{"id":"ba5a8042-13d4-42d1-a60b-b7f5ac645642","arxiv_id":"2510.06648","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Compact Hermitian surfaces with semi-definite Strominger–Bismut-Ricci curvature and vanishing (2,0)-Ricci (or parallel torsion) must be Kähler; a third set of results is conditional on an unproven constant.","lead":"This paper proves Kählerness criteria: compact Hermitian surfaces with semi-definite Strominger–Bismut-Ricci curvature (plus a (2,0)-curvature condition, or parallel torsion) are forced to be Kähler. It extends a 2025 program of X. Yang from the Levi-Civita connection to the torsionful Bismut connection used in string compactifications.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorems 1.1–1.4 hinge on the imported identity (3.32) from a concurrent preprint; a coefficient or sign error there would alter the curvature-to-torsion inequality and could invalidate the main Kählerness conclusion.","rationale":"The reader's weakest assumption correctly identifies the imported identity (3.32) as the load-bearing step. Theorems 1.1–1.4 reduce to a short manipulation of Lemma 3.4, and Lemma 3.4 is an unproved black box from a concurrent preprint. Since the central claim would fail if (3.32) had a sign or coefficient error, this is a genuine correctness risk rather than a stylistic objection. I agree with the reader that this warrants a CONDITIONAL verdict. I also note the Section 6 constant-a issue is a real mathematical gap, but it does not affect the main package Theorems 1.1–1.4, so it does not change the overall verdict. The proposed concrete test — reproducing (3.32) independently or testing it on a concrete Hopf-surface example — would settle whether the concern actually lands. No ad hominem or theatrical framing is needed; the issue is purely evidential and computational.","tokens_in":20885,"tokens_out":12143,"duration_ms":72343,"concrete_test":"Reproduce identity (3.32) from the paper's own established identities (Proposition 3.2 and Lemma 3.3) together with the standard Hermitian-surface identities cited in §2, without invoking [36]. Equivalently, compute both sides of (3.32) (or directly of (3.29)) for an explicit non-Kähler compact Hermitian surface with nonvanishing torsion, e.g. a diagonal Hopf surface with a standard Hermitian metric, using symbolic or high-precision numerical integration. If the equality fails or the coefficient of (|\\bar{\\partial}^*\\omega|^4,1) differs, the derivation of (5.2) and hence the stated curvature condition in Theorem 1.1 is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central chain Theorems 1.1–1.4 rests on Lemma 3.4, whose identities (3.29)–(3.31) are not proved in this paper. They are algebraic reformulations of Yang's identity (3.32), quoted from the concurrent, unrefereed preprint [36]. In particular, Theorem 1.1 combines (3.29) with (5.1) to obtain the decisive estimate (5.2). The correct coefficient of the (|\\bar{\\partial}^*\\omega|^4,1) term and the precise definition of Ric^{(1,1)} in (3.33) are essential: an error there changes the 7/2 coefficient in (1.4) or the sign of the non-positivity condition, and the inequality \\|\\bar{\\partial}\\bar{\\partial}^*\\omega\\|^2 + \\|\\Lambda\\bar{\\partial}\\bar{\\partial}^*\\omega - 3|\\bar{\\partial}^*\\omega|^2\\|^2 \\le 0 would no longer follow. The paper gives the reader no way to check (3.32) without consulting another preprint, so the main theorems are not self-contained at their most load-bearing step. A secondary but distinct gap is Section 6: (6.1) is asserted 'by compactness', but uniform boundedness of \\|R^{SB,C}_{ij}+R^{SB,C}_{ji}-3T_iT_j\\|^2 / (|\\bar{\\partial}^*\\omega|^4,1) is not implied when the denominator vanishes; this affects Theorems 6.1–6.3, though not Theorems 1.1–1.4.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper establishes Kählerness criteria for compact Hermitian surfaces under semidefiniteness assumptions on natural Ricci curvatures of the Strominger–Bismut connection. The main theorems (1.1–1.4) assume that the (2,0)-component of the complexified real Bismut–Ricci curvature vanishes and that a suitable corrected (1,1)-Bismut–Ricci form is non-positive, e.g. Ric^{SB(2)} + (7/2)√−1 ∂̄*ω∧∂*ω ≤ 0. The proofs reduce the hypotheses via integral identities in §3 to the conclusion that a sum of squares involving ∂̄∂̄*ω is ≤ 0; hence ∂̄∂̄*ω=0 and, by (5.3), ∂ω=0. Section 4 derives Chern-number identities, used together with a parallel-torsion hypothesis in Theorem 1.5 to remove the (2,0)-vanishing condition and to classify the Kähler limit as projective or Calabi–Yau. Section 6 states boundedness variants with a constant a.","tokens_in":21217,"tokens_out":16857,"duration_ms":109961,"significance":"If the main identities are correct, the results are strong and geometrically natural: they give Bismut–Ricci analogues of Yang's Riemannian criteria and, conditional on the (2,0)-vanishing hypothesis, rule out non-Kähler compact complex surfaces under explicit semidefinite curvature inequalities. The paper's strengths are the explicit nature of the torsion/curvature identities (Lemma 3.1), the careful bookkeeping of coefficients, and the Bochner-type final argument, which makes the theorems falsifiable and the constants precise. The main correctness risk is the reliance of Lemma 3.4 on the imported identity (3.32) from the concurrent preprint [36]; the local checks I made of the subsequent algebra (e.g., (5.2)) are consistent, but the foundation must be independently verifiable in the present manuscript.","major_comments":[{"comment":"This identity is the engine for Theorems 1.1–1.4, and it is quoted from the concurrent preprint [36] without proof. Equations (3.29)–(3.31), and therefore the decisive estimate (5.2), depend on its exact coefficients and signs. A missing or erroneous term would change the constants 7/2, 3/2, 6, 5 in Theorems 1.1–1.4. Please include a complete proof of (3.32), or a detailed derivation in the notation of §3, rather than sending the reader to [36].","section":"Section 3, Lemma 3.4, Eq. (3.32)"},{"comment":"The line 'It follows from (2.21) and (3.2) that R^{SB,C}_{ij} = T_iT_j = 0' is not immediate. From (3.2), parallel torsion gives ∇_j T_i = 0 and hence R^{SB,C}_{ij} = T_iT_j; the conclusion T_iT_j = 0 needs an additional argument, presumably from the Kähler-like symmetry in [41,42]. Without that, the reduction to (5.10) and the classification in Theorem 1.5 rest on an unstated fact. Please supply the missing argument or give the precise statement in [41,42].","section":"Section 5, proof of Theorem 1.5, after (5.9)"}],"minor_comments":[{"comment":"I do not see the claimed gap concerning vanishing of the denominator. The right-hand side of (6.1) is the global integral (|∂̄*ω|^4,1), not a pointwise denominator. If the integral is positive, compactness gives a finite constant a; if it is zero, then ∂̄*ω=0, M is Kähler, and the left side is also zero. The phrase 'throughout M' is confusing, but the argument is valid.","section":"Section 6, Eq. (6.1)"},{"comment":"There are several typos: the Section 3 heading reads 'Stromonger-Bismut'; the abstract has 'for achieve these results' instead of 'for achieving'; and (1.3) has 'of of'. These should be corrected.","section":"Global"},{"comment":"The sentence '(4.13) follows by (2.2)' appears to refer to the wrong equation; it should cite Lemma 2.2 or equations (2.19)–(2.20).","section":"Lemma 4.3"},{"comment":"The notation Ric^{SB,C}_{(1,1)} + Ric^{SB,C}_{(1,1)} in (1.6), (1.9), and (5.4) is hard to read in the typeset version; if the second term is the conjugate, please make the overline visible. If the two terms are indeed identical, the displayed redundancy should be explained.","section":"Theorems 1.2 and 1.4"},{"comment":"The equality ∥T_iT_j∥² = (|∂̄*ω|⁴,1) is used several times later; stating it explicitly after (3.17) would improve readability.","section":"Proof of Proposition 3.2, Eq. (3.17)"}],"recommendation":"major_revision","confidential_remarks":"The decisive issue is self-containedness: Theorems 1.1–1.4 rest on the unproved identity (3.32) from Yang's concurrent preprint. It may be appropriate to ask the author to include a proof or to coordinate with Yang so that the identity is independently verifiable before acceptance. The Section 6 concern raised in the stress test does not, in my reading, actually land for the reason stated in the minor comments."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know about arXiv:2510.06648. First, the genuinely new content is a set of explicit identities for the Strominger-Bismut connection on Hermitian surfaces — the key one being R^SB,C_{ij} = -∇_j T_i + T_i T_j (Lemma 3.1) — and the reformulation of Yang's Levi-Civita Kählerness criteria in terms of Bismut-Ricci curvatures. Theorems 1.1–1.4 are new statements and the short derivations in §5 check out, provided you grant the imported identity (3.32). Second, that identity is the soft spot: it is quoted from Yang's concurrent preprint [36] and used as a black box at the load-bearing step. A sign or coefficient error in (3.32) would change the constants 7/2, 3/2, 6, 5 in Theorems 1.1–1.4 and the non-positivity conclusion would not follow. The paper gives the reader no way to verify it without reading another preprint. That is the main reason not to accept the paper as is.\n\nWhat is genuinely good: Lemma 3.1 is a real contribution, and the torsion trace computations (T^p_{kj}T^k_{pi}=T_iT_j, T^k_{ij}T_k=0 in dimension 2) are correct. The algebra in §3–§5 is straightforward and mostly verifiable. Theorems 1.1–1.4 are the first Bismut-connection analogues of Yang's criteria, and that is a useful extension.\n\nThe soft spots beyond (3.32): Section 6 asserts the existence of a constant a with ||R^SB,C_{ij}+R^SB,C_{ji}-3T_iT_j||² ≤ a(|∂̄*ω|^4,1) 'by compactness'. The ratio diverges at points where the denominator vanishes, so Theorems 6.1–6.3 are not proved. The proof of Theorem 1.5 also borrows the Gauduchon identity (5.7) without the Gauduchon hypothesis, and the conclusion 'torus or K3' over-reads c1=0 as a real class; Enriques and hyperelliptic surfaces should be handled. These issues do not affect Theorems 1.1–1.4, so the main chain survives if (3.32) is valid.\n\nThe paper will be useful to people working on Hermitian surfaces, the Bismut connection, and non-Kähler geometry. It deserves a serious referee, not a desk reject, but the referee should ask for a proof or full statement of (3.32), a fix for Section 6, and corrections to Theorem 1.5. I would send it to review with those requests.","headline":"New Bismut-connection Kählerness criteria with a genuinely useful torsion identity, but the main theorems lean on an unproved imported identity and Section 6 has a real gap.","tokens_in":21877,"tokens_out":4925,"would_cite":false,"duration_ms":36788,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53C55"],"pacs":[],"model":"deepseek-v4-flash","headline":"If the (2,0) part of the complexified real Strominger-Bismut-Ricci curvature vanishes and the second SB-Ricci curvature obeys a torsion-corrected non-positivity bound, then a compact Hermitian surface is Kähler.","keywords":["Hermitian surfaces","Strominger-Bismut connection","Bismut-Ricci curvature","Kähler surface","torsion","Chern number identities","compact complex surfaces","non-Kähler geometry"],"falsifier":"Take a compact non-Kähler Hermitian surface with an explicitly given metric and numerically compute both sides of the quoted identity (3.32); any discrepancy disproves the main theorems. Alternatively, search for a compact Hermitian surface satisfying Ric^{SB,C}_{(2,0)} = 0 and Ric^{SB(2)} + (7/2)√-1 ∂̄*ω∧∂*ω ≤ 0 that is not Kähler; its existence would directly refute Theorem 1.1.","tokens_in":20639,"feed_emoji":"📐","tokens_out":13127,"duration_ms":88053,"temperature":0.7,"pith_summary":"The paper proves several Kähler criteria for compact Hermitian surfaces: under certain semi-definiteness conditions on the Ricci curvatures of the Strominger-Bismut connection, the surface must actually be Kähler, meaning its torsion vanishes. The central mechanism is a set of explicit identities that tie these Ricci curvatures to the anti-holomorphic piece of the torsion form ∂̄*ω. In the main theorem, vanishing (2,0) Ricci together with non-positivity of Ric^{SB(2)} + (7/2)√-1 ∂̄*ω∧∂*ω forces the squared norm of ∂̄∂̄*ω to be zero, which implies ∂ω = 0. These results reinterpret and extend the curvature–torsion identities previously discovered for the Levi-Civita connection, and if correct they rule out all non-Kähler compact complex surfaces under the stated hypotheses.","feed_headline":"A torsion-corrected Ricci sign forces Hermitian surfaces to be Kähler","feed_subtitle":"With the (2,0) Ricci part zero, the torsion-corrected curvature bound forces every compact Hermitian surface to be Kähler.","key_machinery":"The Strominger-Bismut connection is the unique Hermitian connection with totally skew-symmetric torsion; its real Ricci curvature, after complexification, splits into (2,0), (1,1), and (0,2) parts. The engine of the paper is a curvature–torsion identity (Lemma 3.4) expressing ∥∂̄∂̄*ω∥² + ∥Λ∂̄∂̄*ω∥² as a sum of pairings of the SB-Ricci curvatures with the nonnegative form √-1 ∂̄*ω∧∂*ω, together with squares of the (2,0)-Ricci part and torsion terms. Combining this identity with the assumed inequality converts the geometric curvature condition into an L² estimate that leaves no room for nonzero torsion. The Chern-number identities of Section 4 play the analogous role in the parallel-torsion an","core_discovery":"The paper's central claim is that on a compact Hermitian surface (M,ω), the combination Ric^{SB,C}_{(2,0)} = 0 and Ric^{SB(2)} + (7/2)√-1 ∂̄*ω∧∂*ω ≤ 0 forces the identity ∥∂̄∂̄*ω∥² + ∥Λ∂̄∂̄*ω − 3|∂̄*ω|²∥² ≤ 0, so ∂̄∂̄*ω = 0 and hence ∂ω = 0; therefore (M,ω) is Kähler. Variants with the third and fourth Strominger-Bismut-Ricci curvatures, and with Gauduchon metrics, relax the constant 7/2 to 3/2 or 5/2. A further theorem shows that if the Strominger-Bismut connection has parallel torsion, semi-definiteness of any of the natural SB-Ricci curvatures implies the surface is projective or Calabi-Yau.","pith_inferences":["The proof is essentially an L² gap argument, so the same mechanism might yield Kähler criteria in higher dimensions once an analogue of the curvature–torsion identity is established for Hermitian manifolds of arbitrary dimension.","The constant 7/2 in Theorem 1.1 is likely not optimal; parameterizing the torsion correction in the identity could reveal the sharp threshold beyond which non-Kähler metrics are possible.","A direct numerical check of the imported identity on an explicit compact non-Kähler Hermitian surface (one with an explicit metric) would settle the proof's reliance on the companion preprint independently of geometric intuition.","If these criteria hold, they provide a curvature-only obstruction to non-Kählerity, which could be useful for designing geometric flows that preserve the inequality and converge to a Kähler metric."],"forward_implications":["If Theorem 1.1 is correct, any compact Hermitian surface with vanishing (2,0) part of the complexified real SB-Ricci curvature and with Ric^{SB(2)} + (7/2)√-1 ∂̄*ω∧∂*ω ≤ 0 is Kähler, so its underlying complex surface lies in the Kähler class.","Under the same vanishing (2,0) assumption, the alternative bounds of Theorem 1.2 on Ric^{SB(3)} + Ric^{SB(4)} or on Ric^{SB,C}_{(1,1)} plus its conjugate also force Kählerity.","When the metric is Gauduchon, the torsion-correction constant can be lowered to 3/2 or 5/2 and the conclusion still holds, giving stronger statements for the standard conformal class on compact complex surfaces.","If the Strominger-Bismut connection has parallel torsion, semi-definiteness of any of the four natural SB-Ricci curvatures forces the surface to be either projective or Calabi-Yau (a torus or a K3 surface)."],"fun_headline_variants":["Semi-definite SB-Ricci forces Hermitian surfaces Kähler","Torsion-tuned Ricci sign forces compact Hermitian surfaces Kähler","SB-Ricci semi-definite implies Kähler on compact Hermitian surfaces","Semi-definite torsion-corrected Ricci forces Kähler surfaces"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole argument leans on an L² identity that is quoted from another preprint without proof; a sign or coefficient error in that identity would invalidate the main theorems.","fun_headline_variants_meta":{"raw":{"variants":["Semi-definite SB-Ricci forces Hermitian surfaces Kähler","Torsion-tuned Ricci sign forces compact Hermitian surfaces Kähler","SB-Ricci semi-definite implies Kähler on compact Hermitian surfaces","Semi-definite torsion-corrected Ricci forces Kähler surfaces"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000848,"raw_usage":{"total_tokens":3502,"prompt_tokens":693,"completion_tokens":2809,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":437,"completion_tokens_details":{"reasoning_tokens":2735}},"tokens_in":437,"tokens_out":2809,"duration_ms":15942,"temperature":1.0,"reasoning_tokens":2735,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T11:12:23.653319+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a compact non-Kähler Hermitian surface with an explicitly given metric and numerically compute both sides of the quoted identity (3.32); any discrepancy disproves the main theorems. Alternatively, search for a compact Hermitian surface satisfying Ric^{SB,C}_{(2,0)} = 0 and Ric^{SB(2)} + (7/2)√-1 ∂̄*ω∧∂*ω ≤ 0 that is not Kähler; its existence would directly refute Theorem 1.1.","supporting_citations":[],"review_version":1}