Remarks on Singular K\"ahler-Einstein Metrics

G\'abor Sz\'ekelyhidi; Max Hallgren

REVIEW 2 major objections 4 minor 33 references

Remarks on Singular K\"ahler-Einstein Metrics

T0 review · 2 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Under a synthetic Ricci bound, every rough Kähler-Einstein variety has log terminal singularities.

desk verdict Strong paper, but the RCD proof for flow-limit cones has a real gap in the Sobolev-to-Lipschitz step. read the letter →

arxiv 2505.01943 v1 pith:3H3ONF4Y submitted 2025-05-03 math.DG

Max Hallgren , Gábor Székelyhidi This is my paper · ORCID

classification math.DG MSC 32Q2053C2353C55

keywords singularKähler-EinsteinmetricsRCDspaceslogterminalsingularitiesQ-GorensteinconesKähler-RicciflownoncollapsedlimitsMonge-Ampèreequation

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper asks whether a smooth Kähler-Einstein metric on the regular part of a normal variety, with bounded local potentials, forces the variety to have log terminal singularities. It answers yes whenever the metric completion satisfies a synthetic Ricci lower bound (the RCD condition) together with a quantitative epsilon-regularity condition; such spaces are called rough Kähler-Einstein varieties. The main theorem shows every analytic germ of a rough Kähler-Einstein variety is log terminal, and in the compact or Ricci-flat cone case the metric extends to a singular Kähler-Einstein metric in the Monge-Ampère sense, with algebraic volume ratio in the cone case. Since noncollapsed limits of Kähler-Einstein manifolds and Kähler-Ricci flows automatically meet the rough Kähler-Einstein hypotheses, the two competing notions of singular Kähler-Einstein metric coincide in precisely the settings where these limits appear.

What carries the argument

The central object is the rough Kähler-Einstein variety, defined by a smooth Kähler-Einstein metric on the regular locus with bounded local potentials, local domination of a smooth metric, an RCD metric completion (a synthetic Ricci lower bound together with a dimension bound), and an epsilon-regularity condition. The proof's workhorse is a package of analytic estimates: improved Kato inequalities and Gaussian heat-kernel bounds on the RCD space yield $C^{0}$ and $C^{1}$ control for holomorphic sections of powers of the canonical bundle on the regular locus, after showing the singular set has Hausdorff codimension at least four. L2 estimates for the ∂-operator turn almost-holomorphic peaked sections into genuine holomorphic sections. The peaked sections, concentrated near a prescribed point, imply that a fixed power of the canonical bundle extends across that point, which is exactly the Q-Gorenstein and log-terminal conclusion.

What would settle it

Exhibit a rough Kähler-Einstein variety (smooth Kähler-Einstein metric on the regular locus, bounded local potentials, RCD completion, epsilon-regularity) at some point of which the canonical bundle has infinite index. The theorem asserts no such point exists; finding one would refute the central claim. A practical place to look would be non-log-terminal Kähler cones with Ricci-flat cone metrics, checking whether their metric completions are RCD.

Watch

Extended reading notes

Core claim

The central claim is that the weak and strong notions of singular Kähler-Einstein metric are equivalent under a synthetic Ricci bound. A rough Kähler-Einstein variety, defined by a smooth Kähler-Einstein metric on its regular locus, bounded local potentials, local metric domination, an RCD completion, and an epsilon-regularity condition, has log terminal singularities at every point. In the compact or Ricci-flat cone cases this strengthens to: the variety is Q-Gorenstein, the metric extends to a Kähler current solving the Monge-Ampère equation, and in the cone case the volume ratio is algebraic and the cone is the unique Ricci-flat Kähler cone with its Reeb vector field. Consequently every Ricci-flat metric cone arising as a noncollapsed limit of Kähler-Einstein manifolds or Kähler-Ricci flows satisfies these conclusions.

Load-bearing premise

The load-bearing premise is that the metric completion of the regular locus, with its volume measure, satisfies the RCD condition (a synthetic Ricci lower bound with dimension control) together with the epsilon-regularity condition; if this premise fails, the equivalence of the two notions is not claimed.

Editorial extensions

If this is right

Every Ricci-flat metric cone arising as a noncollapsed limit of Kähler-Einstein manifolds or Kähler-Ricci flows has log terminal singularities and is Q-Gorenstein.
On such spaces, any smooth Kähler-Einstein metric with bounded potentials on the regular locus extends to a singular Kähler-Einstein metric in the Monge-Ampère sense, so the weak and strong notions agree.
For such cones, the volume ratio is an algebraic number, and the cone is the unique Ricci-flat Kähler cone on its underlying variety with the given Reeb vector field.
The local nature of the result means every analytic germ at a singularity of a rough Kähler-Einstein variety is log terminal, not only in the compact or cone cases.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The RCD condition is likely the right synthetic hypothesis for this circle of ideas: it packages the analytic input needed for the section estimates, and the paper suggests that relaxing it would require a wholly different method rather than a small tweak.
The same C^0 and C^1 section estimates could be applied to other Hermitian holomorphic line bundles on singular Kähler spaces, potentially yielding pluricanonical extension theorems beyond the Kähler-Einstein setting.
A testable extension is whether the epsilon-regularity condition is redundant: if it follows from the RCD condition together with bounded potentials, the definition of rough Kähler-Einstein variety could be simplified, widening the class of spaces covered.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

You should know this paper for the C0 and C1 estimates in Section 2, which are a genuinely new tool for constructing sections of powers of the canonical bundle on singular Kähler-Einstein spaces without assuming Q-Gorenstein. The main theorem—rough Kähler-Einstein varieties are log terminal—resolves a conjecture of Sun and a question of Hallgren, and the reduction to the Donaldson–Sun method via RCD heat kernel bounds is well executed. The core is sound.

The soft spot is Proposition 4.2. Theorem 1.6 for Ricci-flow limits rests on proving that the cone C(Z) from an F-limit is RCD(0,2n). The one-paragraph proof invokes Honda's characterization and asserts that C(Z) has the Sobolev-to-Lipschitz property, with no citation or argument. It then verifies only that eigenfunctions (homogeneous harmonic functions) are Lipschitz via Lemma 4.1, which is weaker than Sobolev-to-Lipschitz for arbitrary W^{1,2} functions. Unless Honda's Corollary 3.10 does not require that input—which the text suggests it does—this is a genuine gap: Theorem 1.6 case (B) does not follow from the stated proof. It is likely fixable, by combining the Sobolev inequality cited from [CMZ24] with the doubling and Poincaré structure of the cone, or by locating the property in [Bam21] or [Hal24], but as written it is unsupported. This does not affect Theorems 1.3 and 1.5, and the KE-limit case (A) is fine because RCD stability under GH convergence applies directly.

The paper deserves a serious referee. The main ideas are new, the writing is honest (they mention the independent SWZ25 preprint), and the gap is localized. I would accept for peer review and push for a strengthened proof of Proposition 4.2. I would cite the paper for the rough KE result and the elliptic estimates; the flow-limit conclusion I would cite with a caveat until the gap is closed.

Referee Report

2 major / 4 minor

Summary. The paper introduces the notion of a rough Kähler–Einstein variety (Definition 1.2): a normal Kähler variety with a smooth Kähler–Einstein metric on the regular set, bounded local potentials, local domination, an RCD(λ,2n) condition on the metric completion, and an ε-regularity condition. The main theorem (Theorem 1.3) states that every such variety has log-terminal analytic germs; for compact or conical rough Kähler–Einstein varieties, Theorem 1.5 adds global Q-Gorensteinness, existence of the EGZ singular Kähler–Einstein metric, algebraic volume ratio, and uniqueness of the Ricci-flat cone. Theorem 1.6 applies these results to Ricci-flat cones arising as noncollapsed limits of Kähler–Einstein manifolds or Kähler–Ricci flows, resolving a conjecture from [Sun25] and a question from [Hal24]. The proof combines RCD-space tools (heat-kernel estimates, cutoffs, Moser iteration, improved Kato inequalities) with the Donaldson–Sun construction of peaked holomorphic sections.

Significance. If correct, the paper gives a clean and fairly general bridge between the weak notion of a smooth Kähler–Einstein metric on the regular set with bounded potentials and the stronger EGZ notion, under an RCD hypothesis. The elliptic estimates in Section 2 are carefully written, and the adaptation of Donaldson–Sun to the non-Q-Gorenstein setting is a substantial technical achievement. The case of limits of Kähler–Einstein manifolds (case (A)) appears well supported. However, the advertised application to Kähler–Ricci flow limits (case (B)) rests on Proposition 4.2, whose proof contains an unsubstantiated Sobolev-to-Lipschitz assertion; this is a load-bearing gap. With that gap repaired, the results would be a significant advance.

major comments (2)

[Section 4, Proposition 4.2] The proof asserts that C(Z) satisfies the Sobolev-to-Lipschitz property with no citation or derivation. Honda's characterization [Hon18, Corollary 3.10] appears to require this property as a hypothesis, and the subsequent argument only proves that eigenfunctions of Δ_Z with fixed eigenvalue are Lipschitz, which does not imply Sobolev-to-Lipschitz for arbitrary W^{1,2} functions. Lemma 4.1 handles only dilation-homogeneous harmonic functions on the cone. Consequently the RCD(0,2n) conclusion for case (B) is not established as written, and therefore Theorem 4.3 and Theorem 1.6 do not follow for Kähler–Ricci flow limits. Please supply a proof or a precise reference for the Sobolev-to-Lipschitz property in this setting, or state it as an additional hypothesis if it is not available.
[Section 3, Theorem 1.5(i)] The proof of Theorem 1.5(i) jumps from the local construction of a bounded section to the assertion that X has log-terminal singularities via [EGZ09, Lemma 6.4], and the statement also asserts global Q-Gorensteinness. Log terminality and Q-Cartierness at each point do not by themselves imply that a fixed power of K_X is a line bundle on a compact X, as Remark 1.4 itself notes for noncompact X. Please either cite the standard boundedness of the Cartier index for klt singularities in fixed dimension or provide the compactness argument; otherwise Theorem 1.5(i) is not fully proved.

minor comments (4)

[Definition 1.2] The volume threshold in (v) uses H^{2n}(B(x,r)) ≥ (ω_{2n}−ε)r^{2n}; for consistency with the RCD measure in (iv), please specify explicitly that H^{2n} is the Hausdorff measure of the metric completion and that ω_{2n} is the Euclidean unit-ball volume.
[Proposition 4.2] In the construction of α, the case λ=0 gives α=0, contradicting the displayed 'α>0'. Since eigenvalue-zero eigenfunctions are constant on a connected cross-section, this is harmless but should be noted.
[Lemma 4.1] The heat-kernel time variable is written as 1−t throughout the proof, while earlier sections use r²−t for a scale r. The notational mismatch should be harmonized to avoid confusion.
[Theorem 1.3] The phrase 'argue as in [DS14, Section 3.2.2]' covers a substantial part of the construction of the peaked section v. Since the current setting is not Q-Gorenstein, a slightly longer explanation of which steps of [DS14] carry over verbatim would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular reduction found: Theorem 1.3's log-terminality conclusion is not an input, and the RCD/epsilon-regularity hypotheses are used as assumptions; the Sobolev-to-Lipschitz assertion in Proposition 4.2 is an unproven step (a gap) but not a circular reduction.

full rationale

I walked the derivation chain. Definition 1.2 defines rough Kähler-Einstein varieties by explicitly assuming the RCD condition (iv) and the epsilon-regularity condition (v); Theorem 1.3 concludes log terminality from these assumptions, with the proof constructing Q-Cartier sections via Hörmander L2 estimates and the Donaldson-Sun method. The EGZ criterion [EGZ09, Lemma 6.4] is used as an external theorem, and no parameter is fitted and then renamed a prediction. No statement in the paper is equivalent by construction to its own input. The self-citations [Szé24], [Hal24], and [LS21] are invoked as prior theorems and technical tools (e.g., cutoff constructions, heat kernel estimates, algebraicity of limits), not as premises that already contain the log-terminality conclusion for rough Kähler-Einstein varieties, so they are not load-bearing in the forbidden sense. Per the reviewing rule, I explicitly flag one omitted-support issue: Proposition 4.2 asserts 'C(Z) satisfies the Sobolev to Lipschitz property' with no citation or derivation, and if Honda's characterization [Hon18, Corollary 3.10] requires this property for arbitrary W^{1,2} functions, then the RCD verification for Kähler-Ricci flow limits (case (B)) is incomplete. This is a proof gap or missing justification, not a circular reduction, and therefore it does not raise the circularity score.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim rests on the axioms defining a rough Kähler-Einstein variety, especially the RCD condition and the epsilon-regularity property. No data-fitting parameters appear. The external theorems used (heat kernel, RCD structure, Hörmander estimates) are cited from the literature.

assumptions (5)

domain assumption Definition 1.2(iv): metric completion (X^, d) of (X_reg, ω) is an RCD(λ, 2n) space.
This is the key analytic hypothesis used in Lemma 2.1, Lemma 2.2, Proposition 2.3, and the heat kernel bounds. If false, the proof collapses.
domain assumption Definition 1.2(v): epsilon-regularity: there exists ε > 0 such that volume-almost-Euclidean balls lie in X_reg.
Used to invoke Anderson's epsilon-regularity in Lemma 2.1 and to control the singular set in Proposition 3.1.
domain assumption Definition 1.2(ii),(iii): ω has bounded local potentials and locally dominates a smooth Kähler metric.
Bounded potentials give local Monge-Ampère structure; local domination is used in Lemma 2.2.
standard math External RCD structure theory: De Philippis-Gigli [DPG18], Bruè-Naber-Semola [BNS22], Mondino-Naber [MN19], Honda [Hon18], Ketterer [Ket15].
Used without proof for tangent cone rigidity, codimension bounds, cutoff functions, and cone stability.
standard math EGZ09 Lemma 6.4: existence of an adapted volume form with finite mass implies log terminal singularities.
This criterion is applied at the end of Theorem 1.5(i) to convert the existence of a bounded section into log terminality.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Remarks on Singular K\"ahler-Einstein Metrics." pith.science (2026). https://pith.science/paper/3H3ONF4Y

@misc{pith2026250501943,
  author       = {Pith},
  title        = {Pith review of: Remarks on Singular K\"ahler-Einstein Metrics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3H3ONF4Y}},
  note         = {Machine review of arXiv:2505.01943}
}

read the original abstract

We study two different natural notions of singular K\"ahler-Einstein metrics on normal complex varieties. In the setting of singular Ricci flat K\"ahler cone metrics that arise as non-collapsed limits of sequences of K\"ahler-Einstein metrics or K\"ahler-Ricci flows, we show that an a priori weaker notion is equivalent to the stronger one introduced by Eyssidieux-Guedj-Zeriahi, and in particular the underlying variety has log terminal singularities in this case. Our method applies to more general singular K\"ahler-Einstein spaces as well, assuming that they define RCD spaces.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 25 canonical work pages

[1]

Gioacchino Antonelli, Elia Bru\'e, and Daniele Semola, Volume bounds for the quantitative singular strata of non collapsed RCD metric measure spaces , Anal. Geom. Metr. Spaces 7 (2019), no. 1, 158--178. 4015195

work page 2019
[2]

Luigi Ambrosio, Nicola Gigli, and Giuseppe Savar\'e, Density of L ipschitz functions and equivalence of weak gradients in metric measure spaces , Rev. Mat. Iberoam. 29 (2013), no. 3, 969--996. 3090143

work page 2013
[3]

Anderson, Convergence and rigidity of manifolds under R icci curvature bounds , Invent

Michael T. Anderson, Convergence and rigidity of manifolds under R icci curvature bounds , Invent. Math. 102 (1990), no. 2, 429--445. 1074481

work page 1990
[4]

Bamler, Structure theory of non-collapsed limits of Ricci flows , September 2021

Richard H. Bamler, Structure theory of non-collapsed limits of Ricci flows , September 2021

work page 2021
[5]

, Compactness theory of the space of Super Ricci flows , Invent. Math. 233 (2023), no. 3, 1121--1277

work page 2023
[6]

Elia Bru\`e, Aaron Naber, and Daniele Semola, Boundary regularity and stability for spaces with R icci bounded below , Invent. Math. 228 (2022), no. 2, 777--891. 4411732

work page 2022
[7]

Colding, On the structure of spaces with Ricci curvature bounded below

Jeff Cheeger and Tobias H. Colding, On the structure of spaces with Ricci curvature bounded below. I , Journal of Differential Geometry 46 (1997), no. 3

work page 1997
[8]

Yifan Chen, Shih-Kai Chiu, Max Hallgren, G\'abor Sz\'ekelyhidi, Tat Dat T\^o, and Freid Tong, On K \"ahler- E instein currents , arXiv:2502.09825 (2025)

arXiv 2025

Show all 33 references

[9]

II : L imits with cone angle less than 2 , J

Xiuxiong Chen, Simon Donaldson, and Song Sun, K\"ahler- E instein metrics on F ano manifolds. II : L imits with cone angle less than 2 , J. Amer. Math. Soc. 28 (2015), no. 1, 199--234. 3264767

2015
[10]

Pak-Yeung Chan, Zilu Ma, and Yongjia Zhang, On noncollapsed F -limit metric solitons , arXiv:2401.03387 (2024)

2024 arXiv
[11]

Collins and G \'a bor Sz \'e kelyhidi, Sasaki-- Einstein metrics and K --stability , Geometry & Topology 23 (2019), no

Tristan C. Collins and G \'a bor Sz \'e kelyhidi, Sasaki-- Einstein metrics and K --stability , Geometry & Topology 23 (2019), no. 3, 1339--1413

2019
[12]

Jean-Pierre Demailly, Estimations L^ 2 pour l'op\' e rateur d'un fibr\' e vectoriel holomorphe semi-positif au-dessus d'une vari\' e t\' e k\" a hl\' e rienne compl \`e te , Ann. Sci. \' E cole Norm. Sup. (4) 15 (1982), no. 3, 457--511. 690650

1982
[13]

Demailly, Jean-Pierre , Complex Analytic and Differential Geometry , 2012

2012
[14]

Guido De Philippis and Nicola Gigli, Non-collapsed spaces with R icci curvature bounded from below , J. \'Ec. polytech. Math. 5 (2018), 613--650. 3852263

2018
[15]

213 (2014), no

Simon Donaldson and Song Sun, Gromov- H ausdorff limits of K \"ahler manifolds and algebraic geometry , Acta Math. 213 (2014), no. 1, 63--106. 3261011

2014
[16]

, Gromov-- Hausdorff limits of K \"a hler manifolds and algebraic geometry, II , Journal of Differential Geometry 107 (2017), no. 2

2017
[17]

Evans and Ronald F

Lawrence C. Evans and Ronald F. Gariepy, Measure theory and fine properties of functions, revised ed., Textbooks in Mathematics, CRC Press, Boca Raton, FL, 2015. 3409135

2015
[18]

3, 607--639

Philippe Eyssidieux, Vincent Guedj, and Ahmed Zeriahi, Singular K \"a hler-Einstein metrics , Journal of the American Mathematical Society 22 (2009), no. 3, 607--639

2009
[19]

Nicola Gigli, Andrea Mondino, and Giuseppe Savar\'e, Convergence of pointed non-compact metric measure spaces and stability of R icci curvature bounds and heat flows , Proc. Lond. Math. Soc. (3) 111 (2015), no. 5, 1071--1129. 3477230

2015
[20]

Bin Guo and Jian Song, Nash entropy, C alabi energy and geometric regularization of singular K \"ahler metrics , arXiv:2502.02041 (2025)

2025 arXiv
[21]

Max Hallgren, K\"ahler- R icci tangent flows are infinitesimally algebraic , arXiv:2312.06577 (2024)

2024 arXiv
[22]

Shouhei Honda, Bakry-\' E mery conditions on almost smooth metric measure spaces , Anal. Geom. Metr. Spaces 6 (2018), no. 1, 129--145. 3877323

2018
[23]

Shihoko Ishii, Introduction to singularities, second ed., Springer, Tokyo, 2018. 3838338

2018
[24]

Renjin Jiang, Cheeger-harmonic functions in metric measure spaces revisited, J. Funct. Anal. 266 (2014), no. 3, 1373--1394. 3146820

2014
[25]

44 (2016), no

Renjin Jiang, Huaiqian Li, and Huichun Zhang, Heat kernel bounds on metric measure spaces and some applications, Potential Anal. 44 (2016), no. 3, 601--627. 3489857

2016
[26]

Christian Ketterer, Cones over metric measure spaces and the maximal diameter theorem, J. Math. Pures Appl. (9) 103 (2015), no. 5, 1228--1275. 3333056

2015
[27]

Pure Appl

Gang Liu and G\' a bor Sz\' e kelyhidi, Gromov- H ausdorff limits of K \" a hler manifolds with R icci curvature bounded below II , Comm. Pure Appl. Math. 74 (2021), no. 5, 909--931. 4230063

2021
[28]

Andrea Mondino and Aaron Naber, Structure theory of metric measure spaces with lower R icci curvature bounds , J. Eur. Math. Soc. (JEMS) 21 (2019), no. 6, 1809--1854. 3945743

2019
[29]

Tapio Rajala, Local P oincar\'e inequalities from stable curvature conditions on metric spaces , Calc. Var. Partial Differential Equations 44 (2012), no. 3-4, 477--494. 2915330

2012
[30]

Jian Song, Riemannian geometry of K \"ahler- E instein currents , arXiv:1404.0445 (2014)

2014 arXiv
[31]

Song Sun, Bubbling of K \" a hler- E instein metrics , Pure Appl. Math. Q. 21 (2025), no. 3, 1317--1348

2025
[32]

Song Sun, Jikang Wang, and Junsheng Zhang, On the local topology of non-collapsed R icci bounded limit spaces , preprint, 2025

2025
[33]

G \'a bor Sz \'e kelyhidi, Singular K \"ahler- E instein metrics and RCD spaces , arXiv:2408.10747 (2024)

2024 arXiv

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.

[1] [1]

Gioacchino Antonelli, Elia Bru\'e, and Daniele Semola, Volume bounds for the quantitative singular strata of non collapsed RCD metric measure spaces , Anal. Geom. Metr. Spaces 7 (2019), no. 1, 158--178. 4015195

work page 2019

[2] [2]

Luigi Ambrosio, Nicola Gigli, and Giuseppe Savar\'e, Density of L ipschitz functions and equivalence of weak gradients in metric measure spaces , Rev. Mat. Iberoam. 29 (2013), no. 3, 969--996. 3090143

work page 2013

[3] [3]

Anderson, Convergence and rigidity of manifolds under R icci curvature bounds , Invent

Michael T. Anderson, Convergence and rigidity of manifolds under R icci curvature bounds , Invent. Math. 102 (1990), no. 2, 429--445. 1074481

work page 1990

[4] [4]

Bamler, Structure theory of non-collapsed limits of Ricci flows , September 2021

Richard H. Bamler, Structure theory of non-collapsed limits of Ricci flows , September 2021

work page 2021

[5] [5]

, Compactness theory of the space of Super Ricci flows , Invent. Math. 233 (2023), no. 3, 1121--1277

work page 2023

[6] [6]

Elia Bru\`e, Aaron Naber, and Daniele Semola, Boundary regularity and stability for spaces with R icci bounded below , Invent. Math. 228 (2022), no. 2, 777--891. 4411732

work page 2022

[7] [7]

Colding, On the structure of spaces with Ricci curvature bounded below

Jeff Cheeger and Tobias H. Colding, On the structure of spaces with Ricci curvature bounded below. I , Journal of Differential Geometry 46 (1997), no. 3

work page 1997

[8] [8]

Yifan Chen, Shih-Kai Chiu, Max Hallgren, G\'abor Sz\'ekelyhidi, Tat Dat T\^o, and Freid Tong, On K \"ahler- E instein currents , arXiv:2502.09825 (2025)

arXiv 2025