{"id":"8ba5a5e1-5092-4a0c-85ec-c668b5abaefc","arxiv_id":"1909.02452","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For any Fano manifold, the Donaldson-Futaki invariant of its optimal degeneration is bounded below by -(1-R(X))/R(X) nV, where R(X) is the greatest Ricci lower bound.","lead":"This paper proves a quantitative lower bound for the Donaldson-Futaki invariant of the optimal degeneration produced by the Kähler-Ricci flow on any Fano manifold, expressed through that manifold's greatest Ricci lower bound. The bound yields a uniform finiteness statement for soliton Futaki invariants across all Fano manifolds and a related bound on the infimum of the H-functional.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof's final limit step yields only DF(X_a) ≥ -(1-R)/R nV, not the strict inequality stated in Theorem 1.1, which may actually fail for non-Einstein Kähler-Ricci solitons.","rationale":"The reader's weakest_assumption focuses on strict positivity of the limit soliton's scalar curvature. That concern is not load-bearing: the inequality ρ_Y ≤ n - (1/V)H(X_b) requires only S_{ωY} ≥ 0, and S_{ωY} ≥ 0 is indeed obtained from the maximum principle bound S_{ωt} > inf_X S_{ω0} e^{-t} even when inf S_{ω0} < 0, because the lower bound approaches 0 from below. The more serious problem is the strict inequality in Theorem 1.1. The proof establishes for each r<R(X) that DF(X_a) > -(1-r)/r nV, but the passage to r=R(X) is a limit that preserves only ≥. Moreover, the strict inequality may be outright false for K-semistable manifolds with non-Einstein Kähler-Ricci solitons, where R(X)=1 and the optimal degeneration can be a product test configuration with zero Futaki invariant. This directly threatens the central claim as stated. The paper's applications (Corollaries 1.2 and 1.3) still work with a non-strict bound, so the issue is repairable by changing the theorem's inequality to ≥, which is why I keep the verdict CONDITIONAL rather than REJECT. The reader's overall CONDITIONAL verdict is therefore unchanged, but for a different reason than the one highlighted in weakest_assumption.","tokens_in":9356,"tokens_out":34874,"duration_ms":339179,"concrete_test":"Compute R(X) and DF(X_a) for a Fano manifold admitting a non-Einstein Kähler-Ricci soliton, e.g., the one-point blow-up of CP^2. Use known formulae for toric Fano manifolds (Li, Delcroix) to determine R(X); if R(X)=1, then the claimed strict bound reads DF(X_a)>0. Determine X_a from the Kähler-Ricci flow: when X itself is the limiting soliton, X_a is the product test configuration generated by the soliton vector field. Compute its Donaldson-Futaki invariant (i.e., the Futaki invariant of that vector field). If it is 0, Theorem 1.1 is false as stated. If it is positive, the strict inequality may still hold for that example, but the proof still requires a separate argument to justify the strictness at r=R(X); one should check whether any such argument appears in the manuscript beyond 'letting r↗R(X)'.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Theorem 1.1 claims a strict lower bound DF(X_a) > -(1-R(X))/R(X) nV. In the proof, for each r < R(X), coercivity of the twisted Mabuchi functional gives DF(X_a) > -(1-r)/r nV (Section 3, last paragraph). Letting r ↗ R(X) preserves only DF(X_a) ≥ -(1-R)/R nV, because the right-hand side is continuous in r and a strict inequality for each approximant does not survive the supremum. No additional strictness argument is supplied. The concern is not merely formal: if R(X)=1 and X is a Fano manifold admitting a non-Einstein Kähler-Ricci soliton, then X is K-semistable, so R(X)=1, and the flow can converge to the soliton on X itself. In that case the optimal degeneration X_a is a product test configuration associated to the soliton vector field, whose Donaldson-Futaki invariant is the Futaki invariant of that field, which can be zero. The theorem would then assert 0 > 0, which is false. The reader's identified weakest assumption (strict positivity S_{ωY} > 0) is less load-bearing: the inequality chain ρ_Y ≤ n - (1/V)H(X_b) only needs S_{ωY} ≥ 0, and the maximum principle does deliver S_{ωY} ≥ 0 on Y_reg even when inf_X S_{ω0} < 0, since the lower bound inf_X S_{ω0} e^{-t} tends to 0 from below. Thus the central defect is the unjustified strict inequality in the theorem statement, not the scalar-curvature positivity issue.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the optimal degeneration X_a produced by the normalized Kähler-Ricci flow on a Fano manifold X, in the sense of Dervan–Székelyhidi. Its main result, Theorem 1.1, claims the strict lower bound DF(X_a) > -(1-R(X))/R(X) nV, where R(X) is the greatest Ricci lower bound, V=(-K_X)^n, and DF(X_a) is identified with the integral invariant Fut(W_Y)=∫_Y |∇ρ_Y|^2 ω_Y^n on the Q-Fano limit soliton (Y,ω_Y,W_Y). The proof combines the coercivity of the twisted Mabuchi functional for r<R(X), the non-Archimedean limits of DF, I_NA-J_NA and H, and analytic identities for the limit soliton. The paper derives two applications: a uniform lower bound for Fut(W_Y) over all n-dimensional Fano manifolds (Corollary 1.2), partly generalizing Guo–Phong–Song–Sturm, and an upper bound for inf_{ω∈c_1(X)} H(ω) in terms of R(X) (Corollary 1.3), compared with an inequality of Hisamoto.","tokens_in":9686,"tokens_out":17367,"duration_ms":198312,"significance":"If the main inequality is corrected, this is a useful quantitative complement to Dervan–Székelyhidi's theory of optimal degenerations: it gives an explicitly computable lower bound, in terms of R(X), for an invariant that is otherwise determined implicitly by the soliton vector field. The paper contains no fitted parameters, and the analytic formulas (3.1), (3.2) are correctly sourced. The intended applications, a uniform lower bound for Fut(W_Y) and an inequality for inf H, are plausible and would be of interest. However, the strict inequality in Theorem 1.1 is not established by the given proof and is actually false in simple cases, so the main statement needs revision; the weak form of the theorem appears sufficient for the applications.","major_comments":[{"comment":"The passage from DF(X_a) > -((1-r)/r)nV for every r<R(X) to DF(X_a) > -((1-R(X))/R(X))nV is invalid: letting r tend to R(X) preserves only the weak inequality, because the right-hand side is continuous and a strict inequality for every approximant does not survive the supremum. The strict statement is in fact false for a Kähler-Einstein Fano manifold: in that case R(X)=1, the flow converges to a Kähler-Einstein metric on X, the optimal degeneration is trivial, and DF(X_a)=0, so the asserted inequality would read 0>0. The theorem, the abstract, and Corollaries 1.2 and 1.3 should be reformulated with '≥' (or strictness should be proved by a separate argument); with the weak inequality the applications still follow by choosing the uniform constant F strictly larger than the resulting upper bound.","section":"Section 3, final paragraph; Theorem 1.1"},{"comment":"The claim that the maximum principle bound S_{ω_t} > inf_X S_{ω_0} e^{-t} implies S_{ω_Y}>0 on Y_reg is not justified: if inf_X S_{ω_0}<0, the lower bound tends to 0 from below, and in any case a pointwise limit of positive functions along a sequence of metrics need only be nonnegative. The subsequent inequality ρ_Y ≤ n-(1/V)H(X_b) requires only S_{ω_Y}≥0, and S_{ω_Y}≥0 does follow from the stated maximum principle, so this is a local overstatement rather than a fatal gap, but the text should be corrected.","section":"Section 3, paragraph before Eq. (3.1)"},{"comment":"The reduction 'without loss of generality' from R-degenerations to special degenerations by approximation is asserted rather than proved. Continuity of DF, I_NA-J_NA and H under approximation yields the limiting inequality only in the weak sense, not the strict positivity DF(X_a)+(1-r)(I_NA-J_NA)(X_a)>0 that is used in the proof. Once the theorem is weakened to '≥' this issue disappears, but in the present text it is part of the unjustified strictness of Theorem 1.1.","section":"Section 3, paragraph after Theorem 3.1"}],"minor_comments":[{"comment":"'multipricative property' should be 'multiplicative property'.","section":"Definition 2.3"},{"comment":"The name 'Strum' appears in the text and in [GPSS18, PSS15]; the correct spelling is 'Sturm'.","section":"References and text"},{"comment":"The phrase 'the supremum of the RHS is taken over all special degenerations' should specify whether R-degenerations are included or whether an approximation statement is intended; this matters for the exact meaning of X_a in the later proof.","section":"Section 1, Eq. (1.1)"},{"comment":"The sentence 'By letting r↗R(X), we finish the proof' should explicitly state which inequality is obtained; as written it suggests a strictness that the preceding argument does not supply.","section":"Section 3, proof of Theorem 1.1"}],"recommendation":"major_revision","confidential_remarks":"The paper's core strategy is sound, and the weak form of the main inequality is likely correct and sufficient for the applications. The main issue is the strict inequality in Theorem 1.1, which is both unproved and false in the Kähler-Einstein case; the authors should either prove strictness under suitable nontriviality assumptions or consistently reformulate the statements with '≥'. The scalar-curvature positivity overclaim and the R-degeneration approximation gap are secondary but should be fixed during revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a genuinely new result: a quantitative lower bound on the Donaldson-Futaki invariant of the optimal degeneration in terms of the greatest Ricci lower bound, for every Fano manifold. Dervan-Szekelyhidi had the optimality but no quantitative control, GPSS18 only handled manifolds with solitons, and Hisamoto's inequality needs R > 1/(4π). The proof idea is clear and the applications are natural. No free parameters, no circularity, and the analytic formulas are correctly sourced.\n\nThat said, the theorem overclaims. The coercivity argument gives, for each r < R(X), DF(X_a) > -(1-r)/r nV. Letting r → R(X) is a supremum of strict inequalities and only preserves ≥. The text \"we finish the proof\" hides the loss of strictness. I don't see a separate argument restoring >. The issue is not merely formal: when R = 1 and the manifold is K-semistable, the optimal degeneration can be a product with zero DF; if such a manifold is not K-polystable, Theorem 1.1 would assert 0 > 0. I don't know whether such examples exist among smooth Fano manifolds, but the proof does not rule them out. The clean fix is to state the theorem with ≥; the corollaries still work.\n\nTwo more soft spots, both minor. The reduction to special degenerations is a one-line assertion; it is standard to approximate R-degenerations by test configurations, but the continuity step has to be explicit. And the maximum principle bound on the scalar curvature yields only S_{ωY} ≥ 0 when the initial metric has negative scalar curvature somewhere; but the inequality chain only needs ≥0, so this is a harmless misstatement. The singular regularity input (continuity of ∇_{JV}s on Y) is delegated, which is fine for a preprint.\n\nThis is serious, honest work. I would send it to a strong referee with the explicit question: prove or disprove the strict inequality; if it fails, the weak inequality is still a solid result. The paper deserves a serious referee, not a desk reject.","headline":"Genuinely new quantitative bound for optimal degenerations, but the theorem's strict inequality is not proven; the proof supports only ≥.","tokens_in":10262,"tokens_out":20410,"would_cite":true,"duration_ms":208335,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53C55","14L24"],"pacs":[],"model":"deepseek-v4-flash","headline":"Fano destabilizers obey a computable lower bound.","keywords":["Kähler-Ricci flow","optimal degeneration","Donaldson-Futaki invariant","greatest Ricci lower bound","Q-Fano varieties","Kähler-Ricci solitons","H-functional"],"falsifier":"If one exhibits a Fano manifold $X$ with $R(X)<1$ whose Kähler-Ricci flow limit $(Y,\\omega_Y,W_Y)$ has $S_{\\omega_Y}=0$ at a regular point, then the bound $\\max_Y\\rho_Y\\le n-H(X_b)/V$ and the inequality chain leading to Theorem 1.1 collapse. A concrete place to look is the toric Fano list where $R(X)$ and the soliton metric are explicit: compute $\\inf_Y S_{\\omega_Y}$ on those models; any zero would falsify the strict-positivity step. Running the flow numerically from a metric $\\omega_0\\in c_1(X)$ with $\\inf_X S_{\\omega_0}<0$ and testing whether $S_{\\omega_t}$ becomes eventually bounded below by a positive constant would settle the premise directly.","tokens_in":9031,"feed_emoji":"🌀","tokens_out":9061,"duration_ms":85289,"temperature":0.7,"pith_summary":"This paper establishes a quantitative lower bound for the Donaldson-Futaki invariant of the optimal degeneration that the Kähler-Ricci flow produces from any Fano manifold. In precise terms, for an $n$-dimensional Fano manifold $X$, if $X_a$ is the flow's degeneration and $R(X)$ is the greatest Ricci lower bound, then $DF(X_a)>-(1-R(X))/R(X)\\,nV$, where $V=(-K_X)^n$. The bound is explicit because $R(X)$ is computable in many examples and is known to be uniformly positive in each dimension. The paper draws two consequences: a uniform lower bound on the Futaki invariants of all soliton limits, and an upper bound on the infimum of the $H$-functional that holds without the extra assumption in earlier work. The result matters because it gives general quantitative control in the unstable Fano case, where K-stability tests are hard to evaluate directly.","feed_headline":"Fano destabilizers obey a computable lower bound","feed_subtitle":"The Kähler-Ricci flow's optimal degeneration always has Donaldson-Futaki invariant above an explicit number.","key_machinery":"The load-bearing machine is the limit structure of the normalized Kähler-Ricci flow $d\\omega_t/dt=-\\operatorname{Ric}(\\omega_t)+\\omega_t$. The sequential polarized Gromov-Hausdorff limit is a Q-Fano variety $(Y,\\omega_Y,W_Y)$ with a soliton vector field $W_Y$, and the flow induces a two-step R-degeneration $X\\to\\overline{X}\\to Y$ whose weight decompositions agree, so the invariants $DF$, $H$ and $I_{NA}-J_{NA}$ match on the two steps. The analytic core is the identity $\\Delta_{\\omega_Y}\\rho_Y+\\rho_Y+|\\bar\\partial\\rho_Y|^2=c$ on the regular locus, together with the Hamiltonian-action formula $R(s)=\\sqrt{-1}\\rho_Y s-\\nabla_{JV_Y}s$, which locates a point where the maximum weight is at most $\\rho_Y$, hence $\\lambda_{\\max}\\le\\max_Y\\rho_Y$. The maximum principle $S_{\\omega_t}>\\inf_X S_{\\omega_0}e^{-t}$ is used to get $S_{\\omega_Y}>0$ on $Y_{reg}$, giving $\\max_Y\\rho_Y\\le n-H(X_b)/V$; the coercivity of the twisted Mabuchi functional for $r<R(X)$, passed to non-Archimedean limits, converts this into the theorem.","core_discovery":"The central claim, Theorem 1.1, is that for every $n$-dimensional Fano manifold $X$ the inequality $DF(X_a)>-(1-R(X))/R(X)\\,nV$ holds for the optimal degeneration $X_a$ produced by the Kähler-Ricci flow, where $V=(-K_X)^n$ and $R(X)$ is the greatest Ricci lower bound. The degeneration is a Q-Fano variety $Y$ carrying a singular Kähler-Ricci soliton $(\\omega_Y,W_Y)$, obtained as the unique sequential polarized Gromov-Hausdorff limit of the flow; the algebraic Donaldson-Futaki invariant coincides there with the analytic integral $Fut(W_Y)=\\int_Y|\\nabla\\rho_Y|^2\\omega_Y^n=\\int_Y|W_Y|^2\\omega_Y^n$. The proof compares the flow limit to the non-Archimedean limits of the Mabuchi, $I-J$, and $H$ functionals, and uses coercivity of the twisted Mabuchi functional for all $r<R(X)$ to force the numerical inequality. The corollaries are a uniform bound $Fut(W_Y)>-F(n)$ over all Fano manifolds of fixed dimension and the inequality $\\inf_{\\omega\\in c_1(X)}H(\\omega)\\le(1-R(X))/R(X)\\,nV$.","pith_inferences":["A plausible sharper statement is that the factor $1/R(X)$ is the correct quantitative instability scale: the proof loses exactly this factor when coercivity is passed to the non-Archimedean limit, so any improvement of the bound likely needs finer information about the limit soliton than the greatest Ricci lower bound alone.","The same two-step degeneration mechanism may apply to other geometric flows that have a soliton limit and a coercivity statement for a twisted functional; the template would give explicit lower bounds on stability invariants for other classes of varieties.","Because the bound is monotone in $R(X)$, it should be testable on toric Fano manifolds where $R(X)$ and the soliton metric are explicitly known; one can check numerically whether the inequality is sharp or whether the true destabilizer is strictly more stable."],"forward_implications":["For fixed dimension $n$, the right-hand side $-(1-R(X))/R(X)\\,nV$ is bounded below by a universal constant, so the Futaki invariants of all flow-produced optimal degenerations are uniformly bounded below.","Combining the theorem with the uniform lower bound $R(X)>\\varepsilon(n)$ gives $Fut(W_Y)>-F(n)$ for every $n$-dimensional Fano manifold, generalizing the finiteness result known for Kähler-Ricci solitons to the whole space of Fano manifolds.","The inequality $\\inf_\\omega H(\\omega)\\le(1-R(X))/R(X)\\,nV$ holds for every Fano manifold, with no restriction such as $R(X)>1/4\\pi$; this makes the bound available in the highly unstable regime.","When $X$ itself admits a Kähler-Ricci soliton, applying the argument directly to $X$ yields $Fut(W_X)>-F$, confirming the uniform-bound conjecture previously settled by compactness methods."],"supporting_citations":[{"why":"Supplies the limit-space theory and compactness results needed to define the Gromov-Hausdorff limit of the Kähler-Ricci flow.","marker":"[CW14]"},{"why":"Provides the two-step R-degeneration from $X$ to the soliton limit and the matching of weight decompositions that identifies the invariants on the two steps.","marker":"[CSW18]"},{"why":"Defines the H-invariant and optimal degeneration, proves the optimality identity involving $\\inf H(\\omega)$, and gives $DF\\le H$.","marker":"[DS16b]"},{"why":"Develops the non-Archimedean limits of energy functionals used to pass the coercivity statement to the algebraic invariants.","marker":"[BHJ17]"},{"why":"Gives the asymptotics of energy functionals along geodesic rays that identify the Donaldson-Futaki invariant as a limit of the Mabuchi functional.","marker":"[BHJ19]"},{"why":"Introduces the greatest Ricci lower bound and proves coercivity of the twisted Mabuchi functional for $r<R(X)$.","marker":"[Szé11]"},{"why":"Introduces the H-functional and its role in studying Kähler-Ricci solitons.","marker":"[He16]"},{"why":"Establishes the uniform bound for Kähler-Ricci solitons that the corollary generalizes to all Fano manifolds.","marker":"[GPSS18]"}],"fun_headline_variants":["Kähler-Ricci flow enforces Futaki lower bound","Quantitative Futaki bound from optimal degenerations","Fano manifolds get uniform Futaki invariant bound","Ricci lower bound controls optimal Futaki","Flow limits yield explicit Donaldson-Futaki inequality"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument's load-bearing premise is that the limiting soliton metric $\\omega_Y$ has strictly positive scalar curvature on the regular part of $Y$; the maximum-principle bound $S_{\\omega_t}>\\inf_X S_{\\omega_0}e^{-t}$ stated in Section 3 only yields $S_{\\omega_Y}\\ge 0$ when the initial metric has negative scalar curvature somewhere, so the written proof does not deliver the strict positivity it uses.","fun_headline_variants_meta":{"raw":{"variants":["Kähler-Ricci flow enforces Futaki lower bound","Quantitative Futaki bound from optimal degenerations","Fano manifolds get uniform Futaki invariant bound","Ricci lower bound controls optimal Futaki","Flow limits yield explicit Donaldson-Futaki inequality"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000198,"raw_usage":{"total_tokens":1357,"prompt_tokens":920,"completion_tokens":437,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":536,"completion_tokens_details":{"reasoning_tokens":361}},"tokens_in":536,"tokens_out":437,"duration_ms":4265,"temperature":1.0,"reasoning_tokens":361,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:51:37.746260+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If one exhibits a Fano manifold $X$ with $R(X)<1$ whose Kähler-Ricci flow limit $(Y,\\omega_Y,W_Y)$ has $S_{\\omega_Y}=0$ at a regular point, then the bound $\\max_Y\\rho_Y\\le n-H(X_b)/V$ and the inequality chain leading to Theorem 1.1 collapse. A concrete place to look is the toric Fano list where $R(X)$ and the soliton metric are explicit: compute $\\inf_Y S_{\\omega_Y}$ on those models; any zero would falsify the strict-positivity step. Running the flow numerically from a metric $\\omega_0\\in c_1(X)$ with $\\inf_X S_{\\omega_0}<0$ and testing whether $S_{\\omega_t}$ becomes eventually bounded below by a positive constant would settle the premise directly.","supporting_citations":[],"review_version":1}