{"id":"381eb065-943e-4ed6-822e-3c237880518a","arxiv_id":"2502.02795","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For all n≥3 and p>2, the strong spherical maximal operator is bounded on Lp, resolving the sharp n=3 case of the strong spherical maximal conjecture.","lead":"This paper proves that the strong spherical maximal operator, which averages a function over axis-parallel ellipsoids of every shape, is bounded on Lp for every p>2 in dimensions n≥3. The result improves the previous best range and is sharp in dimension 3, where it resolves the conjecture.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The core volume bound Lemma 2.2 rests on Lemma 3.4, whose O(ρ/t) diameter estimate imports Pawlucki’s quantitative cell decomposition (Lemma 3.7); if that constant is not uniform in t,r,ρ, the δ²/t estimate collapses.","rationale":"I read the proof of Theorem 1.1. The Fourier-analytic reduction, the slicing and duality steps, and the local-to-global passage in §4 are standard and internally consistent. The genuinely novel part is the discretised incidence estimate, where Lemma 2.2 is proved via Lemma 3.2, whose proof uses Lemma 3.4. Lemma 3.4 is the only place where a deep external quantitative result is used as a black box: Pawlucki's Lipschitz cell decomposition. The specific claim that Ω_k(t,r;ρ) is a union of O(1) pieces of diameter O(ρ/t) requires the cell-decomposition constant L to be O(1) uniformly over the parameter ranges, and this uniformity is asserted rather than derived. I agree with the reader's identification of this as the weakest assumption. I found no internal inconsistency: Lemma 3.3 supplies the needed inverse-derivative bound, and the volume/coarea arguments after Lemma 3.2 are correct. The reliance on a cited, published theorem is normal mathematical practice, so I do not see grounds to change the ACCEPT verdict; however, the residual risk is concentrated exactly at the quantitative uniformity of Lemma 3.7, and an independent verification of Lemma 3.4 would substantially increase confidence.","tokens_in":19194,"tokens_out":41136,"duration_ms":398236,"concrete_test":"Independently re-derive Lemma 3.4 for the specific quadratic map Φ_k in (3.4) without invoking the full o-minimal cell decomposition: use Lemma 3.3(ii) and a quantitative inverse function theorem on the open set {|Φ_{t,r}|<ρ/2} to build a cover by O(1) balls of radius Cρ/t, and confirm the constants are independent of t∈(0,2], r∈[1/2,2]^n, and 0<ρ<\\bar c_n t. If the cell count or diameter bound depends on ρ or on the coefficients of Φ_{t,r} beyond O(1), then Lemma 3.2's covering fails and the δ²/t volume estimate in Lemma 2.2 is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central geometric estimate is Lemma 2.2, and its proof relies on Lemma 3.2, which in turn uses Lemma 3.4 to cover the small-Jacobian sublevel set Ω_k(t,r;ρ) by O(1) sets of diameter O(ρ/t). Lemma 3.4 is the only genuinely deep imported step in the incidence-geometry proof: it invokes Pawlucki’s quantitative Lipschitz cell decomposition (Lemma 3.7) to get regular L-cells with L=O(1) inside the image Φ_{t,r}(Ω∩U_m), then pulls them back through the inverse of Φ_{t,r} using the uniform inverse-derivative bound from Lemma 3.3(ii). If the Lipschitz constant in Lemma 3.7 depended on the coefficients of Φ_{t,r} (which vary with t and r) or on ρ, the diameter bound would degrade and the subsequent volume estimate would fail: the dyadic ρ-sum in Lemma 2.2 would no longer produce δ²/(δ+|t1−t2|). The authors do not reprove Lemma 3.7 or verify its quantitative uniformity for this specific semi-algebraic parameter family. This is the most load-bearing assumption in the paper; the surrounding Fourier-analytic and coarea arguments are internally consistent.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves L^p boundedness for p>2 of the strong spherical maximal operator M^st in R^n for all n≥3, improving the previous range p>2(n+1)/(n-1) obtained by Lee, Lee, and Oh and matching the conjectured sharp range when n=3. The proof combines a new discretised incidence-geometry estimate (Theorem 1.3) with the Fourier-analytic L^p-Sobolev estimates of Lee-Lee-Oh and a multiparameter Littlewood-Paley reduction. The core of the paper is a detailed proof of a volume bound (Lemma 2.2) for intersections of thin ellipsoidal annuli after removing exceptional sets of high-order tangencies, using a variable slicing argument, coarea estimates, and a quantitative inverse function theorem.","tokens_in":19428,"tokens_out":14249,"duration_ms":132196,"significance":"If correct, the result resolves the sharp range for the strong spherical maximal function in three dimensions and gives the best known range in all higher dimensions. The geometric part of the proof is substantial and carefully executed: the slicing and duality reductions, the exceptional-set removal, the coarea estimates, and the quantitative inverse function theorem are all developed in detail with explicit dimensional constants. The main novelty, the discretised incidence-geometry volume estimate, is likely to be useful for other multiparameter maximal problems. The proof does rely on one deep imported result, Pawlucki's quantitative Lipschitz cell decomposition (Lemma 3.7), which the authors should state with maximal precision concerning the dependence of the Lipschitz constant; they do, however, cite the theorem and use it in a way that appears consistent with its quantitative content.","major_comments":[],"minor_comments":[{"comment":"In the proof of Lemma 3.4, the sets Φ_{t,r}(Ω_k(t,r;ρ)∩U_m) are not necessarily open because Ω_k includes the closed condition |ω_k|^3 ≥ 2c_n, whereas Lemma 3.7 is stated for open X; please add a sentence explaining that one applies the cell decomposition to the interior of these sets and absorbs the measure-zero boundary into the null set N.","section":"Lemma 3.4 / §3.2"},{"comment":"Lemma 3.7 states that each cell is a regular L-cell for 'some L=O(1)' without specifying the dependence; since the argument in Lemma 3.4 requires L to be bounded by a constant depending only on n and the complexity of X (uniformly in t, r, and ρ), please state this dependence explicitly.","section":"Lemma 3.7 / §3.2"},{"comment":"The text contains the corrupted phrase 'A theorem of /suppress Lojasiewicz'; this should read 'A theorem of Łojasiewicz'.","section":"§3.2, text near Lemma 3.5"},{"comment":"In the paragraph after (2.1), 'a slight abuse of notion' should be 'a slight abuse of notation'.","section":"§2.1"},{"comment":"The passage from the local operator M^loc to the global operator M^st is delegated entirely to [10, §3]; since this is a key step in the proof of Theorem 1.1, it would be helpful to include at least a sketch of the Littlewood-Paley argument and the modifications needed when invoking Proposition 4.1.","section":"§4.2"},{"comment":"Reference [19] (J. Thom) shares the title and page range of [13] (Milnor); please verify the citation and correct the author or title if needed.","section":"References"}],"recommendation":"minor_revision","confidential_remarks":"The paper is mathematically significant and appears sound. The only substantive point of caution is the reliance on Pawlucki's quantitative cell decomposition for the key diameter estimate; if the editorial board is satisfied that the cited theorem indeed provides a Lipschitz constant depending only on dimension and complexity, the proof goes through. The issues I raise are local and can be addressed in a minor revision. I recommend acceptance after these points are clarified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: real advance, and the proof is largely sound. The main thing that worried me—uniformity of the Lipschitz constant in Pawlucki's cell decomposition—turns out not to be a problem.\n\nThe new content is the discretised incidence estimate (Theorem 1.3). The previous best range p>2(n+1)/(n-1) was due to Lee-Lee-Oh; getting down to p>2 for n≥3 is a genuine improvement and is sharp at n=3. The core of the proof, Lemma 2.2 plus Lemma 3.4, is a variable-slicing argument: fix a direction, remove the high-tangency set, then use a quantitative Jacobian bound and a dyadic coarea estimate. Lemma 3.3, which gives |det DωΦ| >~ t^{n-1} and |DωΦ^{-1}| ≤ C/t under a small Φ condition, is worked out explicitly and is believable. The covering-by-refinements step at the end of §3 is a neat way to handle the exceptional sets.\n\nSoft spots: the Fourier-analytic reduction in §4 is mostly delegated to [10] and [17]. Probably fine for a harmonic analysis audience, but a referee should verify the interpolation and kernel domination arguments. Also, the paper imports, not reproves, the quantitative inverse function theorem (Lemma 3.5) and Pawlucki's cell decomposition (Lemma 3.7). That is acceptable, but it means the proof rests on two substantial external tools.\n\nOn the stress-test worry: Lemma 3.4 uses Lemma 3.7 to cover Ω_k(t,r;ρ) by O(1) sets of diameter O(ρ/t). The concern was that Pawlucki's L might depend on t,r. It does not. The semi-algebraic sets to which Lemma 3.7 is applied are images Φ_{t,r}(Ω∩U_m), and these have complexity bounded uniformly in t,r,ρ because the defining polynomials have degree O(1) and bounded coefficients. So Lemma 3.7 indeed gives L=O(1) uniformly, and the diameter estimate survives. That concern does not land.\n\nThe necessary-condition appendix is self-contained and useful.\n\nWho is this for: harmonic analysts working on maximal operators and people interested in incidence-geometric methods in Fourier analysis. It deserves a serious referee. My recommendation: engage with it; likely outcome is acceptance after a careful check of the geometry in §3.","headline":"Genuine sharp-range advance for n=3 with a new incidence-geometric volume estimate; the flagged Pawlucki uniformity worry does not survive contact with the proof.","tokens_in":19990,"tokens_out":3392,"would_cite":true,"duration_ms":32243,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["42B25","14P10"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that the strong spherical maximal operator is bounded on $L^p(\\mathbb{R}^n)$ for every $p>2$ in every dimension $n\\geq 3$, matching the conjectured sharp threshold in dimension three.","keywords":["strong spherical maximal operator","multiparameter maximal averages","L^p estimates","ellipsoidal annuli","discretised incidence geometry","variable slicing argument","semi-algebraic cell decomposition","L2 duality method"],"falsifier":"For a fixed dimension $n$, compute the volume of the intersection $E^{\\delta,k}(0,1)\\cap E^{\\delta,k}(t d_k,r)$ for pairs of ellipsoids whose parameters place them near the higher-order tangency set, and test whether the bound $\\delta^2/(\\delta+t)$ (up to a $\\log\\delta^{-1}$ factor) holds uniformly as $\\delta\\to 0$; a single sequence of configurations violating this bound would falsify Lemma 2.2 and with it the proof of the discretised $L^2$ estimate.","tokens_in":18971,"feed_emoji":"🌐","tokens_out":15917,"duration_ms":137719,"temperature":0.7,"pith_summary":"This paper proves that the strong spherical maximal operator—the maximal average over concentric axis-parallel ellipsoids with arbitrary radii—is bounded on $L^p(\\mathbb{R}^n)$ for every $p>2$ and every dimension $n\\geq 3$. The range matches the conjectured sharp threshold $p>(n+1)/(n-1)$ in dimension three, where the threshold equals $2$. The proof works by isolating a discretised incidence-geometry estimate: after discarding a small set of exceptional higher-order tangencies between pairs of ellipsoids, the intersection of two thin ellipsoidal annuli has volume controlled by $\\delta^2/(\\delta+|t_1-t_2|)$. This estimate is then interpolated with earlier Fourier-analytic $L^p$–Sobolev bounds, improving the previously known range $p>2(n+1)/(n-1)$ and settling the $n=3$ case of the strong spherical maximal conjecture.","feed_headline":"Strong spherical averages are L^p-bounded for every p>2, n≥3","feed_subtitle":"Matches the conjectured sharp range in dimension 3 and improves earlier bounds in all higher dimensions.","key_machinery":"The load-bearing object is the sublevel set $\\Omega_k(t,r;\\rho)$ of the polynomial map $\\Phi_k(\\omega,t,r)$ defined in (3.4), whose zero set encodes the higher-order tangency condition that the gradients $\\nabla F_{0,1}(\\omega)$ and $\\nabla F_{x,r}(\\omega)$ are parallel. The decisive fact is Lemma 3.4: $\\Omega_k(t,r;\\rho)$ can be covered by $O(1)$ sets of diameter $O(\\rho/t)$, obtained by applying the effective inverse function theorem and a quantitative Lipschitz cell decomposition to $\\Phi_k$. That diameter bound feeds into the coarea formula and yields the volume estimate $|E^{\\delta,k}(x_1,r_1)\\cap E^{\\delta,k}(x_2,r_2)|\\lesssim \\log\\delta^{-1}\\,\\delta^2/(\\delta+|t_1-t_2|)$, and the summation argument of [6] converts the pair-wise volume bound into the multiplicity bound behind the discretised $L^2$ estimate.","core_discovery":"The central claim is Theorem 1.1: for every $n\\geq 3$ and every $p>2$, the strong spherical maximal operator $M^{\\mathrm{st}}$ is bounded on $L^p(\\mathbb{R}^n)$, with a constant $C_{n,p}$ depending only on dimension and exponent. The paper proves this through the discretised $L^2$ bound (Theorem 1.3), which is new and independent of the earlier Fourier-analytic machinery: for every $\\varepsilon>0$, the localised supremum over radii in a small cube satisfies $\\|M^\\delta f\\|_{L^2}\\lesssim_\\varepsilon \\delta^{-\\varepsilon}\\|f\\|_{L^2}$. The discretised bound is used to obtain an $L^p$–Sobolev estimate for the local maximal operator, which is then interpolated against the $L^p$–Sobolev estimates of [10] and extended to unrestricted radii by a multiparameter frequency-localisation argument. In dimension three the necessary condition $p>2$ and the sufficiency $p>2$ coincide, so the strong spherical maximal conjecture is resolved there; in higher dimensions the conjecture remains open, and this paper leaves a gap between the proven range $p>2$ and the conjectured range $p>(n+1)/(n-1)$.","pith_inferences":["If the same slicing construction is applied with ellipsoid centres confined to a $k$-dimensional subspace rather than a coordinate line, the argument suggests a hierarchy of multiparameter strong spherical estimates indexed by the dimension of the centre set; this is an extension the paper does not state.","The proof's universal-exceptional-set idea indicates that the $\\log\\delta^{-1}$ factor in the volume bound may be removable by choosing the slicing lines more carefully; testing this on the $n=3$ case would sharpen the constant without changing the main theorem.","The manuscript contains an anomalous inserted token '/suppress' before a cited theorem's name in the proof of Lemma 3.4; the surrounding argument still depends on that cited theorem for the connectivity of the semi-algebraic pieces, so the passage should be checked before relying on the quantitative constant.","A direct numerical evaluation of the volume bound in Lemma 2.2 on explicit pairs of ellipsoids would isolate the geometric mechanism from the Fourier-analytic interpolation and could indicate whether the $p>2$ range is improvable in higher dimensions."],"forward_implications":["The strong spherical maximal conjecture is settled in dimension three: the operator is bounded on $L^p(\\mathbb{R}^3)$ for every $p>2$, exactly the range forced by the necessary condition.","In every dimension $n\\geq 3$, the previously known boundedness range $p>2(n+1)/(n-1)$ is improved to $p>2$.","The discretised $L^2$ estimate (1.4) holds for all $n\\geq 3$ with only a $\\delta^{-\\varepsilon}$ loss, for every $\\varepsilon>0$.","The local maximal operator $M_{\\mathrm{loc}}$ satisfies an $L^p$–Sobolev estimate with a positive power saving for every $p>2$, which is the input that makes the global frequency-localisation argument work."],"supporting_citations":[{"why":"supplies the effective inverse function theorem used to decompose the domain of the polynomial map into $O(1)$ diffeomorphism patches.","marker":"[4]"},{"why":"supplies the quantitative Lipschitz cell decomposition that gives the $O(1)$ intrinsic–extrinsic metric equivalence behind the diameter bound.","marker":"[14]"},{"why":"supplies the $L^2$ duality/summation argument that converts pairwise volume bounds into the multiplicity bound for the discretised maximal operator.","marker":"[6]"},{"why":"provides the $L^p$–Sobolev estimates and the frequency-localisation reduction used to pass from the discretised bound to the full maximal operator.","marker":"[10]"},{"why":"supplies the degree-based volume bound for tubular neighbourhoods of algebraic varieties used to control fibres in Lemma 3.1.","marker":"[21]"},{"why":"provides the kernel-domination device that converts the discretised $L^2$ estimate into an $L^2$–Sobolev bound.","marker":"[17]"},{"why":"backs the quantitative semi-algebraic decomposition facts (connected components, complexity bounds) used in the proof of Lemma 3.4.","marker":"[1]"}],"fun_headline_variants":["L^p bounded for strong spherical maximal: p>2, all n≥3","Strong spherical averages bounded for p>2, sharp in dimension 3","New p>2 proof for strong spherical maximal in every n≥3","p>2 range proven for strong spherical maximal, n=3 sharp"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is the imported quantitative cell-decomposition theorem (Lemma 3.7): it asserts that the relevant real-algebraic pieces can be covered by $O(1)$ cells in which every pair of points can be joined by a curve of length comparable to their straight-line distance, with the comparability constant independent of the scale parameters; if that constant failed to be uniform, the diameter bound behind the volume estimate would collapse.","fun_headline_variants_meta":{"raw":{"variants":["L^p bounded for strong spherical maximal: p>2, all n≥3","Strong spherical averages bounded for p>2, sharp in dimension 3","New p>2 proof for strong spherical maximal in every n≥3","p>2 range proven for strong spherical maximal, n=3 sharp"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000297,"raw_usage":{"total_tokens":1738,"prompt_tokens":976,"completion_tokens":762,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":592,"completion_tokens_details":{"reasoning_tokens":689}},"tokens_in":592,"tokens_out":762,"duration_ms":6816,"temperature":1.0,"reasoning_tokens":689,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T11:07:05.489394+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a fixed dimension $n$, compute the volume of the intersection $E^{\\delta,k}(0,1)\\cap E^{\\delta,k}(t d_k,r)$ for pairs of ellipsoids whose parameters place them near the higher-order tangency set, and test whether the bound $\\delta^2/(\\delta+t)$ (up to a $\\log\\delta^{-1}$ factor) holds uniformly as $\\delta\\to 0$; a single sequence of configurations violating this bound would falsify Lemma 2.2 and with it the proof of the discretised $L^2$ estimate.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the effective inverse function theorem used to decompose the domain of the polynomial map into $O(1)$ diffeomorphism patches."},{"cited_title":"Pawlucki, Lipschitz cell decomposition in O-minimal structures I , Illinois J","cited_arxiv_id":null,"evidence_quote":"supplies the quantitative Lipschitz cell decomposition that gives the $O(1)$ intrinsic–extrinsic metric equivalence behind the diameter bound."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the $L^2$ duality/summation argument that converts pairwise volume bounds into the multiplicity bound for the discretised maximal operator."},{"cited_title":"$L^p$ bound on the strong spherical maximal function","cited_arxiv_id":"2309.15308","evidence_quote":"provides the $L^p$–Sobolev estimates and the frequency-localisation reduction used to pass from the discretised bound to the full maximal operator."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the degree-based volume bound for tubular neighbourhoods of algebraic varieties used to control fibres in Lemma 3.1."},{"cited_title":"Schlag, A generalization of Bourgain ’s circular maximal theorem , J","cited_arxiv_id":null,"evidence_quote":"provides the kernel-domination device that converts the discretised $L^2$ estimate into an $L^2$–Sobolev bound."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"backs the quantitative semi-algebraic decomposition facts (connected components, complexity bounds) used in the proof of Lemma 3.4."}],"review_version":1}