{"id":"b01dc809-a399-42ac-9906-3fb8bf30ecc1","arxiv_id":"2507.09154","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"On weighted Bergman spaces, a uniform bound on L^m norms of S_z1 implies boundedness, and vanishing Berezin transform implies compactness.","lead":"This paper gives conditions on a linear operator's action on normalized Bergman kernels that guarantee the operator is bounded or compact. The compactness result removes an adjoint condition from earlier work at the cost of a larger integrability exponent.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 4's compactness proof uses Lemma 9 to pass from e_S(z)→0 to ||S_z1||_{m,α}→0, but Lemma 9 gives convergence only for t<m, and boundedness of the m-norm does not force its decay.","rationale":"The paper's central new claim is Theorem B / Theorem 4: under a one-sided kernel bound on ||S_z1||_{m,α}, compactness of S is equivalent to vanishing of the Berezin transform. For this to hold, the proof must show that e_S(z)→0 forces the decay of the m-norm of S_z1. The proof does not show this: Lemma 9 explicitly gives only t<m convergence. This is not a stylistic issue or a disagreement with the existing consensus; it is an internal gap in the central argument. The boundedness half (Theorem 1) is a substantial, parameterized estimate and is not the place where the argument breaks. The missing piece is load-bearing because the final finite-rank truncation in Theorem 4 is the only mechanism connecting Berezin decay to compactness. I agree with the reader's identification of this weak point and with the conditional verdict: the manuscript should not be accepted as a complete proof of Theorem B until the m-norm decay is either proved or replaced by an explicitly weaker conclusion. No ad hominem is intended; the concern is purely about the logical chain in Section 3.","tokens_in":12179,"tokens_out":5608,"duration_ms":78108,"concrete_test":"Independently check the exact statement of Lemma 5.3 in Miao–Zheng [8] and its weighted analogue: does it contain t=m or only t<m? If only t<m, attempt to re-run the proof of Theorem 4 replacing the m-norm in the split (3.1) by a fixed t<m; the coefficient estimate then fails exactly at the term ||S_ak1||^p_{m,α}. To decide whether the gap is repairable, search for or construct a bounded linear S on L^p_a(dA_α) with sup_z ||S_z1||_{m,α}<∞ and e_S(z)→0 but limsup_{z→∂D} ||S_z1||_{m,α}>0. If such an example exists, Theorem B is false; if a proof that m-norm decay follows from e_S(z)→0 plus the sup bound can be supplied, the gap is closed.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The proof of Theorem 4 (Section 3) has a genuine gap. After assuming e_S(z)→0, the authors invoke Lemma 9 and then assert: 'we only need to show that if ||S_z1||_{m,α}→0 as z→∂D, then S is compact.' This is the step that carries the whole compactness argument: the subsequent split of the coefficient sum at (3.1) needs ||S_ak1||^p_{m,α} to be small for |ak|>R. But Lemma 9, as stated in the paper, is exactly the equivalence e_S(z)→0 iff ||S_z1||_{t,α}→0 for every t∈[1,m), not for t=m. Boundedness of sup_z ||S_z1||_{m,α} plus convergence of all lower L^t norms does not imply convergence of the L^m norm; analytic examples such as scaled monomials show that lower norms can decay while the m-norm remains bounded away from zero. The paragraph 'By Lemma 2 and the method in Lemma 9...' also asserts, without proof, that the weighted analogue of Lemma 9 holds. Even if that weighted analogue is true, it only supplies t<m convergence, so it does not close the m-norm gap. Thus the proof of the nontrivial implication in Theorem 4 is incomplete as written. The claimed compactness characterization may still be true, but it is not established by the given argument.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies linear operators S on weighted Bergman spaces L^p_a(dA_α) for 0<p<∞. For 1<p<∞ it proves (Theorem 1) that a kernel estimate sup_z ||S_z1||_{m,α} ≤ C, with m > p(2+α)/(1+α) max{1,1/(p-1)}, implies boundedness of S, and it claims (Theorem 4) that under the same condition S is compact if and only if its Berezin transform e_S(z) tends to 0 at the boundary, thereby removing the extra assumption sup_z ||S_z^*1||_m <∞ used by Miao and Zheng. The paper also gives boundedness criteria for 0<p≤1 (Theorem 2 and Corollary 1) and for certain 2<p<∞ cases (Theorem 3).","tokens_in":12538,"tokens_out":17525,"duration_ms":160996,"significance":"If Theorem 4 were fully established, it would be a clean and useful improvement over the Miao–Zheng compactness criterion, and the boundedness criterion in Theorem 1 is of independent interest. The proof of Theorem 1 uses standard atomic-decomposition and Forelli–Rudin techniques and appears essentially correct apart from a sign typo in (2.15); it also supplies a fairly complete proof of the weak-convergence coefficient behavior (Lemma 6). However, the proof of the nontrivial implication in Theorem 4 contains a substantial gap: the bridge from the Berezin transform to the kernel condition (Lemma 9) yields convergence in lower L^t norms only, while the compactness argument requires convergence in the L^m norm. The paper also asserts without proof a weighted analogue of Lemma 9. Thus the main advertised characterization is not established by the present argument, and a nontrivial revision is needed.","major_comments":[{"comment":"The step 'then by Lemma 9, to obtain the desired result, we only need to show that if ||S_z1||_{m,α}→0 as z→∂D, then S is compact' is not justified by Lemma 9. Lemma 9, as stated on page 15, gives the equivalence between e_S(z)→0 and ||S_z1||_{t,α}→0 for every t∈[1,m), with t strictly less than m; it does not give convergence at t=m. The subsequent argument, in particular the split at (3.1), relies on smallness of ||S_{a_k}1||^p_{m,α} for |a_k|>R, which is exactly the m-norm decay that Lemma 9 does not supply. Boundedness of the m-norm together with convergence of all lower t-norms does not force convergence of the m-norm. The unproved weighted analogue of Lemma 9 announced in the preceding paragraph would not close this gap either, since it is also stated only for t<m. Therefore the sufficiency implication in Theorem 4 is not established by the given proof.","section":"Section 3, proof of Theorem 4"},{"comment":"The stated lower bound for m in Theorem 2 does not match the conditions actually used in the proof. The proof requires m > (2+β)/δ and m > p(2+β)/(1+α), i.e. m > max{1+(2+α)/(pδ), ((2+α)+pδ)/(1+α)}. The printed condition, which appears as max{ (2+α)/(pδ+1), 1+pδ/(1+α)+1 } with the second term garbled, does not reduce to the sharp bound m > 2+1/(1+α) in Corollary 1 when δ=(1+α)/p, whereas the proof's conditions do. The statement of Theorem 2 must be corrected to the condition used in the proof, or the proof revised accordingly.","section":"Section 2, Theorem 2"}],"minor_comments":[{"comment":"The exponent on (1-|a_k|^2) in (2.15) has the wrong sign; the estimate that follows from Lemma 2(b) should give the negative exponent -(p-1)(2+α)+p(2+α)/m+pn1, which is what is needed for the cancellation with the factor in (2.12).","section":"Equation (2.15)"},{"comment":"In the final sentence of the proof, '∥Sf_n∥_{p,α} → 0 as n → 0' should read 'as n → ∞'.","section":"Proof of Theorem 4"},{"comment":"The displayed range '0 < q < p/(2+α)' in Theorem 3(b) is much more restrictive than the condition derived in the proof, which is q < p(1+α)/(2+α); please clarify whether this is intentional or a typographical error.","section":"Theorem 3(b)"},{"comment":"The formula for the admissible range of m in Theorem 2 is typeset in a garbled way ('m > max 2 + α pδ + 1, 1 + pδ 1 + α + 1'); it should be rewritten unambiguously with a clear max and parentheses.","section":"Theorem 2 statement"}],"recommendation":"major_revision","confidential_remarks":"The main issue is the gap in the proof of Theorem 4: Lemma 9 only gives lower-norm convergence, and the paper's compactness argument needs the m-norm convergence. The authors may be able to fix this by proving directly that e_S(z)→0 implies ||S_z1||_{m,α}→0 under their hypotheses, or by reworking the compactness argument to use only the t-norms that Lemma 9 controls. If neither is possible, the claimed theorem may still be true but requires a different proof. The boundedness results (Theorems 1 and 3) appear credible, so rejection seems too strong; a major revision is appropriate. The m-condition error in Theorem 2 is a statement-level bug that should also be corrected."},"author_rebuttal":null,"desk_editor":null,"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["30H20","47B32","46E22"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that under one uniform control condition on the conjugated operators $S_z$, a linear operator on a weighted Bergman space is compact exactly when its Berezin transform tends to zero at the boundary.","keywords":["Bergman spaces","linear operators","compact operators","Berezin transform","atomic decomposition","weighted Bergman spaces","unit disk","boundedness"],"falsifier":"Find a bounded linear operator $S$ on $L^p_a(dA_\\alpha)$ satisfying $\\sup_z\\|S_z1\\|_{m,\\alpha}<C$ for the theorem's $m$, with $e_S(z)\\to0$ as $z\\to\\partial\\mathbb{D}$, for which $S$ is not compact. If such an operator exists, Theorem B is false. A direct way to look is to construct $S$ as an infinite sum of rank-one kernel operators with slowly decaying weights near the critical exponent $m$ and compute both $e_S(z)$ and $\\|S_z1\\|_{m,\\alpha}$; failure of the latter to tend to $0$ would pinpoint the weighted Lemma 9 gap.","tokens_in":12018,"feed_emoji":"📐","tokens_out":11266,"duration_ms":116778,"temperature":0.7,"pith_summary":"The paper establishes boundedness and compactness criteria for a linear operator $S$ on weighted Bergman spaces $L^p_a(dA_\\alpha)$ of the unit disk, the spaces of holomorphic functions that are $p$-integrable against $(1-|z|^2)^\\alpha$. The main result is that for $1<p<\\infty$, if the conjugated operators $S_z$ satisfy uniform control $\\|S_z 1\\|_{m,\\alpha}\\le C$ for one sufficiently large exponent $m$, then $S$ is bounded, and $S$ is compact exactly when its Berezin transform $e_S(z)=\\langle S k_z,k_z\\rangle$ tends to $0$ as $z$ approaches the unit circle. This reduces the hypotheses of an earlier compactness theorem [8], which also required a uniform bound on $S_z^*1$. For $0<p\\le 1$ and for the endpoint range $2<p<\\infty$, the same kernel estimates yield boundedness statements with adjusted exponents.","feed_headline":"One kernel condition decides compactness on Bergman spaces","feed_subtitle":"A linear operator is compact exactly when its Berezin transform vanishes at the disk's boundary, dropping a second assumption.","key_machinery":"The load-bearing objects are the conjugated operator $S_z$ and the Berezin transform $e_S(z)$. For each $z$, $S_z$ is the image of $S$ under conjugation by the operator $U_z$ built from the Möbius map $\\varphi_z$ and the normalized reproducing kernel $k_z$; the condition $\\|S_z1\\|_{m,\\alpha}\\le C$ is therefore a single test of how $S$ acts on the family of kernel-normalized constants. The proof mechanism is an atomic decomposition theorem: any $f\\in L^p_a(dA_\\alpha)$ is written as a lattice sum of weighted reproducing kernels, with $\\ell^p$ coefficients. Lemma 7 converts the bound on $\\|S_z1\\|_{m,\\alpha}$ into a pointwise estimate for $|S K_z(w)|$, and then Hölder's inequality plus Forelli–Rudin-type integral estimates bound $\\|Sf\\|_{p,\\alpha}$ by the $\\ell^p$ norm of the coefficients. For compactness, the same estimate shows that if $\\|S_z1\\|_{m,\\alpha}\\to0$ near the boundary, every weakly null sequence $f_n$ is mapped to a norm-null sequence, while the converse uses the weak convergence of normalized kernels.","core_discovery":"Let $1<p<\\infty$ and $\\alpha>-1$, and let $S$ be a linear operator on $L^p_a(dA_\\alpha)$ whose domain contains all reproducing kernels. Define $S_z=U_z S U_z$ through the operators $U_z f=(f\\circ\\varphi_z)k_z$, where $\\varphi_z$ is the Möbius map and $k_z$ is the normalized reproducing kernel. The paper proves that if $\\sup_{z\\in\\mathbb{D}}\\|S_z1\\|_{m,\\alpha}<\\infty$ for some $m>p\\frac{2+\\alpha}{1+\\alpha}\\max\\{1,\\frac1{p-1}\\}$, then $S$ is bounded; and under the same hypothesis, $S$ is compact if and only if $e_S(z)\\to0$ as $z\\to\\partial\\mathbb{D}$. The compactness direction uses the atomic decomposition of Bergman functions over a Bergman-metric lattice, pointwise kernel estimates, and the weak convergence of normalized reproducing kernels; the sharp point is that the adjoint-side hypothesis from [8] is no longer needed. The companion results cover $0<p\\le1$, with the threshold $m>2+\\frac1{1+\\alpha}$, and $2<p<\\infty$, where a smaller exponent $m$ still gives boundedness from $L^p_a(dA_\\alpha)$ into $L^q_a(dA_\\alpha)$ for restricted $q$.","pith_inferences":["If the weighted Lemma 9 gap is closed, the same one-sided criterion would likely extend to other domains with Bergman-type kernels, such as bounded symmetric domains, where atomic decompositions and Berezin transforms exist.","A natural test is to run Theorem B against Toeplitz operators with bounded symbols on weighted Bergman spaces, where both the Berezin transform and compactness are independently understood.","The exact threshold $m=p(2+\\alpha)/(1+\\alpha)\\max\\{1,1/(p-1)\\}$ is not proved sharp; constructing examples where boundedness fails just below it would show the machinery is optimal.","The proof's reliance on lattice decompositions suggests that the criterion may have a formulation in terms of discrete sampling of $\\|S_z1\\|$ on a Bergman-metric lattice, which would be easier to check numerically."],"forward_implications":["A single uniform bound on $\\|S_z1\\|_{m,\\alpha}$ is enough to make compactness on $L^p_a(dA_\\alpha)$ equivalent to boundary vanishing of the Berezin transform, so the adjoint condition in [8] can be removed.","For $\\alpha=0$ and $1<p<3/2$, the required exponent $m$ is smaller than what [8] needs, since $m>p(2+\\alpha)/(1+\\alpha)\\max\\{1,1/(p-1)\\}$ is less than $3/(p-1)$ in this range.","The same pointwise kernel machinery gives boundedness for the quasinormed range $0<p\\le1$, with the threshold $m>2+1/(1+\\alpha)$.","In the gap $p(2+\\alpha)/((p-1)(1+\\alpha))<m\\le p(2+\\alpha)/(1+\\alpha)$ for $2<p<\\infty$, the operator need not be bounded on $L^p$, but it is bounded into $L^q$ for $q$ below $m(1+\\alpha)/(2+\\alpha)$ if $p\\ge m$, or below $p/(2+\\alpha)$ if $p<m$.","Consequently, testing compactness of such operators reduces to checking boundary behavior of the single scalar function $e_S(z)$, once the kernel bound is known."],"supporting_citations":[{"why":"Supplies the earlier compactness criterion and Lemma 5.3 that the paper weakens and extends to weighted spaces.","marker":"[8]"},{"why":"Provides the atomic decomposition, lattice estimates, and weak-convergence tools used throughout the proofs.","marker":"[14]"},{"why":"Gives the Hilbert-space reproducing-kernel-thesis compactness result that the paper's theorem generalizes.","marker":"[9]"},{"why":"Supplies the sharp Forelli–Rudin integral estimates used as Lemma 2 in the kernel bounds.","marker":"[5]"},{"why":"Supplies the weak-continuity criterion for bounded operators used in Lemma 6.","marker":"[4]"},{"why":"Yields the weak convergence of normalized reproducing kernels needed for the compact-to-Berezin direction.","marker":"[12]"}],"fun_headline_variants":["Boundary vanishing Berezin transform equals compactness","Compactness iff Berezin transform vanishes on disk edge","Single kernel condition weakens earlier compactness proofs","Berezin boundary limit decides operator compactness","Compactness decided by Berezin fade at boundary"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The compactness proof assumes without proof that a lemma from the unweighted setting, saying vanishing of the Berezin transform forces $\\|S_z1\\|_t\\to0$ for every $t<m$, still holds on weighted Bergman spaces, and that it gives the full exponent $m$ used in the theorem.","fun_headline_variants_meta":{"raw":{"variants":["Boundary vanishing Berezin transform equals compactness","Compactness iff Berezin transform vanishes on disk edge","Single kernel condition weakens earlier compactness proofs","Berezin boundary limit decides operator compactness","Compactness decided by Berezin fade at boundary"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00021,"raw_usage":{"total_tokens":1391,"prompt_tokens":904,"completion_tokens":487,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":520,"completion_tokens_details":{"reasoning_tokens":425}},"tokens_in":520,"tokens_out":487,"duration_ms":5826,"temperature":1.0,"reasoning_tokens":425,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:04:45.160278+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a bounded linear operator $S$ on $L^p_a(dA_\\alpha)$ satisfying $\\sup_z\\|S_z1\\|_{m,\\alpha}<C$ for the theorem's $m$, with $e_S(z)\\to0$ as $z\\to\\partial\\mathbb{D}$, for which $S$ is not compact. If such an operator exists, Theorem B is false. A direct way to look is to construct $S$ as an infinite sum of rank-one kernel operators with slowly decaying weights near the critical exponent $m$ and compute both $e_S(z)$ and $\\|S_z1\\|_{m,\\alpha}$; failure of the latter to tend to $0$ would pinpoint the weighted Lemma 9 gap.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the earlier compactness criterion and Lemma 5.3 that the paper weakens and extends to weighted spaces."},{"cited_title":"Zhu, Operator Theory in Function Spaces , Second edition","cited_arxiv_id":null,"evidence_quote":"Provides the atomic decomposition, lattice estimates, and weak-convergence tools used throughout the proofs."},{"cited_title":"Mitkovski, B","cited_arxiv_id":null,"evidence_quote":"Gives the Hilbert-space reproducing-kernel-thesis compactness result that the paper's theorem generalizes."},{"cited_title":"Liu, Sharp Forelli-Rudin estimates and the norm of the Bergman projection , J","cited_arxiv_id":null,"evidence_quote":"Supplies the sharp Forelli–Rudin integral estimates used as Lemma 2 in the kernel bounds."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the weak-continuity criterion for bounded operators used in Lemma 6."},{"cited_title":"Zeng, Toeplitz operator on Bergman spaces , Houston J","cited_arxiv_id":null,"evidence_quote":"Yields the weak convergence of normalized reproducing kernels needed for the compact-to-Berezin direction."}],"review_version":1}