Arithmetic hypergeometric $\mathcal {D}$-modules and exponential sums on reductive groups

Chenhan Liu; Lei Fu; Xuanyou Li

REVIEW 1 major objections 5 minor 25 references

Arithmetic hypergeometric $\mathcal {D}$-modules and exponential sums on reductive groups

T0 review · 1 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper constructs an arithmetic hypergeometric D-module whose Frobenius trace is a hypergeometric exponential sum on a reductive group, and proves an explicit exponential-sum bound with a Weyl-chamber constant.

desk verdict Worth a serious referee: the paper builds new arithmetic hypergeometric D-modules with the right Frobenius traces, but a q vs q' normalization slip leaves the proof of Theorem 0.2 incomplete for nontrivial field extensions. read the letter →

arxiv 2608.00470 v1 pith:LM6DNABD submitted 2026-08-01 math.AG math.NT

Lei Fu , Xuanyou Li , Chenhan Liu This is my paper · ORCID

classification math.AGmath.NT MSC 14F1014F3014M2711L07

keywords arithmeticD-modulehypergeometricexponentialsumreductivegroupoverconvergentF-isocrystalFouriertransformationsphericalvarietyWeylchamberNewtonpolytope

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 studies exponential sums of the form sum over g in G(k') of psi(sum_j Tr(A_j rho_j(g))), where G is a reductive group over a finite field and the rho_j are representations. The authors construct an arithmetic hypergeometric D-module whose Frobenius trace at A is exactly this sum, so p-adic cohomology can bound it. Their main theorem states that for every nondegenerate A the absolute value is at most q'^(d/2) d! times an explicit integral over the dominant Weyl chamber, a bound of the same shape as classical weight bounds but with a concrete constant. The construction works through Fourier transform, invariant differential operators, and a formal model that turns the module into an overconvergent F-isocrystal on the nondegenerate locus. A reader should care because it gives a uniform, explicit estimate for a broad family of exponential sums attached to reductive groups.

What carries the argument

The load-bearing object is the arithmetic hypergeometric D-module $\operatorname{Hyp}_{\pi,!}=\pi_{2,!}f^+L_\pi[d+n-1]$, where $L_\pi$ is the Dwork overconvergent $F$-isocrystal on $\mathbb{A}^1$, $f(g,A)=\sum_j \operatorname{Tr}(A_j\rho_j(g))$, and $\pi_2$ is the projection $G\times V\to V$. Its Frobenius trace at $A$ computes the exponential sum. The proof replaces $\operatorname{Hyp}_{\pi,!}$ by a direct summand, the modified hypergeometric D-module $F_\pi(N)$, whose integral models at finite level $m$ are described by invariant differential operators; the associated graded modules have support controlled by the critical locus of $f_{\tau,A}$ on $G\times G$-orbits. The rank estimate comes from the degree of the closure of $G$ in projective space, computed as the displayed Weyl-chamber integral.

What would settle it

Enumerate all $A=(A_1,\dots,A_N)\in\prod_j \operatorname{End}(V_j)(\mathbb{F}_q)$ for a small reductive group such as $G=\mathrm{GL}_2$ with the standard representation and $q\in\{2,3,5,7\}$, identify the nondegenerate ones via the critical-point criterion (0.1.2), and compare $|\sum_{g\in G(\mathbb{F}_q)} \psi(\operatorname{Tr}(A\rho(g)))|$ with $q^{d/2}d!\int_{\Delta_\infty\cap C}\prod_{\alpha\in R^+}\frac{\lambda(H_\alpha)^2}{\rho(H_\alpha)^2}\,d\lambda$; any violation would disprove Theorem 0.2.

Watch

Extended reading notes

Core claim

The paper's central discovery is a p-adic object that packages the entire family of hypergeometric exponential sums: an overholonomic arithmetic D-module $\operatorname{Hyp}_{\pi,!}$ on the space $V=\prod_j \operatorname{End}(V_j)$, with a Frobenius structure, for which $\operatorname{Tr}(F^m, i_A^+ \operatorname{Hyp}_{\pi,!})$ equals $(-1)^{d+n}q^{-1}$ times the sum over $G(k')$. This module is the Fourier transform of the direct image of the structure sheaf from $G$, and a modified version $F_\pi(N)$ admits an integral model whose associated graded module is supported in the critical locus of the Laurent polynomial. Over the nondegenerate open set $V^{\mathrm{gen}}$, $\operatorname{Hyp}_{\pi,!}$ is an overconvergent $F$-isocrystal of rank at most $d!$ times the Weyl-chamber integral; the p-adic weight theorem then bounds Frobenius traces and yields the main estimate.

Load-bearing premise

The bound is conditional on the existence of an integral equivariant compactification $\tilde{Y}\to Y$ over $R$ satisfying Assumptions 2.1 and 4.3—in particular an open piece isomorphic to $(U^+\times U^-)\times S$ with $S$ a toric scheme over the maximal torus—and if such a model does not exist for a representation family, the formal model, overconvergence, and rank bound do not follow.

Editorial extensions

If this is right

For any nondegenerate $A$, the exponential sum is bounded by $q'^{d/2}\, d! \int_{\Delta_\infty\cap C}\prod_{\alpha\in R^+}\frac{\lambda(H_\alpha)^2}{\rho(H_\alpha)^2}\,d\lambda$.
The full family of hypergeometric sums is controlled by one overholonomic D-module, with cohomology concentrated in degree zero over $V^{\mathrm{gen}}$.
Over $V^{\mathrm{gen}}$, $\operatorname{Hyp}_{\pi,!}$ becomes an overconvergent $F$-isocrystal whose rank is at most the explicit Weyl-chamber integral times $d!$.
The Fourier-transform identity $\operatorname{Hyp}_{\pi,!}\simeq \mathcal{F}_\pi(\iota_{k!}\mathcal{O}_G^\dagger)$ ties the construction to Fourier analysis on the affine space of endomorphisms.
The p-adic weight argument supplies the mixedness needed to convert the rank estimate into an exponential-sum bound.
The construction gives a blueprint for bounding other character sums on reductive groups by similar arithmetic D-modules.

Reading between the lines

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

If the main theorem is correct, the same explicit bound should hold for any quasi-finite representation family once an integral equivariant compactification exists; the geometric Assumptions 2.1 and 4.3 are likely an artifact of the method rather than a genuine restriction on the phenomenon.
For the torus $G=\mathbb{G}_m^n$, this construction should reproduce the known p-adic GKZ hypergeometric bounds, giving a concrete consistency check of both the rank formula and the Weyl-chamber integral.
The explicit constant $d!\int_{\Delta_\infty\cap C}\prod_{\alpha}\frac{\lambda(H_\alpha)^2}{\rho(H_\alpha)^2}\,d\lambda$ may be non-optimal; small-case computations could reveal whether the factorial factor can be replaced by a smaller group-theoretic invariant.
A direct computational search over small finite fields for nondegenerate $A$ could test the inequality numerically and calibrate the constant before deeper arithmetic applications.

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

Worth reading: Lei, Xuanyou, and Chenhan construct arithmetic hypergeometric D-modules Hyp_{π,!} and Hyp_{π,+} on V_k whose Frobenius traces match reductive-group hypergeometric sums, and they prove overconvergence on the nondegenerate locus. That is a real step beyond the torus case in [FLWZ] and the l-adic companion [FL]. The rank bound is an explicit Weyl-chamber integral, and the proof is detailed and innocent of fitted parameters.

The soft spot is real and likely fixable. Proposition 1.10 states that the trace of F^m on i_A^+Hyp_{π,!} is q^{-1} times the exponential sum over G(k') with [k':k]=m. Section 1.11 then bounds that trace by q^{(d-2)/2} times the rank, independent of m, and concludes |S_A| ≤ q^{d/2} rank·integral. For m>1 that is far stronger than the q'^{d/2} bound in Theorem 0.2; for d=1 it would give O(q^{1/2}) for a sum over a group with q^m points, contradicting the expected Weil size q^{m/2}. The fix is almost certainly to replace q^{-1} by q'^{-1}=q^{-m} in Prop 1.10, or to make the weight bound degree-aware, (q^m)^{(d-2)/2}. As written, the chain Prop 1.10 → §1.11 does not establish Theorem 0.2 for nontrivial extensions.

Second, the rank bound imports [FL, Theorem 3.10] from the companion preprint. That is a parameter-free geometric computation, not a circular use of the target bound, but it is unpublished and should be checked.

Third, the theorem is conditional on Assumptions 2.1/4.3, the existence of a good equivariant compactification over R. The authors note this holds for almost all primes in a Dedekind setup, but as a blanket hypothesis it is real.

All told, this is not a desk-reject. The construction is novel, the proof is earnest, and the flaw is identifiable and likely repairable. I would send it to a serious referee, expecting a request for revision on the normalization point. I would not cite it in its current form.

Referee Report

1 major / 5 minor

Summary. The paper develops an arithmetic-D-module approach to exponential sums of the form S_A = Σ_{g∈G(k')} ψ(Tr_{k'/k}(Σ_j Tr(A_j ρ_j(g)))), where G is a split reductive group, ρ_j are representations, and A = (A_j) is a tuple of endomorphisms. The authors define arithmetic hypergeometric D-modules Hyp_{π,+} and Hyp_{π,!} on V = ∏_j End(V_j), identify them as Fourier transforms of ι_+O^†_G and ι_!O^†_G, prove overholonomicity under a quasi-finiteness assumption, and show that over the nondegenerate locus V^gen they are overconvergent F-isocrystals whose rank is bounded by d! times an explicit Weyl-chamber integral. From this they derive Theorem 0.2, an explicit Deligne-style bound |S_A| ≤ q'^{d/2} d! ∫_{Δ_∞∩C} ∏_{α∈R_+} (λ,H_α)^2/(ρ,H_α)^2 dλ. The proof follows the Berthelot–Abe–Caro formalism and relies on the existence of a good integral equivariant compactification (Assumptions 2.1 and 4.3).

Significance. If correct, the paper provides a systematic p-adic framework for hypergeometric exponential sums on reductive groups and gives a parameter-free, explicit upper bound with a geometric constant. The construction is substantial: it defines Frobenius-structured arithmetic D-modules, proves their overconvergence away from the degenerate locus, and reduces the exponential-sum estimate to a rank computation plus Abe–Caro weights. The main derivation is detailed and does not use fitted parameters or normalization tricks. The principal weakness is a normalization mismatch between the trace formula and the weight estimate when the extension degree m = [k':k] exceeds 1; as written, the proof does not literally establish the stated q'^{d/2} bound. This issue is local and apparently correctable, so the underlying approach remains credible.

major comments (1)

[§1.9–1.11, Proposition 1.10 and proof of Theorem 0.2] The chain from Proposition 1.10 to Theorem 0.2 does not prove the stated bound for nontrivial extensions. Proposition 1.10 asserts Tr(F^m,i_A^+Hyp_{π,!}) = (-1)^{d+n} q^{-1} S_A, while Theorem 0.2 claims |S_A| ≤ q'^{d/2} d!∫ = q^{md/2} d!∫ for m=[k':k]. Independently, §1.11 bounds |Tr(F^m,i_A^+Hyp_{π,!})| by q^{(d-2)/2} rank, using that H^{-n}(i_A^+Hyp_{π,!}) has weight ≤ d-2. Since weights are measured relative to the q-Frobenius, the m-th iterate has absolute size q^{m(d-2)/2}, not q^{(d-2)/2}. For m>1 the written proof would yield |S_A| ≤ q^{d/2} d!∫, which is stronger than the expected Deligne/Weil size; e.g. for d=1 it gives O(q^{1/2}) for sums over G(k') with |k'|=q^m, contradicting the expected O(q^{m/2}). The normalization must be repaired: either Proposition 1.10 should have q'^{-1}=q^{-m} and §1.11 should use q^{m(d-2)/2}, or an equivalent degree-aware argument must be supplied. As written, the proof of Theorem 0.2 is incomplete for m>1.

minor comments (5)

[Abstract and Introduction] There are several typos, e.g. “exponential sum associated this Laurent polynomial” in the abstract and “degernate locus” in the introduction; a copyedit pass is needed.
[Proposition 1.10] The displayed trace formula omits the inner trace in Σ_j A_j ρ_j(g); the proof in equation (1.10.3) correctly has Tr(A_j ρ_j(g)). The statement should be made consistent with the proof.
[Theorem 0.2 vs. Proposition 4.22] The notation for the volume integrand is inconsistent: Theorem 0.2 writes (λ,H_α)^2/(ρ,H_α)^2 while Proposition 4.22 writes (λ,α)^2/(ρ,α)^2. The co-root/root convention should be fixed throughout.
[§1.9 and §1.11] “Techm¨uller” is a misspelling of Teichmüller; it occurs in at least two places.
[Corollary 4.20] The statement says F_π(N)^{(m)}|_{V^gen} is coherent as an O_{\hat V}-module, while the proof concludes O_{V_k}-coherence on the special fiber; please clarify the completed/module statement.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the exponential-sum bound is derived from independent D-module and geometric inputs, with only a minor overlapping-author citation for a separate volume identity.

full rationale

The central derivation is not circular. The arithmetic hypergeometric D-module Hyp_{π,!} is defined by an explicit six-functor formula with no fitted parameters, and Proposition 1.10 computes its Frobenius trace via base change and the Dwork theta identity rather than by assuming the desired bound. The rank bound in Theorem 1.5 comes from constructing the modified D-module F_π(N), computing its characteristic cycle, and invoking [FL, Theorem 3.10] only for the parameter-free geometric identity [K(G_m×G):K(X')]·deg(X') = d! ∫_{Δ∞∩C} ∏ (λ,H_α)^2/(ρ,H_α)^2 dλ. That identity is a statement about degrees of the compactification, not about exponential sums, so it is independent support for the constant even though [FL] has overlapping authorship. The final weight estimate uses Abe-Caro's p-adic weight theorem. Assumptions 2.1 and 4.3 are geometric existence hypotheses, not the target statement. No fitted input is renamed as a prediction, and no uniqueness theorem is imported from the authors' prior work to force the choice. The m-dependent normalization discrepancy between Proposition 1.10 and Section 1.11 is a correctness matter external to circularity, not evidence that any output is built into an input.

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

The central claim depends on the standard six-functor and weight framework, on Monsky's theta function, and on two imported inputs: the existence of good integral equivariant compactifications (Assumptions 2.1 and 4.3) and the degree formula from the authors' companion paper [FL]. No free parameters were fit and no external entities were postulated.

assumptions (6)

standard math Six-functor formalism, Verdier duality, and weight theory for overholonomic arithmetic D-modules as developed by Abe and Caro ([AC], [A]).
Used throughout Section 1 and Section 1.11 for f^+, f_!, purity, Tate twists, and the weight bound on Frobenius traces.
standard math Berthelot's theory of arithmetic D-modules D^dagger_{X,Q}(dagger T) and the D^dagger-affineness of projective space ([B1], [B3], [NH1], [Ca1]).
Provides the sheaf of rings, overconvergent singularities, the specialization to F-isocrystals, and the affineness used in Proposition 1.3 and Lemma 4.12.
standard math Noot-Huyghe's Fourier transform preserves overholonomicity ([NH2, 5.3.1]).
Proposition 1.7 and Section 4.13 use this to pass from iota_+ O^dagger_G to Hyp_{pi,+} and to define the modified hypergeometric module.
standard math Monsky's theta function theta(x) = exp(pi x - pi x^q) is overconvergent and produces primitive p-th roots of unity ([M, Theorems 4.1-4.4]).
Defines the Frobenius structure on L_pi and the additive character psi used in Proposition 1.10.
domain assumption Good integral equivariant compactifications satisfying Assumptions 2.1 and 4.3 exist for the quasi-finite representation data.
Theorem 0.2 is stated under these assumptions; Remarks 2.2 and 4.4 assert they hold for almost all primes for split reductive groups over Dedekind domains, but this is not proved for every case here.
domain assumption The degree-volume identity from [FL, Theorem 3.10]: [K(G_m times G):K(X')] deg(X'_k) = d! integral over Delta_infty intersect C of product over positive roots alpha of (lambda,alpha)^2/(rho,alpha)^2 dlambda.
Proposition 4.22 quotes this companion result to obtain the rank bound; it is imported rather than proved in this paper.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Arithmetic hypergeometric $\mathcal {D}$-modules and exponential sums on reductive groups." pith.science (2026). https://pith.science/paper/LM6DNABD

@misc{pith2026260800470,
  author       = {Pith},
  title        = {Pith review of: Arithmetic hypergeometric $\mathcal D$-modules and exponential sums on reductive groups},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LM6DNABD}},
  note         = {Machine review of arXiv:2608.00470}
}

abstract

For a finite family of representations of a reductive group, we define a Laurent polynomial on the group. The exponential sum associated this Laurent polynomial is called a hypergeometric exponential sum. We introduce an arithmetic hypergeometric $\mathcal D$-module to study the hypergeometric exponential sum. It is an overholonomic arithmetic $\mathcal D$-module with a Frobenius structure so that the trace of the Frobenius at a rational point is the exponential sum. Over the locus where the Laurent polynomial is nondegenerate, the arithmetic hypergeometric $\mathcal {D}$-module defines an $F$-isocrystal overconvergent along the degenerate locus. As an application, we get an estimation of the hypergeometric exponential sum.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 22 canonical work pages

[1]

Abe, Explicit calculation of Frobenius isomorphisms and Poincar\'e duality in the theory of arithmetic D -modules , Rend

T. Abe, Explicit calculation of Frobenius isomorphisms and Poincar\'e duality in the theory of arithmetic D -modules , Rend. Semin. Mat. Univ. Padova 131 (2014), 89--150

work page 2014
[2]

Abe and D

T. Abe and D. Caro, Theory of weights in p -adic cohomology, Amer. J. Math. 140 (2018), no. 4, 879--975

work page 2018
[3]

Berthelot, Cohomologie rigide et cohomologie rigide \`a supports propres

P. Berthelot, Cohomologie rigide et cohomologie rigide \`a supports propres. Premi\`ere Partie, pr\'epublication, Universit\'e de Rennes (1996)

work page 1996
[4]

Berthelot, Cohomologie rigide et th\'eorie des D -modules , in: p -adic Analysis (Trento, 1989), Lecture Notes in Math., vol

P. Berthelot, Cohomologie rigide et th\'eorie des D -modules , in: p -adic Analysis (Trento, 1989), Lecture Notes in Math., vol. 1454, Springer, Berlin, 1990, pp. 80--124

work page 1989
[5]

Berthelot, D -modules arithm\'etiques

P. Berthelot, D -modules arithm\'etiques. I. Op\'erateurs diff\'erentiels de niveau fini , Ann. Sci. \'Ecole Norm. Sup. 29 (1996), 185--272

work page 1996
[6]

Berthelot, D -modules arithm\'etiques

P. Berthelot, D -modules arithm\'etiques. II. Descente par Frobenius , M\'em. Soc. Math. Fr. (N.S.) No. 81 (2000)

work page 2000
[7]

Bosch, U

S. Bosch, U. G\"untzer, and R. Remmert, Non-Archimedean Analysis, Grundlehren der mathematischen Wissenschaften, vol. 261, Springer-Verlag, Berlin, 1984

work page 1984
[8]

Brion and S

M. Brion and S. Kumar, Frobenius Splitting Methods in Geometry and Representation Theory, Progress in Mathematics, vol. 231, Birkh\"auser, Boston, 2005

work page 2005

Show all 25 references

[9]

Caro, Fonctions L associ\'ees aux D -modules arithm\'etiques

D. Caro, Fonctions L associ\'ees aux D -modules arithm\'etiques. Cas des courbes , Compositio Math. 142 (2006), 169--206

2006
[10]

Caro, D -modules arithm\'etiques surholonomes , Ann

D. Caro, D -modules arithm\'etiques surholonomes , Ann. Sci. \'Ecole Norm. Sup. 42 (2009), no. 1, 141--192

2009
[11]

D. Caro, B. Tsuzuki, Overholonomicity of overconvergent F -isocrystals over smooth varieties, Ann. of Math. 176 (2012), 747-813

2012
[12]

Conrad, Grothendieck Duality and Base Change, Lecture Notes in Mathematics, vol

B. Conrad, Grothendieck Duality and Base Change, Lecture Notes in Mathematics, vol. 1750, Springer, 2000

2000
[13]

Fu and X

L. Fu and X. Li, Hypergeometric -adic sheaves for reductive groups, arxiv: 2411.11215v3

arXiv
[14]

L. Fu, P. Li, D. Wan, and H. Zhang, p -adic GKZ hypergeometric complex, Math. Ann. 387 (2023), no. 3-4, 1629--1689

2023
[15]

Grothendieck, \'El\'ements de g\'eom\'etrie alg\'ebrique

A. Grothendieck, \'El\'ements de g\'eom\'etrie alg\'ebrique. III. \'Etude cohomologique des Faisceaus coh\'erents, Publ. Math. IH\'ES 11 (1961), 5--167

1961
[16]

Grothendieck, \'El\'ements de g\'eom\'etrie alg\'ebrique

A. Grothendieck, \'El\'ements de g\'eom\'etrie alg\'ebrique. IV. \'Etude locale des sch\'emas et des morphismes de sch\'emas , Troisi\`eme partie, Publ. Math. IH\'ES 11 (1966), 5--255

1966
[17]

Monsky, p -adic analysis and zeta functions, Lectures in Mathematics, Kyoto University, Kinokuniya Book-Store Co

P. Monsky, p -adic analysis and zeta functions, Lectures in Mathematics, Kyoto University, Kinokuniya Book-Store Co. Ltd., 1970

1970
[18]

Hartshorne, Algebraic Geometry, Springer-Verlag (1977)

R. Hartshorne, Algebraic Geometry, Springer-Verlag (1977)

1977
[19]

Noot-Huyghe, D^ ( ) -affinit\'e des sch\'emas projectifs, Ann

C. Noot-Huyghe, D^ ( ) -affinit\'e des sch\'emas projectifs, Ann. Inst. Fourier (Grenoble) 48 (1998), no. 4, 913--956

1998
[20]

Noot-Huyghe, Transformation de Fourier des D -modules arithm\'etiques

C. Noot-Huyghe, Transformation de Fourier des D -modules arithm\'etiques. I , in: Geometric Aspects of Dwork Theory, Vol. I, II, Walter de Gruyter, Berlin, 2004, pp. 857--907

2004
[21]

Rittatore, Algebraic monoids and group embeddings, Transform

A. Rittatore, Algebraic monoids and group embeddings, Transform. Groups 3 (1998), no. 4, 375--396

1998
[22]

Serre, Groupes de Grothendieck des sch\'emas en groupes r\'eductifs d\'eploy\'es, Publ

J.-P. Serre, Groupes de Grothendieck des sch\'emas en groupes r\'eductifs d\'eploy\'es, Publ. Math. IH\'ES 34 (1968), 37--52

1968
[23]

Serre, Alg\`ebre locale --- multiplicit\'es, Lecture Notes in Math., vol

J.-P. Serre, Alg\`ebre locale --- multiplicit\'es, Lecture Notes in Math., vol. 11, Springer-Verlag, Berlin, 1975

1975
[24]

The Stacks Project Authors , The Stacks Project, 2025, https://stacks.math.columbia.edu

2025
[25]

Virrion, Dualit\'e locale et holonomie pour les D -modules arithem\'etiques, Bull

A. Virrion, Dualit\'e locale et holonomie pour les D -modules arithem\'etiques, Bull. Soc. Math. France 128 (2000), 1-68

2000

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

Abe, Explicit calculation of Frobenius isomorphisms and Poincar\'e duality in the theory of arithmetic D -modules , Rend

T. Abe, Explicit calculation of Frobenius isomorphisms and Poincar\'e duality in the theory of arithmetic D -modules , Rend. Semin. Mat. Univ. Padova 131 (2014), 89--150

work page 2014

[2] [2]

Abe and D

T. Abe and D. Caro, Theory of weights in p -adic cohomology, Amer. J. Math. 140 (2018), no. 4, 879--975

work page 2018

[3] [3]

Berthelot, Cohomologie rigide et cohomologie rigide \`a supports propres

P. Berthelot, Cohomologie rigide et cohomologie rigide \`a supports propres. Premi\`ere Partie, pr\'epublication, Universit\'e de Rennes (1996)

work page 1996

[4] [4]

Berthelot, Cohomologie rigide et th\'eorie des D -modules , in: p -adic Analysis (Trento, 1989), Lecture Notes in Math., vol

P. Berthelot, Cohomologie rigide et th\'eorie des D -modules , in: p -adic Analysis (Trento, 1989), Lecture Notes in Math., vol. 1454, Springer, Berlin, 1990, pp. 80--124

work page 1989

[5] [5]

Berthelot, D -modules arithm\'etiques

P. Berthelot, D -modules arithm\'etiques. I. Op\'erateurs diff\'erentiels de niveau fini , Ann. Sci. \'Ecole Norm. Sup. 29 (1996), 185--272

work page 1996

[6] [6]

Berthelot, D -modules arithm\'etiques

P. Berthelot, D -modules arithm\'etiques. II. Descente par Frobenius , M\'em. Soc. Math. Fr. (N.S.) No. 81 (2000)

work page 2000

[7] [7]

Bosch, U

S. Bosch, U. G\"untzer, and R. Remmert, Non-Archimedean Analysis, Grundlehren der mathematischen Wissenschaften, vol. 261, Springer-Verlag, Berlin, 1984

work page 1984

[8] [8]

Brion and S

M. Brion and S. Kumar, Frobenius Splitting Methods in Geometry and Representation Theory, Progress in Mathematics, vol. 231, Birkh\"auser, Boston, 2005

work page 2005