":"The passage from the measure ν_NQF on H^{0-}_{ref} to a measure on M is written as 'By Lemma 2.7, we can equivalently consider ν_NQF as a measure on H^{0-}_{ref} ... or on M ...' and the same convention is used for ν_LQF. Even if the local compatibility in Lemma 2.7 is supplied, the paper should spell out the measure-theoretic definition: ν_M should be defined as the pushforward of ν_H under Mγ_ref, and Xγ should be shown to be a measurable inverse in the appropriate almost-sure sense. As written, the identification of the Dirichlet form on L²(M, ν) with the form on L²(H^{0-}_{ref}, ν) is informal, and the subsequent proof of Theorem 1.7 relies on this identification.","section":"Section 2.3 and Section 4.1"}],"minor_comments":[{"comment":"The statement that the LQF volume evolves as a CIR process is only true when f is constant, in particular f = Q_ref; for general f the drift term contains ω_t(f), which is not a function of V_t alone. The wording should be adjusted so that the abstract and the corollary do not overstate the generality.","section":"Abstract and Corollary 1.8"},{"comment":"The proof refers to 'Lemma 2.5' where the intended statement appears to be Proposition 2.5 (the result of Vihko), and later calls Lemma A.1 'a slight generalization of Lemma 2.5'. The numbering should be reconciled.","section":"Section 2.2, Lemma 2.7"},{"comment":"In the proof of Lemma 3.3, the integrals over the grounded space ˚H^{0-}_{ref} contain a stray 'dc' that should be deleted; the measure there is μ_ref, not the ungrounded product measure.","section":"Section 3, Lemma 3.3"},{"comment":"The usual Sobolev space H^s_g and the co-polyharmonic Sobolev space H^s_g (with underline) are visually indistinguishable in the text; please use a distinct symbol such as a different font or an explicit label to avoid confusion.","section":"Definition 2.2 and throughout"},{"comment":"Several equations in the semimartingale analysis are written with ∂_t where the intended meaning is a differential dω_t(h); for example, the display immediately after equation (1.25) should be read as a semimartingale decomposition. This is a presentation issue but may confuse readers.","section":"Section 4.2"}],"recommendation":"major_revision","confidential_remarks":"This is a serious and technically ambitious submission. The referee's main reservation is the unproved gluing step in Lemma 2.7, which is load-bearing for the whole construction. The central computations and the general strategy appear sound, and the missing compatibility argument seems likely to be repairable within the scope of the manuscript. If the author can provide a complete proof of the global reconstruction map, I would be prepared to accept the paper. No circularity concern is apparent; the measures are constructed from CGMC theory and the processes are shown to be symmetrizing."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper is the first existence result for stochastic Q-curvature flows in dimensions n >= 4, and that is a real advance. The Dirichlet-form strategy follows Dubédat–Shen, but the new machinery—co-polyharmonic Gaussian multiplicative chaos, the inversion lemma for non-Lebesgue ground measures, and the NQF/LQF distinction with BESQ/CIR volume dynamics—is genuinely new and mostly solid. I checked the central computations myself: the subcriticality threshold reduces to sigma^2 < 4pi when n = 2, the sphere Q-curvature identity Q_r(1) = (4pi)^(n/2)(n/2-1)! is correct, the NQF volume drift cancels via P_ref(1) = 0, the LQF volume is a CIR process, and the parameter tuning in Section 5.2 reproduces the DSHKS24 measure. Those checks are internally consistent and show the author is not glossing the easy parts.\n\nThe real soft spot is Lemma 2.7. Local inverse maps are built chart-by-chart from Vihko's reconstruction theorem, and the gluing across overlaps is asserted in a single sentence: 'This is clear from the construction in Lemma A.1.' That is not enough. You need equivariance under coordinate changes and under multiplication of the ground measure by a smooth density to obtain a single measurable inverse on M. If that fails, the symmetrizing measures do not live on M, and Theorem 1.7 does not construct the claimed weak solutions. The paper honestly flags local conformal flatness as needed only for this lemma, which is helpful, but it does not fill the gap. I see no contradiction—just a missing proof.\n\nA secondary concern is that the special-standard-core and strong-locality arguments are imported from DS22 by appeal. That is probably fine, but it would be better if the author spelled out the one or two places where higher dimension changes the topology of the core. This is minor compared to Lemma 2.7.\n\nWho is this for? Specialists in stochastic geometric flows and Dirichlet-form methods. It deserves a serious referee: the core idea is right, the computations are credible, and the missing gluing argument is likely fixable. I would not desk reject. I would send to peer review and ask the author to prove the gluing lemma properly, or at least to state it as a separate conjecture with precise hypotheses.","headline":"Genuine extension of stochastic Ricci flow to even-dimensional Q-curvature flows, with sound core computations and one under-proved gluing step that should be fixed before publication.","tokens_in":37858,"tokens_out":1559,"would_cite":true,"duration_ms":15972,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60H15","53C44","60J45","60G60"],"pacs":[],"model":"deepseek-v4-flash","headline":"On even-dimensional closed manifolds satisfying two topological conditions, both the normalized and Liouville Q-curvature flows admit white-noise-driven weak solutions, with explicit volume dynamics.","keywords":["stochastic Q-curvature flow","Dirichlet forms","Gaussian multiplicative chaos","co-polyharmonic Gaussian fields","conformal geometry","weak solutions","CIR process","Polyakov-Liouville measures"],"falsifier":"Compute the covariance bound in Appendix A with a nonconstant smooth ground density $\\lambda$ near 1; if $\\mathrm{Cov}(A_\\varepsilon(x), A_\\varepsilon(x'))$ fails to vanish as $\\varepsilon \\to 0$ for $x \\ne x'$, the local inverse maps cannot be glued and Lemma 2.7 fails.","tokens_in":1811,"feed_emoji":"🎲","tokens_out":6562,"duration_ms":96227,"temperature":0.7,"pith_summary":"On a closed, locally conformally flat manifold of even dimension $n$ satisfying two geometric conditions, the paper constructs two stochastic versions of the $Q$-curvature flow and proves they have weak solutions in the sense of projected SDEs. The solutions are Hunt processes symmetric with respect to measures built from co-polyharmonic Gaussian multiplicative chaos, and they are the first existence results for stochastic $Q$-curvature flows in dimension $n \\ge 4$. The total volume of each solution is itself a Markov process: a square Bessel process for the normalized flow and a CIR process for the Liouville flow. Under extra conditions, the Liouville flow is a stochastic quantization of the even-dimensional Polyakov-Liouville measures.","feed_headline":"Stochastic Q-curvature flows solved in even dimensions","feed_subtitle":"Extends the two-dimensional stochastic Ricci flow to dimensions four and higher, with explicit volume SDEs.","key_machinery":"The central object is the co-polyharmonic Gaussian field (CGF), a log-correlated Gaussian distribution with covariance kernel given by the co-polyharmonic Green function, and its associated co-polyharmonic Gaussian multiplicative chaos (CGMC) measure. The key step that carries the argument is Lemma 2.7, a measurable inversion map that reconstructs the field from its CGMC measure; this map is built by patching local inverses on a finite cover, which is where local conformal flatness enters. The measures $\\nu_{\\mathrm{NQF}}$ and $\\nu_{\\mathrm{LQF}}$ are defined as formal densities interpreted via CGF and CGMC, and Dirichlet form theory converts these measures into Hunt processes whose semimartingale decompositions match the projected SDEs.","core_discovery":"For a closed, locally conformally flat manifold of even dimension $n$ satisfying (A1) and (A2), the paper claims that for any positive $f \\in C^\\infty(M)$ and for noise strength $\\sigma^2 < 2(4\\pi)^{n/2}(n/2-1)!/n$, there exists a weak solution to the normalized $Q$-flow (NQF) with prescribed $Q$-curvature $f$. If $f \\le 0$ and (A2) is replaced by (A2'), then there exists a weak solution to the Liouville $Q$-flow (LQF). The solutions are Hunt processes symmetric with respect to the measures $\\nu_{\\mathrm{NQF}}$ and $\\nu_{\\mathrm{LQF}}$ constructed from co-polyharmonic Gaussian multiplicative chaos, and they solve the projected SDEs obtained by pairing the formal equations with smooth test functions. As immediate corollaries, the total volume $V_t = \\omega_t(1)$ satisfies $dV_t = n\\sigma \\sqrt{V_t} dB_t$ for NQF and $dV_t = -n(\\varrho Q_{\\mathrm{ref}}(1) - \\omega_t(f))dt + n\\sigma \\sqrt{V_t} dB_t$ for LQF.","pith_inferences":["The local conformal flatness assumption is used only in the gluing of local inverse maps; a weaker assumption on the regularity of the covariance remainder would likely extend the theorem to non-flat manifolds.","The inversion lemma in Appendix A already allows smooth non-Lebesgue ground measures close to Lebesgue; expanding it to Hölder densities would widen the class of manifolds without changing the main argument.","The explicit volume dynamics suggest a direct numerical test: simulating the volume SDEs and comparing with the predicted law of the total volume would provide evidence for or against the existence claim.","Uniqueness results for constant $Q$-curvature metrics may be the key input for proving convergence of these stochastic flows to their invariant measures."],"forward_implications":["The total volume of the NQF solution is a square Bessel process, so volume is not preserved in the stochastic flow even though the deterministic flow preserves it.","In the special case $Q_{\\mathrm{ref}} \\le 0$ and $f = Q_{\\mathrm{ref}}$, the LQF volume satisfies a CIR SDE and stays positive almost surely, giving an invariant measure for the flow.","For $Q_{\\mathrm{ref}} < 0$, $f = Q_{\\mathrm{ref}}$, and $\\sigma^2 \\le -2Q_{\\mathrm{ref}}(1)/n$, the LQF is a stochastic quantization of the even-dimensional Polyakov-Liouville measures.","The dimension $n=2$ threshold $\\sigma^2 < 4\\pi$ matches the condition for the two-dimensional stochastic Ricci flow, recovering it as a special case."],"supporting_citations":[{"why":"Supplies the two-dimensional stochastic Ricci flow existence result and the Dirichlet-form strategy that this paper extends.","marker":"[DS22]"},{"why":"Provides co-polyharmonic Gaussian fields, CGMC measures, and the Polyakov-Liouville measures used in Section 5.2.","marker":"[DSHKS24]"},{"why":"Its reconstruction theorem is the basis for Lemma 2.7, the measurable inverse of the CGMC map.","marker":"[Vih24]"},{"why":"Provides the Dirichlet form-Hunt process correspondence used to construct weak solutions.","marker":"[FOT11]"},{"why":"Proves global existence for the deterministic Q-flow and provides conditions (A1) and (A2).","marker":"[Bre03]"},{"why":"Constructs the co-polyharmonic operators $P_g$ used throughout.","marker":"[GJMS92]"},{"why":"Foundational Gaussian multiplicative chaos construction underlying the CGMC moment bounds.","marker":"[Kah85]"},{"why":"Subcritical GMC construction used in Appendix A to extend the inversion argument to non-Lebesgue ground measures.","marker":"[Sha16]"}],"fun_headline_variants":["Even-dimensional stochastic Q-flows now have solutions","Stochastic Q-curvature flow solved in all even dimensions","Volume dynamics of stochastic Q-flows reduce to known SDEs","Even-dimensional Q-curvature flow stochastically quantized"],"cache_read_input_tokens":39552,"weakest_assumption_plain":"The proof depends on being able to reconstruct the conformal factor field from its chaos measure locally and patch those local reconstructions together; if that measurable inversion fails, the symmetrizing measures cannot be defined on the space of volume measures and the main theorem collapses.","fun_headline_variants_meta":{"raw":{"variants":["Even-dimensional stochastic Q-flows now have solutions","Stochastic Q-curvature flow solved in all even dimensions","Volume dynamics of stochastic Q-flows reduce to known SDEs","Even-dimensional Q-curvature flow stochastically quantized"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000837,"raw_usage":{"total_tokens":3644,"prompt_tokens":932,"completion_tokens":2712,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":548,"completion_tokens_details":{"reasoning_tokens":2643}},"tokens_in":548,"tokens_out":2712,"duration_ms":19088,"temperature":1.0,"reasoning_tokens":2643,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:50:10.140999+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the covariance bound in Appendix A with a nonconstant smooth ground density $\\lambda$ near 1; if $\\mathrm{Cov}(A_\\varepsilon(x), A_\\varepsilon(x'))$ fails to vanish as $\\varepsilon \\to 0$ for $x \\ne x'$, the local inverse maps cannot be glued and Lemma 2.7 fails.","supporting_citations":[],"review_version":1}