{"id":"34236742-b239-4ad5-889f-e6bb2a7f9c1a","arxiv_id":"2501.18281","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Small-γ uniqueness of solutions to complex Monge-Ampère mean field equations is established on hyperconvex domains and compact manifolds, partially confirming a Berman-Berndtsson conjecture.","lead":"The paper proves that complex Monge-Ampère mean field equations have a unique solution when the temperature parameter is small, on bounded hyperconvex domains and on compact Kähler and Hermitian manifolds. This partially confirms a conjecture by Berman and Berndtsson and gives a small-γ uniqueness statement in settings where non-uniqueness is known for large γ.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 3.2's auxiliary function ψ is normalized with RHS 2·1_E·f + b·ω^n, which requires ∫_E f ≤ 1/2; the proof only controls ∫_E ω^n, so the main Kähler uniqueness argument has a gap unless an additional estimate is supplied.","rationale":"The paper aims to prove small-temperature uniqueness for complex Monge-Ampere mean field equations by combining refined stability estimates with L∞ bounds. The overall architecture is coherent, and the statements are plausible, but the reader's conditional verdict identifies a real gap in the proof of Theorem 3.2. I agree that this is the load-bearing issue for the main theorem. The auxiliary equation for ψ needs its right-hand side to be a positive probability measure, but the construction only ensures the volume-based condition ∫_E ω^n ≤ 1/2, not the required ∫_E fω^n ≤ 1/2. No available estimate in the paper supplies the missing f-mass control; the L^p bound is insufficient by itself. The coefficient 2 cannot simply be reduced without destroying the strict multiplicative gain used in the domination-principle argument. This is an internal gap in the written proof, not a disagreement with an external consensus, and it may be repairable by a more careful choice of the set E or a stronger smallness statement for {u<v+α}. As written, Theorem 3.2 is not fully established. Theorem 4.2's omitted proof is a secondary issue for the Hermitian extension and does not affect Theorem 1.1. Since the reader already marked the paper CONDITIONAL and identified the same weakest assumption, no adjustment to the verdict is needed.","tokens_in":14057,"tokens_out":14237,"duration_ms":136683,"concrete_test":"Re-derive the normalization in Theorem 3.2 with a general coefficient c in place of 2: solve (ω+ddcψ_c)^n = c·1_E·f·ω^n + b_c·ω^n with b_c = 1 - c∫_E fω^n, and check whether the displayed inequality (1+n^{-1}ε^2log2)^n can be obtained for some c≥1 using only the stated hypotheses. Then attempt an explicit counterexample on X=P^1: take a small set S of volume δ with f=g=N(1+η·χ_S) on S, N=1/δ, compute nearby u,v with sup u = sup v = 0, set α=(sup(u-v)-sup(v-u))/2, and evaluate ∫_{u<v+α} fω^n and ∫_{u<v+α} ω^n; if the f-mass exceeds 1/2 while the ω^n-mass is ≤1/2, the construction of ψ as written fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim Theorem 1.1 is proved via Theorem 3.4, which applies the stability estimate Theorem 3.2 to the densities e^{-γu0-b1}f and e^{-γv0-b2}f. In Step 1 of Theorem 3.2, for E={u<v+α} with ∫_E ω^n ≤ 1/2, the authors define ψ by (ω+ddcψ)^n = 2·1_E·f·ω^n + b·ω^n with b∈[0,1]. For this right-hand side to be a probability measure, one needs 2∫_E fω^n ≤ 1, i.e. ∫_E fω^n ≤ 1/2. The proof only establishes the analogous bound for ω^n, not for the Monge-Ampere density f = (ω+ddcu)^n/ω^n. The L^p bound ||f||_p ≤ B and the small-volume bound ∫_E ω^n ≤ 1/2 do not imply ∫_E fω^n ≤ 1/2: Hölder gives only B·2^{-1/q}, which exceeds 1/2 for large B. The coefficient 2 is not arbitrary: with any smaller coefficient c, the strict gain (1+n^{-1}ε^2log2)^n need not exceed 1, and the domination-principle step loses its force. Thus, unless an extra argument controls the f-mass of E, the uniqueness proof has a genuine gap. Theorem 4.2 is also stated without proof, but the Kähler result does not depend on it.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies uniqueness of continuous solutions to complex Monge-Ampère mean field equations (ω+dd^c u)^n = e^{-γu} f ω^n (and the Euclidean analogue on hyperconvex domains) when the temperature parameter γ>0 is small. It claims three theorems: uniqueness for L^p densities on bounded hyperconvex domains (Theorem 1.2), on compact Kähler manifolds (Theorem 1.1), and on compact Hermitian manifolds with a positive density (Theorem 4.4). The strategy is to combine L∞ estimates (Theorems 2.5 and 3.3, with a new proof) with stability estimates for normalized Monge-Ampère equations (Theorems 3.1 and 3.2, and the quoted Theorem 4.2), then use a contraction argument in γ. The Kähler proof is the core of the paper.","tokens_in":14423,"tokens_out":15132,"duration_ms":130728,"significance":"If correct, the Kähler uniqueness theorem is a significant extension of the known small-γ uniqueness results: it allows L^p densities, requires no smoothness assumptions, and gives a quantitative dependence of γ0 on the data. The local theorem partially confirms a conjecture of Berman and Berndtsson. The paper's stability estimates are refinements of earlier work and the L∞ estimates are self-contained. However, the main proof has a specific gap in Theorem 3.2, and the Hermitian section relies on an unproved stability theorem, so the significance is conditional on repair.","major_comments":[{"comment":"The auxiliary function ψ is defined by (ω+dd^c ψ)^n = 2 1_E f ω^n + b ω^n with b∈[0,1]. For the right-hand side to be a probability measure one needs 2∫_E f ω^n ≤ 1, i.e. ∫_E f ω^n ≤ 1/2. The proof only establishes ∫_E ω^n ≤ 1/2. Hölder's inequality gives ∫_E f ω^n ≤ ||f||_p (ω^n(E))^{1/q} ≤ B 2^{-1/q}, which can be larger than 1/2 for large B, so the L^p bound on f does not supply the missing control. The coefficient 2 is essential: with any smaller coefficient the factor (1+n^{-1}ε^2 log 2)^n would not exceed 1 for small ε, and the subsequent domination-principle step would lose its force. Since Theorem 3.4 and therefore Theorem 1.1 use Theorem 3.2, the central uniqueness proof is not established as written unless an additional estimate for the f-mass of E is supplied.","section":"Theorem 3.2, proof, Step 1"},{"comment":"Theorem 4.2 is a stated stability result that is load-bearing for the Hermitian uniqueness theorem (Theorem 4.4), but its proof is omitted: the text says only that it follows from [LPT21] with minor adjustments. This is not sufficient for a main theorem in a research paper; either a complete proof should be included or the Hermitian claims should be reduced to what is actually proved.","section":"Theorem 4.2"},{"comment":"In the proof of Theorem 2.5, the variational solution u∈E1(Ω) is used to assert ∫_Ω e^{-γu} dμ ≤ A_μ. But A_μ is defined as the supremum over u∈T0, and u is not shown to belong to T0 (T0 consists of bounded E0 functions, while the variational solution is only constructed in E1 and may be unbounded). Without this estimate the chain (dd^c v)^n ≥ (γ^n/n^n) A_μ^{-1} μ is not justified. This L∞ bound feeds Theorem 2.6 and hence the proof of Theorem 1.2, so the gap is load-bearing for the local uniqueness result.","section":"Theorem 2.5, proof"}],"minor_comments":[{"comment":"The name 'Kołodziej' is corrupted to 'Ko/suppress lodziej' in several places, including the references and the main text; this should be fixed.","section":"Throughout"},{"comment":"The expression '21Efωn' should be typeset as '2 1_E f ω^n' for readability; the missing space makes the indicator of E hard to distinguish from a factor 21.","section":"Theorem 3.2, Step 1"},{"comment":"The statement says f,g are 'probability measures', but the proof uses f^{1/n}, g^{1/n} and L^p norms; the authors should clarify that f,g are densities with respect to ω^n.","section":"Theorem 4.2, statement"},{"comment":"The displayed estimate involving ||f||_p^{1/n} is ambiguous; it should state explicitly which L^p norm of (e^{-γu0-b1} - e^{-γv0-b2})^{1/n} f^{1/n} is being bounded.","section":"Theorem 3.4, proof"},{"comment":"The domination principle [GL22, Proposition 2.8] is used in a set-localized form; including the precise statement would make the argument checkable.","section":"Theorem 3.2, domination step"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is serious and within the scope of math.CV. The gap in Theorem 3.2 is the main obstacle; if the authors can supply the missing f-mass estimate, the Kähler theorem is plausibly repairable. The omitted proof of Theorem 4.2 should also be supplied or the Hermitian claims restructured."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take on Lu–Phung (2501.18281). The headline result—uniqueness for small temperature γ in the Kähler and hyperconvex-domain settings with only L^p densities—is genuinely new and worth taking seriously. It goes beyond the 1D result of Gui–Moradifam, the S1-invariant work of Guedj–Kolev–Yeganefar, and the eigenvalue approach of Badiane–Zeriahi. The proof strategy is coherent: refine the stability estimates of [LPT21] and [GLZ18], combine with Kolodziej-type L∞ bounds, then contract.\n\nWhat the paper does well: the stability estimates in Theorems 3.1 and 3.2 are proved in detail rather than assumed; the L∞ estimates (Theorems 2.5 and 3.3) are new proofs using the variational approach; and Example 3.5 correctly shows the small-γ restriction is necessary. The paper is self-contained enough that reliance on the authors' earlier work is minimal.\n\nNow the soft spots. The stress-test and the reader both flag the same issue in Theorem 3.2, and I think it's real. When constructing ψ, the right-hand side 2·1_E·f + b·ω^n must be a probability measure. That requires ∫_E f ω^n ≤ 1/2, but the proof only establishes ∫_E ω^n ≤ 1/2. The L^p bound on f does not close the gap. This is a genuine gap in the proof as written, and since Theorem 3.4 depends on Theorem 3.2, it affects the main Kähler uniqueness theorem. I suspect it's fixable—maybe by choosing α to control the f-mass, or by using a different cut-off—but it needs to be supplied. I would not say the result is wrong; I would say the proof is incomplete.\n\nSecond, Theorem 4.2 is stated without proof. The Hermitian uniqueness theorem (4.4) leans on it, so that part is not yet established. The Kähler result does not depend on it, so the paper's core contribution survives.\n\nWho is this for? Anyone working on complex Monge-Ampère equations, mean field equations, or uniqueness at low temperature. The techniques (stability, L∞ via variational methods) are reusable. It deserves a serious referee, but the referee should ask for the Theorem 3.2 gap to be closed and Theorem 4.2 to be proved or removed. My recommendation: send it to review, and expect a revision.","headline":"Small-γ uniqueness for complex MA mean field equations is a real advance with a clean proof architecture, but the key stability lemma (Thm 3.2) has an unproven normalization step that needs a fix before the main theorem is accepted.","tokens_in":14977,"tokens_out":3111,"would_cite":true,"duration_ms":26347,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32W20","32U05","32Q15","35A23"],"pacs":[],"model":"deepseek-v4-flash","headline":"On compact Kähler manifolds, solutions to the complex Monge-Ampère mean field equation are unique once the temperature parameter $\\gamma$ is small enough; the same holds on bounded hyperconvex domains.","keywords":["complex Monge-Ampère equations","mean field equations","uniqueness","stability estimates","small temperature parameter","hyperconvex domains","Kähler manifolds","Hermitian manifolds"],"falsifier":"Take $E=\\{u<v+\\alpha\\}$ with $\\int_E\\omega^n \\le 1/2$ and choose an $L^p$ probability density $f$ with $\\int_E f\\omega^n = 0.6$; then $2\\,1_E f\\omega^n + b\\omega^n$ has total mass $1.2+b$, so no $b\\in[0,1]$ makes it a probability measure, and the auxiliary function used in the stability estimate is not defined.","tokens_in":13856,"feed_emoji":"📐","tokens_out":17517,"duration_ms":139079,"temperature":0.7,"pith_summary":"This paper proves a low-temperature uniqueness theorem for complex Monge-Ampère mean field equations. On a compact Kähler manifold $(X,\\omega)$ of dimension $n$, for every probability density $f \\in L^p(X,\\omega^n)$ with $p>1$, there is a threshold $\\gamma_0>0$ such that the equation $(\\omega+dd^c\\varphi)^n = e^{-\\gamma\\varphi} f\\omega^n$ has exactly one continuous solution for all $0<\\gamma<\\gamma_0$. The same mechanism gives uniqueness on bounded hyperconvex domains without smoothness assumptions on $f$ or the boundary, partially confirming a conjecture posed for the local mean field equation, and it adapts to compact Hermitian manifolds when $f$ is bounded below by a positive constant. The proof treats $\\gamma$ as a contraction constant: a refined stability estimate bounds the distance between any two solutions by a multiple of $\\gamma$ times that same distance, so the solutions coincide when $\\gamma$ is small.","feed_headline":"Low temperature forces a single Monge-Ampère mean field solution","feed_subtitle":"Small γ makes Monge-Ampère mean field solutions unique on Kähler manifolds and hyperconvex domains.","key_machinery":"The load-bearing mechanism is the refined stability estimate (Theorems 3.1 and 3.2): if $u,v$ are normalized solutions of $(\\omega+dd^c u)^n = f\\omega^n$ and $(\\omega+dd^c v)^n = g\\omega^n$ with $\\sup_X u = \\sup_X v = 0$, then $\\sup_X |u-v|$ is controlled by the $L^p$ distance between $f^{1/n}$ and $g^{1/n}$, up to constants depending on $p$ and uniform bounds on the densities. This estimate is assembled from the mixed Monge-Ampère inequalities, the domination principle for $\\omega$-psh functions, and an $L^\\infty$ estimate (Theorems 2.5 and 3.3) whose proof uses a variational construction and makes the constant $\\gamma_0$ explicit. In the uniqueness argument the same estimate is applied to the two densities of the mean field equation, whose $n$-th root distance is essentially $\\gamma |u-v|$ up to constants; the equation therefore becomes contractive in the sup norm.","core_discovery":"The central claim, Theorem 1.1, is that on a compact Kähler manifold $(X,\\omega)$ normalized by $\\int_X \\omega^n=1$, the equation $(\\omega+dd^c\\varphi)^n = e^{-\\gamma\\varphi} f\\omega^n$ has a unique continuous solution for every probability density $f\\in L^p(X,\\omega^n)$, $p>1$, provided $0<\\gamma<\\gamma_0(X,\\omega,n,p,\\|f\\|_p)$. The proof first establishes uniform boundedness of solutions through an $L^\\infty$ estimate, then rewrites any two solutions in normalized form and applies a stability estimate comparing potentials in terms of the $L^p$ distance of the $n$-th roots of their Monge-Ampère densities. The comparison produces $\\sup_X |u-v| \\le \\gamma C \\sup_X |u-v|$ with $C$ independent of the solution pair, so $\\gamma C<1$ forces $u=v$. In bounded hyperconvex domains the same strategy yields a unique solution in $\\mathrm{PSH}(\\Omega)\\cap L^\\infty$ for small $\\gamma$, and on compact Hermitian manifolds the conclusion holds when $f$ is strictly positive.","pith_inferences":["The threshold obtained is sufficient, not necessary; the projective-space example shows only that uniqueness fails at some larger $\\gamma$, so determining the largest threshold for each manifold and density is a natural open problem.","If the missing $f$-mass bound is supplied, the proof becomes unconditional for arbitrary $L^p$ densities; a natural way is to assume $f\\in L^\\infty$, since then $\\int_E f\\omega^n \\le (\\sup f)\\int_E\\omega^n$.","The Hermitian stability result is stated with a proof left to the reader; completing it would require checking whether the same sublevel-set $f$-mass step recurs there.","The $L^\\infty$ estimates of Theorems 2.5 and 3.3 are derived by solving an auxiliary variational equation; this construction may be reusable for other exponential nonlinearities to obtain explicit bounds rather than qualitative ones."],"forward_implications":["On any compact Kähler manifold, the mean field equation with an $L^p$ density has exactly one continuous solution for all temperatures below an explicit threshold depending only on the manifold, $n$, $p$, and $\\|f\\|_p$.","On bounded hyperconvex domains, uniqueness holds without boundary smoothness and without regularity of $f$, partially confirming the conjecture stated in [BB22].","The same construction yields uniqueness on compact Hermitian manifolds once $f$ is bounded below by a positive constant and $\\gamma$ is small.","The threshold $\\gamma_0$ is quantitative: the local version depends only on $n$, $p$, $\\|f\\|_p$, and the diameter of the domain, so the result can certify uniqueness in concrete problems.","The example on projective space in the paper shows uniqueness cannot hold for every $\\gamma$, so the threshold is genuinely finite and the optimal value remains open."],"supporting_citations":[{"why":"Supplies the uniform $L^\\infty$ bound for normalized Monge-Ampère solutions that makes the stability estimates and constants uniform.","marker":"[Ko/suppress l98]"},{"why":"Provides existence of continuous solutions for small $\\gamma$ on hyperconvex domains and states the conjecture that Theorem 1.2 partially confirms.","marker":"[BB22]"},{"why":"Gives the variational method used to produce solutions of the normalized equation inside the $L^\\infty$ estimate arguments.","marker":"[BBGZ13]"},{"why":"Establishes the stability estimate that Theorems 3.1 and 3.2 refine, and supplies the pattern used for Hermitian manifolds.","marker":"[LPT21]"},{"why":"Provides the uniform exponential-integrability estimate that turns $L^p$ densities into the bound (3.1) needed to make $\\gamma_0$ explicit.","marker":"[Zer01]"},{"why":"Supplies the comparison principle, domination principle, and variational existence results used throughout the proofs.","marker":"[GZ17]"},{"why":"Provides existence and continuity of solutions on compact Hermitian manifolds, used in Theorem 4.1 and the setup for Theorem 4.4.","marker":"[KN19]"}],"fun_headline_variants":["Small γ yields unique Monge-Ampère mean field solutions","Uniqueness at low temperature in Monge-Ampère mean field","Unique solution for Monge-Ampère mean field at small γ","Small temperature parameter ensures uniqueness in mean field","Monge-Ampère mean field equations: uniqueness for small γ"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof of the stability estimate constructs an auxiliary solution from the measure $2\\,1_E f\\omega^n + b\\omega^n$, which is a probability measure only if the $f$-mass of $E$ is at most $1/2$, but the text verifies only that the $\\omega^n$-mass of $E$ is at most $1/2$, not its $f$-mass.","fun_headline_variants_meta":{"raw":{"variants":["Small γ yields unique Monge-Ampère mean field solutions","Uniqueness at low temperature in Monge-Ampère mean field","Unique solution for Monge-Ampère mean field at small γ","Small temperature parameter ensures uniqueness in mean field","Monge-Ampère mean field equations: uniqueness for small γ"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001205,"raw_usage":{"total_tokens":4910,"prompt_tokens":839,"completion_tokens":4071,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":455,"completion_tokens_details":{"reasoning_tokens":3985}},"tokens_in":455,"tokens_out":4071,"duration_ms":24111,"temperature":1.0,"reasoning_tokens":3985,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T00:06:34.747415+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take $E=\\{u<v+\\alpha\\}$ with $\\int_E\\omega^n \\le 1/2$ and choose an $L^p$ probability density $f$ with $\\int_E f\\omega^n = 0.6$; then $2\\,1_E f\\omega^n + b\\omega^n$ has total mass $1.2+b$, so no $b\\in[0,1]$ makes it a probability measure, and the auxiliary function used in the stability estimate is not defined.","supporting_citations":[],"review_version":1}