{"id":"27c9cea3-f12b-4c20-9404-fb9dff10540f","arxiv_id":"2601.20572","paper_version":2,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper gives existence and uniqueness theorems for constant second Chern and Bismut scalar curvature metrics in fixed Hermitian conformal classes, but the proofs contain invalid maximum principle steps.","lead":"This paper proves existence of Hermitian metrics with constant second scalar curvature for both the Chern and Bismut connections within conformal classes on compact complex manifolds, under conditions such as vanishing first Betti number or a sign condition on a conformal invariant. The results are the second-scalar-curvature analogue of the Chern-Yamabe problem, but the proofs contain load-bearing gaps in a key lemma and a maximum principle argument.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 5.3's lower bound f_a >= 0 is false, and the closedness proof of Theorem 1.5 relies on this bound; the same maximum-principle calculation gives only a uniform bound that can be negative, so the C^0 estimate in the continuity method must be revised.","rationale":"I focused on Lemma 5.3 because it is the exact step on which the closedness proof of the continuity method in Theorem 1.5 depends, and the reader's weakest_assumption identifies the same step. The asserted lower bound is genuinely wrong: evaluating (5.9) at a minimum gives only e^f >= (1-a) + a(-S)/C, which can be less than 1. A constant solution with S close to zero gives a concrete counterexample to 0 <= f. I also examined the other flagged issue in Theorem 4.3: the strong maximum principle for Delta psi + a psi <= 0 with a > 0 is not valid as stated, since functions like 1 - cos x on a circle provide counterexamples. However, the nonlinear equation -Delta phi + N1 S phi = N1 mu phi^{q-1} with q > 2 prevents interior zeros by a Taylor-order argument, so that gap is more readily repairable than a false bound. The corrected lower bound in Lemma 5.3 is also straightforward, so the existence results are plausible, but as written the proof of Theorem 1.5 contains a concrete false inequality. This does not change the reader's REJECT verdict.","tokens_in":921,"tokens_out":929,"duration_ms":268082,"concrete_test":"Recompute Lemma 5.3 in a constant model: take u = 0, S(omega') = -delta < 0, lambda = -C, and a = 1. Then f = log(delta/C) is an exact constant solution of (5.6), and inserting it into (5.10) shows that 0 <= f fails when delta < C. Next, verify that replacing the lower bound 0 by the general bound e^f >= (1-a) + a(-S(q))/C, obtained at a minimum point, still yields a uniform C^0 estimate for all a in [0,1]; if it does, the continuity method closes and Theorem 1.5 survives with a corrected lemma.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In the continuity-method proof of Theorem 1.5, Lemma 5.3 claims that every solution f_a of (5.9) satisfies 0 <= f_a(x) <= 1 + min S/lambda. The upper bound follows from evaluating (5.9) at a maximum point. The lower bound does not: at a minimum point q, the inequality only yields C e^{f_a(q)} >= C(1-a) - a S(q) with C = -lambda > 0 and S < 0, i.e. e^{f_a(q)} >= (1-a) + a(-S(q))/C. This quantity is positive but can be strictly less than 1. For example, if u is constant so that S(omega') = -delta and lambda = -C with delta < C, then the a = 1 constant solution f = log(delta/C) solves (5.6) and violates f >= 0. Lemma 5.4 uses the claimed uniform C^0 bound to obtain W^{2,p} and Schauder estimates, so as written the closedness of A and hence Theorem 1.5 are not established. The defect is repairable: the same calculation yields a uniform lower bound involving -S/lambda, so the central claim is plausible, but the submitted proof contains a concrete false step in a load-bearing position.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies Hermitian metrics with constant second scalar curvature on compact complex manifolds. In a fixed Hermitian conformal class, it derives conformal transformation formulas for the curvatures of the Gauduchon connections, defines a new conformal invariant called the second Gauduchon degree, and proves existence and uniqueness theorems for constant second Chern scalar curvature whose sign is controlled by this degree. It also treats a Yamabe-type problem for the second Bismut scalar curvature under the assumption b_1(M)=0, proving that the conformal class contains a metric with constant second Bismut scalar curvature. Finally, under a weak Einstein-type condition involving the third and fourth Chern-Ricci curvatures, it shows that the second Chern scalar curvature is constant, and it gives non-Kahler examples of metrics with constant second Chern scalar curvature.","tokens_in":22739,"tokens_out":23731,"duration_ms":213495,"significance":"If the proofs are completed, the paper would meaningfully extend the Chern-Yamabe theory to second scalar curvatures, introduce a useful conformal invariant, and provide concrete non-Kahler examples. The paper is clearly organized, the computations are explicit, and the use of external results such as Gauduchon's theorem and Yang's Kodaira-dimension criteria is appropriate. However, two load-bearing analytic steps in the main existence proofs are invalid as written, so the significance of the results is currently conditional on a repaired argument.","major_comments":[{"comment":"The two-sided estimate (5.10) is not a consequence of (5.9), and the proof of closedness of the set A therefore breaks. Evaluating (5.9) at a maximum p and a minimum q gives, with S_C^(2)(omega')=lambda e^{-u}, only e^{f_a(p)} <= 1-a+a e^{-u(p)} and e^{f_a(q)} >= 1-a+a e^{-u(q)}. These imply uniform bounds of the form f_a <= max(0, -min_M u) and f_a >= min(0, -max_M u), but they do not imply f_a >= 0 nor the upper bound log(1+min_M S/lambda). The lower bound is in fact false on the local branch through a=0: differentiating F(a,f_a)=0 at a=0 gives (Delta+lambda)w = lambda(e^{-u}-1), where w = df_a/da at a=0. Since lambda<0 and Delta+lambda is negative definite, w is negative wherever u>0, so f_a<0 for small positive a. Lemma 5.4 uses the uniform C^0 estimate from Lemma 5.3 to obtain W^{2,p} and Schauder estimates, so Theorem 1.5 is not established as written. The defect is local: the corrected inequalities above still yield a uniform C^0 bound, so a revision can repair the argument.","section":"Section 5, Lemmas 5.3-5.4"},{"comment":"The proof that the weak minimizer phi_q is strictly positive is invalid. The argument chooses a>0 such that psi=-phi_q satisfies Delta psi + a psi <= 0 and then invokes the strong maximum principle. For an operator with a positive zeroth-order coefficient this conclusion is false: on S^1, psi = -(sin x)^4 is nonpositive with interior zeros and satisfies psi'' + a psi <= 0 for a >= 4. The coefficient in the equation actually satisfied by phi_q has uncontrolled sign, so the maximum principle cannot be applied in this way. A correct proof requires a Harnack-type inequality, a Moser iteration, or a subcritical approximation argument. Since Theorem 1.1 and Corollary 1.3 use the existence of a smooth positive solution of (4.2), the Bismut-Yamabe existence claim is not proved as written. In addition, the passage from the W^{1,2} weak solution to a smooth solution is not the immediate bootstrap claimed, because the right-hand side phi_q^{q-1} initially lies only in L^{q/(q-1)}; an iterative estimate is needed.","section":"Section 4, Theorem 4.3"}],"minor_comments":[{"comment":"The sentence 'Theorem 1.6 admits the following corollary' should refer to Theorem 1.1, not Theorem 1.6.","section":"Section 1, Corollary 1.3"},{"comment":"There are numerous typographical errors, including 'Hermtian', 'compacr', 'ans astheno', and 'Var' in the references; these should be corrected in a revised version.","section":"Throughout"},{"comment":"In equations (7.11) and (7.16), the notation S_C^(1)(omega_1) appears where the metric in Example 7.4 is omega_2; the labels should be made consistent.","section":"Section 7, Example 7.4"}],"recommendation":"major_revision","confidential_remarks":"The two main analytic gaps are load-bearing but localized, and each has a standard repair within the scope of the paper. I therefore recommend major revision rather than rejection. The paper's central claims are plausible, and the conformal formulas, the second Gauduchon degree, and the explicit examples are useful contributions. The author should be asked to rewrite the closedness argument in Section 5 using the corrected C^0 bounds and to supply a valid positivity/regularity argument in Section 4."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague —\n\nI read the paper with the stress-test note in hand, and the note is right. There are two genuine, load-bearing errors.\n\nFirst, Lemma 5.3 claims f_a ≥ 0. The maximum principle at a minimum point only gives e^{f_a} ≥ a(-S)/(-λ)+1-a. With S close to zero this is less than 1, so f_a can be negative; explicit constant solutions of (5.6) show it. The C^0 bound in Lemma 5.4 rests on this, so the closedness of the continuity set A, and hence Theorem 1.5, is not established as written.\n\nSecond, in Theorem 4.3 the positivity of the minimizer is derived from Δψ+aψ≤0 with a>0. That maximum principle does not exist when the zeroth-order coefficient is positive; the equation for ψ does not put the sign in the required place. This step is needed for the Bismut-Yamabe existence claim in Theorem 1.1.\n\nBoth defects are probably repairable. The same calculation in Lemma 5.3 yields a uniform lower bound involving -S/λ, so the continuity method has a reasonable chance with a revised C^0 estimate. The positivity issue should be handled by a subcritical Yamabe-type argument rather than a misapplied strong maximum principle. But a paper with two false steps in the two main existence proofs does not establish its advertised theorems.\n\nCredit where it is due: Section 3 is genuinely useful. The conformal transformation formulas for S_C^(2), S_B^(2), and the third/fourth Ricci curvatures of the Gauduchon connections are new and appear correctly derived. The examples in Section 7 are concrete and include non-Kähler metrics with constant second Chern scalar curvature. The structure is clear, the references are appropriate, and there is no circularity or hidden fitting. The citation pattern stays on the literature.\n\nWho is this for? Researchers working on Chern-Yamabe-type problems in Hermitian geometry. They will want the formulas from Section 3 and the examples even if they do not rely on the main theorems. I would not cite the existence theorems until they are repaired; I might cite the formulas after revision.\n\nMy recommendation: send it to peer review, because the area is active, the formulas and examples deserve circulation, and the flaws are concrete and fixable rather than fatal to the program. But the referee should demand a repaired Lemma 5.3 and a corrected positivity argument before publication.","headline":"New conformal formulas and examples, but the two main existence proofs have real maximum-principle errors; worth peer review but not acceptance as written.","tokens_in":23364,"tokens_out":8702,"would_cite":false,"duration_ms":80560,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53C55","53C21"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper shows that the existence of Hermitian metrics with constant second Chern or Bismut scalar curvature within a conformal class is governed by the sign of a conformal invariant, the second Gauduchon degree, with unique solutions in…","keywords":["second scalar curvature","Hermitian metrics","Bismut connection","Chern connection","Gauduchon metric","conformal class","Yamabe-type problem","second Gauduchon degree"],"falsifier":"Compute the left-hand side of (5.9) at a minimum point $q$ of a solution $f_a$ for $a=1$: the maximum principle gives $S^{(2)}_C(\\omega')(q)-\\lambda e^{f_1(q)}\\ge 0$, i.e. $e^{f_1(q)}\\ge S^{(2)}_C(\\omega')(q)/\\lambda$. Choose a conformal class whose second Chern scalar curvature $S^{(2)}_C(\\omega')$ is negative and nonconstant with $S^{(2)}_C(\\omega')(q)/\\lambda<1$ at the minimum point; then $f_1(q)\\le \\log(S^{(2)}_C(\\omega')(q)/\\lambda)<0$, contradicting the asserted uniform lower bound $f_a\\ge 0$ in Lemma 5.3.","tokens_in":22243,"feed_emoji":"📐","tokens_out":14531,"duration_ms":107302,"temperature":0.7,"pith_summary":"This paper asks whether a compact Hermitian manifold, a complex manifold with a metric that need not be Kähler, can be conformally deformed to a metric whose second scalar curvature is constant, where this curvature is a torsion-sensitive trace of the Chern or Bismut connection. The central claim is that the answer is controlled by a conformal invariant, the second Gauduchon degree: the total integral of the second Chern scalar curvature of the canonical Gauduchon metric in the class. When that invariant is zero, there is a unique conformal metric with vanishing second Chern scalar curvature; when it is negative, a unique conformal metric with constant negative second Chern scalar curvature, and in balanced classes the anti-canonical bundle is not pseudo-effective. For the Bismut connection, the paper proves existence of a constant second scalar curvature metric in every conformal class on manifolds with vanishing first Betti number. A final rigidity result shows that compact pluriclosed Gauduchon metrics satisfying a weak Einstein-type condition have constant second Chern scalar curvature, leading to a dichotomy with Kähler–Einstein metrics.","feed_headline":"Conformal invariant sign dictates constant second scalar curvature","feed_subtitle":"Zero or negative second Gauduchon degree forces existence, and Einstein-type conditions imply Kähler–Einstein rigidity.","key_machinery":"The key object is the second Gauduchon degree $\\Gamma^{(2)}_M(\\{\\omega\\})$, the total integral of the second Chern scalar curvature $S^{(2)}_C$ of the unique Gauduchon metric in the conformal class; it is a conformal invariant that plays the role of a torsion-aware Yamabe invariant. The load-bearing identity is the conformal transformation formula $S^{(2)}_C(\\omega_f)=e^{-f}\\left(S^{(2)}_C(\\omega)-\\Delta^C_\\omega f\\right)$ for the Chern connection, and the analogous formula (3.16) for the second Bismut scalar curvature $S^{(2)}_B$; these convert the geometric problem into solvable semilinear elliptic equations. The Bismut existence proceeds through the functional $Y_q(\\varphi)$ (4.3) and the equation $\\square_{\\omega_B}\\varphi=N_1\\mu_q\\varphi^{q-1}$, while the negative Chern case uses a continuity method with maximum-principle a priori estimates and a uniqueness argument by comparison. The rigidity section uses the sum of the third and fourth Chern–Ricci curvatures $\\Theta^{(3)}(\\omega)+\\Theta^{(4)}(\\omega)$, whose $\\partial\\bar\\partial$-closedness yields a linear elliptic equation $\\Diamond_\\omega f=0$ with only constant solutions on compact pluriclosed Gauduchon manifolds.","core_discovery":"The paper's central discovery is that within a fixed Hermitian conformal class the existence and uniqueness of constant second Chern scalar curvature is decided by the sign of the second Gauduchon degree $\\Gamma^{(2)}_M(\\{\\omega\\})=\\int_M S^{(2)}_C(\\omega_G)\\,\\omega_G^n/n!$, a conformal invariant computed from the Gauduchon representative $\\omega_G$ of the class. If $\\Gamma^{(2)}_M(\\{\\omega\\})=0$, the linear equation $\\Delta^C_{\\omega_G}f=S^{(2)}_C(\\omega_G)$ is solvable, producing a unique (up to scaling) metric $\\omega_a\\in\\{\\omega\\}$ with $S^{(2)}_C(\\omega_a)=0$, and the paper derives that either $\\kappa(M)=-\\infty$ or $\\kappa(M)=0$ with $K_M^{\\otimes m}=\\mathcal{O}_M$ for some $m\\in\\mathbb{Z}_+$. If $\\Gamma^{(2)}_M(\\{\\omega\\})<0$, the nonlinear equation $\\Delta^C_{\\omega'}f=-\\lambda e^f+S^{(2)}_C(\\omega')$, with $\\lambda=\\Gamma^{(2)}_M(\\{\\omega\\})\\mathrm{Vol}(M,\\omega_G)^{-1}$, is solved by a continuity method, giving a unique conformal metric $\\omega_c$ with $S^{(2)}_C(\\omega_c)=\\lambda<0$; in balanced classes the same method yields a metric with that constant as its first Chern scalar curvature and implies $K_M^{-1}$ is not pseudo-effective. For the Bismut connection, the paper proves that when $b_1(M)=0$ the Gauduchon metric is conformal to a balanced metric, and a variational argument on the functional $Y_q$ produces a smooth solution of the corresponding equation, hence a conformal metric of constant second Bismut scalar curvature. Finally, for a pluriclosed Gauduchon metric satisfying the weak second Hermitian–Einstein condition $\\Theta^{(3)}(\\omega)+\\Theta^{(4)}(\\omega)=f\\omega$, the paper proves $S^{(2)}_C(\\omega)$ is constant by showing the defining function $f$ solves a linear elliptic equation whose only solutions are constants; the surface case then forces either $S^{(2)}_C(\\omega)\\equiv 0$ or a Kähler–Einstein metric with negative scalar curvature.","pith_inferences":["The second Gauduchon degree $\\Gamma^{(2)}_M(\\{\\omega\\})$ behaves like a conformal 'charge' whose sign may classify Hermitian conformal classes; the positive case is left open in this paper, and Theorem 5.5 suggests it forces $\\kappa(M)=-\\infty$ and yields constant positive scalar curvature metrics on products with a high-genus curve.","The a priori lower bound $f_a\\ge 0$ in Lemma 5.3 appears incorrect as written: the maximum principle at a minimum point gives $e^{f_a}\\ge aS/\\lambda+1-a$, which allows negative values, so a corrected uniform $C^0$ bound would be needed to make the closedness step of Theorem 1.5 fully rigorous as presented.","The rigidity dichotomy (zero second Chern scalar curvature or Kähler–Einstein) suggests that weak second Hermitian–Einstein metrics form a bridge between Chern–Einstein and Kähler–Einstein geometry; one might test whether the non-positivity of $f$ in Corollary 1.9 can be relaxed to bounded-below $f$.","Because the examples include non-Kähler surfaces with $b_1=1$ (Inoue surfaces), the Bismut theorem's condition $b_1(M)=0$ is sufficient but not necessary; finding the sharp topological condition for Bismut constant scalar curvature would be a natural next step."],"forward_implications":["If $\\Gamma^{(2)}_M(\\{\\omega\\})<0$, every Hermitian conformal class contains a unique constant-negative second Chern scalar curvature metric, and in balanced classes the same constant is realized as the first Chern scalar curvature of a conformal metric while $K_M^{-1}$ is not pseudo-effective.","If $\\Gamma^{(2)}_M(\\{\\omega\\})=0$, the unique conformal metric with $S^{(2)}_C=0$ forces the Kodaira dimension to be either $-\\infty$ or $0$ with a torsion canonical bundle.","On any compact complex manifold with $b_1(M)=0$, including the non-Kähler Calabi–Eckmann manifolds in dimensions $n\\ge 3$, every Hermitian conformal class contains a metric of constant second Bismut scalar curvature.","A compact pluriclosed Gauduchon metric satisfying the weak second Hermitian–Einstein condition must have constant second Chern scalar curvature; on surfaces this gives a strict alternative between $S^{(2)}_C\\equiv 0$ and negative Kähler–Einstein.","The explicit examples (Hopf manifolds, non-Kähler properly elliptic surfaces, Inoue surfaces) show that constant second Chern scalar curvature is not a Kähler-only phenomenon."],"supporting_citations":[{"why":"Supplies the Chern–Yamabe existence and uniqueness strategy (openness, closedness, maximum-principle uniqueness) that the paper adapts to second scalar curvature.","marker":"[1]"},{"why":"Provides the theorem that every Hermitian conformal class contains a unique Gauduchon metric up to scaling, which underlies the definition of the second Gauduchon degree.","marker":"[21]"},{"why":"Provides the torsion and Laplacian identities used in the conformal transformation formulas and in the adjoint computation for the zero-degree linear equation.","marker":"[22]"},{"why":"Gives the curvature relations (2.7) for Gauduchon connections from which the conformal transformation formulas for second scalar curvatures are derived.","marker":"[42]"},{"why":"Supplies the Kodaira-dimension and pseudo-effectivity consequences used in Theorems 1.4, 1.5, and 5.5.","marker":"[44]"},{"why":"Provides the complex Laplacian and conformal scalar curvature identities used to solve the zero-degree equation and to derive conformal formulas.","marker":"[47]"}],"fun_headline_variants":["One sign to rule them all: constant second scalar curvature","How a single sign decides Hermitian metrics' constant curvature","Constant second scalar curvature: it's all about the sign","Sign of Gauduchon degree: gatekeeper to constant curvature","Zero or negative: signs that allow constant second scalar curvature"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The negative-case existence theorem rests on a uniform a priori bound for the approximating solutions of the continuity method, and the paper's derivation of that bound (the lower half of Lemma 5.3) is not correct as written, leaving the closedness step dependent on an unproved estimate.","fun_headline_variants_meta":{"raw":{"variants":["One sign to rule them all: constant second scalar curvature","How a single sign decides Hermitian metrics' constant curvature","Constant second scalar curvature: it's all about the sign","Sign of Gauduchon degree: gatekeeper to constant curvature","Zero or negative: signs that allow constant second scalar curvature"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000801,"raw_usage":{"total_tokens":3612,"prompt_tokens":1127,"completion_tokens":2485,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":743,"completion_tokens_details":{"reasoning_tokens":2403}},"tokens_in":743,"tokens_out":2485,"duration_ms":21210,"temperature":1.0,"reasoning_tokens":2403,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T15:40:12.335395+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the left-hand side of (5.9) at a minimum point $q$ of a solution $f_a$ for $a=1$: the maximum principle gives $S^{(2)}_C(\\omega')(q)-\\lambda e^{f_1(q)}\\ge 0$, i.e. $e^{f_1(q)}\\ge S^{(2)}_C(\\omega')(q)/\\lambda$. Choose a conformal class whose second Chern scalar curvature $S^{(2)}_C(\\omega')$ is negative and nonconstant with $S^{(2)}_C(\\omega')(q)/\\lambda<1$ at the minimum point; then $f_1(q)\\le \\log(S^{(2)}_C(\\omega')(q)/\\lambda)<0$, contradicting the asserted uniform lower bound $f_a\\ge 0$ in Lemma 5.3.","supporting_citations":[{"cited_title":"Angella, S","cited_arxiv_id":null,"evidence_quote":"Supplies the Chern–Yamabe existence and uniqueness strategy (openness, closedness, maximum-principle uniqueness) that the paper adapts to second scalar curvature."},{"cited_title":"Gauduchon, Le th´ eor` eme de l’excentricit´ e nulle, C","cited_arxiv_id":null,"evidence_quote":"Provides the theorem that every Hermitian conformal class contains a unique Gauduchon metric up to scaling, which underlies the definition of the second Gauduchon degree."},{"cited_title":"Gauduchon, La 1-forme de torsion d´ une vari` et` e hermitienne compacte, Math","cited_arxiv_id":null,"evidence_quote":"Provides the torsion and Laplacian identities used in the conformal transformation formulas and in the adjoint computation for the zero-degree linear equation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the curvature relations (2.7) for Gauduchon connections from which the conformal transformation formulas for second scalar curvatures are derived."},{"cited_title":"Yang, Scalar curvature on compact complex manifolds, Trans","cited_arxiv_id":null,"evidence_quote":"Supplies the Kodaira-dimension and pseudo-effectivity consequences used in Theorems 1.4, 1.5, and 5.5."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the complex Laplacian and conformal scalar curvature identities used to solve the zero-degree equation and to derive conformal formulas."}],"review_version":2}