{"id":"09b9dbdd-ae45-49e2-9315-b4f802f3ec85","arxiv_id":"2506.06726","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A bounded operator from ℓ^q to a Banach space is compact exactly when its coordinate shadow under all linear functionals forms a totally bounded set in ℓ^p; for Hilbert-space operators, this is equivalent to total boundedness of the joint numerical range.","lead":"This paper gives conditions that tell exactly when a diagonal operator from a sequence space to any Banach space is compact, and when an infinite list of operators on a Hilbert space defines a compact operator. A reader might care because it ties a classical notion in functional analysis, compactness, to a simple geometric condition on the numerical range of the operators.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; the central characterization in Theorem 3.6 is sound, with only a minor norm typo in Theorem 3.7.","rationale":"I read the paper in good faith and checked the central theorem through the proof. Theorem 3.6 follows from the representation theorem (Theorem 3.5) and Schauder's theorem; there is no hidden assumption at the endpoints because Schauder's theorem holds for all Banach spaces and total boundedness is equivalent to precompactness in the complete spaces ℓ^p and ℓ∞. The Hilbert-space Theorem 5.1 is also sound: the proof that total boundedness of the joint numerical range forces the operator to be wcbs is valid, and for reflexive ℓ^q (1<p<∞) wcbs implies compactness by Corollary 2.9. The only issue I found is a typographical error in Theorem 3.7, where the operator norm of Λ_a is misidentified; this is already flagged by the reader. I partially agree with the reader's weakest_assumption: the endpoint cases are not actually fragile, but the typo is real. Since I find no load-bearing mathematical concern, I do not move the verdict.","tokens_in":14404,"tokens_out":35668,"duration_ms":322917,"concrete_test":"Run a direct endpoint check for p=∞ with A=c0 and a_i=e_i/i: verify that {φ(a):φ∈c0^*,∥φ∥≤1} is totally bounded in ℓ∞ and that Λ_a:ℓ^1→c0 is compact, confirming that the Schauder-based equivalence in Theorem 3.6 holds at the non-reflexive endpoint.","verdict_should_be":"UNCHANGED","load_bearing_attack":"No significant objection identified. The central claim, Theorem 3.6, is correctly derived: Theorem 3.5 represents every bounded Λ:ℓ^q→A as Λ_a, and Schauder's theorem [7, Thm 4.19] reduces compactness of Λ_a to total boundedness of Γ_a(A*_1) in ℓ^p, which is exactly the condition a∈ℓ^p_c(A). The endpoint cases are handled soundly: for p=1 the domain is c0 with dual ℓ^1, and for p=∞ the domain is ℓ^1 with dual ℓ∞, so the adjoint argument remains valid despite non-reflexivity. The Hilbert-space theorem (Theorem 5.1) is also sound: total boundedness of the joint numerical range forces wcbs, and for 1<p<∞ reflexivity of ℓ^q upgrades wcbs to compactness via Corollary 2.9. The only concrete defect is the typographical error in Theorem 3.7, where the operator norm of Λ_a is written as |||a|||_∞ instead of |||a|||_1, and the theorem statement uses ||a||_1 where |||a|||_1 is meant; this does not affect the compactness characterization.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper characterizes compact operators from ℓ^q (1/p+1/q=1, 1≤p≤∞) into an arbitrary Banach space A as diagonal maps Λ_a(β)=Σ β_i a_i induced by sequences a∈ℓ^p_c(A), i.e., those for which the set {φ(a):φ∈A*_1} is totally bounded in ℓ^p. The main theorem (Theorem 3.6) covers 1<p≤∞, and Theorem 3.7 covers the p=1 case with domain c0. The authors also establish a Banach-space isomorphism between ℓ^p_b(A) and B(ℓ^q,A) (Theorem 3.8), and apply the characterization to C(Ω)-valued operators, showing that compactness is equivalent to continuity of the associated function F:Ω→ℓ^p (Theorems 4.2 and 4.3). For B(H)-valued operators, Theorem 5.1 shows that, for 1<p<∞, compactness of the induced operator T:ℓ^q→B(H) is equivalent to total boundedness in ℓ^p of the joint numerical range {⟨Tx,x⟩:∥x∥=1}, with the norm inequality 1/2∥T∥≤ω(T)≤∥T∥. The proofs rely on the Schauder adjoint theorem, Kolmogorov's compactness criterion, and a wcbs (weak*-continuous-on-bounded-sets) framework developed in Section 2.","tokens_in":14614,"tokens_out":29599,"duration_ms":241789,"significance":"The main characterization is clean and useful: for 1<p≤∞, compactness of an operator from ℓ^q to A is exactly the total boundedness in ℓ^p of the dual shadow of the generating sequence. The paper is essentially self-contained, with elementary proofs and explicit constants, and the Hilbert-space numerical-range criterion (Theorem 5.1) is a nice addition that is not obvious a priori. The theorems are stated with precise quantitative inequalities and involve no free parameters. The overall argument is sound; the only defects found are local typos and presentation issues, none of which affect the central compactness characterization.","major_comments":[],"minor_comments":[{"comment":"The norm identity in Theorem 3.7 is misstated: for a∈ℓ^1_b(A) the operator norm of Λ_a:c0→A is |||a|||_1, not ∥a∥_1 (the latter is the ℓ^1(A)-norm, which is generally different). In the proof, 'It is easy to verify that ∥Λ_a∥=|||a|||_∞' should read '∥Λ_a∥=|||a|||_1'. This typo does not affect the compactness characterization, but it should be corrected.","section":"Theorem 3.7"},{"comment":"The assertion that a=(e_1,e_2,...) belongs to ℓ^p_b(A) for every p∈[1,∞) is false for p>2. For instance, with A=ℓ^2, the functional φ=β with β_i=1/√i has ∥β∥_2≤1 but ∥φ(a)∥_p=∥β∥_p=∞ for p>2. The example should be restricted to 1≤p≤2.","section":"Example 3.3"},{"comment":"The abstract contains a typesetting error: '⟨T_1 x, x\\rangel' should read '⟨T_1 x, x\\rangle'. Also, the running title 'onℓ p-spaces' lacks spaces.","section":"Abstract"},{"comment":"In the Introduction, 'Housdorff' should be 'Hausdorff'.","section":"Introduction (page 2)"},{"comment":"In the proof of (4)⇒(1), the conclusion 'We conclude that T is wcbs' should be followed by an explicit invocation of Corollary 2.9: since 1<p<∞, ℓ^p is reflexive, so W(ℓ^q,B(H))=K(ℓ^q,B(H)), hence T is compact. As written, the final step is implicit.","section":"Theorem 5.1 proof"},{"comment":"The statement 'An operator T:ℓ^q→A is compact if and only if the joint numerical range W(T) is a totally bounded subset of ℓ^p' is ambiguous: W(T) is defined for a sequence T=(T_i) in B(H), not for an operator T:ℓ^q→A. Recommend stating the theorem for a sequence T=(T_i) in ℓ^p_b(B(H)) and its induced operator.","section":"Theorem 5.2"}],"recommendation":"minor_revision","confidential_remarks":"The manuscript is sound and the main characterization is correct. The self-citation to [1] for Lemma 2.2 is for a standard fact and does not constitute circularity. The paper is suitable for a functional analysis journal; the needed changes are typographical and presentational. No concerns about scope or citation practices."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick read. The paper is correct and mostly well-written, but the headline theorem is less new than the framing suggests. For 1<p≤∞, the characterization that Λ:ℓ^q→A is compact iff Λ=Λ_a with {φ(a):φ∈A*_1} totally bounded in ℓ^p is, as the author's own proof shows, a direct consequence of Schauder's theorem plus the standard representation of bounded operators on ℓ^q. That makes it a clean formulation of known compactness facts, not a new discovery. The genuinely new part is Section 5: the equivalence between compactness of T:ℓ^q→B(H) and total boundedness of the joint numerical range in ℓ^p. That is a useful reformulation, and the proof is non-tautological: total boundedness of W forces wcbs, and reflexivity of ℓ^q upgrades wcbs to compactness. The polarization argument is standard but applied well.\n\nSoft spots are minor. There is a genuine typo in Theorem 3.7: the norm of Λ_a is equated to |||a|||_∞, while the preceding discussion and the theorem statement require |||a|||_1; the author even says 'it is easy to verify' for the wrong identity. A referee should catch that. More substantial: the paper does not situate Theorem 3.6 in the literature. If it is folklore, say so; if there is a source, cite it. Right now the reader has to do the archaeology. The endpoint cases p=1 and p=∞ are handled carefully, and the C(Ω) section is a nice application. Nothing in the proofs looks fragile.\n\nWho gets value from this? Someone working with operator tuples and numerical ranges who wants a compactness test, or a lecturer looking for a tidy presentation of diagonal operators. It does not restructure anything, but it is not a waste of referee time. I would send it to a referee, ask for a literature check and the typo fix, and expect acceptance after minor revision.","headline":"Correct and clean, but mostly a repackaging of Schauder plus standard facts; the numerical-range criterion for Hilbert-space operator tuples is the one genuinely new piece.","tokens_in":15201,"tokens_out":2874,"would_cite":true,"duration_ms":28618,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46B45","47B37","47A12","47A30"],"pacs":[],"model":"deepseek-v4-flash","headline":"Compact operators from ℓ^q are exactly the diagonal maps whose coefficient sequences have a totally bounded dual shadow.","keywords":["compact operators","ℓ^p spaces","diagonal operators","weak* continuity","joint numerical range","joint numerical radius","total boundedness"],"falsifier":"A direct way to test the endpoint case is to set $A = \\ell^\\infty$ and take any sequence $a = (a_i)$ whose set of entries is totally bounded in $A$; the theorem predicts the induced operator $\\Lambda_a(\\beta) = \\sum \\beta_i a_i$ on $\\ell^1$ is compact. Finding a totally bounded sequence of that kind whose induced operator is not compact would refute the characterization, so a search over such sequences is a concrete falsification test.","tokens_in":14157,"feed_emoji":"📐","tokens_out":20997,"duration_ms":163728,"temperature":0.7,"pith_summary":"The paper establishes a concrete criterion for compactness of bounded linear operators from the sequence space $\\ell^q$ (or from $c_0$) into an arbitrary Banach space $A$. Such an operator is compact exactly when it is a diagonal map $\\Lambda_a(\\beta) = \\sum \\beta_i a_i$ and the set of dual evaluations $\\{\\varphi(a) : \\varphi \\in A^*, \\|\\varphi\\| \\le 1\\}$ is totally bounded in $\\ell^p$, with $1/p + 1/q = 1$. The same principle is then specialized to operators with values in $B(H)$, where compactness is shown to be equivalent to total boundedness of the joint numerical range $\\{\\langle Tx,x\\rangle : \\|x\\| = 1\\}$ in $\\ell^p$. Because compact operators are the tractable ones for spectral and approximation questions, the criterion turns an abstract compactness check into a tail estimate on a sequence in the target space. The endpoint cases $p = 1$ and $p = \\infty$ are covered by separate arguments.","feed_headline":"One tail condition decides compactness of ℓ^q operators","feed_subtitle":"An operator from ℓ^q is compact exactly when its coefficient sequence's dual image is totally bounded in ℓ^p.","key_machinery":"The object that carries the argument is the dual shadow map $\\Gamma_a: A^* \\to \\ell^p$ defined by $\\Gamma_a(\\varphi) = (\\varphi(a_i))$. This map is the adjoint of the diagonal operator $\\Lambda_a$, because $\\beta(\\Gamma_a \\varphi) = \\varphi(\\Lambda_a \\beta)$ for $\\beta \\in \\ell^q$. By the classical theorem that compactness is preserved under taking adjoints, $\\Lambda_a$ is compact precisely when $\\Gamma_a(A^*_1)$ is totally bounded in $\\ell^p$, which is exactly the condition $a \\in \\ell^p_c(A)$. For the Hilbert-space version, the joint numerical range $W(T)$ is a subset of that dual shadow, and a polarization identity together with a weak*-continuity argument shows that total boundedness of $W(T)$ forces $T$ to be compact.","core_discovery":"The central theorem states that for $1 < p \\le \\infty$ and $1/p + 1/q = 1$, a bounded operator $\\Lambda: \\ell^q \\to A$ is compact if and only if $\\Lambda = \\Lambda_a$ for some sequence $a = (a_i)$ in $A$ whose 'dual shadow' $\\{\\varphi(a) : \\varphi \\in A^*_1\\}$ is totally bounded in $\\ell^p$, where $\\varphi(a) = (\\varphi(a_1), \\varphi(a_2), \\ldots)$. Here $\\Lambda_a(\\beta) = \\sum_{i=1}^\\infty \\beta_i a_i$. For $p = 1$ the same equivalence holds for operators on $c_0$. When $A = B(H)$, the theorem takes the form that $T: \\ell^q \\to B(H)$ is compact if and only if its joint numerical range $W(T) = \\{\\langle Tx,x\\rangle : \\|x\\| = 1\\}$ is totally bounded in $\\ell^p$, and then the joint numerical radius obeys $\\frac{1}{2}\\|T\\| \\le \\omega(T) \\le \\|T\\|$.","pith_inferences":["The tail-estimate form suggests a computational test for non-compactness: approximate the supremum over the unit ball of $A^*$ by a finite net of functionals and see whether the tail norm fails to decay; failure of decay would certify non-compactness in concrete examples.","The Hilbert-space theorem can be read as a compactness criterion for infinite tuples of observables: an infinite sequence of bounded operators is jointly compact exactly when its joint numerical range sits as a totally bounded subset of $\\ell^p$, a notion that may be useful in infinite-dimensional quantum measurement settings.","The same dual-shadow construction could plausibly be extended to operators from $\\ell^q$ into non-commutative $L^p$ spaces or operator spaces, replacing the dual unit ball with the appropriate operator-space dual; the paper does not take up that direction."],"forward_implications":["For $1 \\le p < \\infty$, compactness of $\\Lambda_a$ is equivalent to a uniform tail estimate: for every $\\varepsilon > 0$ there is an index $m$ such that $(\\sum_{i>m} |\\varphi(a_i)|^p)^{1/p} < \\varepsilon$ for all $\\varphi$ in the unit ball of $A^*$.","When $A = C(\\Omega)$, compactness of $\\Lambda_F$ is equivalent to continuity of the vector-valued map $F: \\Omega \\to \\ell^p$ and to total boundedness of its image $F(\\Omega)$.","When $A = B(H)$, compactness of $T: \\ell^q \\to B(H)$ is equivalent to total boundedness of the joint numerical range $W(T)$ in $\\ell^p$, and the joint numerical radius satisfies $\\frac{1}{2}\\|T\\| \\le \\omega(T) \\le \\|T\\|$.","The normed spaces $\\ell^p_b(A)$ and $\\ell^p_c(A)$ give an isometric model for bounded and compact operators respectively, so $K(\\ell^q, A)$ is a closed subspace isometrically isomorphic to $\\ell^p_c(A)$.","At the endpoint $p = \\infty$, the criterion says $\\Lambda: \\ell^1 \\to A$ is compact exactly when the set $\\{a_i\\}$ is totally bounded in $A$, a purely geometric condition on the target space."],"supporting_citations":[{"why":"Supplies the theorem that an operator is compact if and only if its adjoint is compact, the bridge between $\\Lambda_a$ and its dual shadow.","marker":"[7, Theorem 4.19]"},{"why":"Supplies the compactness criterion for subsets of $\\ell^p$ in terms of uniform tail decay, used throughout to identify $\\ell^p_c(A)$.","marker":"[4, Theorem 4]"},{"why":"Supplies the weak* continuity result used to prove that compact operators on reflexive domains are weak* continuous on bounded sets.","marker":"[5, Theorem 3.10.1]"},{"why":"Provides the numerical radius inequality $\\frac{1}{2}\\|T\\| \\le w(T) \\le \\|T\\|$ used in bounding the joint numerical radius in the Hilbert-space theorem.","marker":"[3]"},{"why":"Proves that the closed convex hull of a totally bounded set in $\\ell^p$ is compact, used for the $\\ell^\\infty$ endpoint and the $C(\\Omega)$ case.","marker":"[7, Theorem 3.20]"}],"fun_headline_variants":["Compact ℓ^q operators: one total boundedness test","Dual shadow in ℓ^p decides if an ℓ^q operator is compact","Totally bounded coefficient images characterize compact ℓ^q maps","A compactness criterion for operators from ℓ^q to Banach spaces","Joint numerical range total boundedness: compactness for ℓ^q operators"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the characterization remains valid at the two endpoint cases $p = 1$ and $p = \\infty$, where the proof uses separate arguments instead of the standard tail criterion that applies for $1 < p < \\infty$.","fun_headline_variants_meta":{"raw":{"variants":["Compact ℓ^q operators: one total boundedness test","Dual shadow in ℓ^p decides if an ℓ^q operator is compact","Totally bounded coefficient images characterize compact ℓ^q maps","A compactness criterion for operators from ℓ^q to Banach spaces","Joint numerical range total boundedness: compactness for ℓ^q operators"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000438,"raw_usage":{"total_tokens":2300,"prompt_tokens":1092,"completion_tokens":1208,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":708,"completion_tokens_details":{"reasoning_tokens":1110}},"tokens_in":708,"tokens_out":1208,"duration_ms":11601,"temperature":1.0,"reasoning_tokens":1110,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:52:10.232317+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct way to test the endpoint case is to set $A = \\ell^\\infty$ and take any sequence $a = (a_i)$ whose set of entries is totally bounded in $A$; the theorem predicts the induced operator $\\Lambda_a(\\beta) = \\sum \\beta_i a_i$ on $\\ell^1$ is compact. Finding a totally bounded sequence of that kind whose induced operator is not compact would refute the characterization, so a search over such sequences is a concrete falsification test.","supporting_citations":[{"cited_title":"Bhunia, S","cited_arxiv_id":null,"evidence_quote":"Provides the numerical radius inequality $\\frac{1}{2}\\|T\\| \\le w(T) \\le \\|T\\|$ used in bounding the joint numerical radius in the Hilbert-space theorem."}],"review_version":1}