{"id":"85392a1b-8285-4020-8cc4-342ddbc3f8c6","arxiv_id":"2502.09825","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Singular Kähler-Einstein metrics on klt pairs define Kähler currents, and a tame approximation with L^p Ricci control suffices to make the metric completion an RCD space.","lead":"A new proof shows that singular Kähler-Einstein metrics on klt pairs are always Kähler currents, meaning they bound a fixed smooth Kähler metric from below. The same technique yields a new criterion for when such singular spaces are RCD spaces with synthetic Ricci curvature bounds.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 3.1 depends on solving (3.4), a Monge-Ampère equation whose density contains e^{A u_ε}; the paper cites [14, Thm 4.1] for uniform L∞ bounds without checking hypotheses, leaving the existence of the tame approximation (and hence Theorem 1.1) conditional.","rationale":"The paper is a serious advance and the overall strategy (approximate by metrics with L1-controlled negative Ricci, then apply Guo-Phong-Song-Sturm heat-kernel estimates) is coherent. The main theorem for klt pairs is plausible, and the L^p integrability of e^F highlighted by the reader is indeed satisfied for klt pairs, so that specific assumption is not a flaw in the intended application. However, the proof of Proposition 3.1 is the load-bearing construction, and its key analytic input is compressed into one citation asserting a uniform L∞ bound for solutions of (3.4). The equation has an exponential dependence on the unknown u_ε, which is not the standard fixed-density Monge-Ampère setup. A careful referee should verify that [14, Theorem 4.1] covers exactly this equation and that the resulting bound is uniform in ε under the stated hypotheses. This is a verification task, not a known counterexample, so it does not warrant rejection; it reinforces the reader's CONDITIONAL verdict.","tokens_in":19999,"tokens_out":25560,"duration_ms":242018,"concrete_test":"Consult [14, Theorem 4.1] and check whether it applies verbatim to (3.4) for each fixed ε > 0, with A > 0 and density f_ε = e^{-G_ε + g_ε} in L^p, and whether it yields ∥u_ε∥_{L∞} controlled by ∥f_ε∥_{L^p} and the geometry of (Y, π*ω_X + εω_Y). If the theorem only handles equations with a fixed (u-independent) density, then Proposition 3.1 needs a separate existence and uniform L∞ argument for (3.4); providing that argument is the check.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central proof rests on Proposition 3.1, which constructs the tame approximations ω_ε by solving (3.4): (π*ω_X + εω_Y + i∂∂̄u_ε)^n = e^{A u_ε - G_ε + g_ε} ω_Y^n. This is not a standard complex Monge-Ampère equation with fixed density: the right-hand side depends on the unknown u_ε through the exponential e^{A u_ε} with A > 0. The paper asserts that [14, Theorem 4.1] gives a uniform L∞ bound |u_ε| < C, but it does not state the theorem or verify its hypotheses in this setting. The L^p integrability of e^{-G_ε + g_ε} that the reader highlights is necessary, but not sufficient; one also needs a solvability and a priori estimate for the exponentially nonlinear equation. If the cited theorem does not cover this equation, or the L∞ bound depends on ε in an uncontrolled way, then the tame approximation is not produced and Theorem 3.2's heat-kernel argument has no input. Since Theorem 1.1 follows directly from Theorem 1.2, this is the most load-bearing step.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves that singular Kähler-Einstein metrics on klt pairs define Kähler currents (Theorem 1.1). This is derived from a more general statement (Theorem 1.2) for singular Kähler metrics with bounded potential, an L^p integrability condition on the volume density, and a currentwise lower Ricci bound. The proof constructs a 'tame approximation' of the singular metric by smooth metrics on a resolution whose Ricci curvature has negative part controlled in L^1 (Proposition 3.1), and then uses heat-kernel estimates of Guo–Phong–Song–Sturm to obtain a trace bound (Theorem 3.2). The paper also extends the Kähler current property to singular Kähler–Ricci solitons (Theorem 1.3) and gives a criterion for the metric completion to be a non-collapsed RCD space under an L^p bound on the negative part of the approximating Ricci curvature (Theorem 1.5).","tokens_in":20250,"tokens_out":14949,"duration_ms":138381,"significance":"If the proof can be completed, Theorem 1.1 is a substantial improvement over the earlier result of Guedj–Guenancia–Zeriahi [18], which covered only smoothable varieties and the three-dimensional negative/zero Ricci case. The key new ingredient, Proposition 3.1, is an L^1-control approximation lemma that is natural and potentially useful beyond this paper. The extension to Kähler–Ricci solitons and the RCD criterion are also of interest. However, the current manuscript leaves a load-bearing a priori estimate for the approximating Monge–Ampère equations as a citation to [14, Theorem 4.1] without verification, so the central claim is not fully established as written.","major_comments":[{"comment":"The uniform L∞ bound on u_ε is the step that makes the family a tame approximation in the sense of Definition 1.4(b), and it is asserted by the line 'It follows in particular from [14, Theorem 4.1] that we have a uniform L∞ bound |u_ε| < C.' However, (3.4) is not a standard Monge–Ampère equation with fixed density: the right-hand side contains e^{A u_ε} with A > 0, so the cited theorem must apply to equations with exponential dependence on the unknown. The paper does not state [14, Theorem 4.1] nor verify its hypotheses (e.g., normalization of u_ε, conditions on A, and the regularity or integrability of e^{-G_ε+g_ε}). The L^p bound on e^{-G_ε+g_ε} shown earlier is necessary but not sufficient unless the cited theorem indeed covers this nonlinearity uniformly in ε. Since Theorems 1.1 and 1.2 rely on the existence of this tame approximation, please provide the exact statement and verification, or a self-contained proof of the uniform estimate.","section":"Section 3, Proposition 3.1, equation (3.4)"},{"comment":"The same issue occurs in the soliton setting: the density in (4.2) depends on the unknown u_δ through both e^{u_δ} and g_V(m_{ε,u_δ}). The proof asserts 'the C0 estimate is automatic' because |log g_V(m_{ε,u_δ})| is bounded, but this does not yield a δ-uniform L∞ bound for u_δ without further argument; the dependence of the moment map m_{ε,u_δ} on u_δ is exactly the type of nonlinearity that requires an a priori estimate. Since Theorem 1.3 depends on this tame approximation, please either prove the uniform C0 bound or give a precise reference that covers this weighted equation.","section":"Section 4, Proposition 4.1, equation (4.2)"},{"comment":"The proof of Theorem 3.2 is presented as a sketch. In particular, the passage from the inequality ∂_t ∫ Φ H ≥ -2C_2 to the final trace bound is abbreviated by 'first let ε→0, and then t0→0'. Because the heat kernel upper bound C(t) in Theorem 2.2 diverges as t→0, the exchange of limits needs justification: one must show that the ε→0 limit gives a heat kernel on the singular space and that the bound ∫ Φ H(x,y,t0) dy ≤ C_3 controls Φ(x) as t0→0. If this is exactly the argument of [41, Theorem 16], please state the adaptation explicitly; otherwise write out the missing steps.","section":"Section 3, Theorem 3.2"}],"minor_comments":[{"comment":"The phrase 'on any compact set away from π^{-1}(X_reg \\ D)' is self-contradictory; presumably it should be 'on any compact set contained in π^{-1}(X_reg \\ D)' or 'away from π^{-1}(D) ∪ E'.","section":"Section 3, paragraph after (3.6)"},{"comment":"The use of Savin's small perturbation result [38] is only cited; please specify the exact form of the theorem used and explain how it applies to the limiting Monge–Ampère equation with measure-valued right-hand side and to the smooth convergence of u_ε.","section":"Section 3, same paragraph"},{"comment":"Lemma 5.2 states C^\\infty_loc convergence of eigenfunctions, but the proof appears to establish only L^2 convergence of eigenfunctions and convergence of eigenvalues. The local smooth convergence follows from elliptic regularity once the metrics converge smoothly, but this should be stated explicitly.","section":"Section 5, Lemma 5.2"}],"recommendation":"major_revision","confidential_remarks":"The paper builds very heavily on [41] and [24]. The main novelty is Proposition 3.1, so the referee has focused on the missing a priori estimate there. The authors should also make clear, for each theorem, which parts are new and which are imported from [41, Theorem 16] and [24]; this will help the reader and the editor assess the contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the main theorem (singular KE metrics on klt pairs are Kaehler currents) is very likely true, and the paper contains a genuinely new ingredient: the tame approximation with L1 control of the negative part of Ricci. But the written proof of Proposition 3.1 has a gap at exactly the point where it needs to be airtight. I would send it to referees, but I would expect a revision.\n\nWhat is new: Theorem 1.1 resolves an open question from Guedj-Guenancia-Zeriahi, removing the smoothable/three-dimensional restrictions. The approximation lemma in Proposition 3.1 is the real work: it constructs omega_eps with ||(Ric(omega_eps)+2A(omega_eps+omega_X))_-||_{L1} -> 0, using only the L^p density and the Ricci lower bound of the limit. The heat-kernel transport argument from [41] then yields the current lower bound. That is a clean conceptual advance. The RCD result with p > (2n-1)/n is a solid follow-up, and the improved Kato inequality approach is sensible.\n\nThe soft spots. The biggest is in Proposition 3.1. The approximation is defined by solving equation (3.4): the right-hand side depends on u_eps through e^{A u_eps} with A > 0. The paper says, in one line, that [14, Theorem 4.1] gives a uniform L^infinity bound, but it never states the hypotheses of that theorem or verifies that this equation fits them. This is not a routine fixed-density Monge-Ampere equation; for arbitrary A > 0 and an arbitrary smooth Kaehler class there are well-known obstructions to existence for equations of the form (alpha + i d dbar phi)^n = e^{A phi} f. In dimension two this is essentially the Nirenberg/Kazdan-Warner problem. Maybe in the KE application A can be taken to be 1 and the measure has the special adapted form that EGZ treat, but as written the existence and uniform bound are not justified. This is not fatal to the truth of the theorem, but it is load-bearing and needs a real argument, or a precise citation with hypotheses checked.\n\nOther soft spots: Theorem 3.2 is explicitly a sketch; Proposition 4.1 for solitons defers \"exactly as in Proposition 3.1\", so it inherits the same gap; the proof of Lemma 5.2 on eigenvalue convergence is hand-wavy about passing limits to the metric completion. These are minor if the main gap is fixed.\n\nThis paper is for people working on singular Kaehler geometry, non-collapsed RCD limits, and complex Monge-Ampere equations. It deserves a serious referee. I would not desk-reject it. The right call is to send it out and ask the authors to close the gap in Proposition 3.1.","headline":"Strong result with a genuinely new approximation lemma, but the proof of the key tame approximation has a potentially serious gap around citing EGZ for a non-standard Monge-Ampere equation.","tokens_in":20839,"tokens_out":19579,"would_cite":true,"duration_ms":190279,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32Q25","53C55","53C21","32Q20"],"pacs":[],"model":"deepseek-v4-flash","headline":"A singular Kähler–Einstein metric on a klt pair always defines a Kähler current, dominating a fixed smooth Kähler metric by a positive constant.","keywords":["Kähler–Einstein metrics","klt pairs","Kähler currents","singular Kähler metrics","RCD spaces","Ricci curvature bounds","Monge–Ampère equations","heat kernel estimates"],"falsifier":"Find a closed positive current ω = ω_X + i∂∂̄u on a compact normal Kähler space satisfying the three hypotheses of Theorem 1.2, with e^F ∈ L^p for some p>1, but for which the trace of ω_X with respect to ω is unbounded. The theorem asserts no such current exists; a single explicit example of this kind would refute the Kähler-current claim.","tokens_in":19803,"feed_emoji":"📐","tokens_out":11310,"duration_ms":100741,"temperature":0.7,"pith_summary":"The paper proves that every singular Kähler–Einstein metric on a mildly singular algebraic variety (a klt pair) is a Kähler current: it dominates some fixed smooth Kähler metric by a positive constant. This strict positivity was previously known only for smoothable varieties and in three dimensions, so the result is a substantial generalization. The same conclusion is proved for singular shrinking Kähler–Ricci solitons. A second theorem shows that when such a metric can be approximated by smooth metrics whose Ricci curvature has negative part uniformly bounded in a certain L^p norm, the metric completion of the space is a non-collapsed RCD space, meaning it has a synthetic Ricci curvature lower bound.","feed_headline":"Every singular Kähler–Einstein metric is a Kähler current","feed_subtitle":"They dominate some fixed smooth metric by a positive constant, extending a result long known only for smoothable cases.","key_machinery":"The central machinery is a tame approximation of the singular metric ω by smooth Kähler metrics ω_ε on a resolution π:Y→X, built by solving regularized Monge–Ampère equations with cut-off data. The approximations converge locally smoothly away from the singular locus and satisfy uniform L^p bounds; the new point is that their Ricci curvature can be arranged so that ‖(Ric(ω_ε)+2A(ω_ε+π^*ω_X))^-‖_{$L^{1}$(ω_ε)} → 0. Combined with the uniform Sobolev inequality, heat-kernel upper bound, and eigenvalue estimates of [24], this allows the Chern–Lu inequality to be integrated against the heat kernel to give a uniform upper bound on tr_{ω}ω_X, hence the Kähler current property. For the RCD conclusion, an improved Kato inequality for eigenfunction gradients upgrades the L^p bound on the negative Ricci part into a Lipschitz bound on eigenfunctions, which by [41, Proposition 9] is equivalent to being a non-collapsed RCD space.","core_discovery":"On its own terms, the paper establishes Theorem 1.1: if (X,D) is a klt pair and ω is a singular Kähler–Einstein metric, then for any smooth Kähler metric ω_X on X there is a constant ε>0 with ω ≥ εω_X. This follows from Theorem 1.2, a general criterion for positive currents ω = ω_X + i∂∂̄u with bounded potential, smooth outside a divisor, whose volume density e^F lies in L^p for some p>1 and whose Ricci curvature is bounded below on the regular locus. The proof constructs a tame approximation of ω by smooth Kähler metrics on a resolution, chosen so that the negative part of their Ricci curvature tends to zero in $L^{1}$; using the uniform Sobolev and heat-kernel estimates from [24] and the Chern–Lu inequality, this $L^{1}$ control yields a uniform bound on the trace of ω_X with respect to ω, which is equivalent to the desired lower bound.","pith_inferences":["The L^1 control on negative Ricci curvature is likely the operative hypothesis, so the strict-positivity conclusion may survive for wider classes of singular Kähler metrics once a tame approximation with L^1 Ricci control exists, even without the L^p density condition.","Tracking constants in the proof could yield explicit lower bounds for ε in terms of A, p, and ‖e^F‖_{L^p}, which would be useful for studying degenerations and families of singular Kähler–Einstein metrics.","The threshold p > (2n-1)/n in the RCD theorem aligns with the improved Kato inequality; testing the borderline p = (2n-1)/n with explicit families of approximations could reveal whether the exponent is sharp."],"forward_implications":["Every singular Kähler–Einstein metric on a klt pair admits a uniform positive lower bound relative to any smooth Kähler metric, making such metrics amenable to compactness and regularization arguments.","Singular shrinking Kähler–Ricci solitons satisfy the same strict-positivity property, so the results apply to soliton degenerations as well as static KE metrics.","Under an L^p bound on the negative part of approximating Ricci curvature for p > (2n-1)/n, the metric completion is a non-collapsed RCD space, so it carries a synthetic Ricci curvature lower bound in the sense of metric measure geometry.","For Kähler–Einstein spaces with torus symmetry admitting extremal approximations, the metric completion is homeomorphic to the original variety, extending earlier homeomorphism results beyond constant scalar curvature approximations."],"supporting_citations":[{"why":"supplies the tame-approximation framework and the heat-kernel method for deriving the Kähler current property, which the paper extends from constant scalar curvature to L^1 Ricci control.","marker":"[41]"},{"why":"provides the uniform Sobolev inequality, heat kernel upper bounds, and eigenvalue estimates that turn the L^1 Ricci control into a trace bound.","marker":"[24]"},{"why":"establishes the construction of singular Kähler–Einstein metrics and the notion of smooth Kähler metrics on singular spaces used to set up the approximations.","marker":"[14]"},{"why":"proved the Kähler current property in the special cases of smoothable varieties and threefolds, the baseline this paper generalizes.","marker":"[18]"},{"why":"supplies the stability estimates for complex Monge–Ampère equations used to prove local uniform convergence of the approximating metrics.","marker":"[19]"},{"why":"gives the improved Kato inequality for eigenfunction gradients that is the key input in the RCD theorem.","marker":"[37]"}],"fun_headline_variants":["Singular Kähler-Einstein metrics dominate smooth metrics","Every singular KE metric bounds a fixed smooth metric below","Singular KE metrics are currents without smoothable assumption","Kähler-Einstein singular metrics bound smooth ones from below"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof depends on the density of the singular volume with respect to a smooth volume lying in some L^p for p>1; if only $L^{1}$ integrability is available, the tame approximation with controlled negative Ricci curvature is lost and the argument collapses.","fun_headline_variants_meta":{"raw":{"variants":["Singular Kähler-Einstein metrics dominate smooth metrics","Every singular KE metric bounds a fixed smooth metric below","Singular KE metrics are currents without smoothable assumption","Kähler-Einstein singular metrics bound smooth ones from below"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000554,"raw_usage":{"total_tokens":2591,"prompt_tokens":849,"completion_tokens":1742,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":465,"completion_tokens_details":{"reasoning_tokens":1683}},"tokens_in":465,"tokens_out":1742,"duration_ms":12384,"temperature":1.0,"reasoning_tokens":1683,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T20:23:01.170514+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a closed positive current ω = ω_X + i∂∂̄u on a compact normal Kähler space satisfying the three hypotheses of Theorem 1.2, with e^F ∈ L^p for some p>1, but for which the trace of ω_X with respect to ω is unbounded. The theorem asserts no such current exists; a single explicit example of this kind would refute the Kähler-current claim.","supporting_citations":[{"cited_title":"Singular K¨ ahler-Einstein metrics, J","cited_arxiv_id":null,"evidence_quote":"establishes the construction of singular Kähler–Einstein metrics and the notion of smooth Kähler metrics on singular spaces used to set up the approximations."},{"cited_title":"Strict positivity of K¨ ahler-Einstein currents, Forum of Mathematics, Sigma","cited_arxiv_id":null,"evidence_quote":"proved the Kähler current property in the special cases of smoothable varieties and threefolds, the baseline this paper generalizes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the stability estimates for complex Monge–Ampère equations used to prove local uniform convergence of the approximating metrics."},{"cited_title":"On the gradient estimate of Cheng and Yau","cited_arxiv_id":null,"evidence_quote":"gives the improved Kato inequality for eigenfunction gradients that is the key input in the RCD theorem."}],"review_version":1}