{"id":"87d1a19e-99c1-48b4-acc6-31a46ca401cd","arxiv_id":"2501.03050","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Matrix A-infinity weighted Besov and Triebel-Lizorkin type spaces are characterized via phi-transforms, molecules, wavelets, and atoms, with sharp boundedness conditions for almost diagonal and classical operators.","lead":"This paper builds a full theory of weighted function spaces, Besov-type and Triebel-Lizorkin-type, where the weight is a matrix-valued function in the A-infinity class. It proves that these spaces have clean wavelet, molecule, and phi-transform descriptions and that many classical operators are bounded on them, improving several earlier results even in the unweighted case.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 4.5's sharp thresholds rest entirely on the imported quantitative two-cube estimates for A_Q A_R^{-1} (Lemma 2.18); any hidden factor or dimension shift there propagates through (4.4) to every molecular, wavelet, trace, multiplier and CZO result.","rationale":"Read in good faith, the paper is a substantial and mostly careful extension of the authors' program from A_p to A_{p,infty} matrix weights. The proofs of Theorems 4.19 and 4.20 are detailed, and the reduction in (4.23) is legitimate because the weighted averaging norm of a sequence t equals the unweighted norm of u_Q = |A_Q t_Q|. The inhomogeneous reduction in Section 4.3 is also less problematic than it first appears, because the as norm in (3.7) retains the supremum over all P in Q, so extending sequences by zero does not change the norm. The sharpness examples and comparisons with [7] and [9] are coherent. The weakest point is not an internal algebraic error I can identify, but rather that the entire quantitative dimension formalism is imported from a self-cited prior paper without proof or formal verification. The reader flagged this same point, and I agree. Because the concern is about verification rather than a demonstrated contradiction, the appropriate verdict remains CONDITIONAL rather than REJECT or UNVERDICTED. A full derivation of Lemma 2.18, or an independent formal check of it, would settle whether the concern actually lands.","tokens_in":100784,"tokens_out":22201,"duration_ms":200041,"concrete_test":"Independently re-derive a complete proof of Lemma 2.18 from Definition 2.16, and check the claimed bound on the two extremal configurations: (a) ell(R) much smaller than ell(Q) with |x_Q - x_R| comparable to ell(Q), and (b) ell(R) comparable to ell(Q) with |x_Q - x_R| much larger than ell(Q), for a genuinely matrix A_{p,infty} weight that is not in A_p, for example W(x) = diag(|x_1|^{a_1}, ..., |x_m|^{a_m}) with exponents in (-1, infinity) and some a_i > 0. If the derived estimate contains an extra logarithmic factor or requires increasing d1+d2 by any epsilon > 0, recompute ~J and ~s in (4.3) and update the inequalities in (4.4).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is Theorem 4.5: for W in A_{p,infty}, (D,E,F)-almost diagonal matrices are bounded on dot-a_{p,q}^{s,tau}(W) under the strict inequalities (4.4). The constants d_lower and d_upper enter through Lemma 2.18, imported from [11, Lemma 6.7]. That lemma is not proved here, and its sharpness is also delegated to [11, Lemma 7.6]. Every subsequent estimate in Section 4.1, including Lemmas 4.11 and 4.16 and Theorems 4.19 and 4.20, uses Lemma 2.18 to convert reducing-operator ratios A_Q A_R^{-1} into the dimension exponents d1+d2 and d2; the convergence conditions in the k,l sums in (4.18) then become exactly (4.4). Consequently the advertised endpoint conditions, not just qualitative boundedness, are only as secure as the quantitative content of Lemma 2.18. If that estimate carries an additional logarithmic factor, or if the correct spatial decay exponent is d1+d2+epsilon for some epsilon>0, then the thresholds in (4.4) shift and the claimed sharp A_{p,infty} results fail. The manuscript supplies no independent verification; it only states that the estimates are contained in and sharp in [11]. Since the reader's verdict is CONDITIONAL, this is the condition that needs to be discharged.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops inhomogeneous Besov-type and Triebel-Lizorkin-type spaces with matrix A_{p,\\infty} weights, extending earlier homogeneous and Ap-weighted results of the same authors. The main technical axis is Theorem 4.5, which asserts boundedness of (D,E,F)-almost diagonal operators on the homogeneous sequence spaces \\dot a^{s,\\tau}_{p,q}(W) under the sharp-looking conditions (4.4), with thresholds expressed through the A_{p,\\infty}-lower and upper dimensions of W. From this result the paper derives \\varphi-transform characterizations, coincidence of pointwise and averaging weighted spaces, molecular, wavelet, and atomic decompositions, and applications to pseudo-differential operators, trace/extension operators, pointwise multipliers, and Calderon-Zygmund operators. The authors also compare their almost-diagonal conditions with an earlier result of Bownik and Ho and show improvements in the scalar unweighted case.","tokens_in":101045,"tokens_out":13243,"duration_ms":123098,"significance":"If the central theorem is correct, the paper is a substantial advance: it makes matrix-weighted Besov/Triebel-Lizorkin-type spaces with A_{p,\\infty} weights as usable as their Ap-counterparts for most analytic purposes, including molecular decompositions and classical operator boundedness. Notable strengths are the detailed proofs of several core results with explicit constants (Theorems 3.13, 3.23, 3.25, 4.19, 4.20), the construction of independent sharpness counterexamples in Lemmas 4.13 and 4.18, the answer to a question of Volberg about \\dot f^0_{p,2}(W), and the claim of new results even in the scalar unweighted case. The main caveat is that the most load-bearing quantitative input, Lemma 2.18, is imported from the authors' previous paper [11] without proof; the advertised endpoint conditions in (4.4) are only as secure as that imported estimate.","major_comments":[{"comment":"The two-cube estimate for reducing operators is imported from [11, Lemma 6.7] and its sharpness from [11, Lemma 7.6], with no proof. This estimate is the only place in the paper where the exponents d_lower_{p,\\infty}(W) and d_upper_{p,\\infty}(W) are converted into spatial decay rates; it feeds directly into Lemmas 4.11 and 4.16, Theorem 4.20, and the convergence conditions of the k,l-sums in (4.18), which in turn produce the thresholds in (4.4). If the estimate in [11] carried an additional logarithmic factor, or if the correct decay exponent were d1+d2+\\epsilon for some \\epsilon>0, then (4.4) would fail and the claimed sharp A_{p,\\infty} results would not follow. Because the manuscript supplies no independent verification, I cannot certify the central theorem without access to [11]. Please include a proof of Lemma 2.18 and of the sharpness claim, or, if [11] is published, quote the exact statement and confirm that the proof is publicly available.","section":"Section 2.2, Lemma 2.18"},{"comment":"The proof of Theorem 4.20 reduces the supercritical and critical cases to \\dot f^{\\hat s}_{\\infty,\\infty}(A) by invoking [8, Theorem 4.17] 'for any sequence of positive definite matrices A', by inspection. The second half of the proof then identifies \\dot f^{\\hat s}_{\\infty,\\infty}(A)=\\dot f^{\\hat s,1/p}_{p,\\infty}(W), which is fine via Theorem 3.28. But the first reduction is load-bearing: without (4.23), Theorem 4.20 cannot be derived, and hence Theorem 4.5 is incomplete. Since the text itself notes that [8, Theorem 4.17] was originally stated under Ap assumptions, the asserted generalization to arbitrary A must either be quoted verbatim from [8] or proved in the present paper.","section":"Section 4.1, equation (4.23)"},{"comment":"The \\varphi-transform theorem for averaging spaces is presented as a modification of [8, Theorem 3.29], with the sup/inf equivalence in Lemma 3.16 imported from [8, Lemma 3.17] 'by inspection'. These results underlie Theorem 3.30 and all subsequent molecular, wavelet, and atomic characterizations; in particular, the boundedness of T_\\psi in Theorem 3.13 includes a case where the cancellation condition is on only one factor (Corollary 3.8), and the adaptation should be written out. I am not claiming the argument is wrong, but it is not checkable from the present text. Please include the full proof or point to the precise page or display in [8] that justifies each step under the strong-doubling hypotheses.","section":"Section 3.2, Theorem 3.13 and Lemma 3.16"},{"comment":"The trace and extension proofs are outlines; in particular, the claim that [\\ell(Q)]^{1/2}\\theta^\\lambda_Q(\\cdot,0) is a B^{s-1/p,n/(n-1)\\tau}_{p,r}(V)-synthesis molecule is stated as 'one can check', and the lower bound on s involving d_upper_{p,\\infty}(V)/p is new compared with [10]. Since the target space in Theorem 6.3 has no cancellation conditions exactly when s is in the stated range (see Remark 6.4), the molecule verification is essential for the endpoint of the trace theorem. Please provide this verification or a precise reference to the corresponding statement in [10].","section":"Section 6.1, Theorems 6.3 and 6.5"}],"minor_comments":[{"comment":"Corollary 2.15 is applied to shifted dyadic systems Q^\\alpha; as stated, the corollary is for unshifted cubes. A sentence explaining that the shifted version follows from the same proof would remove a small gap.","section":"Section 4.1, Lemma 4.16"},{"comment":"The definition of d_{p'}(W^{-1/(p-1)}) for p\\in(0,1] is informal; either define it properly or state clearly that the term is zero by convention.","section":"Section 4.1, Remark 4.6"},{"comment":"The hypothesis 'whenever W\\in A_{p,\\infty} satisfies d_upper_{p,\\infty}(W)=d_2' is a universal statement over weights; consider rewording it as an implication to avoid ambiguity.","section":"Section 4.1, Lemmas 4.13 and 4.18"},{"comment":"The proof says 'by Holder's inequality' after choosing a=p\\wedge 1; when p<1 this is not Holder's inequality, and the justification should be adjusted.","section":"Section 4.1, Lemma 4.11"}],"recommendation":"major_revision","confidential_remarks":"The paper is heavily self-citing [8,9,10,11], which is natural for a research program, but the referee's main concern is verifiability: Lemma 2.18 is the quantitative engine of the whole paper and is not proved here. Before publication, the journal should confirm that [11] is in final form or publicly available, and ideally the authors should include a self-contained proof of Lemma 2.18 in an appendix. The paper is within the scope of the journal and, conditional on the imported estimate, appears to be a strong contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is the real next step in the matrix-weighted Besov–Triebel–Lizorkin program, and it delivers. The authors move from matrix Ap to the Ap,∞ classes, build the inhomogeneous spaces, and get the expected machinery—φ-transform, molecular/wavelet/atomic decompositions, almost diagonal boundedness, and operator applications—with sharp almost diagonal thresholds, including the new dupper/p term, and an answer to Volberg's p<2 question. The paper is carefully written and mostly proves what it claims, with explicit constants and lemmas in the core sections. Credit also goes to the scalar unweighted consequences: several results remove upper bounds on τ that [104] needed.\n\nThe soft spot is exactly where the reader put it. Theorem 4.5, the load-bearing result, rests on Lemma 2.18, imported from [11, Lemma 6.7], and sharpness is also delegated to [11, Lemma 7.6]. That lemma's quantitative estimates on ||A_Q A_R^{-1}|| in terms of d_lower and d_upper feed into Lemmas 4.11 and 4.16, and from there into (4.4). If that lemma is off by a log factor or an epsilon in the exponent, the advertised thresholds shift. The authors do not reproduce or independently verify it here; they cite their own prior work. That is acceptable practice when the prior result is solid, but here every downstream endpoint condition depends on it. Second, the assembly of Theorem 4.5 is literally “combining Theorems 4.19 and 4.20; we omit the details,” and Theorem 4.20 uses the reduction (4.23) to ˙f∞,∞(A) or ˙f s∞,q(A) by quoting [8, Theorem 4.17]. That reduction is reasonable, but it is another imported step. The molecule verification in Theorem 5.3 is also a “routine verification, omit details.” None of these look like fatal gaps; they look like places where the authors leaned on their own earlier papers and a referee will want the details spelled out.\n\nMy judgement: the central argument holds up as far as I can check, with the caveat that Lemma 2.18 is the unverified hinge. The paper is for harmonic analysts working on matrix-weighted function spaces; it will be cited and used. It deserves a serious referee, and the referee should ask for a self-contained statement and proof or precise reference for Lemma 2.18, plus a fuller proof of Theorem 4.5 and the molecule verification. Conditional acceptance is the right outcome.","headline":"A substantial, carefully built extension of the matrix-weighted Besov/Triebel–Lizorkin program to Ap,∞; the main endpoint theorem hinges on one imported quantitative lemma that the referee should verify.","tokens_in":101639,"tokens_out":1901,"would_cite":true,"duration_ms":55556,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46E35","47A56","42B25","42B20","42C40","35S05","42B35"],"pacs":[],"model":"deepseek-v4-flash","headline":"A sharp almost-diagonal theorem makes matrix $A_{p,\\infty}$-weighted Besov and Triebel-Lizorkin spaces fully tractable.","keywords":["matrix weights","Ap,infty weights","Besov-type spaces","Triebel-Lizorkin-type spaces","almost diagonal operators","molecules and atoms","wavelets","Calderon-Zygmund operators"],"falsifier":"Look for a concrete counterexample to Lemma 2.18: take a weight such as $W(x)=|x|^d I_m$ on $\\mathbb{R}^n$, write down its reducing operators, and test cube pairs with different sizes and centers to see whether $\\|A_Q A_R^{-1}\\|^p$ obeys the claimed bound with $d_1$ near $0$ and $d_2=d$; any pair requiring a larger exponent to keep the right side finite would refute the dimension estimate. Alternatively, choose parameter values satisfying (4.4) and an almost diagonal matrix of the special form $B^{D,E,F}$ from (4.1), and exhibit a sequence in $\\dot a^{s,\\tau}_{p,q}(W)$ whose image has infinite norm, which would disprove Theorem 4.5.","tokens_in":2355,"feed_emoji":"📐","tokens_out":3157,"duration_ms":125393,"temperature":0.7,"pith_summary":"This paper claims that inhomogeneous Besov-type and Triebel-Lizorkin-type spaces weighted by a matrix $A_{p,\\infty}$ weight $W$ are fully analyzable through two numbers attached to $W$: its lower and upper $A_{p,\\infty}$-dimensions. With these dimensions the authors prove a sharp boundedness theorem for almost diagonal matrices on the associated homogeneous sequence spaces, and from that theorem they derive $\\varphi$-transform, molecular, wavelet, and atomic characterizations of the inhomogeneous spaces $A^{s,\\tau}_{p,q}(W)$, together with boundedness of trace, extension, pointwise multiplier, pseudo-differential, and Calderon-Zygmund operators. A sympathetic reading is that matrix $A_{p,\\infty}$ weights, which properly contain matrix $A_p$ weights, now support essentially the same structural toolbox that was previously available only for $A_p$ weights.","feed_headline":"Matrix A∞ weights tame Besov-Triebel-Lizorkin spaces","feed_subtitle":"Sharp almost-diagonal bounds yield wavelets, molecules, traces and singular integrals for these weighted spaces.","key_machinery":"The machinery is the pair of $A_{p,\\infty}$-dimensions of the weight and the sharp estimate they permit for its reducing operators. A reducing operator $A_Q$ is a positive-definite matrix associated to a cube $Q$ such that $|A_Q z|$ is comparable to the $L^p(W)$-average of $z$ over $Q$; Lemma 2.18 bounds $\\|A_Q A_R^{-1}\\|^p$ by a power of the ratio of the side lengths together with $(1+|x_Q-x_R|/\\max(\\ell(Q),\\ell(R)))^{d_1+d_2}$, where $d_1$ and $d_2$ run through the admissible lower and upper dimensions. This converts the continuous weight into a strongly doubling family of matrices, and the averaging spaces $A^{s,\\tau}_{p,q}(A)$ are built directly from that family. Almost diagonal matrices are then the discrete operators whose boundedness, under the dimension conditions, transfers back to the function spaces through the $\\phi$-transform.","core_discovery":"The central claim is that the homogeneous sequence spaces $\\dot a^{s,\\tau}_{p,q}(W)$ for $W\\in A_{p,\\infty}$ admit the same almost diagonal calculus as the unweighted spaces, provided the parameters $D,E,F$ are measured against the quantities $\\tilde J$ and $\\tilde s$ formed from $p,q,\\tau$ and the lower and upper dimensions $d^{\\mathrm{lower}}_{p,\\infty}(W)$, $d^{\\mathrm{upper}}_{p,\\infty}(W)$. Theorem 4.5 states that every $(D,E,F)$-almost diagonal matrix is bounded whenever $D>\\tilde J$, $E>n/2+\\tilde s$, and $F>\\tilde J-n/2-\\tilde s$, with the definitions in (4.3); the term $d^{\\mathrm{upper}}_{p,\\infty}(W)/p$ is the new price paid for leaving the $A_p$ class, and the authors show it is sharp for $p\\in(0,1]$. Carrying this one operator bound through the paper gives the $\\phi$-transform characterization, the equality of pointwise-weighted and averaging spaces, molecular, wavelet, and atomic decompositions, and the classical operator results.","pith_inferences":["Editorial inference: the dimension pair $(d^{\\mathrm{lower}}_{p,\\infty}(W),d^{\\mathrm{upper}}_{p,\\infty}(W))$ is likely the right quantitative object for other problems on these spaces, such as sharp weighted estimates for singular integrals, because all the paper's inequalities degrade exactly through these exponents.","Editorial inference: the removal of the upper bound on $\\tau$ in the scalar unweighted case indicates that the Morrey index is absorbed into the almost diagonal calculus rather than posing a separate obstruction; the same absorption may hold in anisotropic or multi-parameter matrix-weighted settings.","Editorial inference: the averaging-space route, which sidesteps the absence of a vector-valued Fefferman-Stein maximal inequality for $A_{p,\\infty}$, is a reusable strategy: define the space first through discrete matrix averages and prove the pointwise equivalence afterwards."],"forward_implications":["The $\\phi$-transform gives a coefficient characterization of $A^{s,\\tau}_{p,q}(W)$ for every $W\\in A_{p,\\infty}$, so membership in these spaces can be read off from wavelet or $\\phi$-basis coefficients.","Almost diagonal operators bounded on the sequence spaces force molecular and atomic decompositions, so every element of the space has a series expansion whose coefficient norm is equivalent to the original norm.","Pseudo-differential operators of order $u$ map $A^{s+u,\\tau}_{p,q}(W)$ into $A^{s,\\tau}_{p,q}(W)$, and Fourier multipliers are covered automatically without any extra condition beyond $W\\in A_{p,\\infty}$.","Trace into hyperplanes and extension from hyperplanes work for smoothness $s$ above a threshold involving $d^{\\mathrm{upper}}_{p,\\infty}$ of the restriction weight, under a cube-wise comparability condition between the two weights.","Calderon-Zygmund operators are bounded on each $A^{s,\\tau}_{p,q}(W)$; in the scalar unweighted case with $W\\equiv 1$, this removes the upper bound on the Morrey index $\\tau$ that earlier results imposed."],"supporting_citations":[{"why":"Supplies the $A_{p,\\infty}$ weight characterizations, the lower and upper dimensions, and Lemma 2.18, the sharp reducing-operator estimate on which the whole argument rests.","marker":"[11]"},{"why":"Introduced the $A_p$-dimension framework and the homogeneous matrix-weighted sequence spaces whose averaging-space arguments are reused here.","marker":"[8]"},{"why":"Gave the sharp almost diagonal conditions in the $A_p$ case whose proof structure is adapted to $A_{p,\\infty}$, and provides the comparison baseline.","marker":"[9]"},{"why":"Provided the molecular, wavelet, trace and Calderon-Zygmund results in the $A_p$ setting that are here extended to $A_{p,\\infty}$.","marker":"[10]"},{"why":"Is the source of the averaging-space method for matrix weights, which the paper adapts because vector-valued Fefferman-Stein inequalities fail for $A_{p,\\infty}$.","marker":"[39]"},{"why":"Supplies the unweighted Besov-type and Triebel-Lizorkin-type spaces and theorems that the present results improve by removing the upper restriction on $\\tau$.","marker":"[104]"},{"why":"Introduced matrix $A_{p,\\infty}$ weights and raised the question about the equivalence between pointwise and averaging norms that Theorem 3.28 answers for all $p$.","marker":"[95]"},{"why":"Provides the discrete Calderon reproducing formula and $\\phi$-transform machinery underlying the sequence-space characterization.","marker":"[35]"}],"fun_headline_variants":["Almost diagonal operators tame matrix-weighted Besov-Triebel spaces","Matrix A∞ weights: sharp bounds for Besov-Triebel-Lizorkin","Sharp almost-diagonal calculus for matrix-weighted Besov-Triebel","Matrix weights bring traces, wavelets, and singular integrals to BTL"],"cache_read_input_tokens":103552,"weakest_assumption_plain":"The load-bearing premise is the quantitative dimension estimate for matrix weights: for every cube pair, the norm of a product of reducing operators grows at most like the cube-ratio power and distance power governed by the weight's lower and upper $A_{p,\\infty}$-dimensions, and those exponent ranges are sharp; if any $A_{p,\\infty}$ weight violated this estimate, the almost-diagonal theorem and everything built on it would fail.","fun_headline_variants_meta":{"raw":{"variants":["Almost diagonal operators tame matrix-weighted Besov-Triebel spaces","Matrix A∞ weights: sharp bounds for Besov-Triebel-Lizorkin","Sharp almost-diagonal calculus for matrix-weighted Besov-Triebel","Matrix weights bring traces, wavelets, and singular integrals to BTL"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001221,"raw_usage":{"total_tokens":5030,"prompt_tokens":965,"completion_tokens":4065,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":581,"completion_tokens_details":{"reasoning_tokens":3986}},"tokens_in":581,"tokens_out":4065,"duration_ms":32476,"temperature":1.0,"reasoning_tokens":3986,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:57:48.400136+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Look for a concrete counterexample to Lemma 2.18: take a weight such as $W(x)=|x|^d I_m$ on $\\mathbb{R}^n$, write down its reducing operators, and test cube pairs with different sizes and centers to see whether $\\|A_Q A_R^{-1}\\|^p$ obeys the claimed bound with $d_1$ near $0$ and $d_2=d$; any pair requiring a larger exponent to keep the right side finite would refute the dimension estimate. Alternatively, choose parameter values satisfying (4.4) and an almost diagonal matrix of the special form $B^{D,E,F}$ from (4.1), and exhibit a sequence in $\\dot a^{s,\\tau}_{p,q}(W)$ whose image has infinite norm, which would disprove Theorem 4.5.","supporting_citations":[{"cited_title":"Y oussfi, Continuit´ e-Besov des op´ erateurs d´ efinis par des int´ egrales singuli` eres, Manus- cripta Math","cited_arxiv_id":null,"evidence_quote":"Supplies the unweighted Besov-type and Triebel-Lizorkin-type spaces and theorems that the present results improve by removing the upper restriction on $\\tau$."},{"cited_title":"Triebel, Theory of Function Spaces","cited_arxiv_id":null,"evidence_quote":"Introduced matrix $A_{p,\\infty}$ weights and raised the question about the equivalence between pointwise and averaging norms that Theorem 3.28 answers for all $p$."},{"cited_title":"Frazier and B","cited_arxiv_id":null,"evidence_quote":"Provides the discrete Calderon reproducing formula and $\\phi$-transform machinery underlying the sequence-space characterization."}],"review_version":1}