{"id":"0fa21926-dab8-40bb-95d2-dca707ea5df7","arxiv_id":"2501.15801","paper_version":2,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":6,"one_line_summary":"A hybrid subconvex bound for L(1/2, Pi tensor chi) is claimed for prime level P and conductor M with M^{1/5} < P < M^{2/5}, but the final exponent passage in the proof is not justified.","lead":"This preprint claims a hybrid subconvexity bound for GL(4) times GL(1) twisted L-functions, in the prime level and conductor aspects, when the level lies between M^{1/5} and M^{2/5}. A reader might care because it would be the first such bound for a genuine GL(4) L-function, a problem that was previously open.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.1's fourth exponent is not derivable from (3.35): balancing Q^{-δ/2} with Q^δ M^{1/4}/P^{1+μ/2} gives (1-(4+2μ)θ)/(12(4+θ)), not (1-(4+2μ))θ/(12(4+θ)).","rationale":"The reader's REJECT is directionally right: Theorem 1.1 as displayed does not follow from the paper's own estimates. However, the most load-bearing single flaw is not the reader's labelled weakest assumption (the unverified GL(4) Voronoi normalization in Appendix A.1), but the final exponent conversion from (3.35) to (1.1). That conversion is where the advertised bound is defined, and the mismatch is provable from the paper's own equations: the term Q^δ M^{1/4}/P^{1+μ/2} balances to Q^{(1-(4+2μ)θ)/(12(4+θ))}, whereas Theorem 1.1 prints Q^{(1-(4+2μ))θ/(12(4+θ))}. The two differ by Q^{-(1-θ)/(12(4+θ))}, so the printed theorem is stronger than the proof supports. I do not see this as a fatal flaw of the whole method: the remaining terms such as P^{-1/4} and P^{5/8}M^{-1/4} still give genuinely negative exponents over the stated range, so a corrected version would likely retain hybrid subconvexity. For this reason I would move the verdict from outright REJECT to CONDITIONAL: the main theorem must be corrected and the parameter choices in §3.4 re-checked before acceptance. The Appendix's Voronoi correction is a serious unverified premise and should also be independently audited, but it is not where the current text is demonstrably false; hence partial agreement with the reader.","tokens_in":19956,"tokens_out":26494,"duration_ms":228784,"concrete_test":"Re-derive (1.1) from (3.35) at a concrete admissible pair, e.g. θ=1/4, μ=1/10, Q=PM^4. Compute the fourth term of (1.1): the displayed exponent is (1-(4+2μ))θ/(3(16+4θ)) = -(3+2μ)θ/(12(4+θ)) = -0.0157. Now minimize Q^{-δ/2} + Q^{δ + (1-(4+2μ)θ)/(4(4+θ))} over δ; the minimum occurs at δ = -(2/3)(1-(4+2μ)θ)/(4(4+θ)) and has value Q^{(1-(4+2μ)θ)/(12(4+θ))}, which for these parameters is Q^{-0.00098}. If an independent computation reproduces the claimed -0.0157, the concern is resolved; otherwise Theorem 1.1 must be corrected, e.g. by replacing the fourth term with Q^{(1-(4+2μ)θ)/(12(4+θ))} or omitting it.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The advertised main theorem is internally inconsistent with the paper's own final estimate. In (3.35) the relevant term is Q^δ M^{1/4}/P^{1+μ/2}. Writing Q=PM^4, P=Q^{θ/(4+θ)}, M=Q^{1/(4+θ)}, its log-Q exponent is δ + (1-(4+2μ)θ)/(4(4+θ)). Balancing this against the Q^{-δ/2} term from the approximate functional equation forces δ = -2b/3 with b = 1-(4+2μ)θ, so the strongest saving this term can produce is Q^{b/(12(4+θ))} = Q^{(1-(4+2μ)θ)/(12(4+θ))}. Theorem 1.1 instead states Q^{(1-(4+2μ))θ/(12(4+θ))} = Q^{-(3+2μ)θ/(12(4+θ))}. The difference is Q^{-(1-θ)/(12(4+θ))}: the displayed exponent is smaller, i.e. stronger, whenever 0<θ<1, so it is not a consequence of (3.35). For a concrete admissible pair, θ=1/4, μ=1/10 gives the claimed fourth-term exponent -0.0157 but the admissible balanced exponent -0.00098. This is a definite algebraic mismatch in the central statement, not a matter of convention or external bound. The qualitative hybrid subconvexity may survive because the P^{-1/4} and P^{5/8}M^{-1/4} terms already give negative exponents, so the error is repairable; nevertheless Theorem 1.1 as printed asserts a saving that the proof does not support.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies hybrid subconvexity for L(1/2, Π⊗χ), where Π is a normalized Hecke-Maass form on GL(4) of prime level P and χ is a primitive Dirichlet character modulo M. Theorem 1.1 claims a bound of the form Q^{1/4+ε} times a sum of four negative powers of Q = P M^4, for 0<μ<1/2 and M^{1/(4+2μ)} < P < M^{2/5}. Theorem 1.2 is a coefficient-sum estimate obtained by a Holowinsky-Nelson decomposition of χ applied twice, GL(4) Voronoi summation, Poisson summation, and Kloosterman-sum estimates, with parameters T, R, R*, S, μ optimized in §3.4. The appendix claims to correct a missing normalizing factor in Corbett's general Voronoi formula and sketches a proof of the prime-level GL(4) formula used throughout.","tokens_in":20330,"tokens_out":16461,"duration_ms":125986,"significance":"If the proof is completed and corrected, the paper would provide a hybrid subconvexity result for a GL(4)×GL(1) family, a setting where genuine level-aspect subconvexity for GL(4) is largely open. The parameter optimization is explicit and the paper gives a concrete route through standard tools, including a self-contained prime-level Voronoi formula. However, the advertised Theorem 1.1 contains a fourth exponent that is not a consequence of the displayed estimate (3.35); the qualitative subconvexity claim appears repairable by replacing that exponent with the correctly balanced one, since the other three terms already give negative exponents in the stated range, but the theorem as printed asserts a stronger saving than the proof supports.","major_comments":[{"comment":"The fourth exponent in Theorem 1.1 is not derivable from the final estimate (3.35). Writing P = Q^{θ/(4+θ)} and M = Q^{1/(4+θ)}, the term Q^δ M^{1/4}/P^{1+μ/2} in (3.35) has log-Q exponent δ + (1/4 - (1+μ/2)θ)/(4+θ). Balancing this against the Q^{-δ/2} term from the approximate functional equation gives the final exponent (1 - (4+2μ)θ)/(12(4+θ)). The exponent displayed in (1.1) is (1-(4+2μ))θ/(12(4+θ)), which is smaller by (1-θ)/(12(4+θ)) in log-Q scale and is therefore stronger than what the proof establishes whenever 0<θ<1. The theorem must be corrected to the balanced exponent, and the consequences for Corollary 1.3 and the claimed subconvex range should be rechecked in detail.","section":"§1, Eq. (1.1); §3.4, Eq. (3.35)"},{"comment":"Lemma 2.1 is the foundation for the transformations in (3.10), (3.17), and §3.3, but its proof in Appendix A.1 is only a sketch and depends on a claimed correction to Corbett's formula: the factor ∏_{i=2}^{n-2}|t_2⋯t_{n-1}|_p is asserted to be missing on p. 1392 of [6]. The appendix does not give a complete derivation of (4.1), and the subsequent manipulation introduces an undefined parameter L^4 before the change of variables r → rL^4. Because the whole paper collapses if this Voronoi formula is wrong, a complete, verifiable proof (or a precise reference establishing the prime-level formula) is a necessary condition for accepting the results.","section":"§4.1, Eq. (4.1) and Lemma 2.1"}],"minor_comments":[{"comment":"There are several typographical slips, including “whcih”, “Writer →rPι”, “Possion”, and inconsistent capitalization of P in phrases such as “1/p^{1/4−ε}”; these should be cleaned up.","section":"Throughout"},{"comment":"The condition stated before (3.7) is “R< T < M”, while later conditions involve R<S and S+T<R*<TR; the hierarchy of parameters should be stated consistently.","section":"§3.1, display before (3.7)"},{"comment":"The choice R = X/(M^{3/2}P^μ) makes R depend on X, while earlier R is introduced as a free parameter in the decomposition (3.2); the final estimate (3.35) should explicitly explain how the X-dependence cancels after inserting the range of X in Theorem 1.2.","section":"§3.4, Eq. (3.34)"},{"comment":"The appendix would benefit from a list of all substitutions used to pass from (4.2) to (4.3), since several steps are summarized with phrases such as “akin to [11, Eqn. (2.9)]” and are hard to verify.","section":"§4.1"}],"recommendation":"major_revision","confidential_remarks":"The fourth-exponent mismatch in Theorem 1.1 is a genuine algebraic error in the central statement, and the reader's report is right to flag it. However, the corrected exponent still appears to yield a subconvex bound in the same θ-range because the other three terms in (1.1) have negative exponents; hence the issue is repairable within the manuscript's scope. The more serious long-term concern is the completeness of the proof of the prime-level Voronoi formula in Appendix A.1, since the paper's main argument relies on it entirely. I recommend major revision rather than rejection, contingent on a correct restatement of Theorem 1.1 and a complete derivation of Lemma 2.1."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe headline is mixed but clear: this is a real attempt at a genuinely new result, yet the main theorem as printed is not proven. The final exponent in Theorem 1.1 is inconsistent with the paper's own estimate (3.35). Balancing Q^{-δ/2} against Q^δ M^{1/4}/P^{1+μ/2} and converting to Q gives a saving of Q^{(1-(4+2μ)θ)/(12(4+θ))}, not the claimed Q^{(1-(4+2μ))θ/(12(4+θ))}. For θ=1/4, μ=1/10, the two differ by a factor of Q^{-0.0147}, so the advertised bound is materially stronger than the proof supports. This is an internal algebraic mismatch, not a matter of convention.\n\nTo be fair, the qualitative hybrid subconvexity likely survives: the other terms in (3.35) already give negative Q-exponents in the stated range 1/(4+2μ)<θ<2/5, so a corrected optimization would probably still beat convexity. But the exact statement of Theorem 1.1 is unsupported as written, and the endgame needs to be redone.\n\nThe genuinely new content is real: this is the first hybrid level–conductor subconvexity bound for GL(4)×GL(1), and the method — two amplifiers plus two applications of the Holowinsky–Nelson decomposition — is a serious piece of work. The estimates leading up to (3.35) are carefully tracked and mostly standard. The appendix's claimed correction to Corbett's Voronoi normalization is potentially useful, but it is only sketched and it is load-bearing: if the missing factor ∏|t_2...t_{n-1}|_p is wrong, the transformations (3.10)–(3.17) collapse. That needs a complete, readable proof before publication, not a sketch.\n\nThe paper is honest about its limitations: it flags the degenerate-term range loss and the dependence on Langlands parameters. The citation pattern looks fine and covers the relevant subconvexity literature.\n\nWho is this for? Analytic number theorists working on higher-rank subconvexity. They will find the method instructive even with the error. It deserves a serious referee: the mistake is localized, the main idea is sound enough to merit a revision, and the potential correction to a published Voronoi formula is itself worth scrutiny. A referee should insist on a corrected exponent and a complete proof of Lemma 2.1.","headline":"A serious, genuinely new subconvexity attempt for GL(4)×GL(1), but Theorem 1.1's last exponent is algebraically unsupported and the paper needs a corrected endgame.","tokens_in":20965,"tokens_out":3906,"would_cite":false,"duration_ms":30810,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11F67","11F66","11L05"],"pacs":[],"model":"deepseek-v4-flash","headline":"For prime level $P$ and conductor $M$, the paper proves a hybrid subconvex bound for $L(1/2,\\Pi\\otimes\\chi)$ that beats the convexity bound $Q^{1/4+\\varepsilon}$ whenever $M^{1/(4+2\\mu)}<P<M^{2/5}$ for some $0<\\mu<1/2$.","keywords":["subconvexity","hybrid bounds","twisted L-functions","GL(4) Hecke-Maass forms","Voronoi summation","Kloosterman sums","amplification method","Dirichlet characters"],"falsifier":"Compute both sides of Lemma 2.1 numerically for a concrete Hecke-Maass form of a small prime level $P$, a single modulus $c$ coprime to $aP$, and a compactly supported test function $\\omega$; a mismatch of the size of the factor (4.1) would refute the formula. A cheaper check: for $n=4$ the corrected change of variables must produce the extra factor $|t_2t_3|_p$ in (4.1), so a direct $p$-adic computation of the relevant Kloosterman integral will confirm or deny the paper's correction.","tokens_in":19687,"feed_emoji":"🔢","tokens_out":12463,"duration_ms":103528,"temperature":0.7,"pith_summary":"The paper proves a hybrid subconvexity bound for $\\mathrm{GL}(4)\\times\\mathrm{GL}(1)$ twisted $L$-functions: for a normalized Hecke-Maass form $\\Pi$ of prime level $P$ and a primitive Dirichlet character $\\chi$ modulo $M$, the central value $L(1/2,\\Pi\\otimes\\chi)$ is bounded by $Q^{1/4+\\varepsilon}$ times a sum of negative powers of $Q=PM^4$, provided $M^{1/(4+2\\mu)}<P<M^{2/5}$ for some $0<\\mu<1/2$. Since $Q^{1/4+\\varepsilon}$ is the convexity bound, the estimate gives a genuine power saving simultaneously in the level aspect, through $P$, and in the conductor aspect, through $M$. This extends the hybrid subconvexity phenomenon, previously known for Dirichlet and for $\\mathrm{GL}(2)\\times\\mathrm{GL}(1)$ $L$-functions, to a genuine $\\mathrm{GL}(4)$ $L$-function.","feed_headline":"A new bound beats convexity for GL(4) twisted L-functions","feed_subtitle":"For M^{1/5}<P<M^{2/5}, the central value saves a power of Q=PM^4, with an explicit 1/60 saving in a corollary.","key_machinery":"The machinery that carries the proof is the level-$P$ $\\mathrm{GL}(4)$ Voronoi summation formula of Lemma 2.1, applied after the character-decomposition identity (3.2) splits $\\chi(n)$ into a main term and a dual term. Lemma 2.1 rewrites a sum of Fourier coefficients $A_\\Pi(n,1,1)$ weighted by $e(an/c)$ as a sum over divisors $d_1\\mid c$, $d_2\\mid c/d_1$ of hyper-Kloosterman sums $KL_2$ and coefficients $A_\\Pi(m,d_2,d_1)$, with an oscillatory integral transform $\\psi_\\pm(x;\\omega)$ whose main term behaves like $x^{5/8}e(4x^{1/4})$. The appendix derives Lemma 2.1 from the general $\\mathrm{GL}(n)$ Voronoi formula of [6], inserting a corrected normalizing factor in (4.1): a product over $|t_2\\cdots t_{n-1}|_p$ that was missing on page 1392 of [6]. The proof then feeds the Voronoi-transformed sums into Poisson summation and bounds the resulting Kloosterman correlations with the estimate (3.21), after two rounds of amplification by well-chosen prime sets $t$ and $s$.","core_discovery":"The central assertion is Theorem 1.1: for primes $P,M$ with $(P,M)=1$, a normalized Hecke-Maass form $\\Pi$ of level $P$ and trivial nebentypus, and a primitive Dirichlet character $\\chi$ modulo $M$, the central value satisfies $$L(1/2,\\Pi\\otimes\\chi)\\ll $Q^{{1/4+\\varepsilon}}$\\bigl($Q^{{-(2-5\\theta)/(3(32+8\\theta))}}$+$Q^{{-\\theta/(16+4\\theta)}}$+$Q^{{-(1-2\\mu)\\theta/(8+2\\theta)}}$+$Q^{{-(3+2\\mu)\\theta/(3(16+4\\theta))}}$\\bigr)$$ for any $0<\\mu<1/2$, where $\\theta=\\log P/\\log M$ and $Q=PM^4$ is the analytic conductor. Whenever $1/(4+2\\mu)<\\theta<2/5$, every exponent in the parentheses is negative, so the right-hand side is a genuine power saving over the convexity bound $Q^{1/4+\\varepsilon}$. Theorem 1.1 is deduced from Theorem 1.2, which establishes cancellation in the smoothed coefficient sum $S_\\Pi(X,M,P)=\\sum_n A_\\Pi(n,1,1)\\chi(n)n^{-1/2}V(n/X)$ for $X$ near $Q^{1/2}$.","pith_inferences":["The correction in (4.1) is a standalone fact about the general $\\mathrm{GL}(n)$ Voronoi formula: users of the level-modulus formula of [6] in other problems should check whether their normalizations inherit the missing product over $|t_2\\cdots t_{n-1}|_p$.","The restriction $\\theta>1/(4+2\\mu)$ appears to be a technical artifact of the direct bound on the degenerate term; the paper's own remark that a fuller §3.1-style treatment would extend the range to $\\varepsilon<\\theta<2/5$ suggests the barrier is not structural.","A concrete test of the mechanism would be to average the same twisted coefficient sum over characters $\\chi$ or over the level-$P$ family; if the cancellation is as strong as the proof indicates, such averages should recover or improve the exponent $1/4$ outside the parameter range covered here."],"forward_implications":["For any $0<\\mu<1/2$ and $M^{1/(4+2\\mu)}<P<M^{2/5}$, the four negative exponents in (1.1) give a saving over the convexity bound $Q^{1/4+\\varepsilon}$ simultaneously in the level and conductor aspects.","With $P\\asymp M^{2/7+\\varepsilon}$, Corollary 1.3 supplies a family of level-$P$ Hecke-Maass forms for which $L(1/2,\\Pi\\otimes\\chi)\\ll_{\\Pi,\\varepsilon}Q^{1/4-1/60+\\varepsilon}$.","If the degenerate term is treated by repeating the full argument of §3.1 rather than the direct estimate of §3.3, the same method covers the wider range $\\varepsilon<\\theta<2/5$.","The estimate is stated for every normalized Hecke-Maass form of the given level, with implied constants depending only on $\\varepsilon$ and the archimedean parameters of $\\Pi$, so the subconvexity statement is uniform across the level-$P$ family in that sense."],"supporting_citations":[{"why":"Supplies the general $\\mathrm{GL}(n)$ Voronoi formula whose normalizing factor Lemma 2.1 corrects.","marker":"[6]"},{"why":"Provides the character decomposition identity (3.2) used twice in the proof.","marker":"[10]"},{"why":"Gives the explicit Voronoi formula template for newforms that the appendix adapts to $\\mathrm{GL}(4)$.","marker":"[11]"},{"why":"Supplies the Fourier-coefficient second-moment bound used after Cauchy-Schwarz.","marker":"[5]"},{"why":"Provides the Kloosterman-sum correlation estimate (3.21) used to control diagonal contributions.","marker":"[13]"},{"why":"Gives the base $\\mathrm{GL}(n,\\mathbb{Z})$ Voronoi formula that the level-aspect version extends.","marker":"[16]"},{"why":"Contributes the decomposition idea that produces the non-degenerate term in the splitting of $\\chi$.","marker":"[18]"}],"fun_headline_variants":["Beating convexity for GL(4) twisted L-functions","Hybrid subconvexity for GL(4) with Dirichlet twist","Power saving for GL(4) L-functions in level and conductor","Hybrid saving over convexity for GL(4) L-functions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that Lemma 2.1, including the corrected normalizing factor (4.1) inserted into the $\\mathrm{GL}(n)$ formula of [6], is exactly right; if the appended factor were wrong, the Voronoi transformations in (3.10), (3.17), and §3.3 would collapse.","fun_headline_variants_meta":{"raw":{"variants":["Beating convexity for GL(4) twisted L-functions","Hybrid subconvexity for GL(4) with Dirichlet twist","Power saving for GL(4) L-functions in level and conductor","Hybrid saving over convexity for GL(4) L-functions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000695,"raw_usage":{"total_tokens":3405,"prompt_tokens":968,"completion_tokens":2437,"prompt_tokens_details":{"cached_tokens":896},"prompt_cache_hit_tokens":896,"prompt_cache_miss_tokens":72,"completion_tokens_details":{"reasoning_tokens":2358}},"tokens_in":72,"tokens_out":2437,"duration_ms":44761,"temperature":1.0,"reasoning_tokens":2358,"cache_read_input_tokens":896,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T14:00:37.986654+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute both sides of Lemma 2.1 numerically for a concrete Hecke-Maass form of a small prime level $P$, a single modulus $c$ coprime to $aP$, and a compactly supported test function $\\omega$; a mismatch of the size of the factor (4.1) would refute the formula. A cheaper check: for $n=4$ the corrected change of variables must produce the extra factor $|t_2t_3|_p$ in (4.1), so a direct $p$-adic computation of the relevant Kloosterman integral will confirm or deny the paper's correction.","supporting_citations":[{"cited_title":"Corbett, Vorono ˘ ı summation forGLn: collusion between level and modulus , Amer","cited_arxiv_id":null,"evidence_quote":"Supplies the general $\\mathrm{GL}(n)$ Voronoi formula whose normalizing factor Lemma 2.1 corrects."},{"cited_title":"Holowinsky and P","cited_arxiv_id":null,"evidence_quote":"Provides the character decomposition identity (3.2) used twice in the proof."},{"cited_title":"Hou and and G","cited_arxiv_id":null,"evidence_quote":"Gives the explicit Voronoi formula template for newforms that the appendix adapts to $\\mathrm{GL}(4)$."},{"cited_title":"Chandee and X","cited_arxiv_id":null,"evidence_quote":"Supplies the Fourier-coefficient second-moment bound used after Cauchy-Schwarz."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the base $\\mathrm{GL}(n,\\mathbb{Z})$ Voronoi formula that the level-aspect version extends."},{"cited_title":"Munshi, Twists of GL(3) L-functions, In: M¨ uller, W., Shin, S.W., Tem- plier, N","cited_arxiv_id":null,"evidence_quote":"Contributes the decomposition idea that produces the non-degenerate term in the splitting of $\\chi$."}],"review_version":1}