{"id":"6c7165c2-bda9-4199-b8b4-37933abd59e7","arxiv_id":"2608.09871","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The author characterizes L^p to L^q norms of extension operators over hyperbolic rectangles and derives new restriction estimates for degenerate hyperbolic surfaces in R^3.","lead":"This paper proves sharp norm bounds for Fourier extension operators on hyperbolic surfaces over rectangles and uses them to obtain new restriction estimates for surfaces of the form |ξ1|^β1 - |ξ2|^β2. The result extends a known elliptic-surface technique to hyperbolic geometry and improves the exponent range near q > 13/4.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 2.10's O(ℓ²) blurring approximation fails under the stated conventions: the Taylor remainder can be O(σ), not O(ℓ²), leaving the R≲ℓ⁻² case of Theorem 2.2 unsupported.","rationale":"The reader identified the transfer of wave-packet tools to high-eccentricity hyperbolic rectangles as the weakest assumption, and the present concern is in that same part of the argument. However, the sharpest failure is not in the R≫ℓ⁻² wave-packet regime but in the complementary R≲ℓ⁻² reduction (Lemma 2.10), where the paper attempts to blur the high-eccentricity surface into a unit-cube surface. The explicit example h=σℓ⁻²ξ1²ξ2 satisfies the paper's own hyperbolicity condition (1) and makes the claimed O(ℓ²) graph approximation false by a factor ℓ⁻². This is an internal inconsistency, not merely a disagreement with prior literature. Because the reduction (55) is used to obtain the base case for the induction in §2.4, the ℓ-independence asserted in Theorem 2.2 is not proven; Theorem 1.1 inherits this gap. The concern is concrete and checkable, but it does not disprove the main theorem, so the reader's CONDITIONAL verdict remains appropriate. The fix could involve a corrected axis convention, a different approximation of the perturbed hyperbolic phase, or an additional smallness condition on σ relative to ℓ; without such a fix, the central upper bound is incomplete.","tokens_in":24512,"tokens_out":42666,"duration_ms":413694,"concrete_test":"Independently recompute Lemma 2.10 with Q_ℓ=[0,ℓ]×[0,1] and h(ξ1,ξ2)=σℓ⁻²ξ1²ξ2. Verify that h(0)=0, ∇h(0)=0, D²h(0)=0, and H(u,v)=h(ℓu,v)=σu²v satisfies (1) with parameter σ. Then observe that the lemma's \\tilde h is identically zero, so sup_{Q_ℓ}|g−\\tilde g|=σ/2, contradicting the claimed O(ℓ²) error. Next trace the reduction in (55) with R=Cℓ⁻²: the phase difference σ/2 lies outside the C/R Fourier window for all sufficiently small ℓ, so the unit-cube bilinear estimate cannot be invoked uniformly in ℓ. If this computation is correct, Proposition 2.3 fails in the small-scale case and Theorem 2.2's ℓ-independence collapses.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is the reduction in §2.4 for R≲ℓ⁻², which rests on Lemma 2.10. That lemma claims that for g=ξ1²−ξ2²+h over Q_ℓ, the graph Σ_g lies within O(ℓ²) of Σ_\\tilde g, with \\tilde g=ξ1²−ξ2²+h(0,ξ2)+ξ1∂1h(0,ξ2). This is not a consequence of (1). Under the convention forced by the lemma's bound ξ1≤ℓ (ξ1 the short side), take h(ξ1,ξ2)=σℓ⁻²ξ1²ξ2. Then H(u,v)=h(ℓu,v)=σu²v, so (1) holds with parameter σ. But \\tilde h=h(0,ξ2)+ξ1∂1h(0,ξ2)=0, and at ξ1=ℓ, |g−\\tilde g|=|h|=σ|ξ2|, which is O(σ), not O(ℓ²). The displayed proof bounds |∂11h|≲1, whereas (1) only gives ∂11h≈σℓ⁻². If instead ξ1 is the long side, then ξ1≤ℓ is false and the Taylor error is O(1). Thus, under either reading of the axis convention, Lemma 2.10's error estimate fails. Since the R≲ℓ⁻² case of Proposition 2.3 is exactly the regime where this blurring replaces the high-eccentricity surface by a unit-cube surface, the ℓ-independent constant in Theorem 2.2 is not established, and Theorem 1.1 depends on an unproved, as-stated false sublemma.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies Fourier extension operators for perturbed hyperbolic paraboloids over axis-parallel rectangles, aiming to characterize the L^p-to-L^q operator norm in terms of the side lengths. The main results, Theorems 1.1 and 1.2, give sharp two-sided bounds for such norms under the hyperbolicity condition (1), with the upper bounds obtained through a bilinear-to-linear argument and the lower bounds through Knapp examples. The central bilinear estimate, Theorem 2.2, is an extension of Oh's paraboloid bilinear restriction estimate to the hyperbolic phase; its proof uses polynomial partitioning and follows the structure of [O23], with an additional high-eccentricity reduction in Section 2.4. As an application, Proposition 1.5 states new restriction estimates for surfaces |ξ1|^{β1} − |ξ2|^{β2}.","tokens_in":24898,"tokens_out":18558,"duration_ms":185036,"significance":"If the technical gaps in Section 2 are repaired, the paper would make a substantial contribution: Theorems 1.1 and 1.2 provide a clean, apparently sharp description of rectangle extension norms for hyperbolic surfaces over a wide range of exponents, going beyond the elliptic result of [SS21] in the range q > 13/4. The application to surfaces of the form |ξ1|^{β1} − |ξ2|^{β2} is a natural and valuable consequence, and the paper gives both upper bounds and matching Knapp counterexamples. The paper is also honest about its debts: the argument explicitly builds on [O23] and [SS21], and the main line of reasoning is not circular. The presentation is generally well organized, with the rescaling computations and the Whitney-type reductions written out in enough detail to be checkable in those parts.","major_comments":[{"comment":"Lemma 2.10 is false as stated, and the failure is load-bearing for the R ≲ ℓ^{-2} case of Proposition 2.3. Under the convention forced by the proof's bound ξ1 ≤ ℓ, take h(ξ1, ξ2) = σ ℓ^{-2} ξ1^2 ξ2. Then h(ℓu, v) = σ u^2 v, so condition (1) holds with parameter σ, but h̃ = h(0, ξ2) + ξ1 ∂1 h(0, ξ2) = 0, and at ξ1 = ℓ we have |g − g̃| = σ|ξ2|, which is not O(ℓ^2) as ℓ → 0. The displayed proof bounds |∂11 h| ≲ 1, whereas (1) gives only |∂11 h(η, ξ2)| ≤ σℓ^{-2} after accounting for the scaling in the C^N norm. The same scaling issue affects the claim that h̃ is hyperbolic over the unit cube: for example, ∂122 h(0, ξ2) is only controlled by σℓ^{-1}. Since the reduction to the unit-cube surface in (55) requires a graph error of size O(R^{-1}) and R can be as large as ℓ^{-2}, the ℓ-independent constant in Theorem 2.2 and the side-length dependence in Theorem 1.1 are not established by the argument presented.","section":"2.4, Lemma 2.10"},{"comment":"The extension of the [O23] wave-packet machinery to the hyperbolic phase is asserted rather than proved. In particular, the statements that the tube interaction estimates, local constancy, and the Wolff-type tube counting lemmas 'still hold' for the tubes defined in (17), both on Q1 and in the high-eccentricity regime R ≫ ℓ^{-2}, are not accompanied by the required verifications. This is not a purely cosmetic issue: the phase gradients in (17) contain ∂j h, and under (1) these derivatives can be as large as σℓ^{-1} or σℓ^{-2} when ℓ is small, while the hyperbolic surface contains line segments whose interaction geometry differs from the paraboloid. Because Theorem 2.2 is the engine for Theorem 1.1, these deferred checks are central to the proof.","section":"Sections 2.1 and 2.4"},{"comment":"The support-separation condition in Definition 2.1 is inconsistent with the Qℓ convention used in Lemma 2.10. If Qℓ = [−ℓ/2, ℓ/2] × [−1/2, 1/2] with ℓ ≤ 1, then the balls B((±1/2, 0), 1/10) are disjoint from Qℓ for sufficiently small ℓ, so the separated-support hypothesis is vacuous in exactly the high-eccentricity regime that Section 2.4 is designed to analyze. If instead the first coordinate is taken to be the long side so that the balls do meet Qℓ, then Lemma 2.10's key bound ξ1 ≤ ℓ and the Taylor estimate in its proof are invalid. The geometric setup for Theorem 2.2 in the regime ℓ ≪ 1 therefore needs to be restated unambiguously, with the separated caps adapted to the actual rectangle.","section":"Definition 2.1"}],"minor_comments":[{"comment":"The text refers to 'Proposition 1.1' and 'Proposition 1.2' when Theorems 1.1 and 1.2 are meant.","section":"Section 3.3"},{"comment":"There is a typo in 'we keep track track of how the operator norms change'.","section":"Section 3.3"},{"comment":"The outline says 'We also provide more detailed computation of the two-step reduction used to prove Proposition 1.2'; this should refer to Theorem 1.2.","section":"Section 1.1"},{"comment":"In the line 'we have |B11hpη, ξ2q| ≤ ... ≤ 1', the inequality violates the scaling of condition (1); this is part of the issue described in the first major comment, but it should also be fixed in the written proof if the lemma is replaced.","section":"Section 2.4"}],"recommendation":"major_revision","confidential_remarks":"The manuscript relies very heavily on [O23] and [SS21], and the author is a student of a coauthor of [SS21]. I do not see circular reasoning, and heavy dependence on recent work is not by itself a flaw. However, given that the central high-eccentricity reduction appears to be based on a false lemma and that the wave-packet transfer is asserted rather than proved, the editor may wish to seek independent verification of Section 2 before committing to the paper. The result, if correct, is significant enough to justify a major revision rather than a reject, but the current proof is not yet self-contained at a load-bearing point."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a real extension of Schwend–Stovall to hyperbolic surfaces, with a plausible application, but the central bilinear estimate has a load-bearing gap. Lemma 2.10 is false as stated, so Theorem 2.2 is not proved.\n\nWhat is actually new: Theorem 1.1 gives rectangle-dependent operator norms for hyperbolic phases, the first such result in this setting; Theorem 2.2 adapts Oh's bilinear estimate to perturbed hyperbolic paraboloids over highly eccentric rectangles; Proposition 1.5 improves the range for |ξ1|^{β1} − |ξ2|^{β2} from q > 10/3 to q > 13/4. Those are real contributions if the proof holds. The structure is sound—slicing, bilinear-to-linear, Whitney decomposition, polynomial partitioning induction—and the paper is honest about deferring large parts of the argument to [SS21] and [O23].\n\nThe soft spot is not minor. In Section 2.4, the R ≲ ℓ^{-2} case of Proposition 2.3 relies on Lemma 2.10, which claims Σ_g lies in an O(ℓ^2) neighborhood of a unit-cube hyperbolic surface. Under the paper's own normalization (Q_ℓ of size ℓ × 1 and condition (1)), the Taylor computation is wrong. The remainder is (1/2)ξ1²∂11h(η,ξ2); condition (1) only gives ∂11h ≲ σℓ^{-2}, so the remainder is O(σ), not O(ℓ^2). The proof's claim |∂11h| ≲ 1 contradicts the scaling. The stress-test example h(ξ1,ξ2) = σℓ^{-2} ξ1² ξ2 satisfies (1) with \\tilde h = 0 and error σ|ξ2| at ξ1 = ℓ. For R between σ^{-1} and ℓ^{-2}, the graph is not within O(R^{-1}) of \\tilde Σ, so the reduction to the unit cube fails. The ℓ-independent constant in Theorem 2.2 therefore has no support in this regime. This is serious and addressable, not a trivial typo.\n\nOther soft spots are smaller. Section 2.1 asserts that the wave packet interaction and local constancy estimates for the paraboloid transfer to the hyperbolic phase, including high eccentricity; that needs a real proof. Lemma 2.8 is deferred to Oh. Section 5 is explicitly a sketch, which is acceptable if the cited complete proof really covers it, but the dependence should be checked line by line. The heavy reliance on [SS21] and [O23] is disclosed and the citations are appropriate; that is not a problem in itself.\n\nWho should read this: restriction theorists interested in hyperbolic surfaces and polynomial partitioning. It deserves referee time; I would not desk-reject. But I would not accept it in this form. The author needs to fix Lemma 2.10 or replace the blurring argument, and then verify the transferred lemmas. If that happens, the paper should be a solid contribution.","headline":"A genuine hyperbolic analogue of Schwend–Stovall with a real gap: the blurring lemma behind the main bilinear estimate is false as stated, so the eccentricity-independent constant is not proved.","tokens_in":25403,"tokens_out":4905,"would_cite":false,"duration_ms":45178,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["42B10"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves sharp side-length-dependent operator norm estimates for Fourier extension operators over hyperbolic rectangles, and applies them to new restriction bounds for finite-type surfaces |ξ1|^{β1} − |ξ2|^{β2}.","keywords":["Fourier restriction","extension operator","hyperbolic paraboloid","bilinear restriction","polynomial partitioning","rectangle estimates","Knapp examples","degenerate surfaces"],"falsifier":"Compute, for the pure hyperbolic paraboloid g(ξ) = $ξ1^{2}$ − $ξ2^{2}$, the bilinear $L^{{13/4}}$ norm of |Ef1 Ef2|^{1/2} over a ball of radius R ≈ $ℓ^{{−2}}$ with f1, f2 supported in the two separated caps of Q_ℓ, as ℓ → 0; if the optimal constant grows like $ℓ^{{−c}}$ for any c > 0, the eccentricity independence asserted in Theorem 2.2 is false and Theorem 1.1 would need additional side-length factors.","tokens_in":24280,"feed_emoji":"📐","tokens_out":7669,"duration_ms":64613,"temperature":0.7,"pith_summary":"This paper proves a sharp quantitative version of the Fourier restriction problem for surfaces built from a hyperbolic paraboloid: it characterizes the L^p → L^q norm of the extension operator over an axis-parallel rectangle in terms of the rectangle's side lengths. The main theorem covers perturbed hyperbolic phases g = $ξ1^{2}$ − $ξ2^{2}$ + h, with the perturbation allowed to have larger higher derivatives as the rectangle becomes more eccentric. The proof rests on a new bilinear restriction estimate for such perturbed hyperbolic surfaces whose constant is independent of the rectangle's eccentricity, obtained by extending Oh's paraboloid argument and gluing it to a blurring reduction at small scales. As an application, the paper derives new restriction estimates for degenerate surfaces of the form |ξ1|^{β1} − |ξ2|^{β2}, sharp up to the stated exponent region, with the threshold q > 13/4.","feed_headline":"Side lengths fix hyperbolic restriction norms","feed_subtitle":"Bilinear estimates yield exact rectangle bounds and new restriction results for |ξ1|^β1 − |ξ2|^β2.","key_machinery":"The central object is the perturbed hyperbolic paraboloid S = {(ξ, g(ξ)) : ξ ∈ Q_ℓ}, with g = $ξ1^{2}$ − $ξ2^{2}$ + h and h obeying scaled derivative bounds that permit larger higher-order derivatives as the rectangle becomes more eccentric. The carrying mechanism is Theorem 2.2, an eccentricity-independent bilinear restriction estimate for two separated caps on this surface, proved by polynomial partitioning in the style of Oh's paraboloid proof and patched at high eccentricity: for scales R ≲ $ℓ^{{−2}}$ the surface is blurred to a unit-cube hyperbolic surface, while for R ≫ $ℓ^{{−2}}$ wave-packet interaction estimates are transferred from the paraboloid case. This bilinear estimate, interpolated with Lee's $L^{{10/3}}$ mixed-sign result, feeds a bilinear-to-linear argument modelled on Schwend–Stovall's restriction-above-rectangles work, using slicing, Whitney decomposition, rescaled bilinear estimates, and interpolation to produce the rectangle norm formula.","core_discovery":"Theorem 1.1 is the central claim: if ℓ1 ≤ ℓ2 and g is hyperbolic to order N(p,q) over the rectangle Q_ℓ, with phase g(ξ) = $ξ1^{2}$ − $ξ2^{2}$ + h(ξ) and error h satisfying the scaled derivative bounds, then for q > p, q > 13/4, and q = ((4−θ)/(2−θ))p' with 0 < θ ≤ 1, the operator norm satisfies ‖E_ℓ^g‖_{L^p→L^q} ≈ $ℓ1^{{θ p'(1−1/q)}}$; while for q = ((3−θ)/(1−θ))p' with 0 ≤ θ ≤ 1, it satisfies ‖E_ℓ^g‖_{L^p→L^q} ≈ (ℓ1 ℓ2^θ)^{p'(1−1/q)}. The same characterization extends to rotated rectangles whose defining phase has main term ξ1ξ2 (Theorem 1.2). The upper bounds are obtained through a bilinear-to-linear argument; the lower bounds come from standard Knapp examples. As an application, the paper derives Proposition 1.5, giving boundedness for E_β on surfaces |ξ1|^{β1} − |ξ2|^{β2} in the range q > 13/4, q > p, q > 2p', with q/p' ≥ max(1 + 1/(1/2 + 1/max(β1,β2)), 1 + 1/(1/β1 + 1/β2)), and showing the condition is necessary in the stated region.","pith_inferences":["A natural test of the eccentricity-independence claim is to compute the bilinear L^{13/4} norm on the pure hyperboloid h = 0 for rectangles with ℓ → 0; if a logarithmic or power loss in ℓ appears, both the ℓ-independent constant and the sharp rectangle bounds would need correction.","The same bilinear-to-linear machinery could plausibly handle phases with a nonzero linear term or with weaker second-derivative control, since only the scaled derivative bounds and the separation of caps enter the argument; the paper does not pursue this extension.","For the |ξ1|^{β1} − |ξ2|^{β2} application, the sharp exponent region suggests a general template: dyadically decompose a degenerate surface into hyperbolic rectangles and sum the rectangle norms with a Bourgain summation lemma; the open endpoint question is whether the boundary value of q/p' is genuinely unbounded or merely borderline.","The author notes that the θ = 0 endpoint on the scaling line q = 2p' remains open; closing it would merge the two cases of Theorem 1.1 into a single formula along that line."],"forward_implications":["Theorem 1.1 and its rotated version Theorem 1.2 give exact side-length dependence for extension norms over hyperbolic rectangles, so any further restriction estimate on dyadic pieces of a degenerate surface inherits a sharp bookkeeping of scales.","Proposition 1.5 yields new boundedness results for E_β on surfaces |ξ1|^{β1} − |ξ2|^{β2} in the range q > 13/4, q > p, q > 2p', with q/p' ≥ max(1 + 1/(1/2 + 1/max(β1,β2)), 1 + 1/(1/β1 + 1/β2)); the converse Knapp examples show the condition is necessary in the stated region.","Propositions 1.3 and 1.4 extend the rectangle bounds to the L^p-worsening range p ≥ q > 13/4, up to ℓ2/ℓ1 powers that the paper notes cannot be removed by the same argument.","The method also upgrades the elliptic rectangle result of Schwend–Stovall from q > 10/3 to q > 13/4, as the paper remarks after Theorem 1.1."],"supporting_citations":[{"why":"Supplies the improved bilinear restriction estimate for the paraboloid whose polynomial-partitioning proof Section 2 adapts to the hyperbolic phase.","marker":"[O23]"},{"why":"Supplies the restriction-over-rectangles programme: slicing, Whitney decomposition, bilinear-to-linear reduction, and the blurring step for high eccentricity.","marker":"[SS21]"},{"why":"Provides the L^{10/3} bilinear endpoint for surfaces with mixed-sign curvature used in interpolating Theorem 2.2.","marker":"[L06]"},{"why":"Provides the wave-packet decomposition, polynomial Wolff axioms, and wall-case incidence lemmas that the proof reuses.","marker":"[G18]"},{"why":"Supplies the localized multilinear restriction lemma (Lemma 2.7) used to control tangential/transversal wave-packet interactions.","marker":"[B22]"},{"why":"Gives the unit-scale boundedness of E_ℓ^g on hyperbolic surfaces that the Whitney decomposition in Section 3.2 calls on.","marker":"[BMV23]"},{"why":"Provides the perturbed-parabola restriction theorem used along slices to prove Theorem 1.1 in the region T1.","marker":"[T75, Z74]"}],"fun_headline_variants":["Side lengths set Fourier restriction norms on hyperbolic rectangles","Hyperbolic rectangle side lengths fix restriction operator norms","New Fourier restriction bounds from hyperbolic rectangle side lengths","Exact norms for hyperbolic rectangle extension operators","Side-length law governs hyperbolic Fourier restriction"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on the assumption that the paraboloid's wave-packet interaction estimates still hold for the perturbed hyperbolic surface with a constant that does not blow up as the rectangle becomes very narrow; if that transfer fails, the rectangle norm formulas lose their dependence on side lengths.","fun_headline_variants_meta":{"raw":{"variants":["Side lengths set Fourier restriction norms on hyperbolic rectangles","Hyperbolic rectangle side lengths fix restriction operator norms","New Fourier restriction bounds from hyperbolic rectangle side lengths","Exact norms for hyperbolic rectangle extension operators","Side-length law governs hyperbolic Fourier restriction"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000658,"raw_usage":{"total_tokens":2998,"prompt_tokens":919,"completion_tokens":2079,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":535,"completion_tokens_details":{"reasoning_tokens":2011}},"tokens_in":535,"tokens_out":2079,"duration_ms":12882,"temperature":1.0,"reasoning_tokens":2011,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T05:08:22.773134+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute, for the pure hyperbolic paraboloid g(ξ) = $ξ1^{2}$ − $ξ2^{2}$, the bilinear $L^{{13/4}}$ norm of |Ef1 Ef2|^{1/2} over a ball of radius R ≈ $ℓ^{{−2}}$ with f1, f2 supported in the two separated caps of Q_ℓ, as ℓ → 0; if the optimal constant grows like $ℓ^{{−c}}$ for any c > 0, the eccentricity independence asserted in Theorem 2.2 is false and Theorem 1.1 would need additional side-length factors.","supporting_citations":[],"review_version":1}