{"id":"4065e0df-d895-407c-8236-a3d61268996e","arxiv_id":"2506.23206","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Fractional maximal operators map BMO functions into Hausdorff-content BLO spaces, send BMO to VMO via uniform continuity, and preserve vanishing mean oscillation spaces adapted to dyadic Hausdorff content.","lead":"The paper proves new mapping properties of uncentered fractional maximal functions on spaces of functions with bounded and vanishing mean oscillation, generalized through dyadic Hausdorff content. The results sharpen earlier work and provide tools for studying oscillatory behavior of functions on fractal or lower-dimensional sets.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.2's full β range depends on Corollary 3.3, which relies on the capacitary decomposition [4, Thm C] not stated here; a mismatch in its hypotheses would leave only the β<n−α case proven.","rationale":"The reader's conditional verdict is appropriate, and the weakest point is indeed the paper's reliance on external results. I emphasize the [4] dependency because it enters the proof of the headline Theorem 1.2 through Corollary 3.3, whereas the reader's primary accusation targets [19] and Theorem 1.5. The proof of Theorem 3.1 is otherwise detailed and the internal estimates in Section 4 are coherent: Lemma 4.1 is a known published result, the decomposition into g and h is standard, and the cube-selection argument in Lemma 2.8 appears repairable despite a questionable claim about H^n_∞ equalling Lebesgue measure on arbitrary cubes. I found no internal inconsistency that would force rejection. The paper deserves credit for a plausible and genuinely finer formulation, but the unstated dependency on [4] (and, secondarily, [19]) means the central claim is not yet independently verifiable from the manuscript alone. A condition that the authors incorporate or verify the external decomposition theorem is therefore the correct outcome.","tokens_in":34505,"tokens_out":44115,"duration_ms":459349,"concrete_test":"Retrieve arXiv:2501.11412 and check whether [4, Theorem C] and [4, Proposition 2.1] apply verbatim to the function f−essinf^β_{Q′}f with the subordinate dyadic content H^{β,Q′}_∞ used in Theorem 3.1: verify the constants (especially the subadditivity/strong-subadditivity constant in (2.14)) and the covering properties (a)–(d). If any hypothesis fails, attempt a direct dyadic proof of (3.1) without [4]; if (3.1) cannot be recovered, Theorem 1.2's reduction step is unsupported. For Lemma 7.9, also confirm that [19, Theorem 4.15] is stated for δ=β, k=α, and 1<p<β/α; if the range or operator normalization differs, Theorem 1.5 requires a new local estimate.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central theorem is not self-contained at the reduction step: Theorem 1.2 first reduces to 0<β<n−α by invoking Corollary 3.3, whose proof uses the capacitary John–Nirenberg machinery of the companion paper [4, Theorem C and Proposition 2.1] without stating those results. If the decomposition theorem in [4] has hypotheses that are not met in the present setting (e.g., a different dyadic subadditivity constant, an additional measurability/quasicontinuity requirement, or a fixed dyadic grid incompatible with the subordinate content H^{β,Q′}_∞), then the exponential estimate (3.1) fails and Corollary 3.3 collapses. In that case Theorem 1.2 would be established only for β<n−α, losing the advertised full range 0<β≤n. A separate but related gap is Lemma 7.9, where the local oscillation bound invokes [19, Theorem 4.15] for the Euclidean fractional maximal operator on L^p(H^β_∞); since [19] is an unpublished preprint whose statement is not reproduced, the VMO^β preservation result Theorem 1.5 rests on an external estimate that cannot be checked from the present paper alone. The [4] dependency is the more load-bearing for the paper's headline claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the uncentered fractional maximal operator M_α on mean oscillation spaces built from the dyadic Hausdorff content H_∞^β. The main results are: Theorem 1.2, asserting that M_α maps BMO(Q_0) into the finer BLO^β(Q_0) space for 0<β≤n; Theorem 1.3 and Theorem 5.3, giving boundedness of the β_2-dimensional maximal function on BMO^{β_1}; Theorem 1.4, proving uniform continuity of M_α f when f∈BMO(Q_0) with compact support; and Theorem 1.5, asserting preservation of VMO^β under M_α for 0≤α<β. The proofs combine a John–Nirenberg inequality for BLO^β, a capacitary decomposition from the authors' companion work [4], and maximal-function estimates on Choquet spaces, partly taken from the unpublished preprint [19].","tokens_in":34703,"tokens_out":7699,"duration_ms":68848,"significance":"If the results are correct, they improve the Gibara–Kline theorem by placing M_α f not merely in BLO but in the Hausdorff-content-adapted space BLO^β, and they extend the theory to BMO^β, BLO^β, and VMO^β. The BLO^β John–Nirenberg inequality is of independent interest, and the paper contains several careful estimates, notably the local/global decomposition in Sections 6–7. The main caveat is that the proofs rely heavily on external results—the companion paper [4] and the preprint [19]—whose statements are not reproduced, and the α=0 case of Theorem 1.5 is not actually covered by the proof as written.","major_comments":[{"comment":"The reduction in the proof of Theorem 1.2 to the range 0<β<n−α uses Corollary 3.3, whose proof invokes [4, Theorem C] and [4, Proposition 2.1] without stating them. Since [4] is a companion paper that is only listed as 'to appear', the John–Nirenberg-type estimate (3.1) and hence the full range 0<β≤n in Theorem 1.2 cannot be checked from this manuscript alone. Please either state the required capacitary covering and decomposition results explicitly or give a self-contained proof.","section":"§3, proof of Theorem 3.1 and Corollary 3.3"},{"comment":"The bound for the local part of M_α f is the key step in Theorem 1.5, but Lemma 7.9 applies [19, Theorem 4.15] for the fractional maximal operator on Choquet L^p spaces with respect to H_∞^β, with the parameters δ=β, k=α, and 1<p<β/α. This theorem is from an unpublished arXiv preprint and is not stated in the paper, so the scaling and hypotheses behind the estimate O_β(M^loc_{α,λr}f, Q̃) ≤ C λ^{β/p} ℓ(Q_0)^α ω_β(f,3λr) cannot be verified. The preservation of VMO^β therefore rests on an unverifiable external input; please provide the statement or a proof.","section":"§7, Lemma 7.9"},{"comment":"Theorem 1.5 is stated for α∈[0,β), including α=0, but Lemma 7.9 requires a p satisfying 1<p<β/α, which is vacuous when α=0. Remark 1.6 asserts that the α=0 case follows 'by careful observation' without giving the argument. As written, the proof of Theorem 1.5 does not cover α=0. Either remove α=0 from the statement or supply a separate proof for the Hardy–Littlewood maximal operator, for example via the boundedness results in [10].","section":"§7, Theorem 1.5 and Lemma 7.9"}],"minor_comments":[{"comment":"The proof applies the equivalence (6.2) to τ_h f−f, which need not be nonnegative. Please justify this by applying (6.2) to |τ_h f−f| and using Lemma 2.6, or by proving the signed version directly.","section":"§4, proof of Theorem 1.4"},{"comment":"In the statement of Corollary 3.3, 'independent of u' should be 'independent of f'; the variable u is not defined there.","section":"§3, Corollary 3.3"},{"comment":"In the proof of Lemma 7.1, one integral is written as ∫_{Q′} instead of ∫_Q; this is a typographical slip but should be corrected.","section":"§7, Lemma 7.1"},{"comment":"In the sentence 'M_{β_2}φ_j ⊂ L^p(H_∞^{β_1},Q_0)', the symbol ⊂ should be ∈, since the object is a function, not a set.","section":"Appendix A, Theorem A.3"},{"comment":"Theorem 1.1 is credited to 'Kline-Gibara' in the display but to 'Gibara–Kline' elsewhere; please make the attribution consistent.","section":"§1, Theorem 1.1 attribution"}],"recommendation":"major_revision","confidential_remarks":"The central results are plausible and the paper is clearly written, but the two external dependencies ([4] and [19]) are load-bearing and the α=0 case of Theorem 1.5 is a genuine gap in the written proof. These are fixable within the manuscript's scope, so I do not recommend rejection; however, the authors should be asked to state or prove the external results and to either cover α=0 or adjust the statement."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Basak, Chen, and Roychowdhury prove genuinely new results. The headline is that Mα maps BMO(Q0) into BLO^β(Q0), the Hausdorff-content version of bounded lower oscillation, refining Gibara–Kline's classical BLO bound. The companion theorems — boundedness of M^{β2} on BMO^{β1}, uniform continuity of Mα on BMO, and preservation of VMO^β — are also new. The John–Nirenberg inequality for BLO^β in Section 3 is a real addition, and the local-to-global equivalences in Section 2 are handled carefully. The proofs are long but coherent; the local-versus-global decomposition in Section 7 is a sensible engine.\n\nThe soft spots are exactly where the reader's report and stress-test point. Theorem 1.2's advertised full range 0<β≤n rests on Corollary 3.3, which depends on the capacitary decomposition from the authors' companion paper [4, Theorem C and Proposition 2.1]. Those statements are not reproduced, so a referee cannot check whether the hypotheses are satisfied. Lemma 7.9, needed for Theorem 1.5, invokes [19, Theorem 4.15], an unpublished preprint. If either external result has a mismatch, the corresponding theorem does not go through as written. This is not a manufactured concern; both dependencies are load-bearing.\n\nOn the circularity score: I would push back somewhat. The paper does build on the authors' own framework, but the theorems are not corollaries of [4] or [10]. The self-citation is not the problem. The problem is that the external results are unstated. That is also fixable: either state them or have the referee verify them against the sources.\n\nMinor technical issue: the BLO^β definition requires H^β-quasicontinuity, but parts of the proofs work only with local integrability and establish quasicontinuity later. Patchable, but should be cleaned up.\n\nBottom line: this paper deserves a serious referee. I would accept it for peer review, with a conditional verdict. If the external hypotheses check out, the central argument is sound and the contribution is solid within harmonic analysis and capacitary function spaces. The right referee is someone who knows Choquet integrals and maximal operators.","headline":"Genuine refinements over Gibara–Kline, held up by two unstated external dependencies that a referee must verify.","tokens_in":35367,"tokens_out":3470,"would_cite":true,"duration_ms":35098,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46E35","42B35","42B37","32A37","42B25"],"pacs":[],"model":"deepseek-v4-flash","headline":"The uncentered fractional maximal operator $M_\\alpha$ maps BMO into the finer $\\mathrm{BLO}^\\beta$ oscillation spaces built on dyadic Hausdorff content, and preserves the corresponding vanishing-oscillation spaces.","keywords":["fractional maximal function","uncentered maximal operator","dyadic Hausdorff content","Choquet integral","bounded mean oscillation","vanishing mean oscillation","bounded lower oscillation","capacitary function spaces"],"falsifier":"Take $n=1$, $\\alpha=1/2$, $\\beta=1/2$, $Q_0=(0,1)$, and $f(x)=\\chi_{(0,1)}\\log(1/x)$; compute the supremum over cubes $Q\\subset Q_0$ of $\\frac{1}{H^\\beta_\\infty(Q)}\\int_Q (M_\\alpha f - \\mathrm{ess\\,inf}^\\beta_Q M_\\alpha f)\\,dH^\\beta_\\infty$. Theorem 1.2 predicts this is at most $C\\|f\\|_{BMO}$; a cube where the ratio is unbounded as $Q$ shrinks would disprove it. For Theorem 1.5, checking whether the cited Choquet-$L^p$ bound for $M_\\alpha$ truly holds for all $1<p<\\beta/\\alpha$ with the stated constant is the decisive test.","tokens_in":34244,"feed_emoji":"📐","tokens_out":10443,"duration_ms":100503,"temperature":0.7,"pith_summary":"This paper asks where the uncentered fractional maximal operator $M_\\alpha$ sends oscillation functions when the notion of oscillation is adapted to dyadic Hausdorff content. The main theorem establishes that if $0<\\alpha<n$ and $0<\\beta\\le n$, then a compactly supported $f\\in BMO(Q_0)$ satisfies $\\|M_\\alpha f\\|_{\\mathrm{BLO}^\\beta(Q_0)}\\le C\\ell(Q_0)^\\alpha\\|f\\|_{BMO(Q_0)}$, where $\\mathrm{BLO}^\\beta$ is the finer bounded-lower-oscillation scale built on the $\\beta$-dimensional dyadic Hausdorff content. This refines the previously known containment of $M_\\alpha f$ in the classical BLO space. The paper also proves that the $\\beta_2$-dimensional maximal operator maps $BMO^{\\beta_1}$ into $BLO^{\\beta_1}$ for $0<\\beta_1\\le\\beta_2\\le n$, that $M_\\alpha f$ is uniformly continuous for compactly supported BMO data, and that $M_\\alpha$ preserves the vanishing-oscillation spaces $VMO^\\beta$ for $0\\le\\alpha<\\beta$. Why this matters: oscillation spaces control the fine behaviour of rough functions, and showing that fractional maximal operators land in finer scales sharpens the regularity information already encoded in spaces like BMO and VMO.","feed_headline":"Fractional maximal maps gain finer oscillation control","feed_subtitle":"Dyadic Hausdorff content replaces volume: M_α sends BMO into BLO^β and preserves VMO^β.","key_machinery":"The carrier of the argument is the dyadic Hausdorff content $H^\\beta_\\infty$, which measures a set by covering it with dyadic cubes and summing $\\ell(Q)^\\beta$, together with the Choquet integral $\\int f\\,dH^\\beta_\\infty=\\int_0^\\infty H^\\beta_\\infty(\\{f>t\\})\\,dt$. These define the spaces $BMO^\\beta$, $\\mathrm{BLO}^\\beta$, and $VMO^\\beta$, in which the oscillation of $f$ on a cube $Q$ is the $H^\\beta_\\infty$-average of $|f-c|$ over $Q$. The proofs split $M_\\alpha f$ into a local part, controlled through an $L^p(H^\\beta_\\infty)$ bound for $M_\\alpha$ and an exponential-integrability inequality for $BMO^\\beta$, and a global part, controlled by comparing content-averages over nested cubes and using a quantitative oscillation estimate on subsets. A new exponential decay estimate for $\\mathrm{BLO}^\\beta$ is the hinge that yields the inclusion $\\mathrm{BLO}^\\gamma\\subset\\mathrm{BLO}^\\beta$ and the norm bounds.","core_discovery":"The central discovery is that the uncentered fractional maximal operator $M_\\alpha$, whose value at $x$ is the supremum of $\\ell(Q)^{\\alpha-n}\\int_Q|f|$ over all cubes containing $x$, is better behaved on oscillation spaces than previously known. The paper establishes that $M_\\alpha$ maps $BMO(Q_0)$ into $\\mathrm{BLO}^\\beta(Q_0)$ whenever $0<\\alpha<n$ and $0<\\beta\\le n$, with the estimate $\\|M_\\alpha f\\|_{\\mathrm{BLO}^\\beta(Q_0)}\\le C\\ell(Q_0)^\\alpha\\|f\\|_{BMO(Q_0)}$; because $\\mathrm{BLO}^\\beta\\subset \\mathrm{BLO}$, this is a genuine refinement of the classical BLO bound. A second result shows that for $0<\\beta_1\\le\\beta_2\\le n$ the $\\beta_2$-dimensional maximal operator $M^{\\beta_2}$ is bounded from $BMO^{\\beta_1}(\\mathbb{R}^n)$ to $\\mathrm{BLO}^{\\beta_1}(\\mathbb{R}^n)$. The paper further proves that $f\\in BMO$ with compact support implies $M_\\alpha f$ is uniformly continuous and hence in $VMO$, generalizing the earlier VMO-preservation theorem, and that $M_\\alpha$ preserves $VMO^\\beta$ for $\\alpha\\in[0,\\beta)$ under a finite-support condition.","pith_inferences":["Inference: The same local/global decomposition used for Theorem 1.2 should also give boundedness of $M_\\alpha$ from $BMO^\\gamma$ into $\\mathrm{BLO}^\\beta$ for suitable parameter ranges, which the paper does not state.","Inference: Since the argument only uses doubling and strong subadditivity of the dyadic content, it likely transfers to other capacities with those properties, such as general Hausdorff contents on metric spaces.","Inference: The failure at $\\alpha=0$ on hyperplanes suggests the threshold between boundedness and blow-up for $M_\\alpha$ on $BMO^\\beta$ is governed by whether the singular support has positive $H^\\beta_\\infty$ content; testing log-distance functions to lower-dimensional sets would locate the exact boundary.","Inference: Uniform continuity of $M_\\alpha f$ for rough BMO data gives a route to compactness in oscillation spaces that could be useful in PDE arguments where the fractional maximal function appears as a regularizing operator."],"forward_implications":["For any $0<\\alpha<n$, a compactly supported $f\\in BMO(Q_0)$ satisfies $\\|M_\\alpha f\\|_{\\mathrm{BLO}^\\beta(Q_0)}\\le C\\ell(Q_0)^\\alpha\\|f\\|_{BMO(Q_0)}$ for every $0<\\beta\\le n$, so the fractional maximal function has lower oscillation controlled by the BMO norm at every smaller-dimensional scale.","When $0<\\beta_1\\le\\beta_2\\le n$, the operator $M^{\\beta_2}$ is bounded from $BMO^{\\beta_1}(\\mathbb{R}^n)$ to $\\mathrm{BLO}^{\\beta_1}(\\mathbb{R}^n)$, giving a Hausdorff-content analogue of the classical BMO-to-BLO boundedness.","If $f\\in BMO(Q_0)$ has support in $Q_0$, then $M_\\alpha f$ is uniformly continuous on $\\mathbb{R}^n$ and therefore belongs to $VMO(Q_0)$ for $0<\\alpha<n$.","For $0\\le\\alpha<\\beta\\le n$, $M_\\alpha$ preserves $VMO^\\beta(Q_0)$ for compactly supported inputs, including the endpoint $\\alpha=0$.","The exponential-decay estimate for $\\mathrm{BLO}^\\beta$ implies the containment $\\mathrm{BLO}^\\gamma\\subset\\mathrm{BLO}^\\beta$ for $0<\\gamma\\le\\beta\\le n$ with controlled norms."],"supporting_citations":[{"why":"Supplies the prior BMO-to-BLO theorem for $M_\\alpha$ whose bound Theorem 1.2 improves, and the proof scheme the paper adapts.","marker":"[16]"},{"why":"The two-weight maximal inequality behind the $L^p$-to-Choquet bound for $M_\\alpha$ used to control the local part.","marker":"[33]"},{"why":"Defines the $BMO^\\beta$ spaces and provides the exponential-integrability inequality and norm equivalences used throughout.","marker":"[11]"},{"why":"Gives maximal inequalities for the $\\beta$-dimensional operator $M^{\\beta_2}$, used in Theorem 5.3.","marker":"[10]"},{"why":"Cited for the uncentered fractional maximal operator bound on Choquet $L^p$ spaces that Lemma 7.9 depends on.","marker":"[19]"},{"why":"Provides the capacitary covering and exponential-integrability results used in the $\\mathrm{BLO}^\\beta$ exponential-decay proof.","marker":"[4]"},{"why":"Supplies the Riesz-potential bound into BMO and the translation facts used to prove uniform continuity of $M_\\alpha f$.","marker":"[37]"},{"why":"Gives the equivalence between the BMO norm of the Riesz potential and the $L^\\infty$ norm of the fractional maximal function used in Theorem 1.4.","marker":"[1]"},{"why":"Provides a BMO oscillation estimate on subsets used to bound the global part and the $\\alpha=0$ preservation step.","marker":"[35]"}],"fun_headline_variants":["Fractional maximal maps refine BMO into BLO^beta","Dyadic Hausdorff content unlocks finer maximal bounds","M_alpha preserves VMO, maps BMO to BLO^beta","Uncentered maximal operators sharpen oscillation spaces","BLO^beta targets: fractional maximal gains precision"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof that $M_\\alpha$ preserves the fine vanishing-oscillation spaces relies on a boundedness estimate for the uncentered fractional maximal operator on Choquet $L^p$ spaces that is quoted from an unpublished manuscript, not proved here, and if that estimate is wrong the preservation theorem does not go through.","fun_headline_variants_meta":{"raw":{"variants":["Fractional maximal maps refine BMO into BLO^beta","Dyadic Hausdorff content unlocks finer maximal bounds","M_alpha preserves VMO, maps BMO to BLO^beta","Uncentered maximal operators sharpen oscillation spaces","BLO^beta targets: fractional maximal gains precision"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000604,"raw_usage":{"total_tokens":2860,"prompt_tokens":1029,"completion_tokens":1831,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":645,"completion_tokens_details":{"reasoning_tokens":1750}},"tokens_in":645,"tokens_out":1831,"duration_ms":14226,"temperature":1.0,"reasoning_tokens":1750,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T21:51:37.794112+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take $n=1$, $\\alpha=1/2$, $\\beta=1/2$, $Q_0=(0,1)$, and $f(x)=\\chi_{(0,1)}\\log(1/x)$; compute the supremum over cubes $Q\\subset Q_0$ of $\\frac{1}{H^\\beta_\\infty(Q)}\\int_Q (M_\\alpha f - \\mathrm{ess\\,inf}^\\beta_Q M_\\alpha f)\\,dH^\\beta_\\infty$. Theorem 1.2 predicts this is at most $C\\|f\\|_{BMO}$; a cube where the ratio is unbounded as $Q$ shrinks would disprove it. For Theorem 1.5, checking whether the cited Choquet-$L^p$ bound for $M_\\alpha$ truly holds for all $1<p<\\beta/\\alpha$ with the stated constant is the decisive test.","supporting_citations":[{"cited_title":"Gibara and J","cited_arxiv_id":null,"evidence_quote":"Supplies the prior BMO-to-BLO theorem for $M_\\alpha$ whose bound Theorem 1.2 improves, and the proof scheme the paper adapts."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The two-weight maximal inequality behind the $L^p$-to-Choquet bound for $M_\\alpha$ used to control the local part."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the $BMO^\\beta$ spaces and provides the exponential-integrability inequality and norm equivalences used throughout."},{"cited_title":"On Hausdorff content maximal operator and Riesz potential for non-measurable functions","cited_arxiv_id":"2405.12113","evidence_quote":"Cited for the uncentered fractional maximal operator bound on Choquet $L^p$ spaces that Lemma 7.9 depends on."},{"cited_title":"The Capacitary John-Nirenberg Inequality Revisited","cited_arxiv_id":"2501.11412","evidence_quote":"Provides the capacitary covering and exponential-integrability results used in the $\\mathrm{BLO}^\\beta$ exponential-decay proof."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Riesz-potential bound into BMO and the translation facts used to prove uniform continuity of $M_\\alpha f$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the equivalence between the BMO norm of the Riesz potential and the $L^\\infty$ norm of the fractional maximal function used in Theorem 1.4."},{"cited_title":"Shaabani, Maximal operators on BMO and slices , Canad","cited_arxiv_id":null,"evidence_quote":"Provides a BMO oscillation estimate on subsets used to bound the global part and the $\\alpha=0$ preservation step."}],"review_version":1}