Pith. sign in

REVIEW 4 major objections 5 minor 20 references

Local multiplicativity of perverse filtrations

T0 review · 4 major / 5 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 Multiplicative local filtrations occur exactly for elliptic fibrations.

desk verdict A clean local criterion for elliptic fibrations via perverse filtrations, but the proof rests on an analytic extension of prior work that is asserted rather than demonstrated. read the letter →

arxiv 2504.19439 v1 pith:P2YKMCYW submitted 2025-04-28 math.AG

Zili Zhang This is my paper · ORCID
classification math.AG MSC 14C0514D0614F08
keywords perversefiltrationlocalmultiplicativityDouadyspaceHilbertschemeofpointsellipticfibrationKählersurfaceP=Wphenomenon
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 studies the local perverse filtration on the fibers of the map $f^{[n]}: S^{[n]}\to C^{(n)}$ induced by a proper holomorphic fibration $f:S\to C$ from a smooth Kähler surface to a smooth curve, where $S^{[n]}$ is the Douady space of length-$n$ zero-dimensional subspaces of $S$. Its central claim is a dichotomy: at a point of $C^{(n)}$ where the $n$ points are distinct, the local perverse filtration is always multiplicative, while at a point where at least two points collide, it is multiplicative if and only if $f$ is an elliptic fibration. Hence the local perverse filtration is multiplicative on every fiber exactly when the general fiber of $f$ has genus 1. The author presents the result as a local, fiberwise analogue of known global multiplicativity criteria, with consequences for the topology of fibers in Higgs-bundle moduli problems.

What carries the argument

The load-bearing machinery is the local perverse filtration on a fiber, defined as the image of perverse truncation of the direct image $Rf_*\mathbb{Q}_X$ after restriction to the point, together with the Douady-Barlet decomposition of $R\pi_*\mathbb{Q}_{S^{[n]}}$ into shifted symmetric-power pieces. Proposition 4.1 decomposes a fiber over $x=\sum \nu_i x_i$ as a product of fibers of smaller Douady maps, compatibly with the filtration by the Künneth formula. Proposition 4.2 then reduces the diagonal case to the central fiber over $n x$, where multiplicativity is equivalent to multiplicativity of the global perverse filtration over a contractible neighborhood; Theorem 3.3, an analytic counterpart of a global criterion for Hilbert schemes of fibered surfaces, converts this into ellipticity of $f$.

What would settle it

For a non-elliptic example, say a fibration whose general fiber is a genus-2 curve, compute explicitly the local perverse filtration on $F^{[2]}(2p)$ for some $p\in C$ and check the cup product $P_1H^1 \times P_1H^1 \to H^2$; the theorem predicts the image lies outside $P_2H^2$, so finding it inside $P_2H^2$ would refute the claim.

Watch

Extended reading notes

Core claim

The central discovery is Theorem 4.4: for $n\ge 2$, the following are equivalent: $f$ is an elliptic fibration; the local perverse filtration at some point $x\in \Delta$ is multiplicative; the local perverse filtration at all points of $C^{(n)}$ is multiplicative. The point is that a global geometric condition on $f$ — that its general fibers are genus-1 curves — is detected by the cup-product behavior of the cohomology of a single special fiber of the induced Douady-space morphism. Along the way the paper establishes the decomposition of any such fiber into a product of central fibers over points $n x$, with the local perverse filtration obeying the Künneth formula.

Load-bearing premise

The argument assumes that the global multiplicativity criterion proved for algebraic Hilbert schemes still holds for Kähler Douady spaces; the paper states this transfer is straightforward but supplies only a brief justification, and the 'only if' direction depends on it.

Editorial extensions

If this is right

  • At points of $C^{(n)}$ where the $n$ points are distinct, the local perverse filtration is multiplicative for every fibration $f:S\to C$, regardless of the fibers.
  • At a collision point with a repeated point of multiplicity at least 2, local multiplicativity is equivalent to $f$ being an elliptic fibration.
  • Local multiplicativity at one diagonal point forces local multiplicativity at all points of $C^{(n)}$, and conversely.
  • If the global perverse filtration associated with $f^{[n]}$ is multiplicative, then all local perverse filtrations are multiplicative; for local models, the converse also holds.
  • The dichotomy gives a concrete obstruction in Higgs-bundle moduli settings: non-elliptic fibrations must violate local multiplicativity at diagonal points.

Reading between the lines

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

  • If the analytic transfer in Theorem 3.3 holds, the same fiberwise criterion should govern every Douady-space map $S^{[n]}\to C^{(n)}$: elliptic fibrations are the unique case where cup products respect the local perverse grading at collisions.
  • Because a single diagonal fiber detects ellipticity, the theorem offers a computational test: checking multiplicativity on one fiber would certify a global geometric property of $f$.
  • The paper shows local-to-global equivalence only for local models; whether fiberwise local multiplicativity always implies global multiplicativity is a testable next step, and the paper's Conjecture 1.1 predicts it does.
  • In weight-filtration matching contexts, the result suggests that failure of local multiplicativity at diagonal fibers may be a precise obstruction preventing such an identity.
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.

Referee Report

4 major / 5 minor

Summary. The paper studies local perverse filtrations on the fibers of the morphism f^{[n]}: S^{[n]} -> C^{(n)} induced by a proper surjective morphism f: S -> C from a smooth Kähler surface to a smooth curve. The main results, Theorems 4.3 and 4.4, assert that the local perverse filtration is multiplicative at a diagonal point of C^{(n)} if and only if f is an elliptic fibration, and is multiplicative at all points if and only if f is elliptic. The proof strategy is to decompose the fibers over a partition into products of fibers of local models, prove a Künneth formula for local perverse filtrations, and then invoke an analytic extension of the author's earlier algebraic multiplicativity criterion for Hilbert schemes of fibered surfaces.

Significance. If correct, the main theorem gives a clean geometric characterization of local multiplicativity of perverse filtrations for Hilbert schemes of fibered surfaces: the local perverse filtration is multiplicative on every fiber precisely when the general fiber has genus one. The paper also provides a useful reduction of Conjecture 1.1 in this class of examples, showing that a single diagonal point controls the global property when the fibration is an elliptic fibration. However, the proof relies on a substantial unproved analytic extension (Theorem 3.3) and on a neighborhood argument (Proposition 2.6) whose proof is not valid as written. Because these points are load-bearing, the main theorem is not yet established in the present form.

major comments (4)
  1. [Section 3, Theorem 3.3] Theorem 3.3 asserts that the main criteria of [18, Theorems 4.9 and 4.17] hold verbatim for proper holomorphic fibrations of Kähler surfaces and Douady spaces, but no proof is supplied. The sentence 'It is straightforward to check the arguments [18, Section 4.3] does not depend on whether C is algebraic or not' is insufficient, because the argument requires the ring-structure theorem of [11] for the cohomology of Douady spaces and the Nakajima/Virasoro operator relations, and it is not automatic that these extend from projective Hilbert schemes to arbitrary compact Kähler surfaces. Since Proposition 4.2 invokes both directions of Theorem 3.3, and Theorem 4.4 rests on Proposition 4.2, this gap is load-bearing.
  2. [Section 2.3, Proposition 2.6] The proof of Proposition 2.6 is not correct as written. The equality (R^d f_* Q_X)_p = H^d(F) only says the stalk of the direct image is the fiber cohomology; it does not imply that for a fixed neighborhood U the restriction map H^d(f^{-1}(U)) -> H^d(F) is an isomorphism. The additional assertion that the maps H^*(U, P_i) -> H^*(p, i^* P_i) are isomorphisms for each perverse summand P_i in (14) is unjustified and fails in general, for instance for perverse sheaves with monodromy on U\{p}. Because Proposition 4.2 uses Proposition 2.6 to identify the local perverse filtration on F^{[n]}(nx) with the perverse filtration on H^*(T^{[n]}), the main theorem depends on this statement. Please provide a correct proof, for example by using that f^{-1}(U) deformation retracts to F for a proper surface fibration over a curve, or state and prove the precise conditions under which the claimed isomorphism holds.
  3. [Section 4, Theorem 4.3(1)] The proof of Theorem 4.3(1) cites [17, Proposition 4.17] for the multiplicativity of the perverse filtration on an arbitrary fiber of a surface fibration, but [17] is written in the algebraic setting. Please supply the Kähler analogue or explain explicitly why the cited result applies to proper holomorphic fibrations of Kähler surfaces. As written, this step inherits the same type of analytic-extension gap as Theorem 3.3.
  4. [Section 4, Proposition 4.2] In the proof of Proposition 4.2, the argument that the adjunction morphism Rf_{V*} Q_T -> p_* p^* Rf_{V*} Q_T induces an isomorphism H^*(T) -> H^*(F) is terse and conflates Proposition 2.6 with proper base change. Please spell out the isomorphism (20) and its compatibility with the perverse decompositions, since the subsequent claim that ι^{[n]*}: H^*(T^{[n]}) -> H^*(F^{[n]}(nx)) is an isomorphism is central to the proof.
minor comments (5)
  1. [Throughout] The paper contains numerous typos and encoding artifacts, such as 'mulitplicativity' in the title, 'proper holomophic' in the proof of Proposition 4.2, and 'K¨ahler' with an encoding error. These should be corrected before publication.
  2. [Introduction, abstract] The phrase 'Douady space of n points on surface S' should be 'on a surface S' for grammatical correctness.
  3. [Section 4, proof of Proposition 4.1] The displayed diagram contains a garbled arrow artifact 'axisshort/axisshort/arrowaxisright' that should be fixed.
  4. [Section 2.4, proof of Proposition 2.8(1)] The justification 'follows from [17, Lemma 2.9] by taking the stalk at ny' needs a brief explanation that the symmetric-power construction and perverse filtration arguments from the algebraic setting carry over to Kähler manifolds without change.
  5. [References] Reference [15] contains a doubled word 'and and' and should be corrected.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: the main theorem reduces local multiplicativity to the author's prior global criterion [18] via a new local-to-global comparison; the asserted Kähler/Douady extension is a correctness gap, not an input-output identity.

full rationale

I walked the derivation chain of Theorem 4.4. Proposition 4.1 decomposes the fiber of f^{[n]} over a partition point into a product of fibers of Douady-space maps and shows, via the Künneth formula of Proposition 2.7, that local multiplicativity at x is equivalent to local multiplicativity at each diagonal block. Proposition 4.2 treats the diagonal case: it uses Proposition 2.6 and a base-change/cup-product identification to compare the local perverse filtration on F^{[n]}(nx) with the global perverse filtration of the local model f_V^{[n]} over the contractible neighborhood V^{(n)}. That comparison is a theorem proved in the paper, not an identification of the conclusion with an input. Theorem 3.3, which supplies the global criterion 'multiplicative iff elliptic' for the local model, is quoted from the author's earlier peer-reviewed work [18] and asserted to extend to the Kähler/Douady setting in one sentence: 'It is straightforward to check the arguments [18, Section 4.3] does not depend on whether C is algebraic or not.' This is a load-bearing external citation, and the analytic extension is not proved in detail in this manuscript; however, it is not a circular step. [18] is an independent prior theorem with a different statement, the local perverse filtration formalism is new content, and no parameter is fitted to the claimed conclusion, nor is any equation defined in terms of the target result. The omitted proof of the Kähler extension is a completeness or correctness risk, which I flag explicitly, but it does not make the derivation equivalent to its inputs. No self-definitional, fitted-input, renaming, or uniqueness-import pattern is present, so the circularity score is 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper's argument depends on a chain of external results, several of which are the author's own prior theorems. The most fragile is the unproved extension of [18] to the Kähler analytic setting (Theorem 3.3). No fitted parameters or new entities are introduced.

assumptions (4)
  • domain assumption Perverse decomposition theorem holds for proper morphisms between Kähler manifolds, including restrictions to open neighborhoods.
    Invoked in Section 2.2 and Proposition 2.6's proof to decompose Rf_*Q. The paper notes the theorem is true for Kähler manifolds in a footnote, but gives no reference for the open-neighborhood case, which is load-bearing for the local argument.
  • ad hoc to paper The Douady-space decomposition (Proposition 3.1) and the perverse filtration formula (Proposition 3.2) extend verbatim from algebraic Hilbert schemes to Kähler analytic Douady spaces.
    Section 3 asserts that the results of [18] and [17] hold in the Kähler setting and says it is 'straightforward to check' the arguments do not depend on algebraicity. No proof is given; the main theorem's local-to-global bridge uses this.
  • domain assumption The converse part of Theorem 3.3 for local models (multiplicativity of perverse filtration for f^{[n]} implies f is elliptic, and equivalence for local models) holds in the Kähler analytic setting (from [18]).
    Used in Proposition 4.2 to identify local multiplicativity at n·x with global multiplicativity over a contractible neighborhood. The cited result is the author's own prior theorem; its analytic version is asserted, not proved.
  • standard math The perverse filtration on any fiber of a surface fibration over a curve is multiplicative ([17, Proposition 4.17]).
    Used in Theorem 4.3 for the ν_i = 1 factors of the product decomposition.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Local multiplicativity of perverse filtrations." pith.science (2026). https://pith.science/paper/P2YKMCYW

@misc{pith2026250419439,
  author       = {Pith},
  title        = {Pith review of: Local multiplicativity of perverse filtrations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/P2YKMCYW}},
  note         = {Machine review of arXiv:2504.19439}
}
abstract

Let $f:S\to C$ be a proper surjective morphism from a smooth K\"ahler surface to a smooth curve. We show that the local perverse filtration associated with the induced map $S^{[n]}\to C^{(n)}$ is multiplicative on each fiber if and only if $f$ is an elliptic fibration.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 17 canonical work pages

  1. [18]

    Zhang, Multiplicativity of perverse filtration for Hilbert scheme s of fibered surfaces, II , Trans

    Z. Zhang, Multiplicativity of perverse filtration for Hilbert scheme s of fibered surfaces, II , Trans. Amer. Math. Soc. 374 (2021), no. 12, 8573–8602

    work page 2021

  2. [11]

    W. Li, Z. Qin, W. Wang, Vertex algebras and the cohomology ring structure of Hilber t schemes of points on surfaces, Math. Ann. 324 (2002), no. 1, 105–133

    work page 2002

  3. [17]

    Zhang, Multiplicativity of perverse filtration for Hilbert scheme s of fibered surfaces , Adv

    Z. Zhang, Multiplicativity of perverse filtration for Hilbert scheme s of fibered surfaces , Adv. Math. 312 (2017) 636-679

    work page 2017

  4. [1]

    Boalch, Simply-laced isomonodromy systems , Publ

    P. Boalch, Simply-laced isomonodromy systems , Publ. Math. Inst. Hautes ´Etudes Sci. 116 (2012), 1–68

    work page 2012

  5. [2]

    Beilinson, J

    A. Beilinson, J. Bernstein, P. Deligne, Faisceaux pervers.(French) Analysis and topology on singular spaces, I (Luminy, 1981), 5–171. Ast´ erisque, 100. Soci´ et´ e Math´ ematique de France, Paris, 1982

    work page 1981

  6. [3]

    de Cataldo, T

    M. de Cataldo, T. Hausel, L. Migliorini, Topology of Hitchin systems and Hodge theory of character va rieties: the case A1, Ann. of Math. (2) 175 (2012), no.3, 1329–1407

    work page 2012

  7. [4]

    de Cataldo, L

    M. de Cataldo, L. Migliorini, The Douady space of a complex surface , Adv. in Math. (2000) 283–312

    work page 2000

  8. [5]

    de Cataldo, D

    M. de Cataldo, D. Maulik, J. Shen, On the P=W conjecture for SLn, Selecta Math. (N.S.) 28 (2022), no.5, Paper No. 90, 21 pp

    work page 2022

Show all 20 references
  1. [6]

    Dancso, M

    Z. Dancso, M. McBreen, V. Shende, Deletion-contraction triangles for Hausel-Proudfoot var ieties, J. Eur. Math. Soc. (JEMS) 26 (2024), no.7, 2565–2653

    2024

  2. [7]

    Groechenig, Hilbert schemes as moduli of Higgs bundles and local systems , Int

    M. Groechenig, Hilbert schemes as moduli of Higgs bundles and local systems , Int. Math. Res. Not. IMRN (2014), no.23, 6523–6575

    2014

  3. [8]

    Katzarkov, V

    L. Katzarkov, V. Przyjalkowski, A. Harder P=W Phenomena , (Russian) Mat. Zametki 108 (2020), no.1, 33–46; translation in Math. Notes 108 (2020), no.1-2, 39–49

    2020

  4. [9]

    Harder, Z

    A. Harder, Z. Li, J. Shen, Q. Yin, P=W for Lagrangian fibrations and degenerations of hyper-K¨ ahler manifolds , Forum Math. Sigma 9 (2021), Paper No. e50, 6 pp

    2021

  5. [10]

    Hausel, A

    T. Hausel, A. Mellit, A. Minets, O. Schiffmann, P = W via H2, arXiv:2209.05429

    arXiv

  6. [12]

    Maulik, J

    D. Maulik, J. Shen, The P=W conjecture for GLn, arXiv:2209.02568

    arXiv

  7. [13]

    M. A. Nieper-Wißkirchen, Twisted cohomology of the Hilbert schemes of points on surfa ces, Doc. Math. 14 (2009), 749–770

    2009

  8. [14]

    Oblomkov, Z.Yun, The cohomology ring of certain compactified Jacobians , arXiv:1710.05391

    A. Oblomkov, Z.Yun, The cohomology ring of certain compactified Jacobians , arXiv:1710.05391

    arXiv

  9. [15]

    Simpson, Higgs bundles and and local systems, Inst

    C. Simpson, Higgs bundles and and local systems, Inst. Hautes ´Etudes Sci. Publ. Math.(1992), no. 75, 5–95

    1992

  10. [16]

    J. Shen, Z. Zhang, Perverse filtrations, Hilbert schemes, and the P = W conjecture for parabolic Higgs bundles , Alg. Geom. 8 (4) (2021),465-489

    2021

  11. [19]

    Zhang, The P = W identity for cluster varieties, Math

    Z. Zhang, The P = W identity for cluster varieties, Math. Res. Lett. 28 (2021) no. 3, 925-944

    2021

  12. [20]

    Zhang, The P=W identity for isolated cluster varieties: full rank c ase, Mathematische Zeitschrift (2024) 308:3

    Z. Zhang, The P=W identity for isolated cluster varieties: full rank c ase, Mathematische Zeitschrift (2024) 308:3. https://doi.org/10.1007/s00209-024-03568-8. School of Mathematical Sciences, Key Laboratory of Intellig ent Computing and Applications (Min- istry of Education),...

    2024 doi

Pith tools

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