Pith. sign in

REVIEW 4 minor 24 references

A non-holomorphic P=W phenomenon

T0 review · 0 major / 4 minor · reviewed 2026-07-30 · grok-4.5

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

Pith's one-line read The P=W identity holds for every three-dimensional isolated cluster variety, including singular ones whose Lagrangian fibration is not holomorphic.

desk verdict Solid completion of the 3D non-full-rank isolated cluster P=W case; graded dimensions match and kernels identify the filtrations, with Lefschetz failing as advertised. read the letter →

arxiv 2607.26806 v1 pith:BT37AGI4 submitted 2026-07-29 math.AG

Zili Zhang This is my paper · ORCID
classification math.AG MSC 14F0514F4314C3032S60
keywords P=WconjectureclustervarietiesperversefiltrationmixedHodgemodulesLagrangianfibrationintersectioncohomologynon-holomorphic
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

This paper proves that the perverse filtration coming from a real-analytic Lagrangian fibration on a three-dimensional isolated cluster variety coincides with the weight filtration on its cohomology (and on its intersection cohomology). Earlier work already covered the smooth full-rank case; the new result removes the full-rank hypothesis, so the varieties may be singular and the fibration need not underlie any complex structure. On the weight side the author pushes the constant sheaf forward along a natural non-proper algebraic map and reads the mixed Hodge modules; on the perverse side the author builds the truncation by hand from the monodromy and the degeneration of singular fibres. The two filtrations are then identified because both equal the kernel of restriction to the open set where the radial coordinate is not 1. The result shows that the P=W phenomenon survives outside the holomorphic world, even though both the curious hard Lefschetz and the relative hard Lefschetz symmetries fail.

What carries the argument

The real-analytic Lagrangian fibration h together with the open set X°={|z|≠1}. Both the weight filtration (computed via mixed Hodge modules along the algebraic projection to C*) and the perverse filtration (built from monodromy matrices and fibre degenerations) are identified with the kernels of the restriction maps H^i(X)→H^i(X°) for i=2,3; once the kernels coincide, the filtrations coincide.

What would settle it

Pick concrete integers a,b with ord_2(a)=ord_2(b) (so X is singular), compute the weight filtration on H^*(X) by the mixed-Hodge-module push-forward and the perverse filtration by the explicit sheaf decomposition of Rh_*Q, then check whether the two subspaces of H^2 and H^3 that vanish on {|z|≠1} are identical.

Watch

Extended reading notes

Core claim

For every three-dimensional isolated cluster variety X_{a,b} the perverse filtration associated with the real-analytic map h(x,x',y,y',z)=(|x|^2-|x'|^2,|y|^2-|y'|^2,log|z|) (upper-middle perversity) equals the weight filtration: P_k H^*(X,Q)=W_{2k} H^*(X,Q)=W_{2k+1} H^*(X,Q), and the same identity holds for intersection cohomology. The equality is proved by matching both filtrations, graded piece by graded piece, with the kernels of restriction maps to the open set {|z|≠1}.

Load-bearing premise

Both filtrations are completely determined by the classes that vanish when restricted to the open set where the radial coordinate is not equal to 1; if that open set misses a class or if either kernel computation is incomplete, the subspace equality fails even though the graded dimensions still match.

Editorial extensions

If this is right

  • P=W and PI=WI hold for every three-dimensional isolated cluster variety, full rank or not.
  • The phenomenon persists when the variety is singular and the fibration is essentially non-holomorphic.
  • Both curious hard Lefschetz (weight side) and relative hard Lefschetz (perverse side) fail in the non-full-rank setting.
  • The same kernel-to-open-set comparison supplies a template that can be tried in higher-dimensional non-full-rank cases.

Reading between the lines

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

  • The same stratification-and-kernel method may extend P=W to non-isolated cluster varieties once a suitable real Lagrangian fibration is written down.
  • Failure of both Lefschetz symmetries suggests that the numerical coincidence of graded pieces can outlive the geometric symmetries that originally motivated P=W.
  • The explicit monodromy matrices around the axes give a concrete computational test for any proposed higher-dimensional analogue.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

0 major / 4 minor

Summary. The paper proves the P=W and PI=WI identities for all 3-dimensional isolated cluster varieties X_{a,b} (including the non-full-rank, generally singular case). The weight filtration on H^*(X,Q) and IH^*(X,Q) is computed via mixed Hodge modules along the natural non-proper map f:X o C^*, relative Künneth, and resolution of A_1 singularities; the perverse filtration is constructed from the real-analytic Lagrangian fibration h:X o R^3 by explicit fiber geometry, monodromy, specialization maps, and octahedral reorganization of the sheaves R^k h_*Q. Graded dimensions match, and the filtrations are identified by showing both equal ker(H^i(X) o H^i(X°)) for i=2,3 (X°={|z| eq1}). Both curious hard Lefschetz and relative hard Lefschetz fail in the non-full-rank setting.

Significance. This is a genuine advance on the P=W conjecture for cluster varieties: it treats the first singular, essentially non-holomorphic case and supplies a model for higher-dimensional non-full-rank geometry. The W-side MHM computations and the P-side monodromy/specialization analysis are concrete and reusable; the failure of both Lefschetz symmetries is a clean structural observation. The result completes the 3-dimensional isolated case (full-rank already in the author's earlier work) and gives an explicit template for constructing perverse truncations when BBDG does not apply.

minor comments (4)
  1. [throughout] Several typos and phrasing slips: "ths special case", "sutble", "non-holmorphic", "coordinator z", "compacted supported", "mordal", "fibrationhadmits". A light copy-edit pass would remove them.
  2. [2.4] In §2.4 the diagram (11) is stated not to be a fibered product, while (12) is; a one-sentence clarification of the distinction would help the reader.
  3. [4.4–4.5] Proposition 4.14 and Theorem 4.20 rely on Ext-vanishing that is sketched via support dimension; spelling out the Hom spaces one more line would make the octahedral reorganization easier to check.
  4. [1,3,4] The tables of graded pieces (Theorems 1.4, 1.6, 1.7, 3.12, 4.22, 4.23) are clear but would benefit from a uniform caption convention (smooth vs singular, H vs IH).

Circularity Check

1 steps flagged · score 1.0 of 10

No significant circularity: independent geometric computations of both filtrations, matched by dimension then identified via a common topological kernel.

  1. self citation load bearing [Section 1.2 / Corollary 5.5; also Remark 3.13]
    "Since the full rank case has already been treated in [23], it remains to prove the non-full rank case. … In [21], we compute the mixed Hodge structures on IH^*(X_{1,1}) and H^*(X_{1,1}) via compactification …"

    The full-rank 3-fold case and the special MHS of X_{1,1} are imported from the author’s prior papers. These citations are not load-bearing for the new non-full-rank geometry or for the kernel identification that proves P=W; they merely complete the corollary that covers all 3-dimensional isolated cluster varieties. The central computations (monodromy, fiber degenerations, MHM decompositions for general a,b, and Props. 5.1–5.2) stand independently.

full rationale

The paper computes the weight filtration (via MHM pushforwards along the non-proper algebraic map f:X o C*, relative Künneth, resolution of A1 singularities, and Poincaré duality; Theorems 3.7, 3.12, Props. 3.3–3.11) and the perverse filtration (via explicit monodromy matrices, fiber degenerations, specialization maps, and octahedral reorganization of R^k h_*Q; Theorems 4.10, 4.20–4.23) completely independently. Graded dimensions are shown to match (Thms. 3.12/4.22/4.23). The subspaces are then identified because both equal ker(H^i(X) o H^i(X°)) for i=2,3 (Props. 5.1–5.2), where X°={|z| eq1}; each kernel computation uses its own Cartesian diagram and sheaf morphisms and does not presuppose the other filtration. Self-citations to the author’s [21–23] supply only the already-settled full-rank/2D cases and one low-dimensional MHS check; the non-full-rank geometry, monodromy, combinatorial lemma (Cor. 2.16), and kernel arguments are derived afresh and do not reduce to those inputs by construction. No fitted parameters, no uniqueness theorem imported to force the result, and no renaming of a known pattern. Score 1 reflects only the minor, non-load-bearing self-citations.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The argument rests on standard six-functor formalism for mixed Hodge modules and constructible sheaves on real-analytic spaces, the author's prior P=W results for full-rank/2D cluster varieties, and the combinatorial description of the real Lagrangian fibration h. No free parameters. No new physical or geometric entities beyond the already-standard isolated cluster varieties and the fibration h from prior work.

assumptions (6)
  • standard math Six-functor formalism and relative Künneth for mixed Hodge modules (Saito); dualizing complex on smooth n-fold is Q_X[2n](n).
    Used throughout §2.2 and §3 to decompose Rf_! and compute weight filtrations on H_c^* and IH^*.
  • standard math Perverse t-structure on D^b_c of real-analytic spaces for upper-middle perversity p_+(n)=⌈−n/2⌉ (Kashiwara–Schapira).
    Defines the P-filtration in §2.3; choice of upper vs lower middle is load-bearing (Rem 4.24).
  • standard math BBDG decomposition theorem for the resolution of A1 singularities on the threefold.
    Applied in Lemma 3.9 and Thm 3.12 to split Rπ_* Q into IC plus skyscrapers.
  • domain assumption P=W already holds for full-rank isolated cluster varieties and all 2-dimensional cluster varieties.
    Cited as Thm 2.12 / [23] and [22]; used only to reduce the 3D statement to the non-full-rank matrix (a,b).
  • domain assumption The real-analytic map h of (4)/(27) is a proper Lagrangian fibration with the stated fiber and monodromy geometry.
    Taken from the author's prior construction (with a noted typo fix); fiber descriptions are re-proved in §4.1–4.2 but the global existence of h is inherited.
  • ad hoc to paper Kernel of the arc-integration map g:V^0_{Z_a ∪ Z_b}→V^0_{Z_a}⊕V^0_{Z_b} is Q or 0 according as ord_2(a)=ord_2(b) (Cor 2.16).
    Combinatorial lemma proved in §2.5 and used to control H and the singular central fiber when ord_2(a)=ord_2(b).

how reviews work

0 comments
Cite this review

Pith. "Pith review of A non-holomorphic P=W phenomenon." pith.science (2026). https://pith.science/paper/BT37AGI4

@misc{pith2026260726806,
  author       = {Pith},
  title        = {Pith review of: A non-holomorphic P=W phenomenon},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BT37AGI4}},
  note         = {Machine review of arXiv:2607.26806}
}
read the original abstract

We prove the P=W identity for isolated cluster varieties of dimension 3 without the full rank hypothesis. These cluster varieties are generally singular, and the associated Lagrangian fibrations are not complex algebraic. On the P side, we construct the perverse truncation by a detailed analysis of the explicit real-analytic geometry of the Lagrangian fibration. On the W side, we construct a natural non-proper algebraic morphism from the cluster varieties and investigate the decomposition of the derived push-forward of the constant sheaf along this morphism within the derived category of mixed Hodge modules. In the P=W phenomenon for non-full rank isolated cluster varieties of dimension 3, both the curious hard Lefschetz property on the W side and the relative hard Lefschetz property on the P side fail.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 2 linked inside Pith

  1. [1]

    A. A. Beilinson, J. Bernstein, and P. Deligne, Faisceaux pervers, Analysis and Topology on Singular Spaces, I (Luminy, 1981), Ast´ erisque, vol. 100 (Soc. Math. France, Paris, 1982), 5-171

    1981

  2. [2]

    de Cataldo, T

    M. de Cataldo, T. Hausel, L. Migliorini,Topology of Hitchin systems and Hodge theory of character varieties: the case A1, Annals of Mathematics 175 (2012), 1329-1407

    2012

  3. [3]

    de Cataldo, L

    M. de Cataldo, L. Migliorini,The Hodge theory of algebraic maps, Ann. Sci. ´Ecole Norm. Sup. (4) 38 (2005), no. 5, 693–750

    2005

  4. [4]

    de Cataldo, L

    M. de Cataldo, L. Migliorni,Intersection forms, topology of maps and motivic decomposition for resolutions of threefolds, Algebraic cycles and motives. Vol. 1, 102–137, London Math. Soc. Lecture Note Ser., 343, Cambridge University Press, Cambridge, 2007

    2007

  5. [5]

    de Cataldo, L

    M. de Cataldo, L. Migliorini,The decomposition theorem, perverse sheaves and the topology of algebraic maps, Bull. Amer. Math. Soc. (N.S.) 46 (2009), no. 4, 535-633; doi:10:1090/S0273-0979-09-01260-9

    2009

  6. [6]

    Dancso, M

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

    2024

  7. [7]

    Fomin, A

    S. Fomin, A. Zelevinsky,Cluster algebras. I. Foundations,J. Amer. Math. Soc. 15 (2002), no. 2, 497–529

    2002

  8. [8]

    Hartshorne,Algebraic Geometry, Grad

    R. Hartshorne,Algebraic Geometry, Grad. Texts in Math., No. 52, Springer-Verlag, New York-Heidelberg, 1977, xvi+496 pp

    1977

Show all 24 references
  1. [9]

    Hausel, A

    T. Hausel, A. Mellit, A. Minets, O. Schiffmann,P=W viaH 2, arxiv:2209.05429

    arXiv

  2. [10]

    Hotta, K

    R. Hotta, K. Takeuchi; T. Tanisaki,D-modules, perverse sheaves, and representation theory, Progr. Math., 236, Birkh¨ auser Boston, Inc., Boston, MA, 2008, xii+407 pp

    2008

  3. [11]

    Kashiwara, P

    M. Kashiwara, P. Schapira, Sheaves on Manifolds, Grundlehren Math. Wiss., 292[Fundamental Principles of Mathematical Sciences], Springer-Verlag, Berlin, 1990, x+512 pp

    1990

  4. [12]

    Maulik, J

    D. Maulik, J. Shen,The P=W conjecture forGL n, Ann. of Math. (2) 200 (2024), no. 2, 529-556

    2024

  5. [13]

    Maulik, J

    D. Maulik, J. Shen, Q. Yin,Perverse filtrations and Fourier transforms, Acta Math. 234 (2025), no. 1, 1-69

    2025

  6. [14]

    Muller,Locally acyclic cluster algebras,Adv

    G. Muller,Locally acyclic cluster algebras,Adv. Math. 233 (2013), 207–247

    2013

  7. [15]

    T. Lam, D. Speyer,Cohomology of cluster varieties I: locally acyclic case, Algebra and Number Theory, Volume 16, 2022, No 1

    2022

  8. [16]

    Peters, J

    C. Peters, J. Steenbrink,Mixed Hodge structures,Ergeb. Math. Grenzgeb. (3), 52 [Results in Mathematics and Related Areas. 3rd Series. A Series of Modern Surveys in Mathematics], Springer-Verlag, Berlin, 2008. xiv+470 pp. ISBN:978-3-540-77015-2. 46 ZILI ZHANG

    2008

  9. [17]

    Saito,Modules de Hodge polarisables, Publ

    M. Saito,Modules de Hodge polarisables, Publ. Res. Inst. Math. Sci. 24 (1988), no. 6, 849-995

    1988

  10. [18]

    Saito,Mixed Hodge modules, Publ

    M. Saito,Mixed Hodge modules, Publ. Res. Inst. Math. Sci. 26 (1990), no. 2, 221-333

    1990

  11. [19]

    Saito,Decomposition theorem for proper K¨ ahler morphisms, Tˆ ohoku Math

    M. Saito,Decomposition theorem for proper K¨ ahler morphisms, Tˆ ohoku Math. J., 42 (1990), 127-148

    1990

  12. [20]

    Simpson,Higgs bundles and local systems, Inst

    C. Simpson,Higgs bundles and local systems, Inst. Hautes ´Etudes Sci. Publ. Math. (1992), 5–95

    1992

  13. [21]

    Zhang, Z

    Y. Zhang, Z. Zhang,Mixed Hodge Structures of cluster varieties of dimension 3, to appear in J. Alg., arXiv: 2508.13752

    arXiv

  14. [22]

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

    Z. Zhang,The P=W identity for cluster varieties, Math. Res. Lett. Volume 28, Number 3, 925–944, 2021

    2021

  15. [23]

    Zhang,TheP=Widentity for isolated cluster varieties: full rank case, Math

    Z. Zhang,TheP=Widentity for isolated cluster varieties: full rank case, Math. Z. 308 (2024), no. 1, Paper No. 3, 16 pp

    2024

  16. [24]

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

    Z. Zhang,Multiplicativity of perverse filtration for Hilbert schemes of fibered surfaces, II, Trans. Amer. Math. Soc. 374 (2021), no. 12, 8573–8602. School of Mathematical Sciences, Key Laboratory Intelligent Computing and Applications (Tongji University), Ministry of Educatio...

    2021

Pith tools

Reviewed July 30, 2026 · model on record in the stance chip above.