{"id":"ab5a1359-9052-47f2-836e-badf9445cd28","arxiv_id":"2501.10971","paper_version":1,"verdict":"REJECT","confidence":"LOW","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper claims the average fourth moment of holomorphic Hecke cusp forms of weight in [K, K+H], with H = K^(3/4+c), equals 6/pi with a K^(-delta) error, improving Khan's full-length interval result.","lead":"An analytic number theory paper claims to compute the average fourth moment of holomorphic Hecke cusp forms over a short window of weights of length K^(3/4+c), improving the previously known full-interval result. The value matches the random wave prediction from quantum chaos, but the decisive estimate is cited from earlier work rather than proved in this setting.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 5.8 is the load-bearing step: the bound E3 << K^{-1/4+epsilon} K/H is imported from Khan [9, Lemma 3.5] without verifying that the short-interval weight w'(k/K) supported on length H/K satisfies the hypotheses; Remark 5.9 states the Poisson-summed weight has very poor properties.","rationale":"Reading in good faith, the paper is a serious attempt to extend Khan's average fourth-moment result to shorter weight intervals, and the main-term computation through Watson's formula is standard and plausible. The decisive issue is not a disagreement with prior consensus but an internal gap: the claimed improvement in interval length is controlled entirely by Lemma 5.8 (Remark 1.3), and the proof of that lemma is a citation to a full-interval estimate whose hypotheses are not checked against the new support scale. The new weight w'(k/K) is supported on an interval of length H/K, which is K^{-1/4+c} and tends to 0; its derivatives are not O(1), so the stationary-phase and Poisson-summation estimates used in Khan's full-interval setting cannot be invoked without re-derivation. Remark 5.9 confirms this by stating that the transformed function has very poor index properties and that the standard stationary-phase route is difficult. The label confusion in (5.7)/(5.10) and the mislabeled Lemma 5.4 make independent checking harder, but the core problem is the missing derivation of (5.31). If a complete proof of Lemma 5.8 for the new weight were supplied, the theorem would likely be valid; as written, it should be rejected or at most conditionally accepted pending that proof. My recommendation is therefore REJECT, matching the reader's low-confidence rejection.","tokens_in":12382,"tokens_out":8461,"duration_ms":96479,"concrete_test":"Take the statement of [9, Lemma 3.5] and replace its weight h by w' supported on (1, 1+K^{-1/4+c}); re-derive the estimate line by line, keeping track of derivatives of w' and of the dual summation after Poisson summation. In particular, compute sup_j K^{-j} |w'^{(j)}| (or the analogous (K/H)^j-scaled derivative) and check whether the stationary phase derivative at the critical point is still K^epsilon-times larger than the dual summation step. If the resulting bound is only K^{-1/4+epsilon} without the factor K/H, or if it contains a positive power of K/H, then (5.31) fails for H=K^{3/4+c} and Theorem 1.2 does not follow. If the re-derivation yields exactly (5.31), the concern is resolved.","verdict_should_be":"REJECT","load_bearing_attack":"The central claim is Theorem 1.2, and the interval length H=K^{3/4+c} is limited exactly by Lemma 5.8 (Remark 1.3). The proof of Lemma 5.8 consists, after defining w'(k/K) with support (1,(K+H)/K), of the single line 'By [9, Lemma 3.5], we have E3 << K^{-1/4+epsilon} * K/H << K^{-delta}.' This is not a proof of the short-interval transfer. Khan's Lemma 3.5 is proved for a weight of the form h(k/K) supported on (1,2) with derivatives O(1); the new weight w'(k/K) is supported on an interval of length H/K = K^{-1/4+c}, so its derivatives are of size (K/H)^j = K^{j(1/4-c)}. Oscillatory Bessel-sum estimates are highly sensitive to such derivative growth, and the stationary-phase or Poisson-summation argument behind [9, Lemma 3.5] must be re-run with the new scale before the bound (5.31) is available. Remark 5.9 explicitly states that with this smooth function the index properties after Poisson summation are 'very poor', making a power saving by stationary phase difficult. Since the only displayed justification is the cited lemma, Theorem 1.2 is not established as written: if E3 cannot be bounded by a negative power of K, no improvement over Khan's H approximately K result follows.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper claims Theorem 1.2: for fixed 0<c≤1/4 and H=K^{3/4+c}, with the normalization ||F||_2=1, the smoothed average over even weights k∈[K,K+H] of (2/(HW))Σ_k w((k-K)/H)(12/k)Σ_{f∈B_k}||F||_4^4 equals 6/π+O(K^{-δ}) for some 0<δ≤c. The route is standard: Watson's formula reduces the fourth moment to averaged central values of L(1/2,g)L(1/2,sym^2 f×g); the Petersson trace formula produces a diagonal term plus off-diagonal terms; the diagonal gives the constant 6/π. The new ingredient is to run this argument over a short weight interval of length H=K^{3/4+c}, shorter than Khan's H≈K. The paper splits the main off-diagonal term into pieces and bounds them in Lemmas 5.2, 5.4, 5.6 and 5.8. The decisive piece is Lemma 5.8, whose proof is a one-line appeal to Khan's Lemma 3.5 after changing the weight to w'(k/K) supported on (1,(K+H)/K).","tokens_in":12646,"tokens_out":9934,"duration_ms":105601,"significance":"If Theorem 1.2 is correct, it is a genuine improvement: it would confirm the random-wave prediction on average over weight windows of length K^{3/4+c}, shrinking Khan's previous K-length window by a power of K. The main term is parameter-free and emerges cleanly from Watson's formula and the diagonal of the Petersson trace formula, with no fitted constants; the paper is also refreshingly honest in Remarks 1.3 and 5.9, which identify Lemma 5.8 as the exact source of the admissible range of H. However, the main term and the E1/E2 bounds do not by themselves deliver the theorem; the entire shorter-interval gain rests on the E3 bound in Lemma 5.8, and that bound is not proved at the new scale. The paper's significance is therefore conditional on a missing technical argument.","major_comments":[{"comment":"The proof of Lemma 5.8 consists of defining w' supported on (1,(K+H)/K) and then asserting E3 << K^{-1/4+ε} K/H << K^{-δ} 'By [9, Lemma 3.5]'. This does not verify that Khan's Lemma 3.5 applies to the new weight. Khan's lemma is proved for a fixed-scale smooth weight h(k/K) supported on (1,2) with bounded derivatives, whereas w'(k/K) is supported on an interval of length H/K = K^{-1/4+c} and has derivatives of size K^{j(1/4-c)}. Oscillatory Bessel-sum estimates are sensitive to exactly this derivative growth, and the stationary-phase or Poisson-summation method behind [9, Lemma 3.5] must be rerun at the new scale before (5.31) is available. Remark 5.9 explicitly concedes that for this weight the index properties after Poisson summation are 'very poor' and power saving is difficult. Since Remark 1.3 states that the range of H in Theorem 1.2 is determined exactly by Lemma 5.8, this is the load-bearing step: without a proof of (5.31), Theorem 1.2 does not improve on Khan's H≈K result. This is not a routine citation; it is a missing proof.","section":"§5, Lemma 5.8 and Remark 5.9"},{"comment":"The proof of the first part of (5.14) cites Lemma 3.1, but the subsequent bound appears to use only |J|≤1 together with Weil's bound and divisor estimates; if so, the reference to Lemma 3.1 is misleading. In the second part, the claimed decay by partial integration via (3.8) would require derivative bounds of size K^{-j} for a function supported on [K,2K], while w((k-K)/H) has support length H and derivatives of size H^{-j}=K^{-j(3/4+c)}; the stated justification is therefore not directly applicable. The desired decay in that range is more readily obtained from (3.2), since the Bessel argument is O(K^ε), but the proof should say which estimate is being used and where.","section":"§5, Lemma 5.2"}],"minor_comments":[{"comment":"The notation E1,E2 is recycled: E1 and E2 are defined in (5.7) as the two terms of ET, and then (5.10) redefines E2 = E1 + E2 + E3 with new E1 and E2. Lemmas 5.2, 5.4 and 5.6 therefore refer ambiguously to 'E1' and 'E2'. Please rename the three subsums in (5.10), for example D1,D2,D3, throughout Lemmas 5.2-5.8.","section":"§5, (5.7)-(5.10)"},{"comment":"Lemma 5.4 is titled 'For part E1', but in the context of the decomposition (5.10) it appears to concern one of the new subsums of the old E2; the heading should be updated to match the renamed decomposition.","section":"§5, Lemma 5.4 heading"},{"comment":"The support conventions for the weight function are not harmonized: Lemma 3.4 states that h is supported on (0,1), while Lemma 3.5 says 'h as in Lemma 3.4' but its proof defines functions for u∈(1,2) and applies Lemma 3.1 to functions supported on [K,2K]. Please state the exact support and derivative conditions for each lemma.","section":"§3, Lemmas 3.4 and 3.5"},{"comment":"There is a typo: 'Possion summation formula' should be 'Poisson summation formula'.","section":"§5, Remark 5.3"},{"comment":"Remark 5.5 asserts without proof that Lemma 5.4 can be made valid for H=K^{1/2+c} by the same method as Remark 5.3. Since this range is not used in Theorem 1.2, either give the argument or delete the remark to avoid an unverified claim.","section":"§5, Remark 5.5"}],"recommendation":"major_revision","confidential_remarks":"I do not see grounds for outright rejection: the diagonal computation is clean, the main term is correct, and the paper identifies its own critical point accurately. The obstacle is that Lemma 5.8, the one result that makes the shorter interval work, is not proved for the new weight scale, and Remark 5.9 acknowledges the difficulty. This is a substantial but potentially repairable gap: if the author can supply a complete proof of (5.31) under the actual support and derivative conditions of w', the paper would be a solid contribution. I would recommend a major revision rather than rejection, with the revision expected to contain a real proof of Lemma 5.8 and to clarify the notation in §5."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the paper has a real new statement but the step that buys the new interval, Lemma 5.8, is a one-line citation to Khan's paper with a weight that doesn't satisfy the cited lemma's hypotheses, and the author's own Remark 5.9 concedes the standard method fails. As written, Theorem 1.2 is not established.\n\nWhat's new and good: The statement is new: Khan's [9] averages over weights in [K, 2K], so H is about K, while here H = K^{3/4+c}. The main term derivation through Watson's formula, Petersson's trace formula, and the residue computation producing 6/pi is clean and standard. The author is honest: Remarks 1.3, 5.3, 5.7, and 5.9 identify exactly what controls H. Lemma 5.2 and Remark 5.3 show some independent technique.\n\nSoft spots: Lemma 5.8 is load-bearing. The proof is the single line \"By [9, Lemma 3.5]\" applied to a new smooth weight w'(k/K) supported on (1, (K+H)/K), where H/K = K^{-1/4+c}. The cited lemma is for h supported on (1,2) with bounded derivatives; the new weight has derivatives of size (K/H)^j, and oscillatory Bessel estimates are sensitive to that. Remark 5.9 says after Poisson summation the index properties are \"very poor\", which undercuts the stationary-phase route behind the cited lemma. Without a verification that the transfer works, (5.31) is not available and no improvement over Khan follows. This is not a fatal arithmetical error but a missing proof at the exact spot that decides the theorem.\n\nAlso the notation hurts: E1 and E2 are used both as the two pieces of ET in (5.7) and as two of the three pieces of E2 in (5.10); Lemma 5.4 is titled \"For part E1\" when it is actually for the second piece of the split. The reader has to reverse-engineer which bound covers which range; that is fixable but it is real friction.\n\nThe diagonal and the main term are fine, and the overall strategy is coherent. The paper has no fitted parameters and no self-citations; the constant 6/pi comes out of the diagonal.\n\nVerdict: As written it is not publishable, but the gap is specific and probably addressable. This is exactly the kind of paper a referee should see: the intended audience, analytic number theorists working on moments of L-functions, would benefit from a complete proof of Lemma 5.8. Send it to peer review; if the lemma can be made rigorous, the short-interval improvement is a solid, if narrow, result.","headline":"Genuine short-interval improvement claimed, but the step that buys the new interval, Lemma 5.8, is a one-line citation that does not obviously transfer, so the main theorem is not established as written.","tokens_in":13315,"tokens_out":3257,"would_cite":false,"duration_ms":35541,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11F11","11F30","11F66","11F67"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that, for any fixed $0<c\\le 1/4$, the average fourth moment of holomorphic Hecke cusp forms over a weight window of length $K^{3/4+c}$ is $6/\\pi + O(K^{-\\delta})$, matching the random wave prediction and improving the…","keywords":["holomorphic Hecke cusp forms","fourth moment","random wave conjecture","short intervals in weight","automorphic L-functions","Kloosterman sums","Bessel functions","spectral averages"],"falsifier":"Work out the Poisson-summation and stationary-phase analysis of Lemma 5.8 for the specific weight $w'$ supported on $(1,(K+H)/K)$ with $H = K^{3/4+c}$; if the resulting phase function lacks the required non-stationary decay, the estimate $E_3 \\ll K^{-1/4+\\epsilon} K/H$ fails, and the claimed $O(K^{-\\delta})$ error term in Theorem 1.2 would not follow.","tokens_in":12052,"feed_emoji":"🎲","tokens_out":10663,"duration_ms":93952,"temperature":0.7,"pith_summary":"The paper shows that the average fourth moment of holomorphic Hecke cusp forms of weight near $K$ is $6/\\pi + O(K^{-\\delta})$ when the average is taken over a short window of weights of length $H = K^{3/4+c}$, for any fixed $0 < c \\le 1/4$. This matches the value predicted by the random wave conjecture for these arithmetic eigenfunctions and improves on the earlier result that required a window of length comparable to $K$. The proof reduces the fourth moment to central values of automorphic $L$-functions and then estimates the resulting sums of Bessel functions and Kloosterman sums. The limit of the admissible window length is set by one technical estimate for the third error term.","feed_headline":"Random-wave value 6/pi survives on much shorter cusp-form intervals","feed_subtitle":"Hecke cusp forms' average fourth moment is 6/pi over weight windows of length K^{3/4+c}, improving on the prior K-length windows.","key_machinery":"The load-bearing identity is the triple-product formula expressing $|\\langle F^2, G\\rangle|^2$, for $F$ of weight $k$ and $G$ of weight $2k$, as $\\frac{\\pi^3}{2(2k-1)}\\frac{L(1/2,g)\\,L(1/2,\\mathrm{sym}^2 f\\times g)}{L(1,\\mathrm{sym}^2 f)^2 L(1,\\mathrm{sym}^2 g)}$. Combined with the decomposition $\\|F\\|_4^4 = \\sum_{g\\in B_{2k}}|\\langle F^2,G\\rangle|^2$, it turns the fourth moment into average central values of degree-8 $L$-functions. The average over the Hecke basis is then evaluated with the trace formula, which replaces the spectral sum by Kloosterman sums weighted by products of the Bessel functions $J_{k-1}$ and $J_{2k-1}$. The main term $6/\\pi$ emerges from the diagonal contribution, and the off-diagonal pieces are bounded using Weil's bound for Kloosterman sums together with asymptotic lemmas for weighted averages of Bessel functions; the hardest of these is the third error term of Lemma 5.8.","core_discovery":"With $\\|F\\|_2 = 1$, Theorem 1.2 states that for $H = K^{3/4+c}$ the smoothed average $\\frac{2}{HW}\\sum_{k\\equiv 0 \\bmod 2} w((k-K)/H) \\frac{12}{k}\\sum_{f\\in B_k}\\|F\\|_4^4$ equals $6/\\pi + O(K^{-\\delta})$ for some $0 < \\delta \\le c$. Here $w$ is a smooth non-negative function supported on $(0,1)$, $W$ is its integral, and $B_k$ is the Hecke basis of holomorphic cusp forms of weight $k$. This is the fourth moment averaged over weights in $[K,K+H]$, and the constant $6/\\pi$ is the random wave prediction. The result is obtained by the same overall route as the full-interval theorem: the fourth moment is written as a sum of products of central $L$-values, the average over forms is converted to Kloosterman sums, and the error terms are bounded by negative powers of $K$.","pith_inferences":["Because the first two error terms already work for $H = K^{1/2+c}$, a repair of the third error term would plausibly shorten the admissible window to $K^{1/2+c}$; that stronger statement is not established in the paper.","The key oscillatory integral behind Lemma 5.8 could be checked numerically for moderate $K$ and $c=1/4$; a failure of the expected decay would indicate that the improvement over the full-interval result depends on a missing argument.","The same reduction-plus-Bessel-estimate structure may carry over to short-interval averages of other moments or to Maass forms, provided the corresponding analogue of the third error-term estimate is available."],"forward_implications":["The random wave prediction for holomorphic Hecke cusp forms holds on average over weight windows of length $K^{3/4+c}$, which is shorter than the full dyadic range by a factor of $K^{1/4-c}$.","The proof shows that the admissible window length is controlled by the quality of the Bessel-Kloosterman estimates; in particular, the first two error terms are already amenable to $H = K^{1/2+c}$, so the bottleneck is the third error term.","For any fixed $c \\le 1/4$, the error term in the average is $O(K^{-\\delta})$, so the result gives a genuine power saving rather than a slowly decaying error.","The theorem extends the known fourth-moment average from intervals of length comparable to $K$ down to $K^{3/4+c}$, a substantial shrinking of the averaging window."],"supporting_citations":[{"why":"Supplies the triple-product formula that converts the fourth moment into central values of automorphic L-functions.","marker":"[14]"},{"why":"Supplies the approximate functional equations, the Bessel-sum lemmas, and the full-interval fourth-moment theorem that the short-interval result refines.","marker":"[9]"},{"why":"Gives the trace formula used to pass from spectral averages to Kloosterman sums.","marker":"[8]"},{"why":"Provides the lemma on averages of a single J-Bessel function used repeatedly in the error-term estimates.","marker":"[7]"},{"why":"Establishes the decomposition of the L4 norm into a sum over weight-2k forms and the conjecture that the fourth moment tends to 2 under a different normalization.","marker":"[2]"},{"why":"Supplies the GL(3) coefficient identity used to express the main term as a product of symmetric-square L-functions.","marker":"[4]"},{"why":"Gives the bounds on L(1, sym^2 f) that justify discarding most forms via Lemma 2.3.","marker":"[5]"},{"why":"Provides the uniform bounds on J-Bessel functions used to estimate the off-diagonal terms.","marker":"[11]"}],"fun_headline_variants":["Average fourth moment equals 6/pi on shorter weight intervals","6/pi persists for Hecke cusp forms in short weight windows","Shorter intervals keep cusp-form fourth moment at 6/pi","Fourth moment remains 6/pi on shorter cusp-form intervals","Weight windows of K^{3/4+c} still yield fourth moment 6/pi"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole improvement rests on assuming that the bound for the third error term, which was proved for the full-interval weight, transfers unchanged to the new smooth weight supported on $(1,(K+H)/K)$ with $H = K^{3/4+c}$, even though the paper's own remark says the stationary-phase route behind that bound does not plainly apply.","fun_headline_variants_meta":{"raw":{"variants":["Average fourth moment equals 6/pi on shorter weight intervals","6/pi persists for Hecke cusp forms in short weight windows","Shorter intervals keep cusp-form fourth moment at 6/pi","Fourth moment remains 6/pi on shorter cusp-form intervals","Weight windows of K^{3/4+c} still yield fourth moment 6/pi"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00074,"raw_usage":{"total_tokens":3245,"prompt_tokens":826,"completion_tokens":2419,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":442,"completion_tokens_details":{"reasoning_tokens":2322}},"tokens_in":442,"tokens_out":2419,"duration_ms":18359,"temperature":1.0,"reasoning_tokens":2322,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T18:48:43.521562+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Work out the Poisson-summation and stationary-phase analysis of Lemma 5.8 for the specific weight $w'$ supported on $(1,(K+H)/K)$ with $H = K^{3/4+c}$; if the resulting phase function lacks the required non-stationary decay, the estimate $E_3 \\ll K^{-1/4+\\epsilon} K/H$ fails, and the claimed $O(K^{-\\delta})$ error term in Theorem 1.2 would not follow.","supporting_citations":[{"cited_title":"Coefficients of Maass forms and the Siegel zero","cited_arxiv_id":null,"evidence_quote":"Gives the bounds on L(1, sym^2 f) that justify discarding most forms via Lemma 2.3."},{"cited_title":"On the fourth moment of holomorphic Hecke cusp forms","cited_arxiv_id":null,"evidence_quote":"Supplies the approximate functional equations, the Bessel-sum lemmas, and the full-interval fourth-moment theorem that the short-interval result refines."},{"cited_title":"The vanishing of Poincar´ e series","cited_arxiv_id":null,"evidence_quote":"Provides the uniform bounds on J-Bessel functions used to estimate the off-diagonal terms."}],"review_version":1}