Pith. sign in

REVIEW 3 major objections 5 minor 19 references

The Aubert-Zelevinsky involution for $G_2$ and its associated Hecke algebras

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The Aubert–Zelevinsky duality for the p-adic group G2, computed on the principal and intermediate series, coincides with the Kato involution on the associated affine Hecke algebras, verifying several cases of the Bernstein conjecture.

desk verdict A careful, honestly attributed set of explicit Aubert-Zelevinsky duality computations for G2 Bernstein blocks; the Hecke-side matching rests on quoted [AX23] tables, but the case checks are real and the paper deserves refereeing. read the letter →

arxiv 2505.17422 v1 pith:BCHJN4MR submitted 2025-05-23 math.RT

Chuan Qin This is my paper · ORCID
classification math.RT MSC 22E5020C08
keywords Aubert-ZelevinskydualityG2affineHeckealgebrasBernsteinblocksKazhdan-Lusztigtriplesconjectureunitarizabilityp-adicreductivegroups
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 tries to establish that a known duality on representations of the p-adic exceptional group G2—the Aubert–Zelevinsky involution—is the same operation, in the principal and intermediate-series Bernstein blocks, as the Kato involution on the associated affine Hecke algebras. The author computes the duality explicitly on both sides, case by case, using parabolic induction and Jacquet restriction on the representation side and Kazhdan–Lusztig indexing triples on the Hecke side. If the computations are correct, the paper confirms several instances of the Bernstein conjecture for G2: duality preserves unitarizability in the listed cases. A sympathetic reader should care because the result gives concrete evidence for the general principle that such dualities on p-adic groups are visible through their Hecke algebras.

What carries the argument

The central object on the representation side is the Aubert–Zelevinsky duality functor $D_G$ on the Grothendieck group, defined as the alternating sum over subsets of simple roots of $i_{P_I} \circ r_{P_I}$; it is an involution that exchanges parabolic induction and Jacquet restriction. On the Hecke side, the matching mechanism is the Kato involution: Theorem 3.3 defines a twisted action $h \mapsto h^*$ on the extended affine Hecke algebra, and $D[M] = [M^*]$ in the Grothendieck group of finite-dimensional modules. The bridge between the two sides is the indexing of irreducible Hecke modules by Kazhdan–Lusztig triples $(t,e,\rho)$, where $t$ is semisimple, $e$ is nilpotent with $\mathrm{Ad}(t)e = qe$, and $\rho$ is a representation of the component group; the standard modules $M_{t,e,\rho}$ are extracted from the equivariant K-theory of the variety of Borel subgroups containing $t$ and the corresponding unipotent element.

What would settle it

Compute $D_{G2}(\pi(\chi))$ for one of the tabulated cases, for example the ramified-cubic case with $s=1/2$ and $\chi^3=1$, directly from the alternating-sum formula (2.1) using only the Jacquet modules listed in Propositions 3.22 and 3.29, and check whether the result matches the Hecke-side module $M_{t,0,1}$ predicted by the table; a mismatch in any single case would disprove the claimed compatibility.

Watch

Extended reading notes

Core claim

The paper computes the Aubert–Zelevinsky duality functor on the Bernstein blocks of G2 labelled by the principal series and the two intermediate series (whose Levi factors are isomorphic to GL2(F)), and shows that in every computed case the representation-side duality D_{G2} matches the Kato involution on the associated Hecke algebra H(J_s,1). Under the indexing by Kazhdan–Lusztig triples (t,e,ρ), the standard module M_{t,e,ρ} is sent to the corresponding module M_{t,0,ρ}, with explicit identities listed in Propositions 3.12–3.32. Using these pairs together with earlier unitarity results, the paper verifies several cases of the Bernstein conjecture: each unitarizable representation in the listed blocks has a unitarizable Aubert–Zelevinsky dual, and each non-unitarizable one has a non-unitarizable dual.

Load-bearing premise

The load-bearing assumption is that the explicit local Langlands correspondence for G2 and the Kazhdan–Lusztig indexing tables, quoted from the preprint and earlier work, are correct and complete for the blocks treated; if a triple is mislabeled or a block is missing, the Hecke-side identities do not follow from the representation-side computations.

Editorial extensions

If this is right

  • For each Bernstein block treated, the representation-theoretic Aubert–Zelevinsky involution is compatible with the Kato involution on the corresponding affine Hecke algebra.
  • The Bernstein conjecture holds for all the listed representations: unitarizable representations have unitarizable duals, and non-unitarizable ones have non-unitarizable duals.
  • The tables give explicit identifications of standard modules of $H(J_s,1)$ with the irreducible subrepresentations and Langlands quotients of parabolic inductions, such as $\pi(\chi) \leftrightarrow M_{t,e,1}$ and the corresponding quotients with $M_{t,0,1}$.
  • The duality computations reproduce known facts, such as $D_{G2}(\mathrm{St}_{G2}) = 1_{G2}$, serving as a check on the method.

Reading between the lines

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

  • One implication the author leaves implicit is that if the compatibility extends to the remaining Bernstein blocks of G2, the Aubert–Zelevinsky involution for the whole group would be determined by affine Hecke algebra data, making the Bernstein conjecture a finite table-check for each block.
  • A testable extension would be to carry out the same case-by-case strategy for other exceptional p-adic groups once the explicit local Langlands correspondence and triple tables are available; the structure of the argument does not use anything special to G2 beyond the tables.
  • The pairing of standard modules under the Hecke involution suggests a direct test of unitarity: for the blocks studied, checking whether a module carries a definite Hermitian form amounts to verifying positivity of the corresponding standard module data, which could be automated.
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 / 5 minor

Summary. The paper computes the Aubert-Zelevinsky duality functor D_G on several Bernstein blocks of the p-adic exceptional group G2, including the two maximal-parabolic intermediate blocks and the principal series blocks corresponding to various choices of ramified/unramified and quadratic/cubic characters. For each block, the author determines D_G on the irreducible subquotients through explicit Jacquet-module computations based on the geometric lemma, following the method of Muić. The representation-side results are then matched, via the explicit local Langlands correspondence of Aubert–Xu [AX23] and the Kazhdan–Lusztig/Ram indexing of standard modules, with the Kato involution D_Hs on the associated affine Hecke algebra modules. The paper closes with a verification of several instances of the Bernstein conjecture on unitarity under Aubert–Zelevinsky duality for G2.

Significance. If the statements are correct, the paper provides a useful set of explicit, reproducible computations confirming the compatibility of Aubert–Zelevinsky duality with the Kato involution for a range of Bernstein blocks of G2, and gives new evidence for the Bernstein conjecture. The Jacquet-module computations are detailed and are cross-checked through Grothendieck-group equations such as (3.26)–(3.28), (3.41), (3.56), and (3.71); the author also carefully flags a typo in a result of Muić (Remark after Proposition 3.28). The main limitations are that the Hecke-side identifications are quoted rather than proved, the indexing-triple labels are not defined in the paper, and one non-unitarity argument in Section 4 rests on a misapplied converse. These issues are local and repairable.

major comments (3)
  1. [§2.5, §3.1.3, and Propositions 3.12, 3.18, 3.24, 3.25, 3.31, 3.32] The paper never states or proves that the Bernstein equivalence Rep_s(G2) ≅ H_s-mod introduced in Section 2.5 intertwines the Aubert–Zelevinsky duality D_G with the Hecke-side Kato involution D_Hs defined in Section 3.1.3. The Hecke-side identities in the cited propositions are presented as consequences of the G2-side computations, but they follow only if such a compatibility is a known theorem or is proved here. Please cite a precise reference for this compatibility or add a proof for the blocks considered; otherwise the 'deduction' of the Hecke involution is not established.
  2. [§2.5 and §§3.4–3.8] The standard-module labels t_a, t_b, t_c, t_d, t_e, t_g, e_α_, e_β_, p21, p3, and similar expressions are used throughout the tables and propositions without being defined. Section 2.5 recalls the general notion of an indexing triple (s,n,ρ) from [KL87], but the specific labels are neither introduced nor keyed to the tables in [Ram03] and [AX23]. Because the central comparison depends on the correct transcription of each triple, please add a table defining all labels used, with explicit references to the relevant sources.
  3. [§4, non-unitarity paragraph] In the proof of Conjecture 4.1, the claim that Iα(δ(ν^{±1/2}ξ2)) and its dual Iα(ν^{±1/2}ξ2∘det) are not unitarizable is attributed to Proposition 4.1(3). That proposition states only that unitarity of the inducing representation implies unitarity of the induced representation; it does not give the converse needed here. Since this is an essential step in the claimed verification of the Bernstein conjecture, please replace the citation with a correct argument (e.g., using the central character, non-Hermiticity, or a suitable result from [Mui97]).
minor comments (5)
  1. [Abstract and Introduction] The phrase 'mediate series' should be 'intermediate series'.
  2. [Eq. (3.29), Prop. 3.12, Table 5] The label 'Mte,eα_+eα_+2β_,p21q' contains a repeated 'eα_'; please clarify whether this is a typo for 'Mte,eα_+2β_,p21q' (the same issue appears for 'Mte,eα_+eα_+2β_,p3q').
  3. [Definition 2.1] In the definition of the convolution product, 'ϕi, ϕ2' should be 'ϕ1, ϕ2'.
  4. [Proposition 3.21(1)] The notation 'rϕ' in the displayed formula for the Jacquet module of π(χ) is undefined; it should presumably be r_T or r_H, and should be corrected.
  5. [§3.6.1 and §3.7] The phrase 'all reduces' should be 'all reduce' (e.g., the sentence before Corollary 3.7).

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the representation-side D_G computations are self-contained, the Hecke-side identities are external transcriptions, and the Section 4 non-unitarity gap is a correctness issue rather than a circular reduction.

full rationale

The derivation chain is not circular. The representation-side computations in Sections 3.3-3.8 apply the alternating-sum definition D_G (2.1) to Jacquet modules (Lemmas 3.4-3.6) and do not presuppose the Hecke-side results. The Hecke-side duality D_H is defined independently in Section 3.1.3, and Theorem 3.3 (Kato) equates it with the twisted action; the paper cites [Kat93] for this, so the 'or the author's thesis' self-citation is non-load-bearing. The Hecke-side identities (Propositions 3.12, 3.18, 3.24, 3.25, 3.31, 3.32) are quoted from [AX23] tables and [Ram03] tables rather than derived from D_G; relying on an external LLC program creates verification risk if the triples are misquoted, but it is not a circular reduction because neither side is defined in terms of the other. Two correctness gaps are flagged: in Section 4, the non-unitarity of Iα(δ(ν^{±1/2}ξ2)) is said to follow from Proposition 4.1(3), but that proposition only states that unitarity of the inducing representation implies unitarity of the induction, not the converse; and the paper nowhere states a theorem that the Bernstein equivalence intertwines D_G with the Kato D_H, so the 'deduction' of the Hecke-side involution from the representation-side computations is asserted rather than proved. These are unproved bridges or external dependencies, not definitional circularity.

Assumptions & free parameters 0 free parameters · 8 assumptions · 0 invented entities

The paper is a case-by-case computation; it introduces no new free parameters or entities. Its central claims rest on standard theorems (geometric lemma, Langlands quotient theorem, Kato duality) and on a chain of external classifications: Aubert-Xu's explicit LLC for G2 (partly an unpublished preprint) and Ram's tables for rank two affine Hecke algebras. The acknowledged fix of an ambiguity in Proposition 3.28 shows that the internal proofs required repair.

assumptions (8)
  • standard math The geometric lemma and second adjointness for parabolic induction and Jacquet functors (BZ77, Ren10).
    Used throughout Section 3 to compute r_alpha and r_beta images, e.g. equations (3.5)-(3.8).
  • standard math The Aubert-Zelevinsky duality properties (Aub95), including D^2 = Id and behavior under induction.
    Invoked in Section 2.2 and in all duality computations such as (3.26)-(3.28).
  • standard math Rodier's length and multiplicity results for principal series of p-adic groups.
    Used to control the composition series of I(nu^{-1} x xi_2) and I_p(nu chi x chi), e.g. Section 3.4 and Proposition 3.21.
  • standard math Langlands quotient theorem (Kon03).
    Used to identify unique irreducible quotients J_gamma(s, pi); see Section 3.1.1.
  • domain assumption The explicit local Langlands correspondence for G2 and the classification of Bernstein blocks from Aubert-Xu (AX23, AX24).
    The paper reads off the Hecke algebra types, the groups J_s, and the standard module labels from [AX23] tables; [AX23] is a preprint. See Sections 2.5, 3.4-3.8.
  • domain assumption The Kazhdan-Lusztig parametrization of affine Hecke algebra modules by indexing triples (KL87) and Ram's tables for rank two affine Hecke algebras (Ram03).
    The Hecke-side modules M_{t,e,1} and the involution identities are drawn from these tables, e.g. Propositions 3.12, 3.18, 3.24.
  • domain assumption Muic's classification and unitarity results for G2 (Mui97, MT11).
    The Bernstein conjecture verification relies on [Mui97, Section 5,6] for unitarity of the duals; the paper also corrects a typo in [Mui97, Proposition 4.2].
  • standard math Kato's duality theorem for affine Hecke algebras (Kat93).
    Theorem 3.3 states D[M]=[M*]; the paper attributes it to Kato or the author's thesis.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Aubert-Zelevinsky involution for $G_2$ and its associated Hecke algebras." pith.science (2026). https://pith.science/paper/BCHJN4MR

@misc{pith2026250517422,
  author       = {Pith},
  title        = {Pith review of: The Aubert-Zelevinsky involution for $G_2$ and its associated Hecke algebras},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BCHJN4MR}},
  note         = {Machine review of arXiv:2505.17422}
}
abstract

Motivated by the recent work of Aubert-Xu and the techniques in G. Muic's article, we provide examples of computations of the Aubert-Zelevinsky duality functor for the principal and mediate series of the exceptional group $G_2$, and deduce corresponding results regarding the involution on the Hecke algebra side. These computations also allow us to confirm several instances of the Bernstein conjecture for $G_2$. This article is developed from part of the author's PhD thesis.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 18 canonical work pages

  1. [1]

    A-M. Aubert. Dualit \'e dans le groupe de G rothendieck de la categorie des representations lisses de longueur finie d'un groupe r \'e ductif p -adique. Transactions of the American Mathematical Society , 347(6):2179--2189, 1995

    work page 1995

  2. [2]

    Aubert and Y

    A-M. Aubert and Y. Xu. The explicit local L anglands correspondence for G _2 , 2023. preprint, 70 pages, arXiv:2208.12391

    arXiv 2023

  3. [3]

    Aubert and Y

    A-M. Aubert and Y. Xu. Hecke algebras for p -adic reductive groups and local L anglands correspondences for B ernstein blocks. Advances in Mathematics , 436:Paper No. 109384, 45, 2024

    work page 2024

  4. [4]

    Bernstein, P

    J. Bernstein, P. Deligne, and D. Kazhdan. Trace P aley- W iener theorem for reductive p -adic groups . Journal d'Analyse Math \'e matique , 47(1):180--192, 1986

    work page 1986

  5. [5]

    I. N. Bernstein. Le ``centre'' de Bernstein . Repr \'e sentations des groupes r \'e ductifs sur un corps local, Travaux en Cours : 1-32, 1984

    work page 1984

  6. [6]

    C. J. Bushnell and P. C. Kutzko. Smooth representations of reductive p -adic groups: structure theory via types. Proceedings of the London Mathematical Society , 77(3):582--634, 1998

    work page 1998

  7. [7]

    I. N. Bernstein and A. V. Zelevinsky. Induced representations of reductive p -adic groups. I . Annales scientifiques de l' \'E cole Normale Sup \'e rieure , 10(4):441--472, 1977

    work page 1977

  8. [8]

    Casselman

    W. Casselman. Introduction to the theory of admissible representations of p-adic reductive groups. https://www.math.utah.edu/ ptrapa/math-library/casselman/casselman-p-adic-book.pdf, 1995. Draft: 1 May 1995

    work page 1995

Show all 19 references
  1. [9]

    Hazeltine, D

    A. Hazeltine, D. Jiang, B. Liu, C.-H. Lo, and Q. Zhang. On representations of A rthur type and unitary dual for classical groups, 2024

    2024

  2. [10]

    S-I. Kato. Duality for representations of a H ecke algebra. Proceedings of The American mathematical society , 119:941--946, 1993

    1993

  3. [11]

    Kazhdan and G

    D. Kazhdan and G. Lusztig. Proof of the D eligne- L anglands conjecture for H ecke algebras. Inventiones mathematicae , 87:153--216, 1987

    1987

  4. [12]

    T. Konno. A note on the Langlands classification and irreducibility of induced representations of \(p\) -adic groups. Kyushu Journal of Mathematics , 57(2):383--409, 2003

    2003

  5. [13]

    Mui\'c and M

    G. Mui\'c and M. Tadi\'c. Unramified unitary duals for split classical p -adic groups; the topology and isolated representations. In On certain L -functions , volume 13 of Clay Math. Proc. , pages 375--438. Amer. Math. Soc., Providence, RI, 2011

    2011

  6. [14]

    G. Mui \'c . The unitary dual of p -adic G _2 . Duke Mathematical Journal , 90:465--493, 1997

    1997

  7. [15]

    A. Ram. Representations of rank two affine H ecke algebras. In Advances in algebra and geometry ( H yderabad, 2001) , pages 57--91. Hindustan Book Agency, New Delhi, 2003

    2001

  8. [16]

    D. Renard. Repr \'e sentations des groupes r \'e ductifs \(p\) -adiques , volume 17 of Cours Sp \'e cialis \'e s (Paris) . Soci \'e t \'e Math \'e matique de France, Paris, 2010

    2010

  9. [17]

    A. Roche. Types and H ecke algebras for principal series representations of split reductive p -adic groups. Annales Scientifiques de l' \'E cole Normale Sup \'e rieure , 31(3):361--413, 1998

    1998

  10. [18]

    F. Rodier. D \'e composition de la s \'e rie principale des groupes r \'e ductifs p -adiques. In Jacques Carmona and Mich \`e le Vergne, editors, Non Commutative Harmonic Analysis and Lie Groups , pages 408--424. Springer Berlin Heidelberg, 1981

    1981

  11. [19]

    F. Shahidi. Langlands' conjecture on P lancherel measures for p -adic groups. In Harmonic analysis on reductive groups ( B runswick, ME , 1989) , volume 101 of Progress in Mathematics , pages 277--295. Birkh\" a user Boston, Boston, MA, 1991

    1989

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.