{"id":"603d460d-6f36-4ac4-9da9-97498b2fcd21","arxiv_id":"2501.11412","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A packing condition is shown to characterize when an outer capacity is equivalent to its induced Hausdorff content, yielding maximal estimates and John-Nirenberg inequalities for such capacities.","lead":"This mathematics paper extends a famous oscillation inequality, the John-Nirenberg estimate, from ordinary volume to a broad family of geometric set measures called Hausdorff contents, and then to any abstract capacity that satisfies a newly identified packing property. The key result characterizes exactly which capacities behave like Hausdorff contents, so the classical machinery of maximal functions and exponential decay bounds can be transported to them.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1's advertised if-and-only-if is only proved in one direction; the converse (equivalence to H_C^infty implies (P)) is absent, leaving the central characterization overclaimed.","rationale":"Load-bearing concern identified: the paper's principal advertised theorem, Theorem 1, is an equivalence but only one implication is proved. The reader's rationale already notes this gap, though the formal 'weakest_assumption' field emphasizes the doubling condition; my assessment therefore agrees partially. The missing converse is directly about the central claim, whereas doubling is an assumption governing the ball-based extensions. Since the applications only use (P) implies equivalence, the mathematical program remains salvageable, and the honest fix is either to prove the converse (likely true via a Proposition 2.8 type argument) or to weaken Theorem 1 to a one-way statement. The forward proof also has a minor unaddressed point: applying (P) to an infinite collection requires an exhaustion by finite cube families, and the constant in (6.4) should depend on the unspecified A0 from (P) rather than the number 2; both are repairable without changing the qualitative results. A CONDITIONAL verdict is appropriate: the authors must supply the missing converse or explicitly restate the theorem.","tokens_in":25082,"tokens_out":23408,"duration_ms":214973,"concrete_test":"Prove or disprove (ii) implies (i): assume (1/4) H_C^infty(E) <= C(E) <= H_C^infty(E) for all E and show, following the template of Proposition 2.8, that C satisfies (P) with a constant depending only on the equivalence constant. Verify explicitly that the equivalence yields the required hypothesis sum H_C^infty(Q_j) <= A0' H_C^infty(Q') for the relevant non-overlapping collections and that Choquet integrals with respect to C and H_C^infty are comparable without extra assumptions. If the derivation requires doubling or outer regularity of H_C^infty that is not established, Theorem 1's stated iff fails.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Section 6 proves only (i) implies (ii) of Theorem 1; the converse, (ii) implies (i), is neither stated nor proved, yet the abstract and Theorem 1 advertise an 'if and only if' characterization. This is the central new claim of the paper. The applications Theorems A' through D' rely only on the forward direction, so they can be salvaged, but Theorem 1 as written is unsupported. A secondary gap in the forward proof: the selected subfamily {Q_jk} from Proposition 2.1 may be infinite and not contained in any single dyadic cube, so applying (P) requires an exhaustion argument that is omitted; and the constant 2 in (6.4) presumes the (P)-constant A0 equals 2, whereas (P) only supplies an unspecified A0 >= 1, so the final (1/4) bound should in general read 1/(2A0) unless the argument is reworked.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops maximal function estimates, Lebesgue differentiation, Calderón–Zygmund decompositions, and a John–Nirenberg inequality for translation-invariant Hausdorff contents and then extends these results to general outer capacities satisfying a new packing condition (P) together with a doubling condition. The central structural claim is Theorem 1, which asserts that an outer capacity satisfies (P) if and only if it is equivalent to its induced Hausdorff content. The applications Theorems A', B', C', and D' rely on the forward direction of this equivalence plus doubling, and the paper also contains dyadic versions of the maximal and differentiation theorems for general outer measures satisfying (P).","tokens_in":25076,"tokens_out":8975,"duration_ms":84570,"significance":"If the main results are correct, the paper identifies a clean, almost necessary condition on an outer capacity that lets a substantial part of the classical harmonic-analysis package be developed for nonlinear capacitary integrals. The dyadic theory for general outer measures satisfying (P), and the identification of (P) as a useful structural hypothesis, are genuine contributions. However, the advertised equivalence in Theorem 1 is only partially proved, and the proof of the forward direction contains a gap in the application of (P). These issues undermine the central characterization as stated, though they appear to be repairable within the manuscript's framework.","major_comments":[{"comment":"The proof establishes only the implication (i) ⇒ (ii). After deriving C(E) ≤ H^C_∞(E), the rest of the proof assumes (P) and proves the lower bound (1/4)H^C_∞(E) ≤ C(E). There is no argument that the equivalence in (ii) implies the packing condition (P). Yet the abstract and Theorem 1 state an 'if and only if'. The converse direction is a load-bearing part of the paper's central claim and is missing.","section":"Section 6, Proof of Theorem 1"},{"comment":"The application of the packing assumption to the subfamily {Q_{j_k}} is not justified. The subfamily may be infinite and need not be contained in any single dyadic cube Q' ∈ D(Q), whereas (P) is formulated relative to a fixed cube Q'. Moreover, the constant 2 in (6.4) presumes A0 = 2, but (P) only supplies an unspecified constant A0 ≥ 1. An exhaustion argument and a careful tracking of A0 are needed; as written, the lower bound (1/4)H^C_∞(E) is not established.","section":"Section 6, equation (6.4)"},{"comment":"The chain bounding C(U) from below via ∑_k C(Q_{j_k}) and ∑_k H^C_∞(Q_{j_k}) uses the packing inequality in a form that also requires a justification for the ancestors Q̃_k. In particular, the step 'C(U) ≥ C(∪_k Q_{j_k}) ≥ (1/2)∑_k C(Q_{j_k})' relies on (P) with constant 2, but (P) is not verified for the infinite family and does not yield the constant 2. This gap affects the proof of the forward direction of Theorem 1 and hence also the reductions used in Theorems A', B', C', and D'.","section":"Section 6, proof of Theorem 1, lower bound"}],"minor_comments":[{"comment":"There are several typographical errors, including 'Capacit ary' in the title, 'Le besgue' in the abstract, and 'Calder ´ on' in the introduction.","section":"Title and Abstract"},{"comment":"The selection process for the subfamily {Q_{j_k}} and the maximal family of ancestors {Q̃_k} is described informally and would benefit from a more rigorous treatment, especially for infinite families and for the possibility of empty or non-unique maximal families.","section":"Section 2, Proposition 2.1"},{"comment":"The phrase 'we may assume that {Q_i} is maximally disjoint' is not formal; please specify the selection procedure and justify the inclusion (6.5) after this selection.","section":"Section 6, Lemma 6.4"},{"comment":"The sentence 'Therefore, Theorems A', B', C', and D' are applicable to H^C_∞' is confusing because the theorems are stated for the outer capacity C, not for its induced content; please clarify the intended application.","section":"Example 1.5"},{"comment":"The symbol D is used both for the constant in Theorem C' and for the constant in (1.10) in Remark 1.1; this notational overlap should be resolved to avoid confusion.","section":"Remark 1.1 and Theorem C'"}],"recommendation":"major_revision","confidential_remarks":"The paper has a promising core idea and the dyadic results for outer measures satisfying (P) appear sound, but the central Theorem 1 is overclaimed as an if-and-only-if and the forward proof contains a gap in the use of (P). The applications A'-D' rely only on the forward direction, so a corrected proof of that direction plus either a proof of the converse or a reformulation of Theorem 1 would put the paper in good shape. I do not see a reason to doubt the overall approach, and I would encourage the authors to address these points carefully."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The real news here is the packing condition (P) and the observation that it substitutes for linearity in the capacitary integrals. The dyadic maximal, differentiation, Calderón–Zygmund, and John–Nirenberg results for general translation-invariant Hausdorff contents H^φ_∞ are natural extensions of the β-dimensional theory, and the forward direction of Theorem 1—(P) implies equivalence with the induced Hausdorff content—is the correct structural step. The applications Theorems A'–D' follow from that forward direction. That part deserves credit.\n\nThe soft spots are all in Theorem 1 as advertised. The paper states an if-and-only-if, but Section 6 proves only (i)⇒(ii). The converse, that equivalence to H^C_∞ forces the packing condition, is neither stated nor proved. Since the applications only use the forward direction, the main body is probably salvageable, but the central characterization is overclaimed as written.\n\nTwo additional gaps sit inside the forward proof. First, the subfamily {Q_{j_k}} produced by Proposition 2.1 is not contained in any single dyadic cube, while the packing condition (P) as stated applies to a collection subordinate to one Q'. The proof applies (P) directly to the whole infinite family; an exhaustion argument is missing. Second, the step labeled (6.4) writes ∑ C(Q_{j_k}) ≤ 2 C(∪ Q_{j_k}), which assumes the packing constant A0 equals 2. The packing condition only supplies some A0 ≥ 1, so the conclusion should read 1/(2A0) H^C_∞(E) ≤ C(E) unless the argument is reworked. Both look repairable, but they are real gaps in the written proof.\n\nSmaller things: the proof of Theorem a contains a garbled sentence about omitting cubes Q_{j_k} contained in some \tilde{Q}_m, which the reader must reverse-engineer; the constant bookkeeping elsewhere is sloppy. The reliance on the authors' earlier papers for the interpolation lemma, BMO lemmas, and density is not a problem in itself, though it makes the paper less self-contained.\n\nWho should read this: anyone working with Hausdorff contents, capacitary maximal functions, or John–Nirenberg inequalities in nonlinear settings. The packing condition is likely to be taken up by others. The paper deserves a serious referee, but the referee should insist on fixing the converse and the two proof gaps before publication.","headline":"The packing condition is a genuinely useful idea and the dyadic theory for general translation-invariant contents is mostly solid, but Theorem 1's advertised if-and-only-if is only proved in one direction and the proof has a few fixable gaps.","tokens_in":25788,"tokens_out":2902,"would_cite":true,"duration_ms":28466,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46E35","42B35","42B37"],"pacs":[],"model":"deepseek-v4-flash","headline":"An outer capacity satisfies the packing condition exactly when it is equivalent, up to a factor of four, to the Hausdorff content it induces — and this is what lets the whole harmonic-analysis toolbox carry over.","keywords":["capacitary maximal function","Hausdorff content","Choquet integral","John-Nirenberg inequality","packing condition","outer capacity","Lebesgue differentiation","Calderón-Zygmund decomposition"],"falsifier":"Search for an outer capacity $C$ satisfying (P) for which some set $E$ has $C(E) < \\frac14 H^C_\\infty(E)$, or an outer capacity with $\\frac14 H^C_\\infty(E) \\le C(E) \\le H^C_\\infty(E)$ for all $E$ that violates the quasi-additivity in (P); Theorem 1 predicts neither can exist. A concrete place to look is the Sobolev capacity $\\mathrm{Cap}_{1,p}$ on the Cantor-type set of [7, Section 5.3], where the equivalence is known to fail, so (P) must fail there as well.","tokens_in":24696,"feed_emoji":"📐","tokens_out":13040,"duration_ms":111425,"temperature":0.7,"pith_summary":"This paper asks which outer capacities can support the standard harmonic-analysis toolbox: maximal-function bounds, Lebesgue differentiation, Calderón–Zygmund decompositions, and the exponential John–Nirenberg inequality. The authors isolate one structural condition, the packing assumption, that compensates for the nonlinearity of Choquet integrals against a capacity. Their main theorem states that an outer capacity satisfies the packing assumption if and only if it is comparable, up to the universal factor $1/4$, to the Hausdorff content it induces. With the same condition plus a doubling hypothesis, they extend the full toolbox to arbitrary outer capacities. If correct, the packing condition is exactly the substitute for linearity that makes the nonlinear capacitary theory run.","feed_headline":"Capacity equals Hausdorff content under one packing rule","feed_subtitle":"A single quasi-additivity condition unlocks maximal estimates, differentiation, and John–Nirenberg for nonlinear capacities.","key_machinery":"The load-bearing object is the packing assumption (P): for a non-overlapping dyadic subfamily $\\{Q_j\\}$ of $Q'$ with $\\sum_j C(Q_j) \\le A_0 C(Q')$, the Choquet integrals obey $\\sum_j \\int_{Q_j} f\\, dC \\le A_0 \\int_{\\cup_j Q_j} f\\, dC$. This quasi-additivity is what replaces ordinary linearity of the integral in every proof. The dyadic packing lemma (Proposition 2.1) selects from any covering family a subfamily and disjoint ancestors with controlled $\\lambda$-mass; applied to $\\lambda = H^C_\\infty$, it yields exactly the inequality behind the maximal estimates, the decomposition, and the exponential decay. Theorem 1 identifies (P) with the two-sided comparison between $C$ and its induced content, which is what allows the centered ball-based theory to be reduced to the dyadic one.","core_discovery":"For an outer capacity $C$, define the induced Hausdorff content by $H^C_\\infty(E) := \\inf\\{\\sum_i C(Q_i) : E \\subset \\bigcup_i Q_i,\\ Q_i \\in \\mathcal{D}(Q)\\}$. Theorem 1 is the equivalence: $C$ satisfies the packing assumption (P) if and only if $\\frac14 H^C_\\infty(E) \\le C(E) \\le H^C_\\infty(E)$ for every $E \\subset \\mathbb{R}^n$. In the paper's reading, this says that, up to the universal factor $4$, the only outer capacities that pack are those already equivalent to the Hausdorff content they generate. Assuming the doubling condition (1.7), the authors then prove capacitary maximal inequalities, Lebesgue differentiation, a Calderón–Zygmund decomposition, and an exponential John–Nirenberg inequality for general outer capacities (Theorems A$'$, B$'$, C$'$, D$'$), reducing each statement to the dyadic theory of translation-invariant contents $H^\\phi_\\infty$ built from a monotone gauge $\\phi$.","pith_inferences":["Inference: the factor $1/4$ in Theorem 1 comes from the dyadic Whitney argument, and the paper leaves open whether it is optimal; calibrating it on simple self-similar sets would measure the slack in the reduction to dyadic contents.","Inference: the known failure of (P) for Sobolev and Riesz capacities now has a quantitative meaning — their nonlocal character forces the induced content to be strictly smaller, and Theorem 1 says this is the same obstruction as the failure of dyadic quasi-additivity.","Inference: since the dyadic package needs no doubling, a natural testable extension is to replace global doubling by the weaker parent-cube bound (1.10) in the centered theorems; Example 1.6 suggests that the centered results may genuinely require doubling.","Inference: the exponential decay is a step toward a capacitary $H^1$–BMO duality in which $M^\\#_C$ plays the role of the BMO gauge; if such duality holds, it would give a new description of the dual of $L^1(\\mathbb{R}^n; C)$."],"forward_implications":["Every outer capacity satisfying (P) and doubling admits a weak-type $(1,1)$ and strong-type $(p,p)$ bound for the capacitary maximal operator $M_C$.","For such capacities, Lebesgue differentiation holds: $\\frac{1}{C(B(x,r))}\\int_{B(x,r)} |f-f(x)|\\, dC \\to 0$ at $C$-quasi every point for $f \\in L^1(\\mathbb{R}^n; C)$.","A Calderón–Zygmund decomposition exists for doubling (P)-capacities, with the standard two-sided control of average heights over the selected dyadic cubes.","Functions of bounded $C$-mean oscillation satisfy an exponential John–Nirenberg estimate $C(\\{x\\in Q': |f(x)-c_{Q'}|>t\\}) \\le A C(Q') e^{-a t/\\|f\\|_{BMO_C(Q_0)}}$.","The dyadic theorems for translation-invariant contents $H^\\phi_\\infty$ hold without any doubling assumption; Example 1.6 shows they cover gauges that are only locally doubling."],"supporting_citations":[{"why":"Establishes capacitary maximal inequalities and differentiation for spherical Hausdorff contents and supplies the interpolation lemma used to pass from weak to strong type.","marker":"[10]"},{"why":"Proves the John–Nirenberg inequality for $\\beta$-dimensional Hausdorff content and contributes the exponential-decay lemma and sharp maximal function used in Theorems D and D'.","marker":"[12]"},{"why":"Introduces the packing argument for Choquet integrals and Hausdorff content that Proposition 2.1 generalizes.","marker":"[30]"},{"why":"Gives the abstract dyadic-cube construction and strong subadditivity for Hausdorff contents induced by monotone set functions.","marker":"[34]"},{"why":"Provides the Banach space structure and density of continuous functions for $L^1(H^\\lambda)$, used in the differentiation theorems.","marker":"[31]"},{"why":"Supplies the Whitney decomposition used in the proof of Theorem 1 to reduce covers by open sets to dyadic cubes.","marker":"[21]"},{"why":"Records the Cantor-set construction showing Sobolev and Riesz capacities are not equivalent to their induced contents, the canonical family where (P) fails.","marker":"[7]"}],"fun_headline_variants":["Packing condition: capacity is 4-comparable to Hausdorff content","For outer capacities, packing iff Hausdorff-content equivalence","One packing rule: capacity matches Hausdorff content up to 4","Packing condition links capacity to its Hausdorff content"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The full ball-based theory depends on the capacity being doubling in the sense of (1.7); without doubling, the dyadic theorems still hold, but the centered versions for balls and the parent-cube control in the Calderón–Zygmund decomposition are not established, and Example 1.6 stops exactly at that boundary.","fun_headline_variants_meta":{"raw":{"variants":["Packing condition: capacity is 4-comparable to Hausdorff content","For outer capacities, packing iff Hausdorff-content equivalence","One packing rule: capacity matches Hausdorff content up to 4","Packing condition links capacity to its Hausdorff content"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000757,"raw_usage":{"total_tokens":3329,"prompt_tokens":877,"completion_tokens":2452,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":493,"completion_tokens_details":{"reasoning_tokens":2378}},"tokens_in":493,"tokens_out":2452,"duration_ms":21842,"temperature":1.0,"reasoning_tokens":2378,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T18:19:31.591251+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search for an outer capacity $C$ satisfying (P) for which some set $E$ has $C(E) < \\frac14 H^C_\\infty(E)$, or an outer capacity with $\\frac14 H^C_\\infty(E) \\le C(E) \\le H^C_\\infty(E)$ for all $E$ that violates the quasi-additivity in (P); Theorem 1 predicts neither can exist. A concrete place to look is the Sobolev capacity $\\mathrm{Cap}_{1,p}$ on the Cantor-type set of [7, Section 5.3], where the equivalence is known to fail, so (P) must fail there as well.","supporting_citations":[{"cited_title":"↑1, 3, 6, 7, 19, 20 28 R","cited_arxiv_id":null,"evidence_quote":"Proves the John–Nirenberg inequality for $\\beta$-dimensional Hausdorff content and contributes the exponential-decay lemma and sharp maximal function used in Theorems D and D'."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Banach space structure and density of continuous functions for $L^1(H^\\lambda)$, used in the differentiation theorems."},{"cited_title":"Grafakos, Classical Fourier analysis , 3rd ed., Graduate Texts in Mathematics, vol","cited_arxiv_id":null,"evidence_quote":"Supplies the Whitney decomposition used in the proof of Theorem 1 to reduce covers by open sets to dyadic cubes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Records the Cantor-set construction showing Sobolev and Riesz capacities are not equivalent to their induced contents, the canonical family where (P) fails."}],"review_version":1}