{"id":"26a567e8-68a4-481a-86ab-a77556b25a32","arxiv_id":"2608.03184","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Dense coding swapping moves superdense coding advantage between channels using two-qubit unitaries, with exact conditions for three-qubit pure states.","lead":"This paper introduces a quantum communication task called dense coding swapping, in which a two-qubit operation transfers the superdense coding advantage from one receiver to another in a shared multipartite state. The result is a characterization of which quantum states can perform this transfer, with potential use in routing quantum communication and cutting off compromised receivers.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Mixed-state swapping criterion overstates its domain: the Corollary omits the vanishing sender-magnetization assumption on which Theorem 3 and Proposition 1 rest.","rationale":"The paper's central three-qubit pure-state theorem (Theorem 2) appears analytically sound: the Schmidt-coefficient conditions correctly encode initial dense codeability of ρ_AE, non-dense codeability of ρ_AB, and non-ACVENN of ρ_AB, and the generalized W example verifies the boundary. The reader's weakest-assumption diagnosis is accurate and is the most load-bearing issue: the mixed-state sufficient criterion is the main claimed generalization beyond pure states, yet the derivation of both the DC condition (Theorem 3) and the NDC condition (Proposition 1) explicitly requires the sender's local Bloch vector to vanish, and the Corollary applies these thresholds to arbitrary three-qubit states without that restriction. In addition, the NDC threshold is not proven analytically; it is the output of a finite random-state optimization, so even within the m=0 class the sufficiency is numerical rather than rigorous. Neither issue invalidates Theorem 2, but together they justify a conditional verdict: the pure-state contribution is defensible, while the mixed-state characterization and the abstract's broader wording need revision or qualification. Therefore the reader's CONDITIONAL verdict should stand unchanged.","tokens_in":24215,"tokens_out":52989,"duration_ms":411733,"concrete_test":"Sample 10^5 Haar-random three-qubit states, keep those with nonvanishing sender magnetization but with Σ c_i^(1,3) > 2.355, Σ c_i^(1,2) ≤ 1.569, and S(ρ_AB) < 1; for each, numerically maximize C(Tr_E[U_AB ρ_ABE U_AB†]) over a parametrization of SU(4). If any such state fails to yield a final quantum advantage (or any state outside the Corollary's conditions succeeds), the mixed-state criterion is overclaimed. Separately, recompute Proposition 1's β_opt with an exhaustive or analytic search over the m=0 Bloch-parameter polytope to check whether the 1.569 threshold holds rigorously rather than only for the random ensemble.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The pure-state result (Theorem 2) is internally coherent: Lemma 1, Lemma 2, and the ACVENN criterion for rank-two reduced states line up, and the gW example confirms the conditions. The load-bearing weakness is the mixed-state extension. Theorem 3 derives Σ c_i > 2.355 as a sufficient dense-coding condition only for two-qubit states with vanishing sender Bloch vector m, and Proposition 1's NDC threshold Σ c_i ≤ 1.569 likewise assumes m=0 and is obtained by numerical optimization over 10^5–10^6 random states, not by an analytic proof. The Corollary then states a sufficient condition for the 'three-qubit arbitrary state' (Eq. 10) without restating m=0. For a tripartite state whose reduced ρ_AE has nonvanishing sender magnetization, the operator ρ_12 − I/2⊗ρ_2 contains the extra term (1/4)Σ m_i σ_i⊗I, so the trace-distance formulas and thresholds used in Theorem 3 and Proposition 1 are not applicable; the Corollary may certify a non-resource or miss a resource. The numerical-threshold issue also means the NDC bound is not a rigorous theorem even in the m=0 case. The abstract's 'sufficient criterion for mixed states' is therefore not established as stated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces 'dense coding swapping,' a multipartite communication task in which a global unitary on the sender-target pair transfers the dense-coding advantage from one bipartite marginal (A-E) to another (A-B), while the exclusion principle guarantees the original channel becomes non-dense-codeable. The main results are: (i) an if-and-only-if condition for three-qubit pure states in terms of the parameters of a canonical form (Theorem 2); (ii) a sufficient condition for mixed states based on the singular values of bipartite correlation matrices, obtained via Pinsker-type bounds (Theorem 3, Proposition 1, and the Corollary); (iii) an explicit characterization of optimal swapping unitaries for generalized W states and their white/colored-noise admixtures; and (iv) numerical evidence that resource states have low genuine multipartite entanglement, with the required entanglement decreasing with system size. The pure-state derivation is the central analytic contribution; the mixed-state criterion is the part that needs the most careful scrutiny.","tokens_in":24437,"tokens_out":26944,"duration_ms":235354,"significance":"If the pure-state Theorem 2 is correct, it gives a clean and exact operational classification of three-qubit pure states for this new task, together with explicit unitary families that realize the swapping. This is a genuinely useful contribution to the study of conditional-entropy resources and dense coding. The explicit optimum-unitary analysis for generalized W states and the noise-tolerance study are also valuable, and the numerical data on the entanglement requirements is suggestive. The main weakness is that the mixed-state 'sufficient criterion' is stated more broadly than it is derived: it relies on a vanishing sender-magnetization assumption and on a numerically obtained threshold, so the abstract's claim of a sufficient criterion for mixed states is not yet established in full generality.","major_comments":[{"comment":"The mixed-state criterion silently drops the key assumption ⃗m=0. Theorem 3 is explicitly restricted to two-qubit states with vanishing sender Bloch vector, and Proposition 1 has the same restriction. The Corollary, however, states sufficient conditions for 'the three-qubit arbitrary state in Eq. (10)' without restating that the sender's local Bloch vector m^(1) must vanish. For a general Fano state with m^(1)≠0, the operator ρ_12 − I/2⊗ρ_2 contains the extra term (1/4)Σ_i m_i σ_i⊗I, so the trace-distance formulas in Eq. (16) and the thresholds derived from them are not applicable. The abstract's 'sufficient criterion for mixed states' is therefore not established outside the special preparation scenario described in Sec. III B. Please either restrict the Corollary, the abstract, and the concluding remarks to the case m^(1)=0, or extend the derivation to nonvanishing sender magnetization.","section":"§III B, Theorem 3 and the Corollary (Eqs. 10, 12, 14-21)"},{"comment":"The non-dense-codeability threshold Σ c_i ≤ 1.569 is not a rigorous theorem. The proof depends on β_opt = 0.152863 obtained by numerical optimization over 10^5–10^6 Haar-random states, and the text itself states that 'the bound should be regarded as numerical.' As a result, condition 2 of the Corollary is a numerical heuristic, not a proven sufficient condition. In addition, Proposition 1 does not state the necessary condition c_1 < c_2 + c_3 under which the trace distance equals Σ c_i/4. When c_1 ≥ c_2 + c_3, the trace distance is c_1/2, so the inequality Σ c_i ≤ 1.569 does not imply T ≤ 0.392; an example is (c_1,c_2,c_3)=(1,0,0), for which Σ c_i=1 but T=0.5. The proposition and the Corollary must either be rephrased as numerical observations with the c_1 < c_2 + c_3 restriction, or be replaced by an analytic bound.","section":"§III B, Proposition 1 (Eqs. 18-21)"},{"comment":"Lemma 2 is stated as an 'if and only if' characterization of ACVENN for all pure three-qubit states, but as stated it is false. Taking |ψ⟩=|000⟩ gives λ0=1 and all other λ_i=0, so Eq. (8) is not satisfied (the left-hand side equals 1), yet ρ_AB is the pure product state |00⟩⟨00|, which is certainly in the ACVENN class. The intended statement appears to be valid only under the standing hypotheses of Lemma 1 (λ0≠0 and |λ2|>|λ3|), but these assumptions are not mentioned in Lemma 2. Moreover, the proof asserts 'ρ_AB belongs to ACVENN only when S(ρ_E)=1' without a derivation; this is the load-bearing step of the lemma. Because Theorem 2 uses the negation of Eq. (8), this gap needs to be repaired before the pure-state iff can be considered fully rigorous as published.","section":"§III A, Lemma 2 (Eq. 8)"},{"comment":"The Corollary uses the threshold 2.335, while Theorem 3 derives 2√(2 ln 2) ≈ 2.355. If these are meant to be different quantities, the derivation of 2.335 is missing; if it is a typographical error, it must be corrected, since the Corollary is the main mixed-state criterion advertised in the abstract.","section":"§III B, Corollary threshold values"}],"minor_comments":[{"comment":"The parameters λ_i in Eq. (2) are referred to as 'Schmidt coefficients,' but Eq. (2) is not a Schmidt decomposition across any single bipartition; it is the Acín-type canonical form. Please use terminology that distinguishes these canonical amplitudes from the actual Schmidt coefficients of the A:BE cut, as the distinction is important for the proof of Lemma 1.","section":"Throughout, but see Eq. (2)"},{"comment":"The text reports that 'the percentage of states that are suitable is 100% in the case of tripartite states' for Haar-random states. This appears inconsistent with Theorem 2, since the condition |λ2|>|λ3| should select roughly half of the relevant parameter space unless the random generation is conditioned in some way. Please clarify the conditioning used in this numerical statement.","section":"§IV, numerical percentages"},{"comment":"Section VI reuses the theorem number 'Theorem 3' for the generalized-W unitary characterization, even though Theorem 3 already appears in Sec. III B. Please renumber the later theorems.","section":"§VI, Theorem 3"},{"comment":"The sentence describing the correlation matrices appears to mislabel the reduced states: the condition Σ c_i^{(1,3)} > 2.335 is a condition on ρ_AE (parties 1 and 3), while Σ c_i^{(1,2)} ≤ 1.569 is a condition on ρ_AB (parties 1 and 2). The text says these are 'of ρ_AB and ρ_AE respectively,' which is the opposite order and should be corrected.","section":"§III B, Corollary"},{"comment":"The numerical optimization that produces β_opt is described only tersely, and the statement that f(β,T) decreases monotonically in β is followed by a minimization that is hard to parse. A more explicit description of the optimization domain and the meaning of 'the minimum of the optimal β-s' would help reproducibility.","section":"§III B, Proposition 1 proof"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take on Muhuri et al. (2608.03184). The paper has one genuinely solid result: Theorem 2, the Schmidt-coefficient characterization of three-qubit pure states that can swap dense codeability from AE to AB by a two-qubit unitary. I worked through the derivation and it holds: |λ2|>|λ3| and λ0≠0 with the ACVENN condition excluded. That is new—dense coding swapping is distinct from entanglement swapping, and the conditions are usable. The generalized W unitary analysis in Sec. VI is also clean, and the noise tolerance plots are a reasonable practical supplement.\n\nThe soft spots are all in the mixed-state section. Theorem 3 and Proposition 1 are derived for two-qubit states with vanishing sender Bloch vector m=0. The Corollary then states a sufficient condition for the 'three-qubit arbitrary state' without restating that assumption. That is a real overstatement: a three-qubit state with nonzero sender magnetization can have ρ_AE dense codeable and ρ_AB NDC while violating the correlation-matrix inequalities, so the criterion as written can miss or misclassify resources. The paper should either prove the criterion for general m or clearly restrict the statement, and the abstract should say so.\n\nThere is also a numerical threshold issue. Proposition 1's bound Σ c_i ≤ 1.569 comes from a numerical optimization over random states; the paper itself says the bound should be regarded as numerical. That is honest, but it means the NDC criterion is not a theorem. And the Corollary uses 2.335 while Theorem 3 gives 2.355—I can't find where 2.335 comes from. That needs to be reconciled or flagged as a typo. Minor issues: duplicate Theorem 3 numbering, and the N>3 Dicke resource claims don't state how the non-ACVENN condition was verified. No code or data is actually released, just 'available upon request.'\n\nNet: the pure-state result is the contribution and it deserves referee time. The mixed-state claims are not established as stated, but they are fixable. I'd send this to a serious referee and ask for a revision that restricts the mixed-state criterion to m=0 or proves it in general, reconciles the thresholds, and releases the data/code.","headline":"A clean pure-state characterization of dense coding swapping, undermined by a mixed-state extension that overstates its domain and relies on a numerical threshold.","tokens_in":24972,"tokens_out":2861,"would_cite":true,"duration_ms":24994,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P45","81P68"],"pacs":[],"model":"deepseek-v4-flash","headline":"A pure three-qubit state supports dense coding swapping from AE to AB if and only if its Schmidt coefficients obey $|\\lambda_2|>|\\lambda_3|$ with $\\lambda_0\\neq 0$ and $\\rho_{AB}$ is not ACVENN.","keywords":["dense coding swapping","conditional von Neumann entropy","dense coding exclusion principle","ACVENN states","Schmidt coefficients","generalized W states","correlation matrix","genuine multipartite entanglement"],"falsifier":"Prepare (or numerically simulate) the three-qubit generalized W state with $a=1/2$ and small $b>0$; the paper's Lemma 2 places $\\rho_{AB}$ in the ACVENN class, so Theorem 2 predicts that no two-qubit unitary on $AB$ can make it dense codeable. If any $U_{AB}\\in SU(4)$ yields $C(\\rho_{AB})>1$ bit while $C(\\rho_{AE})=1$ bit, the 'if and only if' claim is false.","tokens_in":23986,"feed_emoji":"🔄","tokens_out":16056,"duration_ms":125815,"temperature":0.7,"pith_summary":"The paper introduces a communication task it calls dense coding swapping: inside a multipartite shared state, one sender–receiver pair has a quantum advantage in superdense coding while another pair sharing the same sender does not, and the goal is to transfer that advantage to the second pair by a joint unitary on the sender and the desired receiver, without involving the first receiver. For pure three-qubit states the paper proves this is possible exactly when the Schmidt coefficients satisfy $|\\lambda_2|>|\\lambda_3|$ with $\\lambda_0\\neq 0$ and the reduced state $\\rho_{AB}$ is not in the ACVENN class, the class whose conditional entropy stays non-negative under every global unitary. For mixed states it provides a sufficient criterion based on the singular values of the bipartite correlation matrices, and for generalized W states it identifies the optimal two-qubit unitary that performs the swap. The task matters because it turns the dense coding exclusion principle, normally a limitation, into a controllable way to reroute a communication advantage away from a compromised receiver.","feed_headline":"Which three-qubit states allow dense-coding advantage to be swapped","feed_subtitle":"The protocol reroutes a quantum communication advantage away from a compromised receiver without involving it.","key_machinery":"The object that carries the argument is the conditional von Neumann entropy $S(A|B)=S(\\rho_{AB})-S(\\rho_B)$, whose negativity is the dense-coding quantum advantage. The key identity is $S(A|B)+S(A|E)=0$ for a pure three-qubit state, which forces the two reduced pairs to have opposite signs of conditional entropy and is what makes a swap possible at all. The classification then uses the Schmidt decomposition of the tripartite state, the ACVENN characterisation of when no global unitary can make $\\rho_{AB}$ dense codeable, and the inequality $|\\lambda_2|>|\\lambda_3|$ to decide which pair holds the negative conditional entropy. The explicit swapping tool is the $SU(4)$ unitary $U_{AB}=\\exp[-i(J_x\\sigma_x\\otimes\\sigma_x+J_y\\sigma_y\\otimes\\sigma_y+J_z\\sigma_z\\otimes\\sigma_z)]$, and for generalized W states optimality is pinned down by the condition $\\tan 2J_x^*\\tan 2J_y^*=(1-2b)/(1-2a-2b)$. For mixed states the machinery is the singular-value decomposition of the correlation matrix together with a lower bound on relative entropy in terms of trace distance and a numerical upper bound on relative entropy, which produce the thresholds used in the sufficient criteria.","core_discovery":"The central discovery is that dense codeability can be moved between pairs of a multipartite state by acting only on the sender and the target receiver. For pure three-qubit states, the paper gives a complete classification in Schmidt coefficients: writing the state as $\\lambda_0|000\\rangle+\\lambda_1 e^{i\\phi}|100\\rangle+\\lambda_2|101\\rangle+\\lambda_3|110\\rangle+\\lambda_4|111\\rangle$, the advantage can be swapped from $AE$ to $AB$ if and only if $|\\lambda_2|>|\\lambda_3|$, $\\lambda_0\\neq 0$, and the algebraic condition (8) that would put $\\rho_{AB}$ in the ACVENN class does not hold. Under those conditions some two-qubit unitary $U_{AB}$ makes $\\rho_{AB}$ dense codeable while $\\rho_{AE}$ becomes non-dense codeable. For mixed states, the paper proves that when the sender's local Bloch vector vanishes, a two-qubit state is dense codeable if the sum of the singular values of its correlation matrix exceeds $2\\sqrt{2\\ln 2}\\approx 2.355$, and is non-dense codeable if that sum is at most $1.569$; a corollary adapts this test to a three-qubit mixed state. It also characterizes the optimal swapping unitaries for generalized W states and shows that the resource states tolerate white and colored noise while requiring only small genuine multipartite entanglement.","pith_inferences":["Beyond the paper, the three-qubit theorem looks like the base case of an $N$-party rule: if the sender–target marginal is non-ACVENN and the sender–compromised marginal has negative conditional entropy, the same unitary construction should swap the advantage in larger networks.","Beyond the paper, the mixed-state thresholds should shift when the sender's Bloch vector is nonzero; a biparametric criterion involving $\\beta$ and trace distance $T$ would extend the classification and could be checked numerically by searching states with $\\vec m\\neq 0$.","Beyond the paper, the protocol is directly testable in existing photonic three-qubit superdense-coding setups: prepare a generalized W resource, measure $C(\\rho_{AB})$ and $C(\\rho_{AE})$, apply $U_{AB}$, and verify that the advantage has changed sides."],"forward_implications":["Any pure three-qubit state satisfying the theorem's conditions admits a two-qubit unitary that activates dense coding between Alice and Bob while shutting off the Alice–Evan channel; the advantage is redistributed, not duplicated.","For generalized W states the optimal swapping unitaries obey $\\tan 2J_x^*\\tan 2J_y^*=(1-2b)/(1-2a-2b)$, and the activated capacity reaches $2-H(\\{a,1-a\\})$ bits; the same parameter condition remains optimal when white noise is admixed.","Useful generalized W states tolerate white noise up to a state-dependent probability before losing their swapping capability; with $b$ close to zero the tolerance is around $0.165$, and orthogonal colored noise can in some parameter regions be tolerated even better.","Resource states for dense coding swapping become rare as the number of parties grows (about 3.15% of four-qubit Haar-random states, almost none for five qubits), but asymmetric single-excitation Dicke states continue to support swapping, and the amount of genuine multipartite entanglement they need, measured by the generalized geometric measure, is small and decreases with system size."],"supporting_citations":[{"why":"defines superdense coding, the protocol whose quantum advantage is the resource being swapped.","marker":"[4]"},{"why":"states the dense coding exclusion principle, the constraint that the swapping protocol exploits.","marker":"[39]"},{"why":"defines the ACVENN class; non-membership of $\\rho_{AB}$ is the condition that a unitary can activate its dense codeability.","marker":"[18]"},{"why":"provides the general Schmidt form of three-qubit pure states used in Theorem 1 and Theorem 2.","marker":"[50]"},{"why":"introduces the generalized W states used as the main example family and for explicit optimal unitary calculations.","marker":"[51]"},{"why":"gives the $SU(4)$ parameterization $U_{AB}=\\exp[-i(J_x\\sigma_x\\otimes\\sigma_x+J_y\\sigma_y\\otimes\\sigma_y+J_z\\sigma_z\\otimes\\sigma_z)]$ used for swapping unitaries.","marker":"[52]"},{"why":"supplies the lower bound on relative entropy in terms of trace distance used to derive the dense codeability threshold $\\sum_i c_i>2\\sqrt{2\\ln 2}$.","marker":"[53]"},{"why":"provides the upper bound on relative entropy used with numerical optimization to obtain the non-dense codeability threshold $\\sum_i c_i\\le 1.569$.","marker":"[54]"}],"fun_headline_variants":["Dense coding advantage swapped between channel pairs: exact conditions","Optimal two-qubit gates enable dense coding swapping","Small entanglement suffices to swap dense coding advantage","Dense coding rerouted away from compromised receiver"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The mixed-state characterization assumes the sender's local Bloch vector $\\vec m$ vanishes, justified by a preparation in which Alice's reduced state is maximally mixed; if a candidate state has nonvanishing sender magnetization, the correlation-matrix thresholds are not proven to classify it.","fun_headline_variants_meta":{"raw":{"variants":["Dense coding advantage swapped between channel pairs: exact conditions","Optimal two-qubit gates enable dense coding swapping","Small entanglement suffices to swap dense coding advantage","Dense coding rerouted away from compromised receiver"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00084,"raw_usage":{"total_tokens":3715,"prompt_tokens":1051,"completion_tokens":2664,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":667,"completion_tokens_details":{"reasoning_tokens":2602}},"tokens_in":667,"tokens_out":2664,"duration_ms":17398,"temperature":1.0,"reasoning_tokens":2602,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T14:52:30.063585+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Prepare (or numerically simulate) the three-qubit generalized W state with $a=1/2$ and small $b>0$; the paper's Lemma 2 places $\\rho_{AB}$ in the ACVENN class, so Theorem 2 predicts that no two-qubit unitary on $AB$ can make it dense codeable. If any $U_{AB}\\in SU(4)$ yields $C(\\rho_{AB})>1$ bit while $C(\\rho_{AE})=1$ bit, the 'if and only if' claim is false.","supporting_citations":[{"cited_title":"Nepal, R","cited_arxiv_id":null,"evidence_quote":"states the dense coding exclusion principle, the constraint that the swapping protocol exploits."},{"cited_title":"Horodecki, J","cited_arxiv_id":null,"evidence_quote":"defines the ACVENN class; non-membership of $\\rho_{AB}$ is the condition that a unitary can activate its dense codeability."},{"cited_title":"Prabhu, A","cited_arxiv_id":null,"evidence_quote":"provides the general Schmidt form of three-qubit pure states used in Theorem 1 and Theorem 2."},{"cited_title":"Nepal, R","cited_arxiv_id":null,"evidence_quote":"introduces the generalized W states used as the main example family and for explicit optimal unitary calculations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"gives the $SU(4)$ parameterization $U_{AB}=\\exp[-i(J_x\\sigma_x\\otimes\\sigma_x+J_y\\sigma_y\\otimes\\sigma_y+J_z\\sigma_z\\otimes\\sigma_z)]$ used for swapping unitaries."},{"cited_title":"Shadman, H","cited_arxiv_id":null,"evidence_quote":"supplies the lower bound on relative entropy in terms of trace distance used to derive the dense codeability threshold $\\sum_i c_i>2\\sqrt{2\\ln 2}$."},{"cited_title":"Shadman, H","cited_arxiv_id":null,"evidence_quote":"provides the upper bound on relative entropy used with numerical optimization to obtain the non-dense codeability threshold $\\sum_i c_i\\le 1.569$."}],"review_version":2}