{"id":"1d2a43f8-86c2-4de8-a1ae-5ab094c2565a","arxiv_id":"2509.09259","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The canonical singular Ricci-flat Kähler metric on a compact Kähler log terminal Calabi-Yau variety is orbifold-smooth on the orbifold locus.","lead":"A proof that singular Ricci-flat metrics on compact Kähler Calabi-Yau varieties with mild singularities are smooth on the orbifold part of the space. This removes the projectivity assumption from an earlier known result and gives a regularity statement for canonical metrics on non-projective Calabi-Yau varieties.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Uniform L∞ bound for approximating potentials (item C, proof of Thm 5.5) is asserted but not precisely established; if it fails, strict positivity degenerates and the limiting orbifold-regularity argument collapses.","rationale":"The reader's weakest-assumption analysis identifies the same load-bearing concern I find: the uniform L∞ bound on the approximating potentials φ_t (item C of the proof of Theorem 5.5) is necessary for uniform strict positivity, and the paper's justification for it is not fully rigorous. I have not found a more serious or independent concern; the other ingredients—the locally trivial algebraic approximation from [BGL22], the projective orbifold regularity from [LT19]/[GP24], and the uniform Green-kernel estimates from [GT25]—are cited with enough precision to be checkable, and the internal reasoning elsewhere is coherent. The admitted gap is fillable in principle, but until the sketched simultaneous-resolution argument is written out, the proof is conditional. Therefore I agree with the CONDITIONAL verdict and see no need to change it.","tokens_in":11415,"tokens_out":9143,"duration_ms":101658,"concrete_test":"Independently verify that [DNGG23, Thm 3.4 & 1.1] give a uniform L∞ bound for the family (5.2). Concretely: (1) construct the simultaneous resolution Y→X and fix the hermitian form ω_Y; (2) compute the L^{1+ε} norms of the pulled-back densities (π_t)_* μ_t with respect to ω_{Y_t}^n and confirm they are uniformly bounded in t; (3) trace the output of [DNGG23, Thm 3.4] to confirm it yields sup_t ||φ_t||_{L∞} < ∞. If either step fails, exhibit a one-parameter family of klt Calabi-Yau surfaces t↦X_t for which ε_0(t) from Lemma 3.3 tends to 0 as t→0, demonstrating the degeneracy.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central theorem (Theorem 5.5) proves orbifold regularity for the limiting solution ω_0 by approximating X by projective fibers X_{t_k}. The argument requires uniform control of several constants as t→0. The load-bearing step is item C (page 10): a uniform bound sup_{X_t} |φ_t| ≤ C for the solutions of (5.2). This bound is needed to apply Lemma 3.3 uniformly, yielding the strict positivity constant ε_0 in ω_t ≥ ε_0 ω_{X_t}. If ε_0 degenerates, then the Ricci lower bound in Claim 4.1, inequality (4.7), becomes t-dependent and the Harnack-based uniform bound from Proposition 4.2 no longer follows. The authors state that the precise uniform L∞ statement is 'not explicitly stated' in [DNGG23] and sketch an alternative via a simultaneous resolution Y→X, asserting that the pulled-back densities have uniform L^{1+ε} norm and that [DNGG23, Thm 3.4 & 1.1] then give the desired bound. However, these citations are not verified in the text, and the sketch omits the detailed check that the hypotheses of those theorems are satisfied for the family (5.2). Since the paper's novelty is precisely the non-projective case, and the projective case was already known, a failure of this uniformity would leave the main theorem unsupported. This is a genuine gap, not a fatal error, but it is the weak point on which the proof hinges.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proves Theorem A: if a compact Kähler variety X with log terminal singularities admits a locally trivial algebraic approximation, then the solution omega of the normalized Monge-Ampère equation (1.1) restricts to the orbifold locus X^orb as an orbifold Kähler metric. Combined with the algebraic approximation theorem of BGL22 for compact Kähler Calabi-Yau varieties, this yields Corollary B: every singular Ricci-flat Kähler metric on a compact Kähler klt variety with c1(K_X)=0 has orbifold singularities on X^orb. The proof approximates X by projective fibers X_t and aims to establish quantitative uniformity—of the strict positivity constant, of the orbifold Harnack estimate, and of the Laplacian bound—so that the limiting metric inherits the orbifold regularity. The paper also derives a metric-completion application in Section 5.2.","tokens_in":1489,"tokens_out":1642,"duration_ms":160582,"significance":"If the main theorem is fully supported, this is a meaningful advance: orbifold regularity was previously known for projective X via LT19 and GP24, and the current note extends it to all compact Kähler klt Calabi-Yau varieties by invoking locally trivial algebraic approximation. The quantitative Harnack-type statement in Proposition 4.2 is a useful tool in its own right. The paper is honest about its reliance on recent external results and about the projectivity limitation in Remark 2.5. However, the load-bearing uniformity step in Theorem 5.5—especially the uniform L-infinity bound on the approximating potentials phi_t—is only sketched, and the paper itself acknowledges that the precise statement is not in the cited literature. The result is therefore not yet fully supported as written, but the gap appears fixable.","major_comments":[{"comment":"The uniform bound sup_{X_t}|phi_t| <= C is load-bearing. It supplies the hypothesis phi_t >= -C in Lemma 3.3, which yields a t-independent epsilon0. If epsilon0 degenerates, then inequality (4.7) in Claim 4.1 becomes t-dependent and the Harnack-based argument in Proposition 4.2 collapses. The text states that this bound follows 'essentially' from [DNGG23] but that the precise statement is not explicitly there; the proposed simultaneous-resolution workaround is not sufficiently detailed. In particular, [BL22, Lemma 4.8] is invoked without stating the resolution properties, and the claims that the pull-backs of mu_t to Y_t have uniformly bounded L^{1+epsilon} density and that [DNGG23, Thm 3.4 and 1.1] give the desired sup bound are not checked. Since pi_t^*omega_{X_t} is not a Kähler form in general, the relation between the Monge-Ampère equations (5.2) on X_t and any equations controlled","section":"Section 5.1, item C, proof of Theorem 5.5"},{"comment":"The uniformity of the Green-function constants gamma and G is also asserted rather than proved. The text says it is a consequence of [GT25, Theorem A] applied to a simultaneous resolution, but the exact family version needed for (X_t^reg, omega_t) is not stated. Since the final constant C_U in Proposition 4.2 depends on gamma and G, this is another t-dependent quantity that must be controlled. If the cited theorem indeed covers this family, the verification should be spelled out; otherwise a proof is required.","section":"Section 5.1, item gamma,G, proof of Theorem 5.5"},{"comment":"The passage from the Green-function inequality (4.13) to the L-infinity bound for f relies on the existence of a uniform L^1 bound for f with respect to omega^n and on the choice alpha = gamma/(gamma+1). The argument is correct if the input constants are uniform. However, the proposition as stated requires an upper bound for the integral of omega_V wedge p^*omega^{n-1}, and Remark 4.3 explains that this follows from the L-infinity bound on phi. Thus Proposition 4.2 is only as good as the uniform L-infinity control from Theorem 5.5. This is not an independent flaw, but it means the gap in item C cannot be circumvented by the Harnack estimate alone.","section":"Section 4.2, Proposition 4.2"}],"minor_comments":[{"comment":"The phrase 'énième conséquence' is informal and unclear; it should be replaced by a precise statement such as 'a direct consequence' or a specific reference.","section":"Section 3, proof of Lemma 3.3"},{"comment":"The construction of the cutoff functions tau_k and their G-invariant averaging is only summarized. It would help to spell out that the averaging preserves the integrability estimate (4.11), since the inequality is applied after descending to U.","section":"Section 4, proof of Proposition 4.2"},{"comment":"The norm notation for G_x should specify the measure and the fact that G_x is singular; clarify that the norm is taken with respect to the volume form omega^n on X_reg.","section":"Section 4, equation (4.13)"},{"comment":"The abstract says 'any singular Ricci-flat Kähler metric', while the proof treats the unique solution of (1.1) with a flat hermitian metric. This is standard and likely intended, but the equivalence should be stated explicitly for the reader.","section":"Abstract and Corollary B"}],"recommendation":"major_revision","confidential_remarks":"To the editor: the manuscript is honest about the gap in item C of Theorem 5.5, and the overall argument is likely correct. My recommendation of major_revision is driven by the need to convert the sketched uniform L-infinity bound into a checkable proof. The paper relies heavily on the authors' own preprint [GP24] and on several recent preprints ([CCH+25], [GT25]); it would be prudent to confirm their status at revision time."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a legitimate new extension of the known orbifold regularity result for singular Ricci-flat metrics to non-projective compact Kähler klt Calabi-Yau varieties. The proof is mostly transparent, but one explicitly flagged uniformity estimate is not fully proven, and the paper should be sent to a referee with a request to close that gap.\n\nThe new content is real. For projective varieties, orbifold regularity was proved by Li–Tian and then by Guenancia–Păun. The all-orbifold and isolated-singularity cases were also known. What this note adds is the non-projective case: using the locally trivial algebraic approximation theorem of Bakker–Guenancia–Lehn, they reduce to projective fibers and push uniform estimates to the limit. The quantitative Proposition 4.2, which converts qualitative orbifold regularity into uniform trace bounds, is a nice piece of work. The use of the recent strict positivity theorem of Chen–Chiu–Hallgren–Székelyhidi–Tô–Tong is natural, and the authors are transparent about relying on it.\n\nThe soft spot is item C in the proof of Theorem 5.5. The argument needs a uniform L∞ bound on the approximating potentials φ_t. The authors say this follows 'essentially' from DNGG23 but admit the precise statement is not there; they then sketch an alternative via a simultaneous resolution. The sketch is plausible but omits the verification that the hypotheses of the cited theorems hold for the family. If this uniform bound fails, the positivity constant ε0 degenerates and the Harnack-based estimate collapses, so this is genuinely load-bearing. It is not fatal—the authors flag it and the route to a fix seems credible—but it needs a careful write-up.\n\nThe rest of the proof is in decent shape. The limiting argument in Theorem 5.5 is standard once the uniform bounds are available, and the metric distance application in Section 5.2 is elementary. The self-citations are minor and appropriate.\n\nWho should read this: anyone working on singular Kähler–Einstein metrics, moduli of Calabi-Yau varieties, or orbifold regularity. It deserves a serious referee. My recommendation: send it to peer review, and ask the referee to demand a complete proof of item C. That is the one thing standing between this note and a solid result.","headline":"A genuine extension of orbifold regularity to non-projective klt Calabi-Yau varieties via a quantitative deformation argument, with one explicitly flagged uniformity gap that needs a rigorous fix before the proof is complete.","tokens_in":12240,"tokens_out":2840,"would_cite":true,"duration_ms":30418,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32Q20","32Q25","53C55","32W20"],"pacs":[],"model":"deepseek-v4-flash","headline":"On a compact Kähler variety with log terminal singularities and trivial canonical class, the unique singular Ricci-flat metric is smooth after pullback to local covers of quotient singularities.","keywords":["orbifold regularity","singular Ricci-flat metrics","log terminal singularities","Kähler Calabi-Yau varieties","complex Monge-Ampère equation","locally trivial algebraic approximation","quotient singularities"],"falsifier":"Exhibit a compact Kähler variety X with log terminal singularities and c1(K_X)=0, and a point x in X^orb, such that the pullback of the singular Ricci-flat metric to a local uniformizing cover is not a smooth Kähler metric—e.g., its trace against the Euclidean metric is unbounded. Alternatively, construct a locally trivial algebraic approximation satisfying the paper’s Assumption 5.1 for which sup_{X_t} |phi_t| is unbounded, which would break the uniform positivity step and show the proof’s mechanism fails.","tokens_in":11313,"feed_emoji":"📐","tokens_out":8232,"duration_ms":95228,"temperature":0.7,"pith_summary":"The paper proves a regularity result for singular Ricci-flat Kähler metrics on compact Kähler varieties with log terminal singularities and trivial canonical class: near every finite-quotient singularity, the metric, pulled back to a smooth local cover, extends to a smooth Kähler metric. This is the first proof of this orbifold regularity in the non-projective Kähler setting, extending known projective theorems. The strategy is to degenerate the variety to nearby projective fibers, where the result is already known, and to control all constants uniformly along the degeneration. The engine is a quantitative bound on the trace of the pulled-back metric on local uniformizing covers.","feed_headline":"Singular Ricci-flat metrics are smooth on quotient covers","feed_subtitle":"Extends orbifold regularity from projective to all compact Kähler Calabi-Yau varieties with log terminal singularities.","key_machinery":"The central object is the trace function f = tr_{p^* omega} omega_V on a local uniformizing cover V of a quotient singularity, and the proof’s goal is a uniform L^infinity bound for f. The machinery combines: (1) Chern–Lu inequality plus a recent strict-positivity theorem to get omega >= epsilon omega_X with a uniform epsilon; (2) an elliptic inequality for small powers of a cut-off f, namely Delta_omega (chi f^alpha) >= -C chi f^alpha; (3) a Harnack-type inequality based on Green-kernel estimates that turns an L^1 bound into an L^infinity bound; and (4) a locally trivial deformation to projective fibers, where orbifold regularity is already known and the constants are uniform. A final Evans","core_discovery":"Let X be a compact Kähler space with log terminal singularities and c1(K_X)=0, and let omega be the unique solution of the complex Monge-Ampère equation (omega_X + dd^c phi)^n = mu_h, where h is a flat metric on K_X. The paper establishes that omega restricted to the orbifold locus X^orb has orbifold singularities: for any local finite Galois cover p: V -> U of a quotient singularity, p^* omega extends to a smooth Kähler metric on V. More generally, Theorem A proves the same conclusion whenever X admits a locally trivial algebraic approximation, with no condition on the canonical class; the c1=0 hypothesis enters only through a known algebraic-approximation result. The authors also derive th","pith_inferences":["The uniform trace bound is quantitative and may be reusable to control the degeneration of Ricci-flat metrics in families, for instance yielding uniform diameter or energy estimates along algebraic approximations of Calabi-Yau varieties.","The method suggests that orbifold regularity is a stable property under locally trivial degenerations: if nearby projective fibers admit the regularity with uniform constants, the central fiber inherits it. This may serve as a template for other singular canonical metrics.","The technical gap flagged in item C of the proof—the uniform L^infinity bound on the potentials along the family—could be filled by developing the sketched simultaneous-resolution argument in full detail; if a counterexample to that bound exists, the proof would need a different route, though the conclusion might still hold by other means."],"forward_implications":["The unique singular Ricci-flat Kähler metric in any given class on such a variety is genuinely smooth on the local covers of all quotient singularities; near the orbifold locus there are no singularities beyond those of the quotient structure.","Any compact subset K avoiding the non-orbifold locus has its metric completion homeomorphic to K, and the Ricci-flat distance is bi-Hölder to a fixed Kähler metric.","Because the c1=0 hypothesis is used only to guarantee locally trivial algebraic approximation, any future proof of that approximation property for broader classes would immediately extend the orbifold-regularity theorem to them.","Combined with previously known projective results, the paper implies that all singular Kähler-Einstein metrics (negative, zero, or positive first Chern class) on log terminal compact Kähler varieties have orbifold singularities on the orbifold locus."],"fun_headline_variants":["Ricci-flat Kähler metrics are orbifold regular","Singular CY metrics have orbifold singularities","Canonical metrics are orbifold on quotient covers","Quotient covers smooth Ricci-flat Kähler metrics","All singular Ricci-flat metrics are orbifold"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The proof needs a uniform L^infinity bound on the Monge-Ampère potentials phi_t along the approximating projective fibers; the authors note this follows only “essentially” from a prior theorem without the precise statement, and they give only a sketch via simultaneous resolution, so if this bound fails the uniform strict positivity constant may degenerate and the limit argument collapses.","fun_headline_variants_meta":{"raw":{"variants":["Ricci-flat Kähler metrics are orbifold regular","Singular CY metrics have orbifold singularities","Canonical metrics are orbifold on quotient covers","Quotient covers smooth Ricci-flat Kähler metrics","All singular Ricci-flat metrics are orbifold"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00068,"raw_usage":{"total_tokens":2843,"prompt_tokens":581,"completion_tokens":2262,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":325,"completion_tokens_details":{"reasoning_tokens":2180}},"tokens_in":325,"tokens_out":2262,"duration_ms":21086,"temperature":1.0,"reasoning_tokens":2180,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T19:24:19.130447+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Exhibit a compact Kähler variety X with log terminal singularities and c1(K_X)=0, and a point x in X^orb, such that the pullback of the singular Ricci-flat metric to a local uniformizing cover is not a smooth Kähler metric—e.g., its trace against the Euclidean metric is unbounded. Alternatively, construct a locally trivial algebraic approximation satisfying the paper’s Assumption 5.1 for which sup_{X_t} |phi_t| is unbounded, which would break the uniform positivity step and show the proof’s mechanism fails.","supporting_citations":[],"review_version":1}