{"id":"5b6138bb-6677-4860-bf58-7c110cbaba66","arxiv_id":"1908.03466","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Every 2-positive order zero map between C*-algebras is completely positive, and for unital separable C*-algebras finite decomposition rank can be characterized using 2-positive maps in place of completely positive ones.","lead":"This mathematics paper proves that a weaker positivity condition, 2-positivity, is enough to detect order zero behavior in maps between C*-algebras. It also shows that decomposition rank, a key dimension invariant in the Elliott classification program, can be characterized using only these weaker maps.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the load-bearing Schwarz inequality is valid, including the non-invertible case.","rationale":"The reader identified Proposition 2.5(i) as the hidden load-bearing step. My reading agrees that the Schwarz inequality machinery is the most delicate part, but the step actually used in the proof of Theorem 1.1 is Proposition 2.5(ii), not (i), and that inequality is robust: it follows from commutativity of x and y and complete positivity of positive maps on commutative subalgebras. The non-invertible issue worried by the reader is explicitly addressed by Lemma 2.3 and Corollary 2.4. I checked the surrounding norm estimates in Proposition 3.5 and the passage from positive contractions to arbitrary contractions in Theorem 1.1, and I found no concrete failure. The reduction from 2-positive order zero maps to completely positive maps in Corollary 3.7 is noncircular and uses only previously established results. Since no load-bearing concern survives scrutiny, the reader’s ACCEPT verdict should stand unchanged.","tokens_in":19546,"tokens_out":47896,"duration_ms":504458,"concrete_test":"Independently re-derive Proposition 2.5(ii) without citing [6, Corollary 4.4], using only the commutativity of x and y and the fact that positive maps are completely positive on commutative C*-subalgebras. If this direct derivation fails, the matrix inequality used in Proposition 3.5 would be unsupported and Theorems 1.1 and 1.2 would require revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"After following the proof chain Theorem 1.2 → Corollary 3.7 → Corollary 3.6 → Theorem 1.1 → Proposition 3.5, I found no load-bearing flaw. The most exposed input is Proposition 2.5(ii), the two-variable Schwarz inequality for merely positive maps, which is applied to φ⊗id_{M_2} in Proposition 3.5. This step deserves scrutiny, but it is sound: the hypotheses x = x*, y ≥ 0, and yx = x force xy = x, so x and y commute and generate a commutative C*-subalgebra. Every positive map restricts to a completely positive map on a commutative C*-algebra, so the 2×2 matrix [[φ(x²), φ(x)], [φ(x), φ(y)]] is positive. The generalized inverse of φ(y) is handled by Lemma 2.3 and Corollary 2.4, including the non-invertible case via the support projection. I also checked that the epsilon estimates in Proposition 3.5 are applied to the correct entries of the positive matrix X and that the reduction from positive elements to all contractions is legitimate. No circular dependence, hidden parameter, or inconsistent constant was found. The cited external tools, Choi’s inequality and the Choi–Effros lifting theorem, are standard and apply in the stated settings.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies 2-positive maps with an almost order-zero (disjointness-preserving) property. Theorem 1.1 gives an internal characterization: if a 2-positive contraction φ from A to B satisfies limsup ||φ(a)^2 − φ(a^2)φ(h_λ)|| < δ for a positive contraction a, then φ behaves like h_φπ(·) up to ε, where h_φ is the weak*-limit of φ(h_λ). The proof proceeds by a separable reduction and a Schwarz inequality for positive maps on commutative C*-subalgebras. Corollary 3.7 shows that every 2-positive order zero map is completely positive. Theorem 1.2, the main application, characterizes decomposition rank at most d for unital separable C*-algebras using 2-positive maps instead of completely positive maps in the defining approximation, via a one-sided CPAP characterization of nuclearity (Theorem 5.1) and the Choi–Effros lifting theorem. Section 4 constructs examples of k-positive almost order zero maps that are not (k+1)-positive.","tokens_in":19762,"tokens_out":31872,"duration_ms":314992,"significance":"If the results stand, Theorem 1.2 is a useful simplification: checking decomposition rank requires only 2-positive maps rather than completely positive ones. Theorem 1.1 extends Choi's multiplicative-domain argument and yields a short route to the Winter–Zacharias structure theorem under 2-positivity. Corollary 3.7 is a clean and natural structural statement. The proofs are detailed, self-contained, and rely on standard, independently established tools (Kadison, Choi, Choi–Effros, Tomiyama, Winter–Zacharias). The constants are explicit, and no free parameters, fitted data, or dependence on the author's earlier results enter the proofs of the main theorems. The Section 4 examples usefully delineate k-positivity from (k+1)-positivity in the almost-order-zero setting.","major_comments":[],"minor_comments":[{"comment":"The proof of (iii)⇒(i) produces approximations only on finite sets of unitaries, whereas condition (i) is stated for contractions. The standard Russo–Dye argument (the unit ball is the closed convex hull of the unitaries) should be added to justify the passage to contractions. In the final step, the equality φ∘Q∘ψ(a)=a in the quotient yields the desired approximation only for sufficiently large indices; the proof should explicitly pass from a given μ=(F,ε) to a larger μ′ to obtain the stated error for all x∈F. As written, the sentence 'we conclude that ψ_μ and φ_i,μ satisfy the conditions in (i)' is not literally justified.","section":"§6, proof of Theorem 6.2"},{"comment":"The definition of y in the displayed equation following the Schwarz inequality should read φ(a²+b²) − [φ(a)g_{α2}(φ(h_n))φ(a) + φ(b)g_{α2}(φ(h_n))φ(b)]; the missing bracket makes the formula ambiguous. The asserted bound '‖X‖ε₁ < 5ε₁' is not explained; it follows from the estimate ‖Y‖ ≤ ‖Y₁₁‖ + ‖Y₂₂‖ ≤ 3 for the relevant positive 2×2 block Y, so ‖X‖ ≤ ‖Y‖ < 5. A one-sentence justification would improve readability.","section":"§3, Proposition 3.5"},{"comment":"The estimate ‖φ^{(m)}_λ(x)² − φ^{(m)}_λ(x²)‖ < 6ε is stated without proof. Since this estimate is the point of the example, a brief verification using the fact that the second summand is supported on the (1,1)-corner would be helpful.","section":"§4, Example 4.1"},{"comment":"The proof of Proposition 3.5 assumes ε∈(0,1) at the outset, but the reduction from arbitrary ε>0 to this case is not stated. The reduction is immediate because the expression to be bounded is always at most 2, but it should be mentioned.","section":"§3, Proof of Theorem 1.1"},{"comment":"There are minor typographical issues: 'Schwartz inequality' should be 'Schwarz inequality' in Section 2; in Corollary 2.4(ii) the expression 'ϕ−1/2(y)ϕ(x)' should be 'ϕ(y)−1/2ϕ(x)'; and 'disjointness preserving' appears where 'order zero' is meant in the abstract and introduction.","section":"Throughout"}],"recommendation":"minor_revision","confidential_remarks":"The paper is in good shape. The main technical claims appear correct, and I found no circularity or hidden dependence on the author's prior results. The only issues are local: the proof of Theorem 6.2 needs standard cofinality and Russo–Dye details, and a few displays would benefit from clarification. These can be repaired without further substantive work, so I recommend minor revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Sato proves that 2-positivity suffices in the order zero and decomposition rank framework. The headline result, Corollary 3.7, is that every 2-positive order zero map is completely positive, so replacing complete positivity by 2-positivity in Kirchberg–Winter's decomposition rank costs nothing. Theorem 1.2 states that equivalence cleanly. That is the real content, and it is new, not a repackaging of the CP theory.\n\nThe paper does several things well. Theorem 1.1 gives an internal 'almost order zero' criterion for 2-positive maps using approximate units, extending Choi's multiplicative domain argument. The proof of the structure theorem for 2-positive order zero maps is detailed and follows standard machinery: Kadison's inequality, Choi matrices, separable reduction, and the Choi–Effros lifting theorem. The Section 4 examples, showing k-positive almost order zero maps that are not (k+1)-positive, are a useful calibration. The citation pattern is healthy; the author's own prior work appears only as context, not as a load-bearing input.\n\nThe soft spots are minor and mostly about readability. The epsilon estimates in Proposition 3.5 are intricate, with many interdependent constants; I could not check every inequality line-by-line, but the stress-test note's examination of the exposed Schwarz inequality, including the non-invertible case via support projections, checks out. I do not see a load-bearing gap. The proof of Theorem 1.2 depends on Choi–Effros lifting, which requires nuclearity and separability at that point; the paper states the natural unital separable setting, so that is not a flaw. The writing is terse, and some reductions are compressed, but nothing seems invented or circular.\n\nThis paper is for people working in C*-algebra classification and finite-dimensional approximation. It clarifies the role of 2-positivity and supplies a convenient criterion. I would bring it to a reading group and would cite it in my own work.\n\nRecommendation: send it to a serious referee. Even if one asks for expository improvements, the mathematics deserves referee time. My verdict is accept.","headline":"A genuinely useful 2-positive analogue of the order zero machinery, highlighted by Corollary 3.7 and Theorem 1.2; the proof holds up under scrutiny.","tokens_in":20303,"tokens_out":2019,"would_cite":true,"duration_ms":23323,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46L05","46L35","46L07"],"pacs":[],"model":"deepseek-v4-flash","headline":"For 2-positive maps between C*-algebras, almost order zero is detected by a one-element condition, and decomposition rank at most d is equivalent to approximation by 2-positive order zero maps from finite-dimensional algebras.","keywords":["2-positive maps","order zero maps","almost order zero","orthogonality domain","decomposition rank","complete positivity","C*-algebras","k-positive maps"],"falsifier":"A direct falsifier would be a finite-dimensional counterexample: a unital 2-positive map $\\varphi$ from $M_3$ to $M_3$ such that $\\varphi(a)^2 = \\varphi(a^2)\\varphi(1)$ for every positive contraction $a$, but whose amplification $\\varphi\\otimes\\mathrm{id}_{M_3}$ is not positive. The paper asserts that no such map exists; finding one, for instance by searching over the $9\\times 9$ block matrix of the map's matrix-unit values, would refute the order-zero-to-complete-positivity claim and the decomposition-rank theorem.","tokens_in":19306,"feed_emoji":"🧮","tokens_out":19506,"duration_ms":176952,"temperature":0.7,"pith_summary":"The paper tries to establish that for 2-positive maps between C$^*$-algebras, the property of being (almost) order zero can be checked one positive element at a time, and that this makes 2-positivity sufficient in the definition of decomposition rank. It proves that a 2-positive contraction satisfies the order-zero equation $\\varphi(a)^2 = \\varphi(a^2)\\varphi(1)$ for all positive contractions $a$ exactly when it is order zero, and more generally that a near-order-zero condition on a single $a$ controls $\\varphi(a)\\varphi(b)$ against $h_{\\varphi}\\varphi(ab)$ for every $b$. A direct corollary is that every 2-positive order zero map is completely positive. The second main theorem says that for a unital separable C$^*$-algebra, having decomposition rank at most $d$ is equivalent to the existence of approximations by 2-positive order zero maps from finite-dimensional algebras, so the 'completely positive' in the standard definition can be weakened to '2-positive'. A sympathetic reader would care because this simplifies the verification of decomposition rank and sharpens the boundary between 2-positivity and complete positivity.","feed_headline":"2-positivity is enough for decomposition rank","feed_subtitle":"A single positive contraction captures almost order zero; finite decomposition rank no longer needs complete positivity.","key_machinery":"The machinery is the orthogonality domain $\\mathrm{OD}(\\varphi)$, the subspace of elements $a$ for which $\\varphi(a)\\varphi(b) = \\lim_\\lambda \\varphi(h_\\lambda)\\varphi(ab)$ and $\\varphi(b)\\varphi(a) = \\lim_\\lambda \\varphi(ba)\\varphi(h_\\lambda)$ for all $b$, together with a two-matrix inequality $\\varphi(a^*b)\\varphi(b^*b)^{-1}\\varphi(b^*a) \\le \\varphi(a^*a)$ in the second dual of $B$. The paper shows that for 2-positive $\\varphi$ the orthogonality domain is a C$^*$-algebra that contains the multiplicative domain, and the proof of the one-variable theorem uses the two-matrix inequality to turn the condition $\\varphi(a)^2 \\approx \\varphi(a^2)\\varphi(h_\\lambda)$ into the norm bound that places $a$ in $\\mathrm{OD}(\\varphi)$. That inclusion, together with matrix-entry estimates on an infinite direct sum of copies of $A$, is what carries the argument.","core_discovery":"The central claim is that for a 2-positive contraction $\\varphi$ from a C$^*$-algebra $A$ into $B$, the almost-order-zero condition that $\\limsup_\\lambda \\|\\varphi(a)^2 - \\varphi(a^2)\\varphi(h_\\lambda)\\|$ be small forces the global almost-multiplicativity relation $\\sup_{\\|b\\|\\le 1} \\|\\varphi(a)\\varphi(b) - h_\\varphi\\varphi(ab)\\|$ to be small, where $h_\\varphi$ is the weak$^*$-limit of $\\varphi(h_\\lambda)$. Hence a 2-positive map is order zero exactly when $\\varphi(a)^2 = \\varphi(a^2)\\varphi(1_A)$ for every positive contraction $a$. From this the paper derives that every 2-positive order zero map is completely positive, and it proves the decomposition-rank analogue: for a unital separable C$^*$-algebra $A$, decomposition rank at most $d$ is equivalent to approximation, for each finite set and each $\\varepsilon>0$, by compositions $(\\sum_i \\varphi_i)\\circ\\psi$ in which $\\psi$ is a 2-positive contraction into a finite direct sum of finite-dimensional C$^*$-algebras and each $\\varphi_i$ is a 2-positive order zero contraction.","pith_inferences":["The one-element criterion suggests a practical numerical certificate: checking $\\varphi(a)^2 \\approx \\varphi(a^2)\\varphi(h)$ on a small generating set of positive contractions could certify almost order zero behaviour without constructing the whole multiplicative domain.","If the 2-positive formulation of decomposition rank extends beyond separable unital algebras, decomposition rank could be certified by 2-positive approximations alone, which are often easier to build than completely positive ones in noncommutative probability and quantum information settings.","The examples of $k$-positive almost order zero maps that are not $(k+1)$-positive point to a graded hierarchy between $k$-positivity and complete positivity; the paper shows the hierarchy collapses at the order-zero level, but the behaviour of nearly-order-zero maps at finite $k$ is left unexplored."],"forward_implications":["Finiteness of decomposition rank for a unital separable C$^*$-algebra can be certified by 2-positive maps alone: for each finite set and tolerance one needs one 2-positive contraction into a finite direct sum and 2-positive order zero maps back, with no check over tensor products with arbitrary matrix algebras.","Every 2-positive order zero map admits the classical order-zero structure $\\varphi(a)=h_\\varphi\\pi(a)$ with $\\pi$ a $*$-homomorphism into the bidual, so such maps are automatically completely positive and can be used in place of completely positive order zero maps in dimension computations.","The one-variable criterion gives a practical test for almost order zero: measuring $\\varphi(a)^2 \\approx \\varphi(a^2)\\varphi(h_\\lambda)$ on a positive contraction $a$ controls $\\varphi(a)\\varphi(b)$ against $h_\\varphi\\varphi(ab)$ for every $b$, so no exhaustive family of multiplicative-domain tests is needed.","The equivalence in Theorem 1.2 ties finite decomposition rank to 2-positive order zero approximations passing through nuclearity, so the class of algebras captured is unchanged when complete positivity is relaxed to 2-positivity.","The paper's examples show the phenomenon is graded: for every $k$ there are $k$-positive almost order zero maps that are not $(k+1)$-positive, so the collapse to complete positivity happens exactly at the order-zero level, not for almost order zero maps."],"supporting_citations":[{"why":"It supplies the two-matrix inequality for 2-positive maps that is extended to non-invertible elements and drives the one-variable characterization.","marker":"[6]"},{"why":"It provides the basic inequality for positive contractions used to show that the orthogonality domain is a C$^*$-algebra.","marker":"[5]"},{"why":"It gives the standard structure theorem for completely positive order zero maps that the paper's 2-positive version generalizes and re-proves.","marker":"[38]"},{"why":"It introduces decomposition rank with completely positive order zero approximations, the notion rewritten with 2-positive maps in Theorem 1.2.","marker":"[23]"},{"why":"It supplies the completely positive lifting theorem used to convert the $*$-homomorphism in the proof of Theorem 1.2 into the required approximation.","marker":"[8]"},{"why":"It gives the theorem that nuclear C$^*$-algebras have the completely positive approximation property used in the nuclearity step.","marker":"[22]"},{"why":"It provides the nuclear C$^*$-algebra approximation results that, together with [22], yield the completely positive approximation property used in Theorem 5.1.","marker":"[7]"},{"why":"It supplies the lemma on unitaries and permutations used to build approximately central completely positive maps in the proof of Theorem 5.1.","marker":"[20]"},{"why":"It supplies the criterion for $k$-positivity of trace perturbations on matrix algebras used in the examples of $k$-positive almost order zero maps.","marker":"[34]"}],"fun_headline_variants":["2-positive order zero maps are always CP","Decomposition rank needs only 2-positivity","Almost order zero: 2-positivity suffices","Two-positive maps capture decomposition rank"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the two-matrix inequality for 2-positive maps, $\\varphi(a^*b)\\varphi(b^*b)^{-1}\\varphi(b^*a) \\le \\varphi(a^*a)$ in the second dual, continues to hold without assuming $\\varphi(b^*b)$ is invertible; all the norm estimates that place an element in the orthogonality domain rest on this extension.","fun_headline_variants_meta":{"raw":{"variants":["2-positive order zero maps are always CP","Decomposition rank needs only 2-positivity","Almost order zero: 2-positivity suffices","Two-positive maps capture decomposition rank"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000221,"raw_usage":{"total_tokens":1402,"prompt_tokens":851,"completion_tokens":551,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":467,"completion_tokens_details":{"reasoning_tokens":493}},"tokens_in":467,"tokens_out":551,"duration_ms":5873,"temperature":1.0,"reasoning_tokens":493,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:13:10.481069+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct falsifier would be a finite-dimensional counterexample: a unital 2-positive map $\\varphi$ from $M_3$ to $M_3$ such that $\\varphi(a)^2 = \\varphi(a^2)\\varphi(1)$ for every positive contraction $a$, but whose amplification $\\varphi\\otimes\\mathrm{id}_{M_3}$ is not positive. The paper asserts that no such map exists; finding one, for instance by searching over the $9\\times 9$ block matrix of the map's matrix-unit values, would refute the order-zero-to-complete-positivity claim and the decomposition-rank theorem.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the two-matrix inequality for 2-positive maps that is extended to non-invertible elements and drives the one-variable characterization."},{"cited_title":"Texts and Monographs in Physics","cited_arxiv_id":null,"evidence_quote":"It provides the basic inequality for positive contractions used to show that the orthogonality domain is a C$^*$-algebra."},{"cited_title":"Wolff, Disjointness preserving operators on C∗-algebras, Arch","cited_arxiv_id":null,"evidence_quote":"It gives the standard structure theorem for completely positive order zero maps that the paper's 2-positive version generalizes and re-proves."},{"cited_title":"Kirchberg, C ∗-nuclearity implies CPAP , Math","cited_arxiv_id":null,"evidence_quote":"It introduces decomposition rank with completely positive order zero approximations, the notion rewritten with 2-positive maps in Theorem 1.2."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the completely positive lifting theorem used to convert the $*$-homomorphism in the proof of Theorem 1.2 into the required approximation."},{"cited_title":"Kishimoto, N","cited_arxiv_id":null,"evidence_quote":"It gives the theorem that nuclear C$^*$-algebras have the completely positive approximation property used in the nuclearity step."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the nuclear C$^*$-algebra approximation results that, together with [22], yield the completely positive approximation property used in Theorem 5.1."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the lemma on unitaries and permutations used to build approximately central completely positive maps in the proof of Theorem 5.1."}],"review_version":1}