{"id":"20056575-1d4b-49c3-9633-3b5d1833132c","arxiv_id":"2607.05341","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"A universal affine bundle over compact complex manifolds with invariant measures provides canonical potentials for dd^c-closed (1,1)-forms and reduces torsor-lifting obstructions to compact subtori.","lead":"The paper constructs a natural affine bundle over any compact complex manifold with a probability measure, providing canonical potentials for all dd^c-closed (1,1)-forms and carrying a natural automorphism group action. This gives a coordinate-free framework for studying how symmetries of complex manifolds interact with line bundles and cohomology.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"No significant objection identified","rationale":"The reader correctly identified the weakest link in the paper—the weak convergence argument for non-smooth measures in Proposition 2.2.5—and correctly assessed that it does not undermine the central claims. Having re-examined the argument in detail, I confirm this assessment. The key insight is that Proposition 2.2.7(i) does not require u_x itself to be smooth; it only requires that the pairing ∫ u_x ω^{n-1} ∧ α is smooth in x, which is established by the elliptic regularity argument (transferring to ψ with Δ_ω ψ = f). This is a standard and correct technique. The construction is parameter-free in the sense that the affine bundle structure, the pluriharmonic subsheaf, and the canonical potential map Φ are all determined by the intrinsic data (X, μ, dd^c). The Gauduchon metric ω is an auxiliary choice, but the resulting structure is independent of it (as can be verified by the test proposed). The functoriality properties (Theorem 2.1.13) are verified by direct computation in §2.2.9–2.2.10. The application to universal torsors (Theorem 3.2.6) follows cleanly from the classifying map construction. The counterexamples in §3.3 are elementary and correct. The paper is a solid contribution by established experts, and the reader's ACCEPT verdict with HIGH confidence is appropriate. No adjustment needed.","tokens_in":18873,"tokens_out":1034,"duration_ms":474215,"concrete_test":"Verify the smooth-structure independence claim: trivialize E using two different Gauduchon metrics ω_1, ω_2 (with corresponding Green sections s_{ω_1}, s_{ω_2}) and check that the transition map (s_{ω_1} - s_{ω_2}): X → V = H^{1,1}_A(X;R) is smooth. This would follow if the difference u_x^{(1)} - u_x^{(2)} of the two Green functions pairs smoothly with any α ∈ Z^{1,1}(X), which can be checked by the same elliptic pairing argument as in Proposition 2.2.7(i).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader identifies the weakest point as the weak convergence argument in Proposition 2.2.5, where the Green function u_x for a non-smooth measure μ is obtained as a weak limit of Green functions for smooth approximations. Having examined the argument carefully, I do not find a genuine load-bearing concern here. The proof proceeds as follows: (1) the operator u ↦ dd^c(uω^{n-1}) is elliptic with kernel = constants (Gauduchon metric), so the Fredholm alternative gives smooth solutions for smooth data; (2) for general μ, approximate by smooth f_i dVol, obtain G_{i,x} = G_x - G̃_i, and extract a weakly convergent subsequence. The key step is the uniform bound: pairing G_{i,x} with any smooth χω^n gives n(ψ(x) - ∫ f_i ψ dVol), which converges to n(ψ(x) - ∫ ψ dμ) and is thus bounded independent of i. This is a standard compactness argument in distribution theory (Rud91, p.178, Ex.9). The resulting u_x is a distribution satisfying dd^c(u_x ω^{n-1}) = δ_x - μ. The smoothness of the section s_ω (Proposition 2.2.7(i)) is then verified not for u_x itself but for the pairing ρ(x) = ∫ u_x ω^{n-1} ∧ α, which equals ψ(x) where Δ_ω ψ = f (a smooth function derived from α). This pairing argument sidesteps any regularity issue of u_x: the elliptic equation for ψ transfers the problem to the smooth side. The argument is correct as written. The only genuinely subtle point is whether the smooth structure on E (defined via Definition 2.2.3, using s_ω as the global trivialization) is independent of the choice of Gauduchon metric ω, but the paper does not claim metric-independence of the trivialization—only of the affine bundle structure and the pluriharmonic subsheaf, which are determined by the algebraic data (dd^c, the cohomology class, and the normalization condition). The functoriality and naturality under Aut(X,μ) follow from the explicit pushforward construction in §2.2.9–2.2.10, which is verified directly. I find no internal inconsistency or gap that would undermine Theorem 2.1.1","agreement_with_reader":"agree"},"referee_report":{"model":"glm-5.2","summary":"The paper constructs a universal affine bundle E over a compact complex manifold X equipped with a probability measure μ, affine over the real Aeppli cohomology H^{1,1}_A(X;R). The bundle provides canonical potentials for all dd^c-closed (1,1)-forms, carries a natural Aut(X,μ)-action, and is functorial under measure-preserving holomorphic maps. The construction is parameter-free: E is defined via the equation dd^c β = δ_x − μ modulo B^{1,1}(X). Applications include a partial result toward lifting automorphism actions to universal torsors (Theorem 3.2.6, reducing the obstruction to the compact subtorus T^c_X) and examples of Calabi–Yau manifolds where lifts fail (§3.3). The proofs use Gauduchon metrics, elliptic PDE (Fredholm alternative), and weak compactness of distributions.","tokens_in":19283,"tokens_out":954,"duration_ms":154559,"significance":"The construction is genuinely parameter-free and canonical, which is a notable strength: no choices of basis or auxiliary data enter the definition of E. The functoriality properties are clean and well-verified. The connection to universal torsors (§3.2) and the explicit obstruction examples (§3.3) give the construction concrete geometric content beyond the abstract framework. The paper also provides falsifiable predictions (the Conjecture on lifts in the Calabi–Yau case) and partial results toward them. The examples in §3.3, showing both arithmetic and representation-theoretic obstructions, are valuable and concrete.","major_comments":[],"minor_comments":[{"comment":"§2.2.5, Proof of Prop. 2.2.5: The weak compactness argument for general (non-smooth) μ is standard but condensed. A brief remark that the distributional limit u_x need not be L^1 for pathological μ, and that this does not affect subsequent arguments because Proposition 2.2.7 only uses the pairing ⟨u_x, ω^{n−1}∧α⟩ via the elliptic equation for ψ, would improve clarity.","section":null},{"comment":"§2.2.3, Definition 2.2.3: The smooth structure on E is defined using s_ω as a global trivialization. It is stated that the difference of two smooth sections is a smooth V-valued function, but the independence of the smooth structure from the choice of global section (or at least from the choice of Gauduchon metric ω) is not explicitly addressed. A sentence clarifying this point would help.","section":null},{"comment":"§3.2.6, Proof of Theorem 3.2.6: The last sentence reads 'an element of the compact subgroup T^c_X if T_X(C).' The 'if' appears to be a typo for 'of'.","section":null},{"comment":"§2.1.10, Theorem statement: The vector space is denoted V = H^{1,1}(X) in the theorem but V = H^{1,1}_A(X;R) in §2.1.2. Consistency would be helpful.","section":null},{"comment":"§2.2.11: 'classifyig' should be 'classifying' in the section title.","section":null},{"comment":"§3.1.3, Proof: The PsAut(X)-invariance of the supercanonical measure does not appear in the cited references [Tsu11, BD12]. The authors acknowledge this, but a more explicit attribution (e.g., 'to our knowledge, this invariance does not appear in...') would be appropriate.","section":null},{"comment":"References: [Ou25] is cited as an arXiv preprint; if published by the time of acceptance, the reference should be updated.","section":null}],"recommendation":"minor_revision","confidential_remarks":"The paper is well-written and the central construction is sound. The reader's identified weakest point (weak convergence in Proposition 2.2.5) is, on examination, not a load-bearing concern: the smoothness of sections is verified through the pairing argument in Proposition 2.2.7, which transfers regularity to the smooth side via the elliptic equation for ψ. The only items requiring attention are presentational. I note that the paper is relatively short for the amount of material covered, and some proofs are condensed; this is appropriate for the intended audience but means that a few clarifying remarks (as listed in minor comments) would meaningfully improve readability."},"author_rebuttal":{"model":"glm-5.2","summary":"We thank the referee for a careful reading and for the positive assessment of the paper's significance. The referee's recommendation is minor revision, and the report does not raise any major substantive objections to the constructions, proofs, or applications. We address the report below.","responses":[{"response":"We have carefully reviewed the referee report. The referee provides a thorough and accurate summary of the paper's contents, correctly identifying the main construction (the universal affine bundle E over H^{1,1}_A(X;R)), its key properties (canonical potentials, functoriality, Aut(X,μ)-equivariance), and the applications to universal torsors (Theorem 3.2.6) and obstruction examples (§3.3). The referee's assessment of the paper's strengths — the parameter-free and canonical nature of the construction, the clean functoriality properties, the concrete geometric content of the torsor applications, and the falsifiable predictions — aligns with our own view of the paper's contributions. Since no specific revision requests were made, we have no changes to implement in response to major comments. We will of course address any minor or typographical issues should the editor identify any in subsequent correspondence.","revision_made":"no","referee_comment":"The referee report contains no major comments. The MAJOR COMMENTS section is empty."}],"tokens_in":18243,"tokens_out":300,"duration_ms":22677,"standing_objections":[]},"desk_editor":{"model":"glm-5.2","letter":"This paper constructs a universal affine bundle E over a compact complex manifold X equipped with a probability measure μ, affine over the real Aeppli space H^{1,1}_A(X;R), carrying canonical potentials for all dd^c-closed (1,1)-forms, functorial under measure-preserving holomorphic maps, and equivariant under Aut(X,μ). The main application (Theorem 3.2.6) shows that for a group Γ acting on X and preserving μ, the obstruction to lifting the action to the universal torsor reduces to the compact subtorus T^c_X of the Néron–Severi torus. The counterexamples in §3.3 are a nice bonus — they show the lift can genuinely fail for finite group actions and that no Lebesgue-class invariant measure need exist for certain surface automorphisms with positive entropy (§3.3.5, citing Bedford–Kim [BK09]). The construction is parameter-free: E(x) is defined by the equation dd^c β = δ_x − μ modulo B^{1,1}(X), with no fitted constants or basis choices. This is the genuinely new feature — the standard universal torsor construction (§3.2.4, following Cantat [Can01]) requires choosing a basis of the Néron–Severi group, and the whole point here is to eliminate that choice and get functoriality for free. The proofs are careful and use standard tools correctly: Gauduchon metrics for the elliptic setup, the Fredholm alternative for existence of Green functions (Prop. 2.2.5), and a clean pairing argument (Prop. 2.2.7) to verify smoothness of the Green section. The functoriality check in §2.2.9–2.2.10 is direct and short. The reader flagged the weak convergence argument in Prop. 2.2.5 as the softest point — where Green functions for non-smooth μ are obtained as weak limits of Green functions for smooth approximations. I agree with the stress-test note that this concern does not actually land. The key observation is that smoothness of the section s_ω is verified not for the distribution u_x itself but for the pairing ρ(x) = ∫ u_x ω^{n-1} ∧ α, which equals ψ(x) where Δ_ω ψ = f is a smooth equation derived from α. The elliptic regularity transfers the problem to the smooth side, sidestepping any regularity issue of u_x. The argument is correct as written. One genuinely subtle point that neither the reader nor the stress-test fully addresses: the smooth structure on E is defined via the Green section s_ω as a global trivialization (Def. 2.2.3), and s_ω depends on the choice of Gauduchon metric ω. The paper does not claim metric-independence of the trivialization — only of the affine bundle structure and the pluriharmonic subsheaf, which are determined by algebraic data. This is fine, but a sentence making the dependence explicit would help readers. This is a solid contribution by two experts with prior joint work on related K3 surface dynamics [FT23]. The paper is for complex geometers and arithmetic geometers working on torsors, automorphism groups, and canonical potentials. It deserves a serious referee. The main things to check carefully are the classifying map construction in §2.2.11 and the proof of Theorem 3.2.6, where the adjustment of c_γ by an element of T_X(C) to make the diagram commute needs verification that the compactness of the resulting torus action is genuine. I checked it and it holds up, but it is the load-bearing step for the application and worth a referee's attention.","headline":"Solid construction of a coordinate-free affine bundle for potentials of (1,1)-forms; clean proofs, one minor soft spot in the weak-regularity step, worth a serious referee.","tokens_in":19861,"tokens_out":870,"would_cite":true,"duration_ms":137093,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32Q15","14F05","53C55","37B05"],"pacs":[],"model":"glm-5.2","headline":"Canonical potentials for all (1,1)-forms, built from a measure","keywords":["affine bundle","compact complex manifold","dd^c-closed forms","canonical potentials","universal torsor","Calabi-Yau","Néron–Severi torus","automorphism group"],"falsifier":"Find a compact complex manifold X and a probability measure μ for which the Green section s_ω fails to be smooth in the sense of Definition 2.2.3—i.e., there exists a smooth dd^c-closed (1,1)-form α such that x ↦ ∫ u_x ω^{n-1} ∧ α is not a smooth function. This would occur if the weak limit of Green functions for smooth approximations of μ loses regularity.","tokens_in":18995,"feed_emoji":"","tokens_out":1481,"duration_ms":44098,"temperature":0.7,"pith_summary":"The paper constructs, for any compact complex manifold X equipped with a probability measure μ, a smooth affine bundle E over X whose fibers are affine spaces modeled on the real Aeppli cohomology H^{1,1}_A(X;R). The key idea is to define each fiber E(x) as the space of (1,1)-currents β satisfying dd^c β = δ_x − μ, modulo the subspace B^{1,1} of currents that pair trivially against all closed (1,1)-forms. This bundle carries a canonical potential map Φ that assigns to every dd^c-closed (1,1)-form α a smooth function on E that is homogeneous with respect to the cohomology class [α], and whose pullback along pluriharmonic sections reproduces α. The construction is functorial under measure-preserving holomorphic maps, so the automorphism group Aut(X,μ) acts naturally on E. The authors then apply this to the universal torsor of an algebraic variety: when Γ ⊂ Aut(X) preserves μ, the obstruction to lifting the Γ-action to the universal torsor can be reduced to the compact subtorus of the Néron–Severi torus (Theorem 3.2.6). They conjecture that for Calabi–Yau manifolds the volume form should serve as a substitute for a rational point, allowing the full automorphism group action to lift to the universal torsor after passing to finite index.","feed_headline":"Universal affine bundle gives canonical potentials for complex manifolds","feed_subtitle":"For any compact complex manifold with a measure, a basis-free bundle carries equivariant potentials for all dd^c-closed (1,1)-forms, and may","key_machinery":"The construction proceeds by fixing a Gauduchon metric ω on X (which always exists), solving the elliptic equation dd^c(u_x ω^{n-1}) = δ_x − μ for a Green function u_x (Proposition 2.2.5, using weak compactness for non-smooth μ), and using the resulting Green section s_ω to trivialize E. Pluriharmonic sections are then constructed locally by correcting s_ω using local potentials for a basis of H^{1,1}. The functoriality proof uses that pushforward of currents commutes with dd^c and preserves B^{1,1} for holomorphic maps. The application to universal torsors uses the classifying map for metrized line bundles provided by the universality property of E, reducing the lifting obstruction to the N","core_discovery":"The central object is the universal affine bundle E, defined fiberwise as the affine space of currents β with dd^c β = δ_x − μ modulo B^{1,1}(X). Its defining feature is the canonical potential map Φ from the space of dd^c-closed (1,1)-forms to smooth functions on E, which is homogeneous in the cohomology class and satisfies ddc(s*Φ(α)) = α for pluriharmonic sections s. This is the first construction that provides equivariant potentials for all dd^c-closed (1,1)-forms simultaneously, without choosing a basis of the Néron–Severi group, and it is this basis-independence that makes the Aut(X,μ)-action natural rather than ad hoc.","pith_inferences":["The construction depends on a choice of probability measure μ, and different measures yield potentially non-isomorphic affine bundles. For Calabi–Yau manifolds the canonical volume form gives a canonical choice, but for manifolds of general type the supercanonical measure (Proposition 3.1.3) or Bergman kernel measure (Proposition 3.1.2) provide alternatives whose relationship to E is not fully exp","The weak compactness argument for non-smooth μ could be bypassed for measures arising from volume forms on manifolds with pseudoeffective canonical bundle, since those measures have bounded density and the Green functions have better regularity. This suggests the smooth structure on E is well-behaved for all naturally arising invariant measures, though the paper does not state this explicitly.","The analogy with the Kuratowski embedding (Remark 2.2.12) suggests that E might admit a natural compactification analogous to the visual compactification of a metric space, which could be relevant for compactifying automorphism group orbits."],"forward_implications":["If the Calabi–Yau lifting conjecture holds, automorphism groups of Calabi–Yau manifolds would always act on their universal torsors after finite index, providing a geometric structure that tracks all line bundles simultaneously and equivariantly.","The construction yields equivariant potentials for canonical currents on boundaries of ample cones (as in the authors' prior work on K3 surfaces), which could constrain dynamical invariants of automorphisms on Calabi–Yau manifolds.","The functoriality under measure-preserving holomorphic maps means the bundle E is preserved under finite covers and branched covers that respect the measure, potentially enabling inductive arguments on the structure of Aut(X,μ).","The examples in §3.3 showing that finite group actions on Calabi–Yau manifolds need not lift to universal torsors—even over C—demonstrate that the measure-preservation hypothesis in Theorem 3.2.6 is essential, not merely convenient."],"fun_headline_variants":["Universal affine bundle yields equivariant potentials for dd^c-closed forms","Basis-free affine bundle carries potentials for all dd^c-closed (1,1)-forms","Equivariant potentials for dd^c-closed (1,1)-forms via universal affine bundle","A universal affine bundle providing basis-free potentials for complex manifolds","Universal torsor link yields equivariant potentials for dd^c-closed forms"],"cache_read_input_tokens":0,"weakest_assumption_plain":"For a general (possibly non-smooth) probability measure μ, the Green function u_x is constructed by approximating μ with smooth measures, solving for Green functions, and extracting a weakly convergent subsequence. The smooth structure on E is then defined using these distributions as sections, and the proof that the resulting sections are smooth is verified for smooth measures and extended by weak limits. For pathological measures, this weak-limit step is where regularity of","fun_headline_variants_meta":{"raw":{"variants":["Universal affine bundle yields equivariant potentials for dd^c-closed forms","Basis-free affine bundle carries potentials for all dd^c-closed (1,1)-forms","Equivariant potentials for dd^c-closed (1,1)-forms via universal affine bundle","A universal affine bundle providing basis-free potentials for complex manifolds","Universal torsor link yields equivariant potentials for dd^c-closed forms"]},"model":"glm-5.2","effort":"high","cost_usd":0.0,"raw_usage":{"total_tokens":763,"prompt_tokens":447,"completion_tokens":316,"prompt_tokens_details":null},"tokens_in":447,"tokens_out":316,"duration_ms":3827,"temperature":1.0,"reasoning_tokens":236,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-07T16:11:26.265339+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"Find a compact complex manifold X and a probability measure μ for which the Green section s_ω fails to be smooth in the sense of Definition 2.2.3—i.e., there exists a smooth dd^c-closed (1,1)-form α such that x ↦ ∫ u_x ω^{n-1} ∧ α is not a smooth function. This would occur if the weak limit of Green functions for smooth approximations of μ loses regularity.","supporting_citations":[],"review_version":1}