{"id":"34361726-2372-4424-8e8b-38945a0be43b","arxiv_id":"2411.14767","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper extends Euclidean BMO commutator characterizations to weighted Morrey spaces on spaces of homogeneous type, but the proofs have major gaps.","lead":"This paper claims to characterize BMO spaces through boundedness of commutators of maximal functions on weighted Morrey spaces over spaces of homogeneous type. The proofs contain several false or unjustified steps, so the results are not supported as written.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 4.1's claimed pointwise BMO bound is false; since the proof of Theorem 1.3 relies on it, the sharp-maximal commutator characterization is unproven.","rationale":"Read in good faith, the paper's central claim is that BMO (with b^-∈L∞) is characterized by boundedness of the commutators of the Hardy-Littlewood, sharp, and fractional maximal functions on weighted Morrey spaces. For Theorem 1.3, the sufficient direction must show that b∈BMO implies [M#,b] bounds on L^{q,κ}_ω. Proposition 4.1 is the only proof of this, and its key step is the pointwise oscillation estimate. That estimate is false for unbounded BMO functions: the BMO norm does not control deviations of b(t) from its ball average, and the example log(1/|x|) shows the supremum in question is infinite. The paper's proof therefore contains a genuine mathematical falsehood at a load-bearing point, not merely a missing constant or a typographical issue. This validates the reader's REJECT verdict. The reader also flags other invalid steps (the localization equality in Proposition 3.1, the weight condition in Lemma 2.3, and the constant value of M#(χ_Q)); those reinforce the same conclusion, but the Proposition 4.1 pointwise bound is the single most load-bearing objection: it is demonstrably false and is required for the sharp-maximal commutator characterization. As submitted, the derivation does not support the theorem's central claim.","tokens_in":14602,"tokens_out":21616,"duration_ms":211050,"concrete_test":"Compute the left side of the claimed estimate in Proposition 4.1 for X=R, b(x)=log(1/|x|), r'=2, B=(-1,1), and t=2^{-N}. The integral (1/2)∫_{-1}^1 |log(1/|y|)-N log 2|^2 dµ(y) grows like (N log 2)^2, so the supremum over N is infinite, whereas ||b||_{BMO} is finite. This directly falsifies the pointwise estimate and invalidates the sufficiency proof of Theorem 1.3. If the estimate held, this computation would have to produce a finite bound independent of N.","verdict_should_be":"REJECT","load_bearing_attack":"The load-bearing step is in the proof of Proposition 4.1: the inequality sup_{B∋t} (1/µ(B) ∫_B |b(y)-b(t)|^{r'} dµ(y))^{1/r'} ≤ C ||b||_{BMO} for every t. This cannot hold for BMO functions that are unbounded at a point. BMO controls oscillation around the ball average b_B, not around the pointwise value b(t). For a concrete instance, take X=R with Lebesgue measure and b(x)=log(1/|x|), a standard BMO function. Fix B=(-1,1) and t=2^{-N}. Then b(t)=N log 2, while the average of b(y) over B is finite and independent of N, so the average of |b(y)-b(t)| over B is at least N log 2 minus a constant and tends to infinity with N. The r'-version is even larger. Thus the supremum is infinite, contradicting Proposition 4.1. This bound is exactly what converts the pointwise commutator expression to M_r(f)(t) in Proposition 4.1 and hence is necessary for the sufficiency direction (i)=>(ii) of Theorem 1.3. With this step false, the claimed BMO characterization via [M#,b] is not established by the manuscript.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper claims characterizations of BMO spaces on spaces of homogeneous type via boundedness of commutators of the Hardy-Littlewood maximal function, the sharp maximal function, and the fractional maximal function on weighted Morrey spaces. Four theorems are stated: Theorem 1.1 for [M_p,b], Theorem 1.2 for the maximal commutator C_b, Theorem 1.3 for [M^#,b], and Theorem 1.4 for [M_γ,b]. The proofs follow a common template: boundedness of the commutator on a weighted Lebesgue space is transferred to a ball-local estimate, which is then converted into a BMO condition by testing on characteristic functions. The paper also proves the reverse implications by deriving BMO bounds and L^∞ bounds for b^- from the assumed Morrey norm inequalities.","tokens_in":14885,"tokens_out":8455,"duration_ms":88305,"significance":"If the results were correct, they would extend a well-known line of Euclidean results (Bastero-Milman-Ruiz, Zhang-Wu) to the general setting of spaces of homogeneous type and weighted Morrey spaces, which is a natural and potentially useful contribution. The paper does clearly place the problem in context and cites the relevant literature. However, the central proofs contain several false assertions that directly affect the main theorems, and at least one condition in Theorem 1.3 is proved in a different form from the one stated. The claimed characterizations therefore are not established by this manuscript. The paper does not contain machine-checked proofs or reproducible code; its main value would be as a genuine extension, but that extension is not demonstrated here.","major_comments":[{"comment":"The identities M_p(b f)(t) = ω(B)^{-κ/q} M_p(b h)(t) and M_p(f)(t) = ω(B)^{-κ/q} M_p(h)(t) for t∈B are false for the global Hardy-Littlewood maximal function, because the supremum on the left is taken over all balls containing t, which may extend outside B, whereas the right-hand side uses the global maximal function of the un-truncated h. Only a local version with balls constrained to B would give such an identity. Since Proposition 3.1 uses these identities to pass from a weighted L^q bound for [M_p,b] to a Morrey bound, the proof of Proposition 3.1, and hence the implication (i)=>(ii) of Theorem 1.1, is invalid as written.","section":"Section 3, Eq. (3.10)"},{"comment":"The proof of Proposition 4.1 contains the step sup_{B∋t} (μ(B)^{-1} ∫_B |b(y)-b(t)|^{r'} dμ(y))^{1/r'} ≤ C ||b||_{BMO}, which is false for general BMO functions. For example, on X=R with Lebesgue measure, b(x)=log(1/|x|) belongs to BMO, but for B=(-1,1) and t=2^{-N}, the average of |b(y)-b(t)| over B grows like N, so the supremum is infinite. BMO controls oscillation around ball averages, not around arbitrary points. This bound is exactly what converts the commutator expression into a multiple of M_r(f)(t), so Proposition 4.1 and the sufficiency part of Theorem 1.3 are not proved.","section":"Section 4, Proposition 4.1"},{"comment":"The argument produces a bound for sup_Q ||(b-2M^#(bχ_Q))χ_Q||, but the stated condition (1.3) involves M^#(b), not M^#(bχ_Q). The identity M^#(χ_Q)=1/2 only yields a bound for |(1/2)b - M^#(bχ_Q)|, and no comparison between M^#(bχ_Q) and M^#(b) is supplied. Since the proof of (iii)=>(i) also uses M^#(bχ_Q), the equivalence in Theorem 1.3 is not established for the condition as stated.","section":"Section 4.1, proof of (ii)=>(iii) of Theorem 1.3"},{"comment":"The equality M_γ(bχ_Q)(x)=M_{γ,Q}(b)(x) for x∈Q is not generally true for the unrestricted fractional maximal function: balls containing x but not contained in Q appear on the left and are not excluded on the right. At best a doubling inequality with an extra constant could hold, but the proof of (ii)=>(iii) of Theorem 1.4 relies on this equality to identify the commutator norm with the expression in (1.4). This gap affects a central implication of Theorem 1.4.","section":"Section 4.2, Eq. (4.14)"}],"minor_comments":[{"comment":"There are numerous typographical errors, including 'homoegenous', 'differentiation', 'charaterization', and inconsistent use of 'space of homogeneous type'; a careful proofreading pass is needed.","section":"Throughout"},{"comment":"In the proof, ||M_p(f)||_{L^{q,κ}_ω(X)} is written as a single ball average, but the Morrey norm is a supremum over balls; the display should use ≤ sup_B or be rewritten.","section":"Lemma 2.3"},{"comment":"The proof uses an auxiliary exponent r with p<r<q, but p is not defined in the statement of Proposition 4.1; specify the range of r explicitly.","section":"Section 4, Proposition 4.1"},{"comment":"The Hölder step introducing ω(t)^{1/q}ω(t)^{-1/q} and the subsequent A_q estimate are correct, but the notation M_{p,B} for the local maximal function is introduced earlier and should be used consistently: in Eq. (3.10) the subscript B is missing, which contributes to the ambiguity in the main proof.","section":"Section 3.1, Eq. (3.11)"}],"recommendation":"reject","confidential_remarks":"The manuscript shows signs of being an early draft with many typographical errors and a heavy reliance on a single template proof. The false assertion in Proposition 4.1 about pointwise BMO oscillation is not a small gap: it is a fundamental misunderstanding of the BMO condition. Even if the main theorems are true and provable by other means, the submitted proofs are not correctable by local revisions. I also note that reference [5] is a self-citation of the author's own prior work, which is used as a black box; it would be helpful to specify exactly which result from [5] is being invoked."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here is the short version. The paper states the natural weighted Morrey analogues of the Bastero–Milman–Ruiz and Zhang–Wu commutator characterizations on spaces of homogeneous type. That target is legitimate, but the proofs as written rest on several false claims, so the theorems are not established.\n\nThe specific characterizations of BMO via [M_p,b], [M^#,b], and [M_γ,b] on weighted Morrey spaces over spaces of homogeneous type do not appear in the cited literature, so the statement of the results is a genuine contribution, if only a modest one. The paper is organized, it engages the prior work, and the overall strategy—reduce Morrey estimates to weighted L^p estimates through localization, then use known BMO bounds—is the standard approach in this area.\n\nThe soft spots are serious. Proposition 3.1 uses the localization identity (3.10): for f = hχ_B/ω(B)^{κ/q}, it claims M_p(bf)(t)=ω(B)^{-κ/q}M_p(bh)(t) for t∈B. That is not an identity: the global maximal function at t sees balls that extend outside B, and b h χ_B has extra zeros, so the two sides differ. This breaks the sufficiency direction of Theorem 1.1.\n\nProposition 4.1 is worse. It uses sup_{B∋t} (1/μ(B)∫_B |b(y)-b(t)|^{r'} dμ)^{1/r'} ≤ C||b||_{BMO}, pointwise in t. That is false for unbounded BMO functions; BMO controls oscillation around ball averages, not around arbitrary points. The stress-test example b(x)=log(1/|x|) on R is correct, and it shows the supremum is infinite. Since this bound is what converts the sharp-maximal commutator estimate into the Morrey estimate, Theorem 1.3 is unproven.\n\nLemma 2.3 has a separate gap: it applies Theorem 2.2 at exponent q/p while only assuming ω∈A_p, but the theorem needs the weight in A_{q/p}, which does not follow from A_p. The lemma might be true under stronger hypotheses, but it is not supported as stated. And in the proof of Theorem 1.3, M^#(χ_Q)=1/2 is used as an exact equality (it is at best an upper bound), and M^#(bχ_Q) is slipped in where condition (iii) has M^#(b). These are not cosmetic.\n\nIn sum, the paper has a plausible target and a sensible outline, but the load-bearing steps do not hold. I would send it to a referee who knows weighted Morrey spaces, because a fix may be nearby, but I would not cite it and I would not make it required reading. If the choice is mine, the verdict is reject with a clear roadmap: fix the localization, fix the BMO pointwise bound, and repair Lemma 2.3.","headline":"Plausible extension with genuine new statements, but the proofs contain several load-bearing false inequalities; needs major revision.","tokens_in":15341,"tokens_out":7554,"would_cite":false,"duration_ms":68154,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["42B25","43A85"],"pacs":[],"model":"deepseek-v4-flash","headline":"Bounded commutators of maximal functions characterize BMO functions on weighted Morrey spaces over spaces of homogeneous type.","keywords":["maximal function commutator","BMO","weighted Morrey space","space of homogeneous type","sharp maximal function","fractional maximal function","Muckenhoupt weight","doubling measure"],"falsifier":"Set $X=\\mathbb{R}$ with Lebesgue measure and the trivial weight, take the nonnegative BMO function $b(x)=|\\log|x||$, and fix the ball $B=(-1,1)$. In the pointwise estimate used in Proposition 4.1, let $t=\\varepsilon$ tend to $0$; then the quantity $(1/|B|\\int_B |b(y)-b(t)|^{r'}\\,dy)^{1/r'}$ grows like $\\log(1/\\varepsilon)$, while the BMO norm of $b$ stays finite. That directly disproves the estimate on which the sufficiency proof of Theorem 1.3 rests.","tokens_in":14353,"feed_emoji":"📐","tokens_out":22020,"duration_ms":194237,"temperature":0.7,"pith_summary":"This paper tries to establish that, on spaces of homogeneous type, the boundedness of maximal-function commutators is exactly what separates BMO functions from more general locally integrable functions. Concretely, it claims that a real-valued locally integrable function $b$ is in BMO with bounded negative part $b^{-}$ if and only if the commutator of the Hardy-Littlewood maximal function (and likewise of the sharp maximal and fractional maximal functions) is bounded on the appropriate weighted Morrey spaces. The equivalence is proved by showing that commutator boundedness is equivalent to a uniform ballwise oscillation condition, a Morrey-space analogue of the classical characterization of BMO. If the claim is right, it extends a Euclidean theory of maximal commutators to metric measure spaces with doubling measures and gives a nonlinear operator-theoretic test for BMO.","feed_headline":"Maximal-function commutators characterize BMO on Morrey spaces","feed_subtitle":"On metric spaces with doubling measures, commutator boundedness becomes a ball-by-ball oscillation test.","key_machinery":"The central object is the commutator $[T,b]f = bTf - T(bf)$, where $T$ is the Hardy-Littlewood maximal function $M_p$, the sharp maximal function $M^{\\sharp}$, or the fractional maximal function $M_\\gamma$. These operators are studied on weighted Morrey spaces, whose norm is a supremum over balls of a weighted $L^p$ average divided by a power of the weight of the ball. The proofs are carried by three mechanisms: decomposition of $b$ into positive and negative parts with an inequality that controls the difference between $[T,b]$ and $[T,|b|]$; known pointwise domination bounds for maximal commutators; and testing the bounded commutator on characteristic functions, which forces the local oscillation condition. Muckenhoupt $A_p$ and $A_{p,q}$ assumptions on the weight make the weighted Hölder estimates uniform in the ball, so the oscillation condition implies BMO.","core_discovery":"The central claim is an if-and-only-if characterization. For weights in the appropriate Muckenhoupt classes, the following are equivalent for a real-valued locally integrable $b$ on $(X,d,\\mu)$: $b$ belongs to BMO$(X)$ and $b^{-}$ is bounded; the commutator $[M_p,b]$ is bounded on the weighted Morrey space $L^{q,\\kappa}_{\\omega}(X)$ for $p<q$; and the ballwise quantity $\\sup_B \\|(b-M_{p,B}b)\\chi_B\\|_{L^{q,\\kappa}_{\\omega}(X)}/\\|\\chi_B\\|_{L^{q,\\kappa}_{\\omega}(X)}$ is finite. The paper states the same three-way equivalence for the sharp maximal commutator on $L^{q,\\kappa}_{\\omega}(X)$ with $\\omega\\in A_1$, and for the fractional maximal commutator as a bounded map between $L^{p,\\kappa}_{(\\omega^p,\\omega^q)}(X)$ and $L^{q,\\kappa q/p}_{\\omega^q}(X)$, with $1/q=1/p-\\gamma$. In each case, testing the commutator on characteristic functions yields the oscillation condition, and Muckenhoupt weight conditions convert weighted Hölder integrals back into the unweighted BMO norm.","pith_inferences":["A natural extension would be to test whether the real-valued assumption can be dropped; the proofs split $b$ into its positive and negative parts, so a complex-valued version would need a different decomposition and is not an immediate corollary.","The same test-on-characteristic-functions scheme could be applied to dyadic maximal operators on spaces of homogeneous type; one would then expect dyadic BMO versions of these equivalences.","The ballwise oscillation condition is computable from local averages, which suggests it could serve as a numerical indicator for whether a function belongs to BMO on an arbitrary metric measure space.","A further question is whether, in the fractional case, the explicit assumption of a bounded negative part can be relaxed to a condition tied to the weight classes, since that assumption enters through a pointwise domination inequality."],"forward_implications":["On any space of homogeneous type with a doubling measure, boundedness of the Hardy-Littlewood maximal commutator on the weighted Morrey space forces the symbol to be in BMO with bounded negative part, so the operator cannot be bounded for merely locally integrable functions.","The maximal commutator theorem gives a converse: the maximal commutator operator is bounded on the weighted Morrey space precisely for BMO symbols, making boundedness of a nonlinear operator a membership test for BMO.","The fractional maximal result pins down the exact pair of weighted Morrey spaces, with the exponent on the target space determined by $1/q=1/p-\\gamma$, between which a bounded fractional maximal commutator can act.","All equivalences pass through a uniform ballwise oscillation condition, so membership in BMO is checked locally, without Fourier analysis or translation invariance."],"supporting_citations":[{"why":"Defines weighted Morrey spaces and supplies the Hardy-Littlewood and fractional maximal operator bounds on them that the proofs invoke as Lemmas 2.2 and 2.6.","marker":"[4]"},{"why":"Gives the Euclidean $L^p$ characterization of the commutators with the maximal and sharp maximal functions that this paper extends to weighted Morrey spaces on spaces of homogeneous type.","marker":"[2]"},{"why":"Supplies the pointwise estimate $C_b f \\leq C\\|b\\|_{BMO} M_2 f$ used to prove the maximal commutator theorem.","marker":"[1]"},{"why":"Provides the two-weight commutator bound for $[b,M]$ on spaces of homogeneous type, adapted through the identity $M_p f=(M(|f|^p))^{1/p}$.","marker":"[5]"},{"why":"Gives the pointwise inequality used in the fractional commutator proof that controls the commutator by the maximal commutator plus a term involving the negative part.","marker":"[7]"},{"why":"Gives the domination of the fractional maximal commutator by a sum of compositions of maximal operators, used to transfer boundedness from maximal operators to commutators.","marker":"[35]"},{"why":"Defines the $A_{p,q}$ weight classes and the weight properties used to pass from the Morrey oscillation condition to BMO.","marker":"[8]"},{"why":"Provides the Euclidean fractional maximal commutator result that the paper reformulates in the weighted Morrey setting.","marker":"[12]"}],"fun_headline_variants":["Maximal commutators on weighted Morrey spaces characterize BMO","Commutator boundedness on Morrey spaces is equivalent to BMO","BMO characterized by maximal commutators on Morrey spaces","Testing commutators on balls gives BMO on Morrey spaces","Weighted Morrey commutators characterize BMO"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The sufficiency proof for the sharp-maximal commutator assumes that a BMO function's oscillation around any single point inside a ball is controlled uniformly by the BMO norm, which is false for unbounded BMO functions because BMO only controls oscillation around the ball's average.","fun_headline_variants_meta":{"raw":{"variants":["Maximal commutators on weighted Morrey spaces characterize BMO","Commutator boundedness on Morrey spaces is equivalent to BMO","BMO characterized by maximal commutators on Morrey spaces","Testing commutators on balls gives BMO on Morrey spaces","Weighted Morrey commutators characterize BMO"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001488,"raw_usage":{"total_tokens":5924,"prompt_tokens":842,"completion_tokens":5082,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":458,"completion_tokens_details":{"reasoning_tokens":4998}},"tokens_in":458,"tokens_out":5082,"duration_ms":34611,"temperature":1.0,"reasoning_tokens":4998,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:56:15.244654+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Set $X=\\mathbb{R}$ with Lebesgue measure and the trivial weight, take the nonnegative BMO function $b(x)=|\\log|x||$, and fix the ball $B=(-1,1)$. In the pointwise estimate used in Proposition 4.1, let $t=\\varepsilon$ tend to $0$; then the quantity $(1/|B|\\int_B |b(y)-b(t)|^{r'}\\,dy)^{1/r'}$ grows like $\\log(1/\\varepsilon)$, while the BMO norm of $b$ stays finite. That directly disproves the estimate on which the sufficiency proof of Theorem 1.3 rests.","supporting_citations":[{"cited_title":"Komori and S","cited_arxiv_id":null,"evidence_quote":"Defines weighted Morrey spaces and supplies the Hardy-Littlewood and fractional maximal operator bounds on them that the proofs invoke as Lemmas 2.2 and 2.6."},{"cited_title":"Bastero, M","cited_arxiv_id":null,"evidence_quote":"Gives the Euclidean $L^p$ characterization of the commutators with the maximal and sharp maximal functions that this paper extends to weighted Morrey spaces on spaces of homogeneous type."},{"cited_title":"Agcayazi, A","cited_arxiv_id":null,"evidence_quote":"Supplies the pointwise estimate $C_b f \\leq C\\|b\\|_{BMO} M_2 f$ used to prove the maximal commutator theorem."},{"cited_title":"A note on two weight commutators of maximal functions on spaces of homogeneous type","cited_arxiv_id":"2012.00575","evidence_quote":"Provides the two-weight commutator bound for $[b,M]$ on spaces of homogeneous type, adapted through the identity $M_p f=(M(|f|^p))^{1/p}$."},{"cited_title":"Zhang, J","cited_arxiv_id":null,"evidence_quote":"Gives the pointwise inequality used in the fractional commutator proof that controls the commutator by the maximal commutator plus a term involving the negative part."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the domination of the fractional maximal commutator by a sum of compositions of maximal operators, used to transfer boundedness from maximal operators to commutators."},{"cited_title":"Muckenhoupt and R","cited_arxiv_id":null,"evidence_quote":"Defines the $A_{p,q}$ weight classes and the weight properties used to pass from the Morrey oscillation condition to BMO."},{"cited_title":"Zhang, J","cited_arxiv_id":null,"evidence_quote":"Provides the Euclidean fractional maximal commutator result that the paper reformulates in the weighted Morrey setting."}],"review_version":1}