Pith. sign in

REVIEW 3 major objections 4 minor 17 references

Non-archimedean topological mirror symmetry for $SL_n$ and $PGL_n$ Higgs bundles

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This paper proves that gerbe-twisted p-adic integrals on moduli spaces of $SL_n$ and $PGL_n$ Higgs bundles agree for arbitrary degree, removing the coprime assumption from earlier proofs.

desk verdict A genuine extension of non-archimedean mirror symmetry to arbitrary degree, with a sign-convention typo that needs fixing but doesn't sink the central idea. read the letter →

arxiv 2507.02588 v1 pith:KT6BUNS2 submitted 2025-07-03 math.AG

Elsa Maneval This is my paper · ORCID
classification math.AG MSC 14D2014H6014F2014G1011S80
keywords Higgsbundlestopologicalmirrorsymmetryp-adicintegrationgerbesHitchinfibrationPrymvarietiesintersectioncohomologySYZduality
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 establishes non-archimedean topological mirror symmetry for $SL_n$ and $PGL_n$ Higgs bundles in arbitrary degree, removing the coprime restriction that appeared in previous proofs. It shows that for $D = K_C$ or $\deg D > 2g-2$, the gerbe-twisted p-adic integral over the $SL_n$ moduli space equals the gerbe-twisted p-adic integral over the $PGL_n$ moduli space, up to an explicit Tate-pairing phase depending on the determinant line bundle. The result completes the coprime-case topological mirror symmetry for odd-degree twists and, in the meromorphic case, relates the $SL_n$ integral to intersection cohomology traces. If correct, the coprime assumption is not intrinsic to mirror symmetry for Higgs bundles, and the p-adic integrals become a common tool linking the coprime, non-coprime, and conjectural BPS-cohomology formulations.

What carries the argument

The key mechanism is the torsor of trivialisations $\mathrm{Triv}^0$ of a $\mu_n$-gerbe, upgraded to a $G_m$-gerbe of $n$-torsion, which converts a gerbe on one Hitchin fibre into a Prym torsor on the dual fibre. On the $PGL_n$ side the gerbe $\alpha_N$ is $\Gamma$-equivariant, and its trivialisations are controlled by the relative group scheme $\Gamma_N$ over $C$ built from the Poincar\'e line bundle and the Abel-Jacobi map. These identifications let the author express gerbe functions through Tate pairings, and the p-adic measure is constructed from translation-invariant volume forms pulled back along the self-dual isogeny $\mathrm{Prym}^0 \to \mathrm{Prym}^0/\Gamma$, whose cotangent-volume ratio is controlled by [GWZ20, Lemma 6.15].

What would settle it

Compute the two p-adic integrals of Theorem 4.4 explicitly for a non-coprime example, for instance genus 2 with $n=2$, $D=K$, and even degree, by counting $O_F$-points on the two moduli spaces over a p-adic field and comparing with the predicted gerbe-twisted volumes; a mismatch, or a direct check that the torsor isomorphism in Proposition 3.3 fails for a spectral curve with a non-trivial $N$-class, would falsify the claim.

Watch

Extended reading notes

Core claim

The central claim is Theorem 4.4, an equality of p-adic integrals over $O_F$-points of the good moduli spaces. For any degrees $d,e$ and determinant line bundle $L$, the integral of the gerbe function $f^{e'}_\alpha$ on the $SL_n$ side equals the integral of $f^{d'}_{\alpha_N}$ times $\exp(2\pi i \langle \beta_{L_e}(-), \delta(L_0)\rangle_\Gamma)$ on the $PGL_n$ side, with $d' = d + \deg D$, $e' = e + \deg D$, $L_0 = L \otimes D \otimes N^{-d'}$, and $D = \bigotimes_{i=0}^{n-1} O_C(-iD)$. The proof reduces to the smooth Hitchin fibres, where the $SL_n$ fibre is a product of Prym torsors dual to the $PGL_n$ fibre, via the two SYZ-type symmetries: $\mathrm{Triv}^0(\alpha^{e'} \mid h^{-1}_{SL_n}(a)) \cong h^{-1}_{PGL_n}(a)$ and $\mathrm{Triv}^0(\alpha_N^{d'} \mid h^{-1}_{PGL_n}(a)) \cong \mathrm{Prym}^{N^{d'}}$. Gerbe functions become Tate pairings on the fibres, and the canonical p-adic volumes of the dual abelian varieties $\mathrm{Prym}^0$ and $\mathrm{Prym}^0/\Gamma$ are equal.

Load-bearing premise

The load-bearing premise is the second SYZ-type symmetry (Proposition 3.3): the torsor of $\Gamma$-equivariant trivialisations of the gerbe $\alpha_N$ on a $PGL_n$ Hitchin fibre is isomorphic to the $\mathrm{Prym}^{N^d}$-torsor, a statement that depends on the compatibility of the group scheme $\Gamma_N$ with the Abel-Jacobi normalisation of the Poincar\'e bundle; if this isomorphism failed, the product formula for the $SL_n$ fibre and the Tate-pairing computation would break.

Editorial extensions

If this is right

  • Coprime topological mirror symmetry holds for odd-degree $D$ with the corrected isotypical character twist, so the earlier parity assumption was not needed.
  • For meromorphic $D > K$ and odd prime $n$, the gerbe-twisted $SL_n$ p-adic integral counts intersection cohomology of the moduli space, extending p-adic integration to singular non-coprime spaces.
  • The equality holds for arbitrary degrees, so any future BPS sheaf should reproduce these integrals as Frobenius traces, as formulated in Conjecture 5.9.
  • The refinement in Theorem 4.7 gives a Fourier-type relation between $SL_n$ integrals over degree-shifted determinants and single $PGL_n$ integrals over a fixed specialization locus, providing finer isotypic information.

Reading between the lines

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

  • A plausible next step, not taken in the paper, is to extend the equality to all effective divisors $D$ by treating the non-reduced spectral locus, since the fibrewise argument covers the smooth locus and the remaining strata have measure zero.
  • The parity correction to the coprime case suggests that endoscopic decompositions for odd-degree meromorphic twists should be revisited wherever an equivariant local system is normalized by a basepoint.
  • The BPS conjecture can be tested directly by computing the left-hand integral for a small-rank example where the BPS cohomology is known, since the formula predicts the Frobenius trace on the untwisted part.
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

3 major / 4 minor

Summary. The paper proves a non-archimedean topological mirror symmetry statement for moduli spaces of SL_n and PGL_n Higgs bundles on a curve, allowing arbitrary degree and thus going beyond the coprime case treated by Groechenig–Wyss–Ziegler. The main result, Theorem 4.4, is an equality of p-adic integrals over the integral points of the good moduli spaces, with a phase factor given by a local Tate pairing between the class of the SL_n determinant line bundle and a torsor class on the PGL_n side. The proof goes through two SYZ-type symmetries (Propositions 3.1 and 3.3), conversion of these symmetries into identities of gerbe functions via Tate duality, and then a fibrewise p-adic integration argument following the strategy of [GWZ20] and [COW24]. The paper also derives the coprime-case stringy twisted E-polynomial equality, relates the p-adic integrals to intersection cohomology for prime n and deg D > 2g-2, and proposes a conjecture linking them to BPS cohomology.

Significance. If the main theorem is correct, it is a substantial advance: it removes the coprimality assumption that was central to previous non-archimedean proofs of the Hausel–Thaddeus conjecture, and it gives a clean conjectural framework for the non-coprime case in both the classical and meromorphic settings. The paper is largely self-contained and does not fit parameters to data; the central equality is derived internally from the SYZ-type symmetries and p-adic integration. The refinements and cohomological consequences, especially the correction of a parity condition in earlier work, are potentially significant. However, the proof contains a sign inconsistency in the Tate-pairing phase and a key torsor isomorphism whose proof is only sketched, both of which need to be resolved before the main claim can be taken as established.

major comments (3)
  1. [§4.2, Lemma 4.2 and the display after (4.3)] The derivation of Lemma 4.2(2) has a sign inconsistency that propagates to Theorem 4.4. The displayed computation after (4.3) uses the term ⟨δ(L0), β(y)⟩_Γ, whereas (4.3) and Lemma 4.2(2) state ⟨β(y), δ(L0)⟩_Γ. For the self-dual finite group Γ = Pic^0(C)[n], the local Tate pairing on H^1(F,Γ) is skew-symmetric, so the two expressions differ by a sign; the paper never states its pairing convention. Since Theorem 4.4's phase factor exp(2πi⟨β_{Le}(−), δ(L0)⟩_Γ) is taken from Lemma 4.2(2), the displayed proof does not justify the stated equality unless one of the occurrences is corrected. Please state the convention and make the order consistent throughout, then re-check the sign in the fibrewise argument of Theorem 4.4.
  2. [§3.2, Proposition 3.3] The proof of the second SYZ-type symmetry is too compressed for a result on which the main theorem depends. In particular, the construction of the Γ_N-equivariant structure on L'_{L1}, the identification N' = N via the normalization of the pulled-back Poincaré sheaf at Nm^{-1}(N'), and the final conclusion by torsor rigidity are asserted rather than proved. Since (4.1) uses Proposition 3.3 to identify the SL_n Hitchin fibre with a product of Prym-torsors, the proof should be expanded or a precise reference should be given for each of these steps.
  3. [§4.1, Lemma 4.1] The verification of the [COW24] measure conditions is summarized rather than fully checked. The codimension computation for the strata S_{k,n1,n2} is quoted without derivation, and the existence of O_F-point lifts is only referenced to [Bay+21, Lemma 21.22] and [AHH23, Theorem A8]. Because the canonical measures µ_can are used in the statement of Theorem 4.4, I would like a fuller verification or an explicit statement of which theorem in [COW24] applies verbatim to the SL_n and PGL_n stacks.
minor comments (4)
  1. [§4.3, proof of Theorem 4.4] In the case where both fibres have a rational point, the sentence "the integrands can be expressed in terms of Tate pairing against the dual fibre, exp(2πi⟨−, h^{-1}_{SLn}(a)(F)⟩) (resp. exp(2πi⟨−, h^{-1}_{SLn}(a)(F)⟩))" repeats h^{-1}_{SLn} twice; the second occurrence should presumably be h^{-1}_{PGLn}(a).
  2. [§4.2, first use of Tate pairing] The paper should state explicitly, at the point where the local Tate pairing ⟨−,−⟩_Γ is first used, whether it is defined to be alternating and in which argument order it is taken; this is directly connected to the sign issue raised in the major comments.
  3. [Abstract] There is a typo in the abstract: "Thesep-adic volumes" should read "These p-adic volumes".
  4. [§2.4.1, Lemma 2.11] The description of the Γ-action on the non-rigidified stack is dense; a short example or a diagram displaying the 1-commutativity condition would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main p-adic equality is derived from independently proved SYZ-type torsor isomorphisms and Tate duality, not from its own statement.

full rationale

Theorem 4.4 is not obtained by assuming itself or by fitting a parameter. The derivation chain is: Propositions 3.1 and 3.3 establish torsor isomorphisms between gerbe-trivialisation torsors and Hitchin fibres; Section 4.2 converts these into Tate-pairing identities via (4.3) and the self-duality of the isogeny Prym0 -> Prym0/Γ; and Theorem 4.4 then integrates the resulting gerbe functions fibrewise using the [COW24] measure. The external results [GWZ20], [COW24], [MS22], and [MS23] are used as one-way ingredients and do not contain Theorem 4.4 as an input. There is no fitted parameter relabelled as a prediction, and no load-bearing self-citation chain is present. The apparent sign/order mismatch in the display preceding Lemma 4.2, where ⟨δ(L0), β(y)⟩ appears while Lemma 4.2(2) and Theorem 4.4 use ⟨β(y), δ(L0)⟩, is a consistency or correctness concern for a potentially alternating Tate pairing rather than a circular reduction. Similarly, the terse conclusion in Proposition 3.3 that two Prym0-torsors are isomorphic needs justification over non-algebraically closed fields, but that is also a completeness or correctness issue, not a circularity. The central equality has independent content and is not equivalent to its inputs by construction.

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

The paper is a proof-based mathematics preprint. It introduces no fitted numerical parameters: the auxiliary choices (line bundles N, L_e) are not fitted, and the fermionic shifts F_γ are determined by geometry. The central claim rests on a set of established external results: the BNR spectral correspondence, the moduli theory of Higgs bundles (including normality/smoothness and the canonical p-adic measure of [COW24]), self-duality of Prym and Tate duality, and, in Section 5, the intersection-cohomology mirror symmetry and χ-independence results of [MS22] and [MS23]. Several of these are cited as black boxes, and the paper itself notes that the BPS interpretation is speculative and that the cohomological interpretation is proven only under extra hypotheses (n prime, D > K even).

assumptions (5)
  • domain assumption Existence and geometric properties (normal, rational singularities, stabilizer-free open of codimension at least 2) of the Higgs bundle moduli spaces used to define the canonical p-adic measures.
    Invoked in Lemma 4.1, referring to [Sim94], [Ver24], [COW24] for the details; the verification is summarized rather than fully proven in this paper.
  • standard math BNR spectral correspondence identifying Hitchin fibres with Picard schemes of spectral curves.
    Used in Section 2.5 and Proposition 3.1 to identify h^{-1}_SLn(a) with a Prym-torsor and h^{-1}_PGLn(a) with its quotient, citing [BNR89].
  • standard math Self-duality of the finite group scheme Γ=Pic^0(C)[n] and the self-dual isogeny Φ:Prym0 → Prym0/Γ, together with the relevant Tate pairing.
    Used in the conversions of Section 4.2, citing [GWZ20, Prop 7.10] and [Mil06].
  • domain assumption The canonical p-adic measure of [COW24] exists under the conditions verified in Lemma 4.1.
    The measure is not constructed in this paper; Lemma 4.1 only checks the hypotheses of [COW24].
  • domain assumption Maulik-Shen's isomorphism (5.1) and their χ-independence theorem [MS23], plus Hansen-Scholze relative perversity [HS23], are correct as stated.
    These are external results used in Proposition 5.6 to convert the p-adic integral into an intersection-cohomology trace; the paper flags that one of these results originally carried an implicit parity hypothesis, now corrected in [MS25b].

how reviews work

0 comments
Cite this review

Pith. "Pith review of Non-archimedean topological mirror symmetry for $SL_n$ and $PGL_n$ Higgs bundles." pith.science (2026). https://pith.science/paper/KT6BUNS2

@misc{pith2026250702588,
  author       = {Pith},
  title        = {Pith review of: Non-archimedean topological mirror symmetry for $SL_n$ and $PGL_n$ Higgs bundles},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KT6BUNS2}},
  note         = {Machine review of arXiv:2507.02588}
}
abstract

The Hausel-Thaddeus conjectures concern topological mirror symmetry between moduli spaces of $SL_n$ and $PGL_n$ Higgs bundles on a curve. A non-archimedean approach was introduced by Groechenig, Wyss and Ziegler, proving the conjecture for coprime rank and degree. This article is concerned with its generalisation to the non-coprime case. We treat both the classical ($D=K$) and meromorphic ($D>K$) settings. We prove an equality of $p$-adic volumes twisted by gerbes between moduli spaces of $SL_n$ and $PGL_n$ Higgs bundles of arbitrary degree. In the meromorphic case, building on results of Maulik and Shen, we show that these twisted $p$-adic volumes are related to intersection cohomology. We also conjecture a connection between these $p$-adic volumes and $BPS$ cohomology.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

17 extracted references · 15 canonical work pages

  1. [1]

    Twisted Bundles and Admissible Covers

    [ACV03] Dan Abramovich, Alessio Corti, and Angelo Vistoli. “Twisted Bundles and Admissible Covers”. In: Communications in Algebra 31.8 (2003), pp. 3547–3618. DOI: 10.1081/AGB- 120022434 . URL: https: //doi.org/10.1081/AGB-120022434. [AHH23] Jarod Alper, Daniel Halpern-Leistner, and Jochen Heinloth. “Existence of moduli spaces for algebraic stacks”. In: In...

    work page doi:10.1081/agb- 2003

  2. [8]

    Rational Singularities for Moment Maps of Totally Negative Quivers

    URL: https : / / www . sciencedirect . com / science / article / pii / 0550321396004348. [Ver24] Tanguy Vernet. “Rational Singularities for Moment Maps of Totally Negative Quivers”. In:Transformation Groups (Aug. 2024). ISSN : 1531-586X. DOI: 10.1007/s00031-024-09873-0 . URL: http://dx.doi. org/10.1007/s00031-024-09873-0

    work page doi:10.1007/s00031-024-09873-0 2024

  3. [11]

    Motivic mirror symmetry and $\chi$-independence for Higgs bundles in arbitrary characteristic

    arXiv: 2205.15393 [math.AG] . URL: https://arxiv.org/abs/ 2205.15393. [Kli24] Tyson Klingner. Spectral Data of Special Orthogonal Higgs Bundles and Hecke Modification

    work page Pith review arXiv

  4. [12]

    Spectral Data of Special Orthogonal Higgs Bundles and Hecke Modification

    arXiv: 2312.05711 [math.AG]. URL: https://arxiv.org/abs/2312.05711. [LW21] Franc ¸ois Loeser and Dimitri Wyss. “Motivic integration on the Hitchin fibration”. In: Algebraic Geometry (Mar. 2021), pp. 196–230. ISSN : 2214-2584. DOI: 10.14231/ag-2021-004. URL: http://dx.doi.org/ 10.14231/ag-2021-004. [Lor07] Falko Lorenz. Algebra. en. 2008th ed. Universitext...

    work page Pith review arXiv 2021

  5. [14]

    Endoscopic decompositions and the Hausel-Thaddeus conjecture

    arXiv: 2008.08520 [math.AG]. URL: https://arxiv.org/abs/2008.08520. [MS25b] Davesh Maulik and Junliang Shen. On the intersection cohomology of the moduli of SLn-Higgs bundles on a curve

    work page Pith review arXiv 2008

  6. [15]

    On the intersection cohomology of the moduli of $\mathrm{SL}_n$-Higgs bundles on a curve

    arXiv: 2103.01285 [math.AG]. URL: https://arxiv.org/abs/2103.01285. [Mau21] Mirko Mauri. “Topological mirror symmetry for rank two character varieties of surface groups”. In: Abhandlungen aus dem Mathematischen Seminar der Universit¨ at Hamburg91.2 (Aug. 2021), pp. 297–303. ISSN : 1865-8784. DOI: 10.1007/s12188-021-00246-y . URL: http://dx.doi.org/10.1007...

    work page Pith review arXiv 2021

  7. [16]

    Generalised Prym Varieties as Fixed Points

    [NR75] M. S. Narasimhan and S. Ramanan. “Generalised Prym Varieties as Fixed Points”. In:The Journal of the Indian Mathematical Society 39.1-4 (1975). ISSN : 2455–6475. [Ngˆo10] Baochˆau Ng ˆo. “Le lemme fondamental pour les alg `ebres de Lie”. fr. In: Publications Math´ ematiques de l’IH ´ES 111 (2010), pp. 1–169. DOI: 10.1007/s10240-010-0026-7 . URL: ht...

    work page arXiv 1975

  8. [100]

    Duality for Commutative Group Stacks

    Ast´erisque. Paris: Soc. Math. France, 1982, pp. 5–171. [Bro19] Sylvain Brochard. “Duality for Commutative Group Stacks”. In: International Mathematics Research Notices 2021.3 (Dec. 2019), pp. 2321–2388. ISSN : 1687-0247. DOI: 10.1093/imrn/rnz161 . URL: http://dx. doi.org/10.1093/imrn/rnz161. [Bu+25] Chenjing Bu, Ben Davison, Andr ´es Ib ´a˜nez N ´u˜nez, ...

    work page doi:10.1093/imrn/rnz161 1982

Show all 17 references
  1. [1968]

    Mirror symmetry for moduli spaces of Higgs bundles via p-adic integration

    REFERENCES 33 [GWZ20] Michael Groechenig, Dimitri Wyss, and Paul Ziegler. “Mirror symmetry for moduli spaces of Higgs bundles via p-adic integration”. In: Inventiones Mathematicae 221 (July 2020), pp. 505–596. DOI: 10.1007/ s00222-020-00957-8 . URL: https://doi.org/10.1007/s00...

    2020 doi

  2. [1999]

    [HL24] Victoria Hoskins and Simon Pepin Lehalleur

    arXiv: math/9907034. [HL24] Victoria Hoskins and Simon Pepin Lehalleur. Motivic mirror symmetry and χ-independence for Higgs bundles in arbitrary characteristic

    arXiv

  3. [2006]

    P=W conjectures for character varieties with symplectic resolution

    [FM22] Camilla Felisetti and Mirko Mauri. “P=W conjectures for character varieties with symplectic resolution”. en. In: Journal de l’ ´Ecole polytechnique - Math´ ematiques9 (2022), pp. 853–905. DOI: 10.5802/jep.196. URL: https://www.numdam.org/articles/10.5802/jep.196/. [Gir7...

    2022 doi

  4. [2007]

    Endoscopic decompositions and the Hausel–Thaddeus conjecture

    [MS21] Davesh Maulik and Junliang Shen. “Endoscopic decompositions and the Hausel–Thaddeus conjecture”. In: Forum of Mathematics, Pi 9 (2021), e8. DOI: 10.1017/fmp.2021.7. [MS22] Davesh Maulik and Junliang Shen. “On the intersection cohomology of the moduli of SLn-Higgs bundle...

    2021 doi

  5. [2011]

    Prym varieties of spectral covers

    arXiv: 1102.1717 [math.AG]. [HP12] Tam´as Hausel and Christian Pauly. “Prym varieties of spectral covers”. en. In: Geom. Topol. 16.3 (Aug. 2012), pp. 1609–1638. [HT03] Tam´as Hausel and Michael Thaddeus. “Mirror symmetry, Langlands duality, and the Hitchin system”. In: Inventi...

    2012 arXiv

  6. [2017]

    [CS86] Gary Cornell and Joseph H Silverman, eds

    arXiv: 1708.08124 [math.AG]. [CS86] Gary Cornell and Joseph H Silverman, eds. Arithmetic Geometry. New York, NY: Springer New York,

    arXiv

  7. [2023]

    URL: https: //arxiv.org/abs/2109.06766

    arXiv: 2109.06766 [math.AG]. URL: https: //arxiv.org/abs/2109.06766. [Hau11] Tamas Hausel. Global topology of the Hitchin system

    arXiv

  8. [2024]

    URL: https://arxiv.org/abs/2409.17358

    arXiv: 2409.17358 [math.AG]. URL: https://arxiv.org/abs/2409.17358. [Gro62] Alexander Grothendieck. Fondements de la g´ eom´ etrie alg´ ebrique.fr. Extraits du S´eminaire Bourbaki 1957-1962

    1957 arXiv

  9. [2025]

    BPS invariants from p-adic integrals

    arXiv: 2502 . 04253 [math.AG]. URL: https : / / arxiv . org / abs / 2502 . 04253. [COW24] Francesca Carocci, Giulio Orecchia, and Dimitri Wyss. “BPS invariants from p-adic integrals”. en. In: Compos. Math. 160.7 (July 2024), pp. 1525–1550. [CW17] Sebastian Casalaina-Martin and...

    2024

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.