{"id":"a3dd114b-a12a-4c4e-a487-928db1f1c908","arxiv_id":"2505.05998","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A one-parameter family of genuine multipartite entanglement measures is defined as the geometric mean of bipartite α-concurrences over all bipartitions, shown to satisfy the standard monotonicity axioms and to rank GHZ states above W states.","lead":"The authors define a new family of numbers that quantify genuine multipartite quantum entanglement, one for each value of a tunable parameter, using the geometric mean of bipartite entanglement values across every way of splitting a system. This gives researchers another tool for ordering quantum states, including cases where older measures assign the same value.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Continuity claim fails at α=0: Gα is discontinuous at product states despite Theorem 2 claiming continuity for all 0≤α≤1/2.","rationale":"The paper's central claim is that GαC is a well-defined GME measure for 0≤α≤1/2 with the additional property of continuity for pure states. After independent review, the construction for 0<α≤1/2 appears internally coherent: the proof of Theorem 1 follows from the bipartite monotonicity of [33], Mahler's inequality, and the standard convex-roof inheritance; the W-state formulas are consistent with the correct reduced density eigenvalues because the coherence in the one-excitation subspace yields a single k/n eigenvalue, not k separate 1/n eigenvalues. The genuinely load-bearing flaw is the endpoint α=0, where the α-concurrence becomes rank-1, destroying continuity. The reader's weakest_assumption identified exactly this issue, and our explicit example with GHZ-like states confirms it. Because the error is confined to the endpoint and removable by restricting to α>0, the reader's CONDITIONAL verdict is appropriate; no change is needed.","tokens_in":8801,"tokens_out":38498,"duration_ms":351415,"concrete_test":"Evaluate G_0 on the family |ψ(ε)⟩=√(1-ε)|000⟩+√ε|111⟩ for a sequence ε_k→0, and compare with G_0(|000⟩). For every ε>0, compute the reduced density matrix of one qubit (and of two qubits) for |ψ(ε)⟩; each has rank 2, so C_0=1 for every bipartition and G_0=1. At ε=0, the state is |000⟩, all reduced ranks are 1, and G_0=0. This demonstrates the discontinuity and falsifies Theorem 2 at α=0. If the authors intend the open interval, the test should instead verify that Lemma 1's bound (9) holds for α>0 by checking a sequence approaching a product state and observing |C_α|→0 as ε^α→0.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central measure is defined for 0≤α≤1/2, and Theorem 2 (with Lemma 1) asserts continuity of Gα on pure states for this full range. At α=0, however, C_0(|ψ⟩)=rank(ρ_A)-1, which is not a continuous function of the state: any arbitrarily small entanglement makes the reduced density matrix rank 2, so C_0 jumps from 0 to 1. Concretely, take |ψ(ε)⟩=√(1-ε)|000⟩+√ε|111⟩ in a three-qubit system. For every ε>0, each single-qubit reduced density has eigenvalues 1-ε and ε, hence rank 2, so C_0=1 for every bipartition and G_0(|ψ(ε)⟩)=1. At ε=0, |ψ(0)⟩=|000⟩ is product, all reduced ranks are 1, and G_0=0. Thus G_0 is discontinuous at ε=0, contradicting Theorem 2. The proof of Lemma 1 also collapses at α=0 because the bound |p_i^α-r_i^α|≤ε_i^α becomes trivial (ε_i^0=1), yielding only |C_0(ψ1)-C_0(ψ2)|≤d, which does not vanish as ε→0. This is not a peripheral technicality: the abstract explicitly advertises continuity for the parameter range including α=0. A straightforward remedy is to restrict the definition and all theorems to 0<α≤1/2, where C_α is continuous.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a one-parameter family of genuine multipartite entanglement (GME) measures G_α, defined as the geometric mean, over all bipartitions, of the bipartite α-concurrence, with a convex-roof extension to mixed states. The main claims are: (i) Theorem 1 that G_α satisfies the standard axioms M1–M5 for all 0 ≤ α ≤ 1/2; (ii) Theorem 2 that G_α is continuous on multipartite pure states over the same parameter range; (iii) analytic formulas for n-qubit GHZ and W states showing GHZ > W in this measure; and (iv) explicit examples in which G_α distinguishes states that other GME measures (GMC, GGM, concurrence fill, G_3C) do not. The construction is a parameterized analogue of the geometric-mean concurrence (GBC) and geometric-mean q-concurrence (GqC) measures.","tokens_in":9134,"tokens_out":8209,"duration_ms":91711,"significance":"If the stated results hold for 0 < α ≤ 1/2, the paper provides a clean, tunable family of GME measures with exact formulas for GHZ and W states and illustrative examples of ordering differences. The analytic GHZ/W formulas and the explicit distinguishing examples are concrete assets, and the construction uses no fitted parameters. The main caveat is that the advertised full parameter range includes α = 0, where the continuity claim is false, and the continuity proof as written has a dimension-dependent gap. The monotonicity of the building block is imported from Ref. [33]. With the α = 0 endpoint removed and the proof gaps addressed, the central construction is plausible and would be a useful addition to the GME-measure toolbox.","major_comments":[{"comment":"The continuity claim fails at α = 0, which is included in the stated range 0 ≤ α ≤ 1/2. For the three-qubit family |ψ(ε)⟩ = √(1−ε)|000⟩ + √ε|111⟩, every single-qubit reduced density has rank 2 for ε > 0, so C_0 = rank − 1 = 1 for every bipartition and hence G_0(|ψ(ε)⟩) = 1, while G_0(|000⟩) = 0. The bound in Lemma 1 also degenerates at α = 0 because |p_i^0 − r_i^0| ≤ ε_i^0 = 1 gives only |C_0(ψ1) − C_0(ψ2)| ≤ d, which does not vanish as ε → 0. The proof can be repaired by restricting the definition, Theorem 1, Theorem 2, and the abstract to 0 < α ≤ 1/2; alternatively, separate treatment of α = 0 would have to confront the genuine discontinuity.","section":"Section III, Lemma 1 and Theorem 2"},{"comment":"The continuity proof assumes that all α-concurrence values lie in (0,1), but this is false for subsystems of dimension at least 3. For a maximally entangled qutrit pair, C_α = 3^{1−α} − 1 > 1 for α ≤ 1/2 (e.g., α = 1/3 gives 3^{2/3} − 1 ≈ 1.08). The inequality |x_1···x_k − y_1···y_k| ≤ Σ_i |x_i − y_i| used in Eq. (14) is not valid when the factors exceed 1. This does not disprove continuity, but it invalidates the stated proof for arbitrary multipartite pure states. A corrected proof needs a dimension-dependent uniform bound on C_α (or on each factor) in the product-difference estimate, or the theorem should be restricted to systems where all C_α ≤ 1.","section":"Section III, proof of Theorem 2, Eqs. (13)–(14)"},{"comment":"The proof of the GME axioms is terse at points that carry real weight. For (M3) and (M5), the manuscript imports from Ref. [33] the monotonicity and average monotonicity of bipartite α-concurrence without stating the precise assumptions under which those results hold (pure vs. mixed states, and whether α = 0 is included). The passage from pure-state monotonicity to the convex-roof measure G_α is also only asserted: it should be justified explicitly, for instance by invoking the standard convex-roof argument for strongly monotone pure-state measures, or by citing the specific theorem in Ref. [33] that covers mixed states. For (M1), the positivity of G_α on genuine multipartite entangled mixed states requires a compactness or closedness argument for the set of biseparable states; the phrase 'directly implies' is not sufficient as written.","section":"Section III, proof of Theorem 1, items (M1), (M3), (M5)"}],"minor_comments":[{"comment":"There are several typographical issues: 'duet to' should be 'due to', 'wit h respect' should be 'with respect', 'Cresàro' should be 'Cesàro', and 'R´enyi' should be 'Rényi'.","section":"Throughout"},{"comment":"The summation notation in the continuity bound is ambiguous for even n: the upper limit appears to be (n−1)/2 in all cases, but for even n the sum over bipartitions must run to n/2 − 1 plus the special central-bipartition contribution, consistent with the definition of c(β). Please clarify the piecewise form.","section":"Theorem 2, Eq. (12)"},{"comment":"In the displayed formula after Example 1, 'Gα|GHZn⟩)' should read 'Gα(|GHZn⟩)'.","section":"Example 1"},{"comment":"The statement 'The last equality holds when all ε_i = ε/d' is not quite accurate: equality in the Hölder-type bound requires both Σ ε_i = ε and equal ε_i, and the condition on the r_i should be stated if the equality is meant exactly.","section":"Section III, after Lemma 1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript relies heavily on Ref. [33] by two of the present authors for the LOCC monotonicity and strong monotonicity of α-concurrence. Since the present paper does not re-derive those results, the editor may wish to have the validity of Ref. [33] for mixed states and for the endpoint α = 0 checked independently. The main technical claims of the present paper are otherwise straightforward to verify for 0 < α ≤ 1/2, provided the continuity proof is corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nHere's my take on arXiv:2505.05998. The paper defines GαC, the geometric mean of bipartite α-concurrences over all bipartitions, as a one-parameter family of genuine multipartite entanglement measures. The construction is exactly what it appears to be: a routine extension of GBC [42] and GqC [43] to the α-concurrence of Ref. [33]. That's not damning—the W-state formulas and the ordering examples are new, and the family gives you a tunable measure with analytic results for GHZ and W states. For 0<α≤1/2, the M1-M5 proof is mostly sound, modulo some hand-waving in the convexity and the convex-roof inheritance steps.\n\nThe problem is the stated range. The paper defines α-concurrence for 0≤α≤1/2 and Theorem 2 claims continuity for all pure states in that range. At α=0, C_0(|ψ⟩)=rank(ρ_reduced)-1, which is 0 for any product state and 1 for any state with full-rank single-system reductions. The GHZ-type family |ψ(ε)⟩=√(1-ε)|000⟩+√ε|111⟩ has G_0=1 for every ε>0 and G_0=0 at ε=0. So G_0 is discontinuous. The lemma proof collapses because |p^0-r^0|≤ε^0=1 gives a trivial bound. This is a real error, but it has an obvious fix: restrict the definition and all claims to 0<α≤1/2.\n\nThere's a second, subtler gap in the continuity proof that survives for α>0. The proof of Theorem 2 uses inequality (14), |x_1...x_k-y_1...y_k|≤Σ|x_i-y_i|, which they state for x_i,y_i∈(0,1). But α-concurrence can exceed 1—for a maximally entangled state of dimension d it's d^{1-α}-1, which is >1 for d large enough. So the bound as written doesn't apply. The continuity claim is probably true anyway, but the proof needs a different bound or a normalization.\n\nThe distinguishing-power claims are supported only by examples, as stated. That's fine, but the abstract's \"can distinguish different GME states that other GME measures fail to\" is a bit strong. The reliance on Ref. [33] for LOCC monotonicity is acceptable—it's the standard source, and two of the authors are behind it—but it would be better if they restated the needed result as a lemma.\n\nBottom line: this is a useful paper for the entanglement-measure community, with a fixable but genuine flaw. It deserves a serious referee, not a desk reject, and the referee should require the α>0 restriction and a patched continuity proof. I'd bring it to a reading group and would cite it if I worked on GME measures, after the correction.","headline":"A competent but routine one-parameter extension of geometric-mean GME measures; fix the α=0 endpoint and the continuity proof and it's refereeable.","tokens_in":9700,"tokens_out":5398,"would_cite":true,"duration_ms":50441,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper constructs a one-parameter family of genuine multipartite entanglement measures, $G_\\alpha$, as the geometric mean of bipartite $\\alpha$-concurrences and proves that it satisfies the standard axioms, is continuous on pure…","keywords":["genuine multipartite entanglement","$\\alpha$-concurrence","geometric mean","convex roof extension","LOCC monotonicity","GHZ states","W states","entanglement ordering"],"falsifier":"Compute $G_0$ on the family $|\\psi(\\epsilon)\\rangle=(|000\\rangle+\\epsilon|111\\rangle)/\\sqrt{1+\\epsilon^2}$: for every $\\epsilon>0$ each bipartition has a rank-2 reduced state, so $G_0=1$, while at $\\epsilon=0$ the state is separable and $G_0=0$; this makes $G_0$ discontinuous at a pure state, directly contradicting the pure-state continuity claim if the stated range includes $\\alpha=0$.","tokens_in":8598,"feed_emoji":"🔗","tokens_out":14022,"duration_ms":138098,"temperature":0.7,"pith_summary":"The paper introduces a one-parameter family of genuine multipartite entanglement measures, denoted $G_\\alpha$, built as the geometric mean of bipartite $\\alpha$-concurrences over every possible bipartition of an $n$-partite system. Its central claim is that $G_\\alpha$ satisfies the standard axioms of a multipartite entanglement measure for all $0\\le\\alpha\\le 1/2$, is continuous on pure multipartite states, and admits exact formulas for $n$-qubit GHZ and W states. If correct, this gives a tunable quantitative scale for genuine multipartite entanglement that can distinguish states other established measures rank as equal. Such a scale matters for tasks such as quantum secret sharing and teleportation, where the amount of genuine multipartite entanglement is the resource being consumed.","feed_headline":"New measure ranks multipartite entanglement with one tunable knob","feed_subtitle":"Geometric mean of α-concurrences yields a genuine-entanglement family that separates GHZ from W states.","key_machinery":"The central object is the $\\alpha$-concurrence of a bipartition, $C_\\alpha^{S|\\bar S}(|\\psi\\rangle)=\\operatorname{Tr}(\\rho_S)^{\\alpha}-1$ for $0\\le\\alpha\\le 1/2$, where $\\rho_S$ is the reduced state on one side of the bipartition; it interpolates between rank-like counting and ordinary concurrence-type behavior. The multipartite measure multiplies such concurrences over all pairwise complementary bipartitions, takes the root given by the number of bipartitions, and extends to mixed states by convex roof. The load-bearing steps are the strong monotonicity of each bipartite factor and the geometric-mean inequality that lets the root of a product pass through a sum of ensemble averages.","core_discovery":"On its own terms, the paper's discovery is that the convex-roof extension of the geometric mean of bipartite $\\alpha$-concurrences is a genuine multipartite entanglement measure: $G_\\alpha(\\rho)$ vanishes exactly on biseparable states, is convex, does not increase under local operations and classical communication, is invariant under local unitaries, and satisfies strong monotonicity on average under LOCC. The proof carries the monotonicity property from each bipartite $\\alpha$-concurrence to the product, using the geometric-mean inequality that converts a product of sums into a sum of products. The paper also proves a continuity bound for pure multipartite states and derives exact closed forms: $G_\\alpha(|GHZ_n\\rangle)=2^{1-\\alpha}-1$, and a product formula for $|W_n\\rangle$ that is always smaller, so by this measure GHZ states are more genuinely entangled than W states.","pith_inferences":["Because $\\alpha$-concurrence approaches a rank-based count as $\\alpha\\to0$, the $G_\\alpha$ family interpolates between Schmidt-rank-like and entropy-like entanglement quantifiers; this suggests choosing $\\alpha$ according to which operational resource a protocol consumes.","The proof structure is generic: any bipartite entanglement monotone that is strongly monotone under LOCC could replace $\\alpha$-concurrence in the geometric mean and yield a valid GME measure, so the construction is a template rather than a single measure.","The continuity bound for $\\alpha>0$ suggests a practical estimation route: if a prepared state is certified to trace-norm error $\\epsilon$, its $G_\\alpha$ value is known to within $O(\\epsilon^\\alpha)$, which may make the measure accessible without full tomography.","One could test experimentally whether the predicted GHZ-over-W ordering survives depolarizing noise; because $\\alpha$ is tunable, the measure may reveal a noise threshold where the ordering reverses."],"forward_implications":["For every $\\alpha$ in the stated range, $G_\\alpha$ is a legitimate genuine multipartite entanglement measure, so it can be assigned to arbitrary mixed multipartite states in resource-theoretic protocols.","$G_\\alpha(|GHZ_n\\rangle)=2^{1-\\alpha}-1$ for any number of qubits, while $G_\\alpha(|W_n\\rangle)$ is always smaller, giving a quantitative sense in which GHZ states carry more genuine multipartite entanglement than W states.","Absolutely maximally entangled states, whose half-system reductions are maximally mixed, attain the maximal value of $G_\\alpha$ for every allowed $\\alpha$.","The parameter $\\alpha$ yields inequivalent orderings: explicit three- and four-qubit examples show pairs of states with equal values of the genuine multipartite concurrence (GMC) or the generalized geometric measure (GGM) but different $G_{1/2}$, and ranges of parameters where the ordering is reversed.","The continuity bound means that, for $\\alpha>0$, a pure state moved by trace-norm distance $\\epsilon$ changes $G_\\alpha$ by at most a constant times $\\epsilon^\\alpha$, so the measure is stable under small perturbations."],"supporting_citations":[{"why":"Establishes the LOCC nonincrease and strong monotonicity of bipartite $\\alpha$-concurrence, which Theorem 1 imports for the monotonicity axioms.","marker":"[33]"},{"why":"Supplies the geometric-mean-of-bipartite-concurrence construction that $G_\\alpha$ generalizes.","marker":"[42]"},{"why":"Introduces the parameterized geometric mean of $q$-concurrence, the direct precursor of the $\\alpha$-parameter family.","marker":"[43]"},{"why":"Defines genuine multipartite concurrence and provides the axiomatic setting used for comparisons.","marker":"[5]"},{"why":"States the axioms (M1)-(M5) that a well-defined entanglement measure must satisfy.","marker":"[4]"},{"why":"Provides the unitarily invariant norm inequalities used in the continuity lemma for bipartite $\\alpha$-concurrence.","marker":"[44]"},{"why":"Gives the earlier result that GHZ states are more genuinely entangled than W states, which the paper's formulas reproduce.","marker":"[50]"},{"why":"Supplies the companion ordering result that supports the GHZ-versus-W comparison.","marker":"[51]"}],"fun_headline_variants":["α-concurrence family quantifies genuine multipartite entanglement","Geometric mean of α-concurrences: new GME measure","One-parameter GME measure: GHZ beats W states","Genuine entanglement measure with tunable α"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The construction leans on the previously established fact that bipartite $\\alpha$-concurrence never increases, on average, under local operations and classical communication for every $\\alpha$ in the stated range; if that fact has an exception at $\\alpha=0$ or for mixed states, the measure's defining axioms lose their foundation.","fun_headline_variants_meta":{"raw":{"variants":["α-concurrence family quantifies genuine multipartite entanglement","Geometric mean of α-concurrences: new GME measure","One-parameter GME measure: GHZ beats W states","Genuine entanglement measure with tunable α"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000178,"raw_usage":{"total_tokens":1269,"prompt_tokens":891,"completion_tokens":378,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":507,"completion_tokens_details":{"reasoning_tokens":309}},"tokens_in":507,"tokens_out":378,"duration_ms":4203,"temperature":1.0,"reasoning_tokens":309,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:52:30.266848+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute $G_0$ on the family $|\\psi(\\epsilon)\\rangle=(|000\\rangle+\\epsilon|111\\rangle)/\\sqrt{1+\\epsilon^2}$: for every $\\epsilon>0$ each bipartition has a rank-2 reduced state, so $G_0=1$, while at $\\epsilon=0$ the state is separable and $G_0=0$; this makes $G_0$ discontinuous at a pure state, directly contradicting the pure-state continuity claim if the stated range includes $\\alpha=0$.","supporting_citations":[{"cited_title":"Wei, S.-M","cited_arxiv_id":null,"evidence_quote":"Establishes the LOCC nonincrease and strong monotonicity of bipartite $\\alpha$-concurrence, which Theorem 1 imports for the monotonicity axioms."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the geometric-mean-of-bipartite-concurrence construction that $G_\\alpha$ generalizes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the parameterized geometric mean of $q$-concurrence, the direct precursor of the $\\alpha$-parameter family."},{"cited_title":"Ma, Z.-H","cited_arxiv_id":null,"evidence_quote":"Defines genuine multipartite concurrence and provides the axiomatic setting used for comparisons."},{"cited_title":"Bhatia, Matrix Analysis Springer, Berlin, 2013","cited_arxiv_id":null,"evidence_quote":"Provides the unitarily invariant norm inequalities used in the continuity lemma for bipartite $\\alpha$-concurrence."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the earlier result that GHZ states are more genuinely entangled than W states, which the paper's formulas reproduce."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the companion ordering result that supports the GHZ-versus-W comparison."}],"review_version":1}