{"id":"aeeaf7f3-c258-44ad-bba4-0f9940d15e05","arxiv_id":"2507.06394","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Special values of Bessel-Speh functions for finite general linear groups are exactly given by new exotic matrix Kloosterman sums, which factor into Hall-Littlewood polynomials at Frobenius roots.","lead":"This mathematics paper defines new 'exotic matrix Kloosterman sums' and shows they exactly compute special values of Bessel-Speh functions attached to Speh representations of finite general linear groups. It also expresses the new sums as products of Hall-Littlewood polynomials evaluated at Frobenius eigenvalues, giving explicit formulas and new identities.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 4.6 is only as secure as the imported gamma-factor equality Theorem 4.4 from [5,44], which is not re-derived here and has no independent check for non-principal-series generic representations.","rationale":"The paper is internally coherent: the definitions of exotic matrix Kloosterman sums, non-abelian exotic Gauss sums, and Bessel–Speh functions align; the sign and normalization checks in the principal-series and c=1 limits work; and the Hall–Littlewood computations appear consistent. The single load-bearing step that is not established in this paper is Theorem 4.4, exactly as the reader's weakest-assumption analysis says. Because the main theorem is a short reduction to that imported equality, acceptance should be conditional on verification of Theorem 4.4 in a genuinely new case; the appendix conveniently provides the character-theoretic data needed for such a check. This is not an accusation of error, but a precise statement of where the central claim's security is concentrated.","tokens_in":42986,"tokens_out":19693,"duration_ms":223679,"concrete_test":"Run the smallest non-principal case: q=3, k=c=2, with τ the irreducible cuspidal representation of GL_2(F_3) attached to a regular character α:F_4^×→C^×. Compute K_{τ,ψ}(h) for every h∈GL_2(F_3) directly from the character-averaging formula in §4.3.1, using the Appendix A character table for tr Δ(τ,2). Then form Γ_GK(π∨×τ∨,ψ) from the definition in §4.3.3 for each irreducible π of GL_2(F_3) and compare with ε_0(π∨×τ∨,ψ) from §4.1. A complete match independently verifies Theorem 4.4 in the first non-principal case and closes the dependency gap; any mismatch refutes the central claim.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim Theorem 4.6 is proved by reducing Bessel–Speh special values to Ginzburg–Kaplan gamma factors and then invoking Theorem 4.4, the equality γ_GK(π×τ,ψ)=ε_0(π×τ,ψ) for every irreducible π and every irreducible generic τ. This imported identity is the only bridge from the representation-theoretic object K_{τ,ψ} to the exponential-sum side. If it failed in any non-principal-series case, Theorem 4.6 would fail. The present paper contains no proof of Theorem 4.4, and the cited proofs in [5] and [44] rely on heavy, non-machine-checked machinery; [44] is a preprint. Internal consistency checks do pass in known limiting regimes: for principal series λ=(1^k) the sign factor (-1)^{(k+s)c} becomes (-1)^{2kc}=1 and the identity reduces to formula (1), and for c=1 it reduces to Curtis–Shinoda. The untested regime is non-principal τ with c≥2, exactly where the paper claims new content.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces exotic matrix Kloosterman sums, generalizing both Katz's exotic Kloosterman sums and the twisted matrix Kloosterman sums of earlier work, and proves two main results. First, for an irreducible generic representation τ of GL_k(F) with cuspidal support determined by regular characters α_j, the special values of the associated Bessel–Speh function are shown to equal explicit sign and q-power multiples of these exotic matrix Kloosterman sums (Theorem 4.6, restating Theorem 1.1). Second, exotic matrix Kloosterman sums are expressed as products of modified Hall–Littlewood polynomials evaluated at the roots of the L-function of an exotic Kloosterman sheaf (Theorems 5.3 and 1.2). The proofs use Shintani's norm map, Kondo's Gauss sums, Macdonald's characteristic maps, and previously established Ginzburg–Kaplan gamma factors. Applications include identities for Bessel functions, bounds for the special values, and a comparison with Casselman–Shalika formulas.","tokens_in":43134,"tokens_out":19416,"duration_ms":174264,"significance":"If the results are correct, they provide a substantial bridge between exponential sums and representation theory of finite general linear groups, unifying and extending results of Curtis–Shinoda, Katz, and the author's earlier work. The Macdonald characteristic map computation in Section 5 is original and appears to be the first explicit connection of this type with gamma-factor theories. The paper also gives concrete applications, including a generating-function identity and effective bounds. The dependence on an unproved external gamma-factor identity is a concern, but the Hall–Littlewood part is proved independently and is a genuine contribution.","major_comments":[{"comment":"The unnormalized formula in Theorem 5.3 states Kl(α,ψ,h) = (-1)^{(k-1)c} q^{(k-1)c^2} ∏ \tilde{H}_{μ_j}(ω_{1,[ξ_j]},...,ω_{k,[ξ_j]}; q^{a_j}). This is inconsistent with the proof. From the normalized identity Kl^* = ∏ \tilde{H}_{μ_j}((-1)^{(k-1)a_j} ω^*_{1,[ξ_j]},...,ω^*_{k,[ξ_j]}; q^{a_j}) and the definitions Kl^* = q^{-(k-1)c^2/2} Kl (Section 3.5) together with \tilde{H}(ω^*) = q^{-(k-1)c/2} \tilde{H}(ω), one obtains Kl = (-1)^{(k-1)c} q^{(k-1)c(c-1)/2} ∏ \tilde{H}(ω), not q^{(k-1)c^2}. For c = 1, the stated exponent gives q^{k-1} ≠ 1, which contradicts Remark 3.9, where Kl equals the classical exotic Kloosterman sum. The normalized version appears correct and is what is used later, but the unnormalized statement as printed is false and must be corrected.","section":"Theorem 5.3 (and Theorem 1.2)"},{"comment":"The central identity Theorem 4.6 is derived by reducing Bessel–Speh special values to Ginzburg–Kaplan gamma factors and then invoking Theorem 4.4, the equality γ_GK(π×τ,ψ) = ε_0(π×τ,ψ) for all irreducible π and all irreducible generic τ. This equality is not proved in this paper; it is imported from the author's prior work [5] and the preprint [44]. Since [44] is an unpublished arXiv preprint and the equality is used precisely in the regime of non-principal-series generic τ with c ≥ 2, where the paper claims new content, Theorem 4.6 is conditional on an external result whose correctness has not been independently verified here. The internal consistency checks (principal series λ=(1^k), and c=1 reducing to Curtis–Shinoda) do not cover this regime. Please either include a proof of the required cases of Theorem 4.4, or explicitly state Theorem 4.6 as contingent on the external equality and indicate the status of [44].","section":"Section 4.4, Theorem 4.6"}],"minor_comments":[{"comment":"In the sentence 'Given a character α = α_1 × ... × α_s → C', the domain is missing; it should be α : F_λ^× → C^×.","section":"Section 1.4"},{"comment":"The sentence 'Carmon proved in [3, Theorem 6.18] that for any irreducible generic representation τ of GL_{kc}(F)' should read 'of GL_k(F)', since the Speh representation Δ(τ,c) is a representation of GL_{kc}(F).","section":"Section 4.2.3"},{"comment":"The subscript in 'p^{[ξ]}_{kdegθ/degξ}' is ambiguous; please clarify the intended notation, for instance by writing p^{[ξ]}_{k deg(θ)/deg(ξ)} or adding a parenthetical explanation.","section":"Section 5.2.6, equation (15)"},{"comment":"The remark that the normalized version 'is not exactly the same normalized version as in [46]' is useful, but it would be clearer to state the precise difference in conventions, since a mismatch here can affect comparisons with the earlier paper.","section":"Section 2.3.2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is substantial and the independent part (Section 5) is well executed, but the unnormalized statement of Theorem 5.3 contains a clear exponent error that must be fixed, and the main relation Theorem 4.6 depends on an external equality from a preprint [44] that is not re-proved. The author should be asked to correct the q-exponent in Theorem 5.3 (and Theorem 1.2) and to clarify the status of the dependence on [44]. I do not see grounds for rejection, but the paper is not ready for acceptance as it stands."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThis one is worth reading if you work on finite-field Kloosterman sums or Bessel functions. The paper defines exotic matrix Kloosterman sums, which genuinely generalize both Katz's exotic sums and the author's earlier twisted matrix sums, and then proves explicit formulas for Bessel-Speh special values in terms of them for all irreducible generic representations of GL_k(F). The Hall-Littlewood factorization (Theorem 5.3) is new, and its proof via Macdonald characteristic maps is independent of the gamma-factor route and is clean. The applications include new identities for Bessel functions and bounds; those look correct.\n\nThe writing is clear and honest. The paper explicitly says most proofs are short because they rely on heavy machinery (Shintani, Kondo, Silberger-Zink, Macdonald, and the author's own previous work). That is accurate, and it is not a flaw by itself.\n\nThe real soft spot is Theorem 4.6. It is the bridge from Bessel-Speh values to the exponential-sum side, and it depends on Theorem 4.4: the Ginzburg-Kaplan gamma factor equals the epsilon_0 factor for every irreducible pi and irreducible generic tau. That equality was proved in [5] and [44], but is not re-derived here. [5] is in press, but [44] is an arXiv preprint. If that equality fails for some non-principal-series tau with c>=2, the main theorem fails. The checks in the paper only cover principal series and c=1 (Curtis-Shinoda), so the genuinely new regime is exactly the one not independently verified in this paper. That is a vulnerability worth flagging to the referee, but it is not an internal error and not a reason to desk-reject.\n\nThe paper also leans heavily on the author's own prior work, but that is justified here: those are the papers that proved the ingredients. The self-citation pattern does not look like padding.\n\nMy bottom line: the central claim is new, the derivation is coherent, and the dependence on imported results is explicit. The right referee should check [44] carefully and ask whether the gamma-factor equality is unconditionally proved there. If yes, accept; if [44] has gaps, the main theorem needs support. Either way, this deserves a serious peer review.\n\nBest.","headline":"A serious and largely convincing paper: the new exotic matrix Kloosterman sums are genuinely useful, and the main identity is derived rather than assumed; the main risk is a single imported gamma-factor theorem that the author does not re-prove.","tokens_in":43697,"tokens_out":2535,"would_cite":true,"duration_ms":27856,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["20C33","11L05","11T24"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that special values of Bessel-Speh functions for all irreducible generic representations of general linear groups over finite fields are exotic matrix Kloosterman sums up to explicit scalar factors.","keywords":["matrix Kloosterman sums","exotic Kloosterman sums","Bessel-Speh functions","Speh representations","finite general linear groups","Hall-Littlewood polynomials","Ginzburg-Kaplan gamma factors","Shintani norm map"],"falsifier":"Pick a small finite field, for instance $\\mathbb{F}_3$, take $k=2$ with $\\tau$ the cuspidal representation of $\\operatorname{GL}_2(\\mathbb{F}_3)$ attached to a regular character of $\\mathbb{F}_9^\\times$, choose $c=2$, and let $h\\in\\operatorname{GL}_2(\\mathbb{F}_3)$ be a regular elliptic matrix. Compute $K_{\\tau,\\psi}(h)$ directly from the character formula for Speh representations and compute the exotic matrix Kloosterman sum on the right-hand side of Theorem 4.6 by summation over $\\operatorname{GL}_2(\\mathbb{F}_9)$; any mismatch between the two sides would settle the theorem negatively.","tokens_in":42722,"feed_emoji":"🧮","tokens_out":15223,"duration_ms":137573,"temperature":0.7,"pith_summary":"This paper proves an exact formula for the special values of Bessel–Speh functions, the distinguished matrix coefficients attached to Speh representations of general linear groups over finite fields. The formula says that for every irreducible generic representation, each such value is equal to a newly introduced exponential sum, an exotic matrix Kloosterman sum, multiplied by an explicit sign and a power of $q$. These new sums unify the previously studied twisted matrix Kloosterman sums and the classical exotic Kloosterman sums. The paper also proves that any exotic matrix Kloosterman sum can be written as a product of modified Hall–Littlewood polynomials evaluated at the Frobenius eigenvalues of an exotic Kloosterman sheaf. In this way representation-theoretic data are converted into arithmetic data, giving a uniform description of Bessel–Speh values across all irreducible generic representations.","feed_headline":"Bessel-Speh values are exotic matrix Kloosterman sums","feed_subtitle":"New formula covers every irreducible generic representation over a finite field, up to explicit sign and q-power.","key_machinery":"The carrier of the argument is the exotic matrix Kloosterman sum $\\operatorname{Kl}(\\alpha,\\psi,h)$, defined for a single extension field $\\mathbb{F}_k$ by summing over $x\\in\\operatorname{GL}_c(\\mathbb{F}_k)$ whose norm-map conjugacy class contains $h$, weighted by $\\alpha(\\det x)\\psi_k(\\operatorname{tr}x)$, and for a partition by convolving such sums over factor products. Two independent mechanisms prove the two main theorems. Theorem 4.6 is derived from the equality, imported from previous work, between the Ginzburg–Kaplan gamma factor and the tensor-product epsilon factor; a character-averaging formula converts the Bessel–Speh value into a trace of this gamma factor, and Schur orthogonality then recovers the exponential sum. Theorem 5.3 is obtained through the characteristic maps that identify the ring of class functions with a ring of symmetric functions; these send the global class function $h\\mapsto\\operatorname{Kl}(\\alpha,\\psi,h)$ to an explicit product of $L$-functions of the exotic Kloosterman sheaf, and expanding that product under the Cauchy identity yields the modified Hall–Littlewood polynomials.","core_discovery":"The paper's central claim is that the representation-theoretic side and the exponential-sum side coincide exactly. For an irreducible generic representation $\\tau$ of $\\operatorname{GL}_k(F)$ with cuspidal support $\\{\\tau_1,\\dots,\\tau_s\\}$, where each $\\tau_j$ corresponds to a regular character $\\alpha_j$ of $\\mathbb{F}_{k_j}^\\times$, Theorem 4.6 gives $$K_{\\tau,\\psi}(h)=(-1)^{(k+s)c}$q^{{-(k-1)c^2}}$\\operatorname{Kl}(\\$alpha^{{-1}}$,\\psi,(-1)^{k-1}$h^{{-1}}$)$$ for every $h\\in\\operatorname{GL}_c(F)$, where $K_{\\tau,\\psi}(h)$ is the Bessel–Speh value at $\\operatorname{diag}(I_{(k-1)c},h)$ (with $k=1$ treated separately by $\\tau(\\det h)\\psi(\\operatorname{tr} h^{-1})$) and $\\operatorname{Kl}$ is the newly defined exotic matrix Kloosterman sum. The paper also proves a second, more explicit identity (Theorem 5.3): for $h$ written as a product of generalized Jordan blocks $J_{\\mu_i}(h_{\\xi_i})$, the sum $\\operatorname{Kl}(\\alpha,\\psi,h)$ factors into modified Hall–Littlewood polynomials evaluated at the Frobenius roots of the exotic Kloosterman sheaf at the eigenvalues $\\xi_i$. Together these results give a complete arithmetic description of Bessel–Speh special values.","pith_inferences":["The resemblance highlighted in the paper's last section suggests a dictionary in which normalized Frobenius roots of exotic Kloosterman sheaves play the role of Satake parameters for finite-field generic representations; making this dictionary precise could transfer Casselman–Shalika formulas from unramified local representations to Bessel–Speh values.","Because the Hall–Littlewood formula expresses every exotic matrix Kloosterman sum through one $k$-element multiset of roots, the identity can be tested by comparing direct exponential-sum evaluations with sheaf-side traces for small $q$, which would also isolate where the imported gamma-factor equality enters.","A different proof or a strengthening of the imported gamma-factor equality would automatically upgrade the main theorem to other families of representations or other models, since no other step in Section 4 depends on the specific shape of the Speh representation."],"forward_implications":["Every Bessel–Speh value for an irreducible generic $\\tau$ is explicitly determined by exponential-sum data, not merely by abstract matrix coefficients.","Exotic matrix Kloosterman sums inherit two multiplicativity properties: factorization over block-diagonal matrices with disjoint spectra, and a unipotent-average identity.","The Hall–Littlewood expression turns these sums into quantities computable from the Frobenius characteristic polynomial, and Deligne's Weil bound gives absolute-value bounds for them.","New Bessel-function identities follow, including a generating function over Jordan blocks whose coefficients are elementary and complete homogeneous symmetric polynomials in the Frobenius roots.","The formula specializes to the earlier $c=1$ exotic Kloosterman formula and to the twisted matrix Kloosterman formula for principal series, so previously separate cases become one statement."],"supporting_citations":[{"why":"Defines the Ginzburg–Kaplan gamma factors and proves their equality with tensor-product epsilon factors; this is the input Theorem 4.6 invokes.","marker":"[5]"},{"why":"Proves the gamma-factor result for level-zero supercuspidal Speh representations that underlies the imported equality.","marker":"[44]"},{"why":"Gives the c=1 exotic Kloosterman formula whose matrix generalization is the paper's target.","marker":"[8]"},{"why":"Supplies the norm map from twisted conjugacy classes to conjugacy classes and the lift relation used to define exotic matrix sums.","marker":"[36]"},{"why":"Determines the cuspidal support of the lifts, enabling the reduction of non-abelian exotic Gauss sums to products.","marker":"[38]"},{"why":"Computes the classical non-abelian Gauss sum, the base case for the reduction.","marker":"[27]"},{"why":"Provides the characteristic maps and Hall–Littlewood machinery for the product formula.","marker":"[31]"},{"why":"Establishes the earlier matrix Kloosterman/Hall–Littlewood identity that Theorem 5.3 generalizes.","marker":"[46]"},{"why":"Defines tensor-product epsilon factors and their multiplicativity, used in Proposition 4.1.","marker":"[41]"}],"fun_headline_variants":["Bessel-Speh values equal exotic Kloosterman sums","Exotic Kloosterman sums capture Bessel-Speh values","Bessel-Speh to Kloosterman: exact identity","Special Bessel-Speh values are exotic sums"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the Ginzburg–Kaplan gamma factor equals the tensor-product epsilon factor for every irreducible $\\pi$ and every generic $\\tau$; this equality is imported from earlier work and is not reproved here, and Theorem 4.6 collapses if it fails.","fun_headline_variants_meta":{"raw":{"variants":["Bessel-Speh values equal exotic Kloosterman sums","Exotic Kloosterman sums capture Bessel-Speh values","Bessel-Speh to Kloosterman: exact identity","Special Bessel-Speh values are exotic sums"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000285,"raw_usage":{"total_tokens":1748,"prompt_tokens":1087,"completion_tokens":661,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":703,"completion_tokens_details":{"reasoning_tokens":591}},"tokens_in":703,"tokens_out":661,"duration_ms":6677,"temperature":1.0,"reasoning_tokens":591,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T19:06:09.847616+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Pick a small finite field, for instance $\\mathbb{F}_3$, take $k=2$ with $\\tau$ the cuspidal representation of $\\operatorname{GL}_2(\\mathbb{F}_3)$ attached to a regular character of $\\mathbb{F}_9^\\times$, choose $c=2$, and let $h\\in\\operatorname{GL}_2(\\mathbb{F}_3)$ be a regular elliptic matrix. Compute $K_{\\tau,\\psi}(h)$ directly from the character formula for Speh representations and compute the exotic matrix Kloosterman sum on the right-hand side of Theorem 4.6 by summation over $\\operatorname{GL}_2(\\mathbb{F}_9)$; any mismatch between the two sides would settle the theorem negatively.","supporting_citations":[{"cited_title":"Carmon and E","cited_arxiv_id":null,"evidence_quote":"Defines the Ginzburg–Kaplan gamma factors and proves their equality with tensor-product epsilon factors; this is the input Theorem 4.6 invokes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the c=1 exotic Kloosterman formula whose matrix generalization is the paper's target."},{"cited_title":"Shintani","cited_arxiv_id":null,"evidence_quote":"Supplies the norm map from twisted conjugacy classes to conjugacy classes and the lift relation used to define exotic matrix sums."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Determines the cuspidal support of the lifts, enabling the reduction of non-abelian exotic Gauss sums to products."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Computes the classical non-abelian Gauss sum, the base case for the reduction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the characteristic maps and Hall–Littlewood machinery for the product formula."},{"cited_title":"Zelingher","cited_arxiv_id":null,"evidence_quote":"Establishes the earlier matrix Kloosterman/Hall–Littlewood identity that Theorem 5.3 generalizes."},{"cited_title":"Ye and E","cited_arxiv_id":null,"evidence_quote":"Defines tensor-product epsilon factors and their multiplicativity, used in Proposition 4.1."}],"review_version":1}