{"id":"cc6c14fe-0f03-4b0f-96a6-0205e4616d3d","arxiv_id":"2509.09442","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Uniform K-stability for models implies existence and uniqueness of a cscK metric in any Kähler class, including transcendental (non-projective) classes.","lead":"This mathematics paper proves that a stability condition called uniform K-stability for models guarantees the existence of a unique constant scalar curvature Kähler (cscK) metric in every Kähler class. It extends a result known for projective manifolds to all compact Kähler manifolds using non-Archimedean pluripotential theory.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 9.2.1's approximation argument falsely identifies valuation images of refined dual complexes with the original, so the entropy convergence in (3)⇒(2) is unjustified.","rationale":"The reader's verdict identifies the dependence on the external framework [MP24] as the weakest assumption, and I agree the proof of Theorem A is conditional on that framework being correct. However, the present pass finds a more specific internal gap: the proof of Theorem 9.2.1, which establishes the equivalence between uniform K-stability for models and uniform bK-stability, contains a concrete false assertion about the valuation images of refined dual complexes. This is not merely a missing detail; it is an incorrect statement that, if not corrected, invalidates the approximation argument (3)⇒(2). The example of a blow-up of a point in the central fiber shows that i_{X_j}(Δ_{X_j}) ≠ i_X(Δ_X), so the measures μ_j need not be supported on the original dual complex, and the application of Lemma 7.2.4 to obtain Ent(μ_j)→Ent(μ) is unjustified. Without that convergence, the equality of the non-Archimedean Mabuchi functionals along the approximation is not established, and the bridge to [MP24, Theorem A] is broken. The paper also contains other places where detailed proofs are delegated to '[BJ22]' or 'the proof is exactly the same', but the dual complex issue is the most load-bearing because it is a concrete false assertion in the central reduction. I therefore maintain a conditional verdict: the paper should be accepted only after the approximation argument in Theorem 9.2.1 is repaired, either by ensuring the supports of the approximating measures can be chosen within the original dual complex, or by proving entropy convergence directly without relying on the false set equality.","tokens_in":37153,"tokens_out":23152,"duration_ms":273836,"concrete_test":"Let X be a compact Kähler surface and consider the trivial model X×P^1 with X0=X×{0}. Let X' be the blow-up of X×P^1 at a point p∈X0, with E the strict transform of X0 and F the exceptional divisor. Compute the divisorial valuations v_F and v_E on a local function x with ord_E(x)=0 but x(p)=0: ord_F(x)=1. Since i_X(Δ_X) (here a single point) only contains scalar multiples of v_E, v_F∉i_X(Δ_X). This directly falsifies the set equality claimed in the proof of Theorem 9.2.1. To test whether the approximation step can be repaired, run the same argument for φ≡0 supported on the original point complex and check whether any μ_j charges F; if it does, the entropy convergence step fails as stated.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The proof of (3)⇒(2) in Theorem 9.2.1 constructs PL functions f_j on refinements X_j of X and asserts that because Δ_{X_j} is a subdivision of Δ_X, the sets i_{X_j}(Δ_{X_j}) and i_X(Δ_X) coincide as valuations in X^na. This is false. For a concrete example, blow up a point p in the interior of a smooth central-fiber component E of the trivial model X×P^1. The exceptional divisor F defines a divisorial valuation v_F. In local coordinates with E={x=0} and p=(0,0), ord_E(x)=0 but ord_F(x)=1, so v_F is not a monomial valuation with respect to the original SNC divisor; hence v_F ∉ i_X(Δ_X). Consequently, the measures μ_j := MA_A(P_A(f_j)), which by Proposition 4.2.2 are supported on the vertices of Δ_{X_j}, can charge valuations outside i_X(Δ_X). Lemma 7.2.4 only yields convergence of projected entropies Ent(μ_{j,X}) for a fixed model X, not of the full entropies Ent(μ_j). The claimed convergence M_A(P_A(f_j))→M_A(φ) therefore lacks a valid justification. Since Theorem 9.2.1 is the bridge from uniform K-stability for models to uniform bK-stability, this gap directly threatens Theorem A.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops non-Archimedean pluripotential theory for arbitrary compact Kähler manifolds and proves two main theorems. Theorem A asserts that uniform K-stability for models implies the existence and uniqueness of a cscK metric in any Kähler class, extending Chi Li's algebraic theorem to the transcendental setting. Theorem B is a non-Archimedean Calabi–Yau theorem: the Monge–Ampère operator is a homeomorphism between sup-normalized finite-energy potentials and finite-energy measures. The paper also proves continuity of envelopes, orthogonality, a regularity theorem for solutions whose Monge–Ampère measure is supported on a dual complex, and a valuative criterion for K-stability for models with an explicit β-invariant formula. The proof of Theorem A proceeds by proving an equivalence between uniform K-stability for models and uniform bK-stability (Theorem 9.2.1), then invoking [MP24, Theorem A].","tokens_in":37505,"tokens_out":16242,"duration_ms":201407,"significance":"If correct, the paper constitutes a substantial advance: it removes projectivity and rationality assumptions from a central YTD-type implication, and it substantially develops transcendental non-Archimedean pluripotential theory. The envelope continuity, orthogonality, and the non-Archimedean Monge–Ampère theorem are valuable tools beyond the cscK application. The paper is honest about its dependence on the framework of [MP24] and [DXZ23], and it gives to a large extent detailed proofs rather than mere references. However, the bridge theorem (Theorem 9.2.1) contains a specific gap in the approximation argument that is load-bearing for Theorem A; until that gap is repaired, the main theorem is not established as written.","major_comments":[{"comment":"The proof asserts: 'Since the dual complex Δ_{X_j} is a subdivision of Δ_X, as valuations they coincide i_{X_j}(Δ_{X_j}) = i_X(Δ_X)⊆X^na.' This is false. For example, let X be the trivial model X×P^1 and let p be a point in the interior of a smooth central-fiber component E. Blow up p and let F be the exceptional divisor. In local coordinates E={x=0}, take y with y(p)=0 transverse to E; then ord_E(y)=0 but ord_F(y)=1, so v_F is not a monomial valuation with respect to the original SNC divisor, hence v_F∉i_X(Δ_X). Therefore the measures μ_j=MA(P_A(f_j)), supported on vertices of Δ_{X_j}, may charge valuations outside i_X(Δ_X). Lemma 7.2.4 only gives convergence of the projected entropies Ent(μ_{j,X}) for a fixed model X, not of the full entropies Ent(μ_j). The claimed convergence M_A(P_A(f_j))→M_A(φ) is thus unjustified. Since Theorem 9.2.1 is the bridge from uniform K-stability for model","section":"§9.2, proof of (3)⇒(2)"},{"comment":"The proof of Theorem A relies entirely on the implication (3)⇒(2) in Theorem 9.2.1 and then on [MP24, Theorem A]. Because the approximation step in (3)⇒(2) is not valid as written, the claimed equivalence between uniform K-stability for models and uniform bK-stability is not established. Without this equivalence, Corollary 9.2.2 does not follow. A repair is likely possible, for instance by a more careful choice of approximating measures or by controlling the difference between full and projected entropy for refined dual complexes, but it is not present in the manuscript.","section":"§9.2 / Corollary 9.2.2"}],"minor_comments":[{"comment":"The definition of the dual complex with the condition ∑_{i∈J} w_i b_i ≤1 appears nonstandard: for a single component this gives an interval rather than a vertex. Please clarify the normalization and explicitly identify how the vertices correspond to the divisorial valuations v_{E_i}.","section":"§1.3.4"},{"comment":"Two typos: 'coursest' should be 'coarsest' and 'Propoerties' should be 'Properties'.","section":"§5, §1.3.7"},{"comment":"The comparison principle is stated as 'the proof is exactly the same' as in the algebraic case and then only sketched. In particular, the step showing that if v_{E_i}∈U then D_1 and G coincide in a neighborhood of E_i is abbreviated. Since Theorem 8.1.1 is used in Corollary 8.1.3 and Theorem 8.2.1, which in turn feed Theorem 9.2.1, a fuller proof or a precise reference with verified hypotheses would be helpful.","section":"§8.1"},{"comment":"The use of [DXZ23, Proposition 3.1] is justified by saying its proof 'applies as is' to the transcendental setting. Given that this is a key step in the explicit β-formula, please spell out the necessary hypotheses and the transcendental version of the statement.","section":"§10.1.1"},{"comment":"References [BJ25] and [DZ25] are listed as 'in preparation'. The discussion in §1.6 should make clear whether any statement from these works is used, rather than merely contextual.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses an important problem and contains substantial new results in non-Archimedean pluripotential theory. However, the main application depends on the approximation argument in Theorem 9.2.1, where the identification of refined dual complexes with the original one as subsets of X^na is incorrect. This is a load-bearing gap, not a presentation issue. I recommend major revision: the authors should either prove a correct entropy-convergence statement for their approximation or supply a different argument for (3)⇒(2)."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nThis paper is an ambitious attempt to prove the uniform YTD conjecture in the fully transcendental setting: uniform K-stability for models implies a unique cscK metric. The architecture is sound. They first prove continuity of envelopes and the orthogonality property in the non-Archimedean setting for arbitrary compact Kähler manifolds. From that they get a non-Archimedean Calabi–Yau theorem (Theorem B) and the strong-topology machinery. That part is a genuine, useful extension of Boucksom–Jonsson's algebraic theory, and the paper is well written.\n\nThe serious problem is in the proof of Theorem 9.2.1, specifically (3) implies (2). The argument approximates φ∘p_X by PL functions f_j on refined models X_j, then claims the images of the dual complexes i_{X_j}(Δ_{X_j}) and i_X(Δ_X) coincide because Δ_{X_j} is a subdivision of Δ_X. That is false. A blow-up of a point in the interior of a component E creates an exceptional divisor F whose associated divisorial valuation is not monomial with respect to the original components, so it is not in i_X(Δ_X). The measures μ_j are supported on the vertices of Δ_{X_j}, so they can charge valuations outside Δ_X. Lemma 7.2.4 only gives convergence of the projected entropies Ent(μ_{j,X}) for a fixed model X, not of the full entropy. Therefore the convergence M_A(P_A(f_j))→M_A(φ) is unjustified. Since this equivalence is the bridge that yields Theorem A via [MP24, Theorem A], the main theorem is not proven as it stands.\n\nThis looks fixable—if the approximations are chosen as pullbacks of PL functions on the original complex, the support issue disappears—but it needs to be worked out. The paper also leans heavily on the first author's previous preprint and several 'proof is exactly the same' delegations; that is a secondary concern.\n\nI would still send this to a serious referee. The envelope theory and the NA Calabi–Yau theorem are significant and may survive independent of whether the cscK application goes through. A referee should focus on Theorem 9.2.1 first.","headline":"Solid NA pluripotential theory, but the bridge from model stability to bK-stability has a gap in the approximation argument, so Theorem A is not established as written.","tokens_in":38001,"tokens_out":5759,"would_cite":true,"duration_ms":66334,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32Q26","32U15","14G22","53C55"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that if a compact Kähler manifold is uniformly K-stable for models, then its Kähler class contains exactly one constant scalar curvature metric.","keywords":["constant scalar curvature Kähler metric","uniform K-stability","non-Archimedean pluripotential theory","tropical analytification","Monge–Ampère equation","Calabi–Yau theorem","valuative criterion","transcendental Kähler geometry"],"falsifier":"The most direct falsifier would be a compact Kähler manifold and Kähler class that is uniformly K-stable for models but has no cscK metric, or two distinct cscK metrics. Concretely, one could check a non-projective complex torus or K3 surface: compute the β-invariant over all divisorial measures for a chosen class; if the infimum of β/E is positive but an a priori estimate shows the Mabuchi functional is unbounded below, the central claim fails. Conversely, a class that fails the β-criterion but is known to admit a cscK metric would also signal a gap.","tokens_in":37044,"feed_emoji":"📐","tokens_out":5651,"duration_ms":58129,"temperature":0.7,"pith_summary":"The paper proves that if a compact Kähler manifold with a Kähler class is uniformly K-stable for models—a numerical inequality on all big test configurations—then the class contains exactly one constant scalar curvature (cscK) metric. This was previously known only for projective algebraic manifolds; the new proof works for arbitrary compact Kähler manifolds, where there are no algebraic line bundles to carry the stability condition. The route goes through a non-Archimedean Calabi–Yau theorem: on the tropical analytification of the manifold, the Monge–Ampère operator is a homeomorphism between finite-energy potentials and finite-energy measures. Why care: it removes a major hypothesis from a central sufficiency direction of the conjecture relating canonical metrics to stability, and supplies a computable valuative criterion for the stability condition.","feed_headline":"Stability condition guarantees a unique cscK metric","feed_subtitle":"New proof extends stability-to-metric result to all compact Kähler manifolds via tropical analysis.","key_machinery":"The central object is the tropical analytification X^na of a compact Kähler manifold, a compact Hausdorff space whose points are semivaluations on coherent ideal sheaves; it replaces the Berkovich space in the non-algebraic setting. On it the paper defines A-psh functions, a non-Archimedean Monge–Ampère operator, and spaces of finite-energy potentials and measures. The load-bearing identities are the Continuity of Envelopes Property and the Orthogonality Property for the envelope P_A(f), which together imply that the Monge–Ampère operator is a homeomorphism between sup-normalized finite-energy potentials and finite-energy probability measures.","core_discovery":"On the paper's own terms, the discovery is that the transcendental (not necessarily projective) analogue of non-Archimedean pluripotential theory is strong enough to carry the stability-to-existence argument: Theorem A, the existence and uniqueness of a cscK metric under uniform K-stability for models, holds for every compact Kähler manifold and Kähler class. The bridge is an equivalence between uniform K-stability for models and uniform bK-stability, together with a non-Archimedean Calabi–Yau theorem that makes the identification possible by solving Monge–Ampère equations for measures supported on dual complexes and by proving continuity of the solutions. The authors also extract a finitely","pith_inferences":["The authors do not state it, but the homeomorphism of the non-Archimedean Monge–Ampère operator likely carries over to weighted or twisted Monge–Ampère equations, giving a non-Archimedean route to other canonical metrics.","A testable extension: the β-invariant formula reduces stability checking to finitely many restricted-volume computations for a given finite set of divisors; one could implement this numerically for toric or low-dimensional examples to search for destabilizing measures.","The openness of uniform bK-stability in the Kähler class suggests the full set of classes admitting cscK metrics is open; proving this directly would give a transcendental necessity direction the paper does not pursue.","If the algebraic necessity of K-stability for cscK existence holds in the transcendental setting, then combining it with Theorem A would complete the conjecture; this implication is not claimed here."],"forward_implications":["Uniform K-stability for models is sufficient for cscK existence in all Kähler classes on all compact Kähler manifolds, not just projective ones.","The non-Archimedean Calabi–Yau theorem gives a way to solve Monge–Ampère equations on tropical spaces, showing that finite-energy measures are exactly Monge–Ampère measures of sup-normalized potentials.","The valuative criterion reduces the stability condition to checking a β-invariant on divisorial measures, computable from log discrepancies and restricted volumes.","The equivalence between the two stability notions shows that geodesic-ray stability is no stronger than model stability in the general Kähler setting.","Uniform bK-stability is an open condition in the Kähler class, implying that the set of Kähler classes admitting cscK metrics is open."],"fun_headline_variants":["Uniform K-stability forces cscK metric on all Kähler manifolds","Transcendental Calabi-Yau theorem extends cscK existence","Non-Archimedean geometry proves cscK metric uniqueness","From K-stability to cscK: now for every compact Kähler class","K-stability for models guarantees cscK metric in general"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The result rests on the validity of the prior transcendental non-Archimedean pluripotential framework for arbitrary compact Kähler manifolds and on the previously established fact that uniform bK-stability produces a unique cscK metric.","fun_headline_variants_meta":{"raw":{"variants":["Uniform K-stability forces cscK metric on all Kähler manifolds","Transcendental Calabi-Yau theorem extends cscK existence","Non-Archimedean geometry proves cscK metric uniqueness","From K-stability to cscK: now for every compact Kähler class","K-stability for models guarantees cscK metric in general"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000863,"raw_usage":{"total_tokens":3578,"prompt_tokens":741,"completion_tokens":2837,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":485,"completion_tokens_details":{"reasoning_tokens":2739}},"tokens_in":485,"tokens_out":2837,"duration_ms":23004,"temperature":1.0,"reasoning_tokens":2739,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T19:04:07.585449+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The most direct falsifier would be a compact Kähler manifold and Kähler class that is uniformly K-stable for models but has no cscK metric, or two distinct cscK metrics. Concretely, one could check a non-projective complex torus or K3 surface: compute the β-invariant over all divisorial measures for a chosen class; if the infimum of β/E is positive but an a priori estimate shows the Mabuchi functional is unbounded below, the central claim fails. Conversely, a class that fails the β-criterion but is known to admit a cscK metric would also signal a gap.","supporting_citations":[],"review_version":1}