{"id":"e27f1c26-d51c-4fb4-a7fb-b8d634403ea6","arxiv_id":"2507.06963","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Matching of relative orbital integrals on PGL(3) and metaplectic orbital integrals on the triple cover of SL(3) is established for the full spherical Hecke algebras.","lead":"This paper proves a new 'Fundamental Lemma' for the cubic Shimura lift, matching orbital integrals on two different groups, PGL(3) and a cubic cover of SL(3), for every function in the full spherical Hecke algebras. The result is a key local step toward proving a global Shimura lift from the triple cover of SL(3) to PGL(3), a higher-rank case that has been open.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 5.11 as written does not prove the coefficient list it states: its Case 7 repeats the indices of Case 5 with a different value, and the required sixth coefficient (2m+n−2, m+2n−2)=q^2\\bar{g} is never computed.","rationale":"I read the paper as an explicit computational proof of a new Fundamental Lemma: Theorem 3.1 asserts an algebra isomorphism and a matching of orbital integrals, and the argument reduces everything to the evaluation tables in Propositions 6.17, 7.9, 8.17, and 9.6, assembled in Section 10. The most load-bearing component is therefore correct evaluation of the expansions and the orbital integrals, not any single number-theoretic input. The proof of Proposition 5.11 is the place where the expansion coefficients for the metaplectic side are established, and it contains a clear textual problem: Case 7 repeats the indices of Case 5 but attributes a different value, while the sixth nonzero coefficient required by the proposition is not computed at all. This is more concrete than the reader's weakest-assumption concern about the cited identity (21), which I do not see contradicted internally and which is supported by the authors' prior work. The paper has real independent support: the basis construction and algebra isomorphism are standard, the transfer factors are explicit and checkable, and the final comparison is a direct table check. However, the unverified coefficient in a load-bearing expansion is enough to keep the paper conditional pending a computation rather than fully verified. The proposed test settles the issue directly: recompute the one missing integral and, if needed, re-run the affected table entries in Proposition 8.17.","tokens_in":82394,"tokens_out":11038,"duration_ms":114637,"concrete_test":"Recompute Ψ'_{m,n}(2m+n−2, m+2n−2) directly from Lemma 5.12 and the two subdomains of Lemma 5.5(7), using the same cocycle evaluations as in the surrounding cases of Proposition 5.11. If the result is not q^2\\bar{g}, replace the sixth line of Proposition 5.11, re-derive Corollary 5.13, and recheck the affected entries of Proposition 8.17 to determine whether the generic matching still holds.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The big-cell matching in Theorem 3.1 depends on the Whittaker-coefficient expansion Corollary 5.13, which is built directly from Proposition 5.11. In the proof of Proposition 5.11, after Case 6 the text reads 'Case 7: i=2m+n−2, j=m+2n−1' — the same pair already treated in Case 5, where the value q^2 was obtained. That repeated Case 7 concludes q^2 g instead. This is an internal inconsistency in the proof as written. The index pair that Lemma 5.5 actually assigns to the two-subdomain case used in Case 7 is (2m+n−2, m+2n−2), i.e. the sixth nonzero coefficient in the proposition, which should be q^2\\bar{g}; that computation is absent. Since Corollary 5.13 enters Proposition 8.17 with a coefficient q^2\\bar{g} attached to J^l(2m+n−2,m+2n−2), an error in this coefficient would propagate through the big-cell orbital-integral table and break the comparison with Proposition 6.17 in Section 10. This is a concrete, localizable gap, not merely a matter of relying on the cited identity (21).","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper establishes a Fundamental Lemma for Hecke correspondences between the spherical Hecke algebra of PGL_3(F) and the spherical Hecke algebra of anti-genuine functions on the cubic metaplectic cover of SL_3(F), for a local field F containing cube roots of unity and residual characteristic greater than 3. The matching is stated in Theorem 3.1 for all elements of the spherical Hecke algebras, extending the authors' prior unit-element result. The method is fully computational: the authors expand each orbital integral in terms of translates of the unit element via Iwasawa decompositions, compute the Whittaker coefficients of the Hecke basis elements, then evaluate the resulting relative and metaplectic orbital integrals by decomposing the domains of integration and using relations between cubic exponential sums and Kloosterman sums. The paper concludes by comparing the resulting tables for the generic, one-parameter, and isolated orbits.","tokens_in":82587,"tokens_out":6762,"duration_ms":71135,"significance":"If correct, this is a substantial result: it provides the first full spherical-Hecke-algebra Fundamental Lemma in a relative rank two setting, and it supplies a key local input for a relative trace formula approach to the conjectural cubic Shimura lift to PGL(3). The authors are explicit that the central computation is new and that the unit-element case from [FO] is only the base case. The paper also gives concrete, falsifiable identities: the orbital-integral evaluations in Propositions 6.17 and 8.17 are stated with explicit coefficients and can in principle be checked case by case. However, the proof as written contains an internal inconsistency in the computation of the Whittaker coefficients on the metaplectic side, and that computation is load-bearing for the big-cell matching. Because that gap is local and potentially repairable, the appropriate disposition is major revision rather than rejection.","major_comments":[{"comment":"The proof of Proposition 5.11 is internally inconsistent and omits a required coefficient. In the proof, Case 7 is labeled 'i=2m+n−2, j=m+2n−1', the same index pair already treated in Case 5 (and listed as item (4) of the proposition), yet it concludes q^2 g instead of the q^2 obtained in Case 5. The index pair (2m+n−2, m+2n−2), which Lemma 5.5 assigns to the two-subdomain case and which appears as item (6) of Proposition 5.11 with value q^2\\bar{g}, is never computed. This is not a cosmetic issue: Corollary 5.13 contains the term q^2\\bar{g} J^l(x;2m+n−2,m+2n−2), and that term is used in the derivation of Proposition 8.17 in Section 8.8. The proof of Theorem 3.1 therefore rests on an unproved and, as written, contradicted coefficient evaluation. The authors should correct the case labeling and supply the missing computation for (2m+n−2, m+2n−2), or otherwise justify the value q^2\\bar{g}.","section":"5.2.3"},{"comment":"Proposition 5.6 is load-bearing for the entire relative side: Corollary 5.7, and hence all formulas in Sections 6 and 7, depend on the listed Whittaker coefficients. The proof ends with 'The evaluation of this integral is straightforward using Lemma 5.5 and is omitted here.' In view of the inconsistency found in the parallel coefficient computation of Proposition 5.11, this omission is no longer acceptable as a routine detail. Please provide the complete evaluation, or at least the integrals over each Γ_{µ,λ} listed in Lemma 5.5, so that the coefficient list in Proposition 5.6 can be verified directly.","section":"5.1.4"}],"minor_comments":[{"comment":"The displayed vector for ξ[3] appears to have nine coordinates, while ξ is an element of F^8. Please check the entries and correct the typo.","section":"2.6"},{"comment":"In the last two displayed cases, '|b_{n+2n}|' should presumably read '|b_{m+2n}|'. The same subscript error appears in the corresponding line of Proposition 6.17 or its surrounding text.","section":"6.7.4"},{"comment":"Identity (21) is cited from [FO, Corollary 7.5] and then used at many points in Section 8 to rewrite Kloosterman sums as cubic exponential sums. Since this identity is a major input to the proof of Proposition 8.2 and Proposition 8.17, the paper would be easier to certify if (21) were stated as a numbered theorem with a proof sketch or with a precise reference to the exact statement in [FO].","section":"4.5"}],"recommendation":"major_revision","confidential_remarks":"The paper is a long and dense computational proof, and the main obstacle is the specific missing and contradictory Whittaker-coefficient computation in Proposition 5.11. If that computation is supplied and the consequent cases in Proposition 8.17 are rechecked, the result is likely sound. The editors may also wish to ask the authors to make the proof of Proposition 5.6 fully explicit."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a serious paper with a real result in its sights, but the proof as written has a hole in a load-bearing spot. The genuinely new content is the extension of the authors' earlier unit-element matching to the full spherical Hecke algebras for the cubic Shimura lift to PGL(3). They compute Whittaker expansions, evaluate the orbital integrals on both sides, and the final tables in Propositions 6.17 and 8.17 appear to match. That is a substantial, honest computational achievement.\n\nThe soft spot is Proposition 5.11. The stated coefficient list has six nonzero Whittaker coefficients, but in the proof Case 7 repeats the index pair of Case 5, (2m+n–2, m+2n–1), and assigns it a different value, q^2 g, instead of the q^2 already obtained. The sixth listed coefficient, (2m+n–2, m+2n–2) = q^2 \\bar{g}, is never computed under the written indices. If that pair was meant to be the index in Case 7, the computation still ends with q^2 g, not q^2 \\bar{g}. Either way, Corollary 5.13 is built on a coefficient that is either missing or contradictory. Since Corollary 5.13 feeds directly into Proposition 8.17, and the final matching in Section 10 relies on Proposition 8.17 matching Proposition 6.17, Theorem 3.1 is currently unsupported at that point.\n\nThe rest of the computation looks careful, and the paper is transparent about what is omitted, such as the Whittaker evaluation in Proposition 5.6 and some cases in Corollary 6.12. The dependency on identity (21) from the authors' prior work is real, but it is a cited result and not by itself a flaw; it just needs checking in review.\n\nBottom line: this paper deserves a serious referee and a request for revision. The gap is concrete and local, likely fixable by correcting a sign or an index, but it must be fixed and the propagation through Corollary 5.13, Proposition 8.17, and the final comparison must be verified before the result is trusted.","headline":"Impressive computational scope, but Proposition 5.11 as written does not prove its coefficient list, and the big-cell matching hangs on that proposition.","tokens_in":83156,"tokens_out":6859,"would_cite":false,"duration_ms":61256,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11F70","11F25","11F27","11F67","11F72","22E50","22E55"],"pacs":[],"model":"deepseek-v4-flash","headline":"Over a local field containing cube roots of unity, this paper proves that the spherical Hecke algebra of $\\mathrm{PGL}_3(F)$ is isomorphic to that of anti-genuine functions on the cubic cover of $\\mathrm{SL}_3(F)$, and that all relevant…","keywords":["Shimura correspondence","cubic metaplectic cover","relative trace formula","Hecke algebra isomorphism","Fundamental Lemma","orbital integrals","minimal representation","Kloosterman sums"],"falsifier":"Take $F=\\mathbb{Q}_7$, whose residual characteristic is 7 and which contains cube roots of unity, fix the standard additive character and cubic Hilbert symbol, and compute both sides of identity (21) for $|a|=|c|=|d|=q^2$ with any $t$ whose valuation is not divisible by 3; a single unequal pair would falsify the key identity and, through Propositions 8.17 and 6.17, the Fundamental Lemma.","tokens_in":82128,"feed_emoji":"⚖️","tokens_out":10241,"duration_ms":113058,"temperature":0.7,"pith_summary":"Over a nonarchimedean local field $F$ that contains cube roots of unity and has residual characteristic greater than 3, this paper proves a local distribution matching, a Fundamental Lemma, for the cubic Shimura correspondence between $\\mathrm{PGL}(3)$ and the triple cover of $\\mathrm{SL}(3)$. It constructs an algebra isomorphism between the spherical Hecke algebra of $\\mathrm{PGL}_3(F)$ and the spherical Hecke algebra of anti-genuine functions on the cubic cover of $\\mathrm{SL}_3(F)$, and shows that the relative orbital integrals attached to the $\\mathrm{SO}(8)$ minimal-representation period match the metaplectic Kuznetsov orbital integrals on the covering group, for every Hecke function and with explicit transfer factors. This extends the earlier matching for the unit element of the Hecke algebras to the full spherical Hecke algebra. The result is the geometric core of a planned relative trace formula comparison that would yield a new global Shimura lift from genuine automorphic forms on the triple cover of $\\mathrm{SL}(3)$ to automorphic forms on $\\mathrm{PGL}(3)$, and would characterize the image of that lift by a period.","feed_headline":"Cubic Shimura lift: Hecke matching proved at all levels","feed_subtitle":"A local distribution matching over the full spherical Hecke algebra opens the path to a relative trace formula and a global lift.","key_machinery":"The argument rests on two mechanisms. First, the Iwasawa decomposition expands each orbital integral as a finite linear combination of orbital integrals of unit-element translates, with coefficients equal to Whittaker coefficients; those coefficients are evaluated via Cartan decompositions, and on the cover this requires the block-compatible cocycle and the splitting of the maximal compact subgroup. Second, identity (21), relating the cubic Kloosterman sum $K(t;c,d)$ to the cubic exponential sum $C(-3a,c^{-1}d^{-1}a^3;0)$, converts all metaplectic big-cell contributions into the same cubic exponential sums that appear on the relative side. The final comparison is a case-by-case identification of the 24 nonvanishing formulas.","core_discovery":"The central claim is Theorem 3.1: for every $f$ in the spherical Hecke algebra of $\\mathrm{PGL}_3(F)$, the relative orbital integral $O(\\xi,\\omega(f)\\phi_0)$ at each relevant orbit equals, up to an explicit transfer factor, the metaplectic orbital integral $O'(x,\\mathrm{Sh}(f))$ on the cubic cover $G'$ of $\\mathrm{SL}_3(F)$. The map $\\mathrm{Sh}$ sends a basis element $f_{m,n}$ to $q^{-2(m+n)}f'_{m,n}$ and is an algebra isomorphism. The matching covers the generic family of orbits, two one-parameter families, and three isolated orbits; the transfer factors involve the cubic Hilbert symbol and additive characters evaluated at the parameters. The proof is a direct computation: both sides are expanded into translates of the unit element, all resulting integrals are evaluated in closed form, and 24 nonzero cases on each side are compared.","pith_inferences":["Because the transfer factors multiply to 1 over all places, the local matching is globally compatible; the full global relative trace formula still requires matching at ramified places, which this paper does not address.","The one-parameter and isolated orbit matchings could likely be derived from the generic orbit by studying asymptotics of relative orbital integrals, but doing so would require a theory of relative Shalika germs that is not developed here.","The heavy dependence on identity (21) suggests that the method will not automatically extend to higher-degree covers, where the analogous relation between higher Kloosterman sums and exponential sums is not known."],"forward_implications":["For every basis element $f_{m,n}$, both classes of orbital integrals are evaluated in closed form, and the evaluations coincide under the map $f_{m,n}\\mapsto q^{-2(m+n)}f'_{m,n}$.","The matching therefore holds for all functions in the spherical Hecke algebras, not just the unit elements matched in the authors' previous work.","This supplies the geometric comparison needed for a relative trace formula whose spectral side would realize a global Shimura lift from genuine automorphic forms on the triple cover of $\\mathrm{SL}_3$ to automorphic forms on $\\mathrm{PGL}_3$.","The same relative trace formula would show that the image of the lift is detected by nonvanishing of the period built from the minimal representation of $\\mathrm{SO}_8$.","The local transfer factors have product $1$ over all places, so the local theorem can be inserted consistently into global comparisons."],"supporting_citations":[{"why":"Supplies the matching for unit elements, the metaplectic splitting lemma, and the Kloosterman–exponential identity (21) at arbitrary level that the full-Hecke computation relies on.","marker":"[FO]"},{"why":"Gives the unramified local map, taking powers on Satake parameters, that the algebra isomorphism Sh extends linearly.","marker":"[KP86]"},{"why":"Provides the metaplectic Cartan decomposition and the computation of Whittaker coefficients used to evaluate the coefficients in the expansion on the cubic cover.","marker":"[BF99]"},{"why":"Proves the first-level case of the Kloosterman–cubic-exponential identity used as a base for the arbitrary-level version.","marker":"[DI93]"},{"why":"Conjectures the SO8 minimal-representation period that detects the cubic Shimura lift, framing why the relative orbital integrals are the right distributions.","marker":"[BFG01]"},{"why":"Establishes the minimal representation of SO8 via a theta lift, used to set up the relative orbital integrals on PGL3.","marker":"[KR94]"}],"fun_headline_variants":["Hecke matching at all levels for cubic lift","New fundamental lemma for Shimura lift","Cubic cover Hecke isomorphism matches distributions","Full distribution matching for PGL(3) lift","Shimura lift: all Hecke elements matched"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole comparison depends on identity (21), which equates a cubic Kloosterman sum to a cubic exponential sum when $|a|=|c|=|d|>q$; if that identity were false for some level or character, the big-cell orbital integrals would not match and Theorem 3.1 would collapse.","fun_headline_variants_meta":{"raw":{"variants":["Hecke matching at all levels for cubic lift","New fundamental lemma for Shimura lift","Cubic cover Hecke isomorphism matches distributions","Full distribution matching for PGL(3) lift","Shimura lift: all Hecke elements matched"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000279,"raw_usage":{"total_tokens":1664,"prompt_tokens":957,"completion_tokens":707,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":573,"completion_tokens_details":{"reasoning_tokens":638}},"tokens_in":573,"tokens_out":707,"duration_ms":7981,"temperature":1.0,"reasoning_tokens":638,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:50:36.249592+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take $F=\\mathbb{Q}_7$, whose residual characteristic is 7 and which contains cube roots of unity, fix the standard additive character and cubic Hilbert symbol, and compute both sides of identity (21) for $|a|=|c|=|d|=q^2$ with any $t$ whose valuation is not divisible by 3; a single unequal pair would falsify the key identity and, through Propositions 8.17 and 6.17, the Fundamental Lemma.","supporting_citations":[],"review_version":1}