{"id":"6664251c-2771-4646-8fab-96e36ed7d57a","arxiv_id":"1908.01833","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For every p in (1, infinity), the maximal modulation of the line-restricted Hilbert transform along the parabola is bounded on Lp, uniformly over all lines.","lead":"A harmonic analysis paper proves uniform Lp bounds for maximal modulations of the Hilbert transform along the parabola, after restricting the multiplier to lines. This is partial progress on an open parabolic Carleson problem from work by Guo, Pierce, Roos and Yung.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 2's middle-interval estimate depends on the maximally truncated polynomial Carleson theorem covering the comparison operator in (9); the paper does not verify the cited theorem's exact scope, and this is the least externally secured step.","rationale":"The reader identified the same weakest assumption: the applicability of the maximally truncated polynomial Carleson theorem to the operator arising after Taylor approximation. I agree that this is the most load-bearing external step. The paper's internal TT* estimates are plausible and the algebra in Section 2, while compressed, appears repairable; the strongest risk is that the cited polynomial Carleson theorem may not, as stated, cover the exact maximal-truncation form needed for the comparison in (9). Because the standard reduction from the p.v. version to the maximally truncated version is well known and likely applies, this concern does not change the overall accept verdict, but it should be checked explicitly before the proof is regarded as complete.","tokens_in":18709,"tokens_out":60881,"duration_ms":596411,"concrete_test":"Retrieve Lie (arXiv:1105.4504v3) and Zorin-Kranich (arXiv:1711.03524v5) and verify whether either theorem bounds sup_{P∈P_5} sup_{0<ε<R} |∫_{ε<|t|<R} f(x-t)e^{iP(t)} dt/t| on L^p, uniformly in P, ε, and R. If only the non-maximal p.v. operator is stated, check the reduction for (9): show that for each x the tail ∫_{|t|>r(x)} |f(x-t)|/|t| dt is bounded by C Mf(x) uniformly in r(x), and that the p.v. operator dominates the restricted integral; if this reduction fails for measurable r(x), the middle-interval bound in Part 2 is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on Theorem 2. In Part 2, after the Taylor approximation (9), the phase is replaced by the degree-five polynomial P_{b(x)}(t) and the resulting local integral is declared bounded by the maximally truncated polynomial Carleson theorem from [14,23]. This is load-bearing: if that theorem only bounds the non-truncated principal-value operator, the paper must still supply the standard reduction to restricted intervals, and if its maximal truncations do not cover finite intervals with upper cutoff r(x)=b(x)^{-1/6}, the comparison is not controlled. The paper cites [14,23] without quoting their theorem statements, so this step is the least externally secure. The approximation error in (9) itself is sound: on |t|≤b^{-1/6} the Taylor remainder satisfies b|t|^6≤1, and the error integral is bounded by C Mf(x). The remaining TT* estimates and the k,j summations are internally coherent, so the issue is specifically the scope of the external polynomial Carleson result, not the oscillatory-integral machinery.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proves uniform Lp estimates (1<p<∞) for the maximal modulation operators C_{a,b}f(x)=sup_N |T_{a,b}(e^{iN⋅}f)(x)|, uniformly in the line parameter a (including a=+∞) and in the intercept b, where T_{a,b} are the one-dimensional operators whose multipliers are the restriction of the parabolic Hilbert transform multiplier m_2(ξ,η) to the line (aη+b,η). The proof proceeds in two steps: Proposition 1 reduces the original two-parameter problem to the boundedness of a model operator C^R with phase b(x)[t+1]^{1/2} and a linear modulation e^{iN(x)t}; Theorem 2 then proves the required uniform bound for C^R. The proof of Theorem 2 splits into a small-b regime (b(x)≤10), handled by the classical maximally truncated Carleson theorem plus TT* estimates with exponential decay in dyadic scales, and a large-b regime (b(x)>10), where the phase is replaced by its degree-five Taylor polynomial on the interval |t|≤b(x)^{-1/6}, invoking the polynomial Carleson theorem, while the remaining layers are treated by TT* estimates. The paper also proves (Proposition 5) that no fixed finite-degree polynomial approximation can replace the Taylor step uniformly in b, so the degree-five choice is in a sense necessary.","tokens_in":18926,"tokens_out":14745,"duration_ms":140271,"significance":"If the proof is correct, this is a substantive advance on the parabolic Carleson problem of Question 1: it gives uniform Lp bounds for maximal modulations of the full family of line restrictions of the parabolic Hilbert transform, going beyond the partial results of Roos [18] and the restricted-phase results of Guo–Pierce–Roos–Yung [9]. The method is a clean and instructive combination of the polynomial Carleson theorem with TT* oscillatory estimates, and the explicit exponential decay in Propositions 3 and 4 is presented in sufficient detail to be checkable. Proposition 5 is an elegant and rigorous obstruction result. The paper is honest about its reliance on external results, though that reliance should be spelled out more precisely.","major_comments":[{"comment":"The claim that the local integral in Eq. (9) is bounded by a maximally truncated polynomial Carleson operator of degree ≤5 in the sense of [14,23] is load-bearing for Theorem 2 and is not fully verified. The manuscript does not quote the exact theorem it relies on. Please state the theorem (or a precise corollary) and explicitly check that it applies to: (i) the x-dependent truncation radius b(x)^{-1/6} (if the cited theorem is stated only for constant truncation R, the standard reduction to x-dependent cutoffs must be supplied); (ii) polynomial coefficients that are measurable functions of x after the Kolmogorov–Seliverstov linearization; (iii) the full phase P_{b(x)}(t)+N(x)t of degree 5, with maximal truncation. Without this verification, Eq. (9) leaves a gap in the proof of Theorem 2.","section":"Section 3, Part 2, Eq. (9)"}],"minor_comments":[{"comment":"There are several typographical errors ('Hilber t', 'P ar abola', 'POL YNOMIAL', 'OSCILLA TOR Y', 'Ackowledgements'); please correct them.","section":"Title and Abstract"},{"comment":"The displayed definition of φ_{b(x)}(t) has garbled summation limits; please rewrite it with clear bounds, for instance j ranging from (2−1/6)⌊log_2 b(x)⌋ to 2⌊log_2 b(x)⌋−3 (or whatever is intended).","section":"Section 3, Part 2"},{"comment":"The Taylor remainder bound is stated for t∈[−1/2,1/2], but the integration interval is [−b(x)^{-1/6}, b(x)^{-1/6}], which exceeds [−1/2,1/2] when b(x)<64. Since b(x)^{-1/6}<1, the estimate still holds with an absolute constant, but the interval should be adjusted and the constant justified.","section":"Near Eq. (9)"},{"comment":"The reference 'Proposition 2 in Chapter VIII of [20]' should be given a precise number or title to avoid collision with Proposition 2 of this paper.","section":"Lemma 2"},{"comment":"The phrase 'both bounded in Lp' should explicitly state that the two choices are [u]^{1/2}=|u|^{1/2} and sign(u)|u|^{1/2}.","section":"Proposition 2"},{"comment":"References [14] and [23] are listed as preprints; if published versions exist, they should be updated.","section":"References"},{"comment":"Several changes of variables in the reduction are stated without showing Jacobians and interval endpoints; expanding them would improve readability, though I did not find an error in the stated equivalences.","section":"Section 2"}],"recommendation":"major_revision","confidential_remarks":"This paper is a good fit for math.CA and the main result is likely correct. My main concern is the unquoted invocation of the polynomial Carleson theorem at Eq. (9), which is load-bearing and should be closed by either quoting the exact theorem or supplying a short reduction. This is fixable, and I expect a positive outcome after revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a genuine, plausible result that deserves a serious referee. The main theorem — uniform Lp bounds for maximal modulations of the Hilbert transform along the parabola restricted to lines — is new, and the proof strategy mixes polynomial Carleson and TT* decay in a way I haven't seen before.\n\nThe reduction in Section 2 is the densest part. It's sketched rather than fully expanded, but the comparisons are standard and the uniformity in a seems to hold. Theorem 2 is the heart, with a sensible small-b/large-b split. The TT* estimates in Propositions 3 and 4 have the right decay, and Proposition 5 is a clean, convincing argument: any good uniform polynomial approximation to sqrt(t+1) must have degree tending to infinity. That's a nice observation.\n\nThe soft spot is the middle interval in Part 2 of Theorem 2, exactly where the stress test points. The operator is compared to a maximally truncated polynomial Carleson operator of degree ≤ 5, and the paper cites Lie and Zorin-Kranich without quoting their theorems. A referee should verify that the cited result covers maximal truncations with measurable polynomial coefficients on finite intervals with upper cutoff b(x)^{-1/6}. I suspect it does — Zorin-Kranich's theorem is built for maximal polynomial modulations — but the paper should state the theorem and check the hypotheses. If there is a gap, it's patchable, not fatal.\n\nThe math is otherwise coherent. There are some typos and OCR artifacts, and Section 2 compresses a few algebra steps, but nothing load-bearing. The citation pattern looks fine: primary sources, no self-citation inflation. The paper also honestly says it does not resolve Question 1, which keeps the claim proportioned.\n\nI'd bring this to the attention of a harmonic analyst working on Carleson operators. It's the kind of paper that might prompt further work on the full parabolic Carleson problem. My recommendation: send it to peer review, with instructions to check the scope of the cited polynomial Carleson theorem in the middle-interval comparison.","headline":"A genuinely new and plausible line-restricted bound for the parabolic Carleson problem, with one external step that needs a referee's scrutiny.","tokens_in":19450,"tokens_out":3259,"would_cite":true,"duration_ms":30905,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["42B20","42B25","44A12"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves uniform L^p bounds for maximal modulations of the Hilbert transform along the parabola when restricted to lines.","keywords":["Hilbert transform along the parabola","Carleson operator","polynomial Carleson theorem","maximal modulations","oscillatory integrals","TT* method","singular integrals","uniform L^p bounds"],"falsifier":"The decisive check is the stationary-phase estimate in Lemma 2 at the boundary $|\\xi'| = 2^{-j/100}$, $h=1$, with $b(x)=b(y)=10$: the determinant lower bound used to prove $|Q(s')| \\gtrsim 2^{j/3}$ must have an implied constant independent of $h$; a direct computation showing the constant grows with $j$ would make the claimed $2^{-j/9}$ decay false and would collapse the proof of Theorem 2.","tokens_in":18503,"feed_emoji":"📐","tokens_out":10476,"duration_ms":101522,"temperature":0.7,"pith_summary":"This paper tries to establish uniform $L^p$ bounds, for every $1<p<\\infty$, for maximal modulations of the Hilbert transform along the parabola when the multiplier is restricted to lines. Concretely, Theorem 1 asserts that the operator norm of $C_{a,b}f(x)=\\sup_{N\\in\\mathbb{R}}|T_{a,b}(e^{iN\\cdot}f)(x)|$ is bounded uniformly over all slopes $a$ and intercepts $b$ of the restricting line. The argument reduces this two-parameter problem to a single family of one-dimensional oscillatory singular integrals with phase $b(x)[t+1]^{1/2}$, and proves those are bounded uniformly in the truncation $R$ by combining the polynomial Carleson theorem with oscillation decay obtained through the $TT^*$ method. If correct, this is concrete progress toward the open question of bounding the full parabolic Carleson operator, and it exhibits where one-dimensional polynomial-phase technology is sufficient.","feed_headline":"Line-restricted parabola Hilbert transform has uniform L^p bounds","feed_subtitle":"Maximal modulations of each one-dimensional restriction are bounded independently of the slope and intercept.","key_machinery":"The load-bearing object is the linearized phase operator $C_R$: via Kolmogorov--Seliverstov linearization, the suprema over $N$ and $b$ become a single operator with measurable functions $N(x)$ and $b(x)$, and the phase is $b(x)[t+1]^{1/2}$. The proof splits $|t|$ into dyadic scales and applies two mechanisms: for $|t|\\le b(x)^{-1/6}$ it replaces the phase by its degree-five Taylor polynomial $P_{b(x)}(t)=b(x)\\left(1+\\tfrac{t}{2}-\\tfrac{t^2}{4}+\\tfrac{3t^3}{8}-\\tfrac{15t^4}{16}+\\tfrac{105t^5}{32}\\right)$, reducing the operator to a maximally truncated polynomial Carleson operator of degree at most five; for larger $|t|$ it uses the $TT^*$ method, bounding $S_j(S_j)^*$ through oscillatory integrals whose phase is controlled by the vector $Q=(\\varphi'',-\\tfrac{2}{3}\\varphi''')$, represented as a matrix times a vector so that determinant and norm bounds yield stationary-phase decay like $2^{-j/200}$.","core_discovery":"The paper's central discovery is that maximal modulations of each one-dimensional restriction of the parabolic Hilbert transform are uniformly $L^p$-bounded. In symbols, for the multiplier $m_2$ of the parabolic Hilbert transform and its line restrictions $m_{a,b}(\\eta)=m_2(a\\eta+b,\\eta)$ (with $m_{+\\infty,b}(\\eta)=m_2(\\eta,b)$), the operators $C_{a,b}f(x)=\\sup_{N\\in\\mathbb{R}}|\\mathcal{F}^{-1}(m_{a,b}\\widehat{M_N f})(x)|$ satisfy $\\sup_a\\|\\sup_b C_{a,b}\\|_{p\\to p}<\\infty$ for $1<p<\\infty$. The route is a reduction (Proposition 1) to the truncated operators $C_R f(x)=\\sup_{N,b}\\left|\\int_{-R}^R f(x-t)e^{iN t}e^{ib[t+1]^{1/2}}\\frac{dt}{t}\\right|$, followed by a proof (Theorem 2) that these are bounded independently of $R$: on a shrinking interval around the origin the phase is compared to a degree-5 Taylor polynomial, so the polynomial Carleson theorem applies, while on the remaining intervals oscillation is strong enough that $TT^*$ and stationary phase give summable dyadic decay.","pith_inferences":["A natural next test is whether the two-regime split persists in a full time-frequency proof of the open parabolic Carleson question; the line restriction may be exactly the regime where a one-dimensional polynomial Carleson theorem suffices, and a two-dimensional analogue would be needed only for genuinely two-dimensional strips.","The asymptotic threshold $d\\ge 4$ suggests a trade-off between polynomial degree and interval size: choosing a higher-degree Taylor polynomial shrinks the Carleson interval and moves more of the analysis into the oscillatory regime, which might be optimizable for better decay exponents.","One could test the conjectured general principle numerically for smooth compactly supported $f$ and slowly varying $b(x)$: if the constants in the model operator $C_R$ drift with $R$ in simulations, the uniform $R$ statement in Theorem 2 would be suspect even though the proof's dyadic bounds appear summable."],"forward_implications":["Theorem 1 gives uniform $L^p$ bounds for maximal modulations along every line in the multiplier plane, so the constant in the $L^p$ estimate does not depend on the slope or intercept of the line.","If the same argument can be extended from lines to arbitrarily thin strips with constants independent of width, the limit argument in Section 4 would yield $L^2$ bounds for the full parabolic Carleson operator $C_2$.","The dyadic split at $b(x)^{-1/6}$ and the degree-five Taylor phase make concrete the paper's suggested principle: maximal operators with phases $N\\cdot t + b\\,\\eta(t+1)$ should be $L^p$ bounded whenever $\\eta$ is smooth away from the origin with controlled derivatives.","For monomial curves $(t,t^m)$, the reductions are expected to carry through with phase $[t+1]^{1/m}$, giving line-restricted uniform bounds for higher-order parabolic Hilbert transforms.","Proposition 5 shows a fixed-degree polynomial approximation of $\\sqrt{t+1}$ over a fixed interval forces the degree to grow with $b$, so the shrinking interval in the proof is not an artifact: some cutoff is necessary in this approach."],"supporting_citations":[{"why":"Provides the polynomial Carleson theorem used to bound the degree-5 phase approximation on the middle interval.","marker":"[14]"},{"why":"Extends the polynomial Carleson theorem to maximally truncated maximal polynomial modulations, which is exactly the comparison operator in (9).","marker":"[23]"},{"why":"Introduces the dyadic decomposition with TT* and oscillatory integral estimates that the paper's outer-interval analysis follows.","marker":"[22]"},{"why":"Supplies the stationary-phase proposition applied in Lemmas 2 and 4 to get $2^{-j/9}$ and related decay.","marker":"[20]"},{"why":"Source of the vector-matrix determinant method for lower-bounding second and third derivatives of the phase.","marker":"[9]"},{"why":"Bounds the quadratic Carleson operator, used to dispose of horizontal lines ($a=+\\infty$) in Theorem 1.","marker":"[13]"}],"fun_headline_variants":["Uniform L^p for line-restricted parabolic Hilbert transform","Line-restricted parabolic Hilbert transform: uniform L^p maximal modulations","Maximal modulations along lines: uniform L^p for parabolic Hilbert transform","Parabolic Hilbert transform on lines: uniform L^p for maximal modulations","TT* and Carleson yield uniform L^p for line-restricted parabolic Hilbert"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the maximally truncated polynomial Carleson theorem covers degree-5 operators whose coefficients are measurable functions of $x$ (the linearized phase $b(x)$), after the phase $b(x)\\sqrt{t+1}$ is replaced by its Taylor polynomial on $[-b(x)^{-1/6}, b(x)^{-1/6}]$; if that theorem does not apply to measurable-coefficient operators, or if the error estimate (9) fails uniformly, Theorem 2 does not follow.","fun_headline_variants_meta":{"raw":{"variants":["Uniform L^p for line-restricted parabolic Hilbert transform","Line-restricted parabolic Hilbert transform: uniform L^p maximal modulations","Maximal modulations along lines: uniform L^p for parabolic Hilbert transform","Parabolic Hilbert transform on lines: uniform L^p for maximal modulations","TT* and Carleson yield uniform L^p for line-restricted parabolic Hilbert"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001314,"raw_usage":{"total_tokens":5325,"prompt_tokens":890,"completion_tokens":4435,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":506,"completion_tokens_details":{"reasoning_tokens":4337}},"tokens_in":506,"tokens_out":4435,"duration_ms":32327,"temperature":1.0,"reasoning_tokens":4337,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:03:45.917281+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The decisive check is the stationary-phase estimate in Lemma 2 at the boundary $|\\xi'| = 2^{-j/100}$, $h=1$, with $b(x)=b(y)=10$: the determinant lower bound used to prove $|Q(s')| \\gtrsim 2^{j/3}$ must have an implied constant independent of $h$; a direct computation showing the constant grows with $j$ would make the claimed $2^{-j/9}$ decay false and would collapse the proof of Theorem 2.","supporting_citations":[{"cited_title":"The Polynomial Carleson Operator","cited_arxiv_id":"1105.4504","evidence_quote":"Provides the polynomial Carleson theorem used to bound the degree-5 phase approximation on the middle interval."},{"cited_title":"Maximal polynomial modulations of singular integrals","cited_arxiv_id":"1711.03524","evidence_quote":"Extends the polynomial Carleson theorem to maximally truncated maximal polynomial modulations, which is exactly the comparison operator in (9)."},{"cited_title":"Stein, S","cited_arxiv_id":null,"evidence_quote":"Introduces the dyadic decomposition with TT* and oscillatory integral estimates that the paper's outer-interval analysis follows."},{"cited_title":"Stein, Harmonic Analysis: Real variable methods, Orthogonality, and Oscilatory integrals, Prince- ton University Press, Princeton, NJ, 1993","cited_arxiv_id":null,"evidence_quote":"Supplies the stationary-phase proposition applied in Lemmas 2 and 4 to get $2^{-j/9}$ and related decay."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Source of the vector-matrix determinant method for lower-bounding second and third derivatives of the phase."},{"cited_title":"Lie, The (weak-L2) boundedness of the quadratic Carleson operator","cited_arxiv_id":null,"evidence_quote":"Bounds the quadratic Carleson operator, used to dispose of horizontal lines ($a=+\\infty$) in Theorem 1."}],"review_version":1}