{"id":"1c10b88d-476b-45a7-b730-af9a3c7c8d1e","arxiv_id":"2505.24772","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A symmetry-based construction of unentangled densities is used to define and compute a mixed-state concurrence-like measure Cs for thermal states of symmetric spin clusters.","lead":"The authors define a symmetry-guided procedure to estimate the entanglement of thermal mixed states in small symmetric spin systems, by mixing in specially constructed unentangled density matrices. The resulting quantity Cs reproduces the exact two-qubit concurrence and gives finite-temperature curves for three-spin clusters, which may be useful for frustrated magnet studies.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Cs is only an upper bound on the convex roof: the Step 5 optimization fixes the pure states inside each degenerate multiplet, so equality needs a proof that no lower decomposition of ρ_i exists.","rationale":"Good-faith summary: the paper aims to make the convex-roof minimization tractable for symmetric thermal states by restricting decompositions to symmetry-adapted states and separable symmetric densities. This is a legitimate strategy. What would have to be true for the central claim is that the reduced family of decompositions contains an optimal one. The weakest part is not only the enumeration of separable η_k; it is that the procedure never optimizes over the pure states inside a degenerate multiplet. The objective assigns each retained ρ_i its average pure-state concurrence, but an arbitrary rank-2 density's convex roof is generally lower than the average of two fixed orthogonal pure states in its support. The two-qubit tests are reassuring but nonprobative because no entangled degenerate multiplet survives in those cases. A concrete low-temperature three-spin doublet provides a direct numerical test. If the exact convex roof of that doublet matches C_i, the specific concern would be resolved, although a general proof would still be needed. Therefore I agree with the reader's conditional verdict, but the stated condition should be expanded: the fixed-multiplet ansatz must be shown to be optimal, in addition to the η_k list being complete.","tokens_in":15395,"tokens_out":8805,"duration_ms":118496,"concrete_test":"Choose a three-spin-line parameter set where the thermal density at very low T is dominated by a single entangled doublet, e.g. J1 = -1, J2 = -0.8, J3 = 0.5, J4 = 0, with T much smaller than the gap (the upper curve of Fig. 5). Compute the exact convex roof of ρ_ground = (|Ψ3+⟩⟨Ψ3+|+|Ψ3-⟩⟨Ψ3-|)/2 using a numerical search over two-state decompositions of this rank-2 density; if the result is strictly less than C3 = C(Ψ3±), the Cs reported in Section 5 overestimates the concurrence. Run the same search on the two-qubit control state (|Φ+⟩⟨Φ+|+|Φ−⟩⟨Φ−|)/2, where the correct convex roof is 0, to validate the numerical routine.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that Cs(ρ), defined in Section 3 Step 5 and Section 4, equals the convex-roof concurrence (6). The optimization varies coefficients p'_i for the fixed multiplet densities ρ_i = (1/d_i) Σ_j |Ψ_ij⟩⟨Ψ_ij| and coefficients q_k for fixed separable densities η_k. This is a strict subset of all decompositions, so Cs is an upper bound on the true concurrence. Equality requires either (i) an optimal ensemble can always be chosen using exactly these states, or (ii) each ρ_i itself has convex roof equal to (1/d_i)Σ_j C(Ψ_ij). Neither is shown. The two-qubit checks in 3.1.1–3.1.2 do not test this: in those cases every surviving multiplet is either non-degenerate or has a degenerate density that is itself one of the η_k, so no entangled rank-2 multiplet is ever evaluated. For the three-spin systems the ρ_i are rank-2 mixtures of time-reversed eigenstates, and a rank-2 mixture can have convex roof strictly below the average of the pure-state concurrences (a two-qubit example is (|Φ+⟩⟨Φ+|+|Φ−⟩⟨Φ−|)/2, whose concurrence is 0 although each pure state has concurrence 1). Nothing in the manuscript rules out the same effect for ρ_3 or ρ_4. This concern is independent of whether the η_k list is complete: adding all separable extreme points would only allow subtraction of separability, not replacement of an entangled doublet by a less entangled decomposition of the same density.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a symmetry-based method to estimate the concurrence of thermal mixed states of symmetric spin systems. The authors group degenerate eigenstates into multiplet densities ρ_i, construct a set of symmetric separable densities η_k from product states along local axes, and then solve a linear program that re-expresses the thermal density as ρ = Σ_i p'_i ρ_i + Σ_k q_k η_k. The resulting quantity Cs(ρ) = Σ_i (1/d_i) p'_i C(Ψ_i) is put forward as the concurrence of the state. The method is illustrated on two-spin and three-spin (line and triangle) systems, with two-spin results checked against Wootters' formula and three-spin results presented as heat maps over coupling parameters.","tokens_in":15771,"tokens_out":6059,"duration_ms":71651,"significance":"If the central claim were established, the method would provide a rare tractable route to a quantitative entanglement measure for highly symmetric mixed states, with potential applications to frustrated spin systems such as spin ice. The explicit construction of symmetric separable densities and the reduction to a linear program are genuinely useful ideas, and the two-spin examples are verified exactly against Wootters' formula. However, the paper's main assertion that Cs equals the convex-roof concurrence is not proved for the three-spin systems: the optimization is restricted to a fixed set of pure states within each degenerate multiplet and to a heuristic list of separable densities, so Cs is at best an upper bound on the true concurrence. The three-spin results therefore need either additional proof or a careful reframing as a variational upper bound.","major_comments":[{"comment":"The quantity Cs is defined by minimizing over the restricted family of decompositions ρ = Σ_i p'_i ρ_i + Σ_k q_k η_k, where each ρ_i = (1/d_i)Σ_j |Ψ_ij⟩⟨Ψ_ij| is kept intact. Since the convex roof in Eq. (6) minimizes over all decompositions, Cs is an upper bound on the true concurrence unless one proves either that an optimal convex-roof ensemble can always be chosen in this restricted form, or that the convex roof of each ρ_i equals the average (1/d_i)Σ_j C(Ψ_ij). Neither proof is given. This is load-bearing: for the three-spin line (Section 3.1.3), ρ_3 and ρ_4 are rank-2 mixtures of time-reversed eigenstates, and for rank-2 states the convex roof can be strictly smaller than the average pure-state concurrence (e.g., the equal mixture of |Φ+⟩ and |Φ−⟩ has concurrence 0 although each Bell state has C=1). The manuscript contains no argument excluding this effect for ρ_3 or ρ_4, so the plotted Cs in Section 5 may overestimate the entanglement.","section":"Section 3, Step 5; Eq. (6)"},{"comment":"The lists of separable symmetric densities η_j (nine for the line, eight for the triangle) are obtained by enumerating separable product states on local axes and deleting those states expressible as positive sums of others. The paper does not prove that the remaining set is complete, i.e., that it contains all extreme points of the convex hull of symmetric separable density matrices for the relevant symmetry-invariant subspace. Without such a proof or a dual linear-programming certificate, the optimization in Section 4 may subtract less separable weight than is actually possible, again pushing Cs upward. This issue is independent of the restricted-multiplet concern in the previous comment: even a complete η set would not fix the unproven treatment of the degenerate multiplets.","section":"Sections 3.1.3 and 3.2"},{"comment":"The three-spin results are not benchmarked against any independent estimate of the true convex-roof concurrence. The two-spin checks in Sections 3.1.1 and 3.1.2 are valuable, but they do not exercise the problematic case: every entangled multiplet there is non-degenerate or has a density that coincides with one of the η_k, so no rank-2 entangled multiplet is ever evaluated. For the three-spin line and triangle, no comparison is made with exact values at special parameter points or with numerical lower bounds from another method (e.g., a semidefinite relaxation or a direct search over rotations of the doublet states). This leaves the magnitude of the possible over-estimation identified above unquantified and makes the physical interpretation of the heat maps in Figs. 3, 6, 7, and 8 unclear.","section":"Section 5"}],"minor_comments":[{"comment":"The name Wootters is misspelled as 'Wootter' in Sections 3.1.1 and 3.1.2; it should be 'Wootters'.","section":"Throughout"},{"comment":"The word 'temperatute' in the last paragraph of the Introduction is a typo for 'temperature'.","section":"Introduction"},{"comment":"The caption of Fig. 3 states that the right column is T=0.01 and the left column is T=0.2, but the text in Section 5.1 refers to 'the higher temperature results (right column of Fig. 3)'. These statements are inconsistent; the text should refer to the left column if the caption is correct, or the caption should be corrected.","section":"Section 5.1, Fig. 3 caption and text"},{"comment":"The notation for the example separable states in the η_j list (e.g., '|+ ++⟩ zxx') is not explained in the text; a brief definition of the axis subscripts in this context would improve readability.","section":"Section 3.1.3, Eq. (24)-(27)"},{"comment":"The linear program is described in terms of expressing each η_j as a linear combination of η_1 and the ρ_i, but it is not stated whether the coefficients c_ji and c'_j are unique or how degeneracies in the representation are handled; a short clarification would help reproducibility.","section":"Section 4"},{"comment":"In the sentence 'The lower curve the ground state has zero concurrence', the word 'in' is missing after 'curve'; it should read 'The lower curve, the ground state has zero concurrence'.","section":"Section 5.1"}],"recommendation":"major_revision","confidential_remarks":"The core difficulty is that the manuscript defines Cs as a restricted optimization and then calls it 'the concurrence', but the equality with the convex roof is not established. I do not see this as an immediately fatal flaw because the method may be reframed as a variational upper bound, and the two-spin verification is solid. However, if the authors wish to retain the strong claim, they need to supply either a proof of the restricted-decomposition property for the specific multiplets or a comparison with exact convex-roof values. The paper may also be considered for a venue that values computationally tractable entanglement estimates over exactness, but as submitted the central quantitative claim is not supported."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the good news: the symmetry-guided construction of separable densities plus the LP optimization is genuinely new as far as I can tell, and the two-qubit checks match Wootters exactly. That is real, careful work. But the central claim — that Cs is the convex-roof concurrence — has a load-bearing gap. The optimization in Section 3, Step 5 mixes the fixed multiplet densities ρ_i with the η_k. It never recombines pure states inside a degenerate multiplet. For a doubly-degenerate pair of time-reversed entangled eigenstates, the density (|Ψ+><Ψ+| + |Ψ-><Ψ-|)/2 can have convex roof strictly below the average of the two pure-state concurrences. A two-qubit example is (|Φ+><Φ+| + |Φ-><Φ-|)/2, which is the maximally mixed state of two qubits: separable, so concurrence 0, despite each Bell state having concurrence 1. Nothing in the manuscript rules out the same effect for ρ_3 or ρ_4 in the three-spin cases. So Cs is an upper bound on the true concurrence, not the concurrence itself, unless the authors prove that an optimal ensemble can always be chosen using exactly these fixed densities. That proof is not present. The separate completeness issue for the η_k list also matters, but it is independent: adding more separable extreme points would only allow subtracting separability, not lowering the entanglement of a degenerate doublet. The stress-test concern survives reading the paper. What the paper does well: the two-spin algebra is clean, the explicit construction of symmetric separable densities is a useful idea, and the three-spin heat maps are plausible. I also appreciate the candid scalability remarks in Section 6. But the conclusion overstates the result: calling Cs \"the concurrence\" without resolving the multiplet issue is not justified. The honest fixes are (i) prove that an optimal ensemble can be chosen with those fixed multiplet densities, or (ii) benchmark against exact convex roof results for small systems. This deserves a serious referee: the idea is good enough that a referee should ask these questions, not reject on sight. For me, I would not yet cite it as the concurrence, but I would read a revised version closely.","headline":"A genuinely new symmetry-based construction, but Cs is only an upper bound on the convex roof until the degenerate-multiplet issue is resolved.","tokens_in":16211,"tokens_out":2280,"would_cite":false,"duration_ms":27821,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["03.67.Mn","75.10.Jm"],"model":"deepseek-v4-flash","headline":"The paper argues that for thermal states of symmetric spin systems, the true concurrence is obtained by a linear program that optimally mixes explicitly constructed separable densities into the state.","keywords":["entanglement","spin systems","symmetry","density matrix","concurrence","thermal states","convex roof","frustrated magnetism"],"falsifier":"Run a numerical search over decompositions, such as a semidefinite relaxation of the convex roof, for the three-spin triangle at parameters where $C_s>0$; if any valid decomposition gives a strictly lower concurrence than $C_s$, the separable list is incomplete. A direct enumeration of the extreme points of the symmetric separable polytope would settle the same question.","tokens_in":15205,"feed_emoji":"🧲","tokens_out":7029,"duration_ms":76716,"temperature":0.7,"pith_summary":"Thermal mixtures of symmetric spin systems can hide their entanglement: although the eigenstates of a symmetric Hamiltonian are often entangled, the Boltzmann-weighted mixture can be re-expressed as a sum of less entangled or separable pieces. This paper argues that for such systems the true mixed-state concurrence, the minimum over all decompositions, can be found by symmetry alone. The authors explicitly construct a finite list of separable density matrices that share the Hamiltonian's symmetry, then linearly optimize how much of each can be mixed into the thermal state. The resulting quantity $C_s(\\rho)$ reproduces the exact two-qubit concurrence, reduces to the ground-state concurrence as $T\\to 0$, and vanishes at high temperature. This gives a practical route to quantifying entanglement in small symmetric spin systems, including frustrated geometries relevant to spin ice.","feed_headline":"Symmetry turns thermal spin entanglement into a linear program","feed_subtitle":"A new decomposition reproduces two-qubit concurrence exactly and maps frustration in three-spin systems.","key_machinery":"The load-bearing object is the finite set of separable symmetric densities $\\eta_k$: for each geometry, the paper enumerates all product states of the form $|\\pm\\rangle_\\alpha \\otimes |\\pm\\rangle_\\beta \\otimes \\cdots$ with each spin quantized along its local $x$, $y$, or $z$ axis, averages them over the symmetry group, and keeps only the resulting densities that are not positive sums of others, nine for three spins on a line and eight for three spins on a triangle. These $\\eta_k$ are then used as free variables in a linear program: starting from the thermal density $\\rho=\\sum_i p_i \\rho_i$, one maximizes the weight $q_k$ of each separable density that can be subtracted, reducing the coefficients $p'_i$ of the entangled eigenstate densities. The minimized value of $\\sum_i (1/d_i) p'_i C(\\Psi_i)$ is the proposed concurrence $C_s(\\rho)$.","core_discovery":"The central claim is that the concurrence of a thermal density $\\rho$ of a symmetric spin system is the minimum of $C_s(\\rho)=\\sum_i (1/d_i)\\, p'_i\\, C(\\Psi_i)$ over decompositions $\\rho=\\sum_i p'_i \\rho_i + \\sum_k q_k \\eta_k$, where the $\\rho_i$ are the symmetry-grouped eigenstate densities, the $\\eta_k$ are explicitly constructed separable densities that are invariant under the same symmetry group, and $C(\\Psi)$ is the multipartite concurrence of Eq. (4). By construction $C_s$ equals the convex-roof infimum restricted to this family of decompositions, so the paper identifies $C_s$ with the concurrence of $\\rho$. The method is demonstrated on two-spin systems, three spins on a line, and three spins on a triangle; in the two-spin cases it matches the exact two-qubit concurrence, and in the three-spin cases it produces heat maps of $C_s$ versus coupling constants at low temperature, with sharp boundaries from level crossings and smooth onset from degeneracy lifting.","pith_inferences":["If the enumerated $\\eta_k$ list is incomplete, $C_s$ overestimates the true convex roof; a certificate of completeness, such as an extreme-point enumeration of the symmetric separable polytope, would turn $C_s$ from an upper bound into the exact concurrence.","The same optimization could be applied to other convex-roof entanglement measures by replacing $C(\\Psi)$ with a different pure-state monotone, since the separable $\\eta_k$ construction only uses symmetry, not the specific concurrence formula.","The authors' scalability remark suggests replacing Eq. (4) with the $I$-concurrence to cut the $2^N$ term count; a concrete test would be whether the heat-map structure survives on a 16-site pyrochlore cluster.","A direct experimental signature: for the triangle at $J_1>0$, the central point with all $J_2=J_3=J_4=0$ is predicted to have zero thermal concurrence at any temperature, while any small perturbation produces a finite value, a sharp and testable crossover."],"forward_implications":["For two spins with full or axial rotational symmetry, $C_s$ reproduces the exact two-qubit concurrence, including the kink where a coefficient crosses $1/2$.","In the zero-temperature limit $C_s$ equals the concurrence of the ground-state multiplet; at high temperature it vanishes, matching the known separable high-temperature limit.","For three spins on a line and on a triangle, $C_s$ produces phase-like maps in the coupling-constant plane: sharp boundaries where level crossings change the ground state, and smooth onset of entanglement where degeneracies are lifted.","The method extends to any symmetric arrangement of qubits or spins and is proposed as a quantitative measure for states with large ground-state degeneracy, such as frustrated spin systems."],"supporting_citations":[{"why":"Supplies the convex-roof definition of mixed-state entanglement that the paper aims to evaluate.","marker":"[1]"},{"why":"Provides the fidelity and concurrence formulation for conjugated states that underlies the convex roof.","marker":"[2]"},{"why":"Gives the multipartite concurrence measure of Eq. (4) used for pure states throughout the paper.","marker":"[3]"},{"why":"Classifies all ensembles with a given density matrix, justifying the left-unitary re-decomposition step.","marker":"[4]"},{"why":"Supplies the exact two-qubit concurrence formula that the new method must reproduce.","marker":"[5]"},{"why":"Gives the entanglement-of-formation benchmark for two qubits used as a check in the two-spin examples.","marker":"[6]"},{"why":"Motivates the finite-temperature spin-ice problem that the method is designed to address.","marker":"[7]"},{"why":"Provides the prior 16-site pyrochlore concurrence study that the present method aims to generalize.","marker":"[8]"}],"fun_headline_variants":["Thermal spin entanglement solved via symmetry","Linear program unlocks thermal entanglement","Symmetry makes mixed-state entanglement computable","Symmetry turns thermal spin states into linear programs","Exact concurrence from symmetry for thermal spin systems"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The method assumes that the finite list of separable symmetric densities it constructs, after dropping those that are positive sums of others, contains every extreme point of the convex hull of symmetric separable states; if any separable symmetric density is missing, the optimized value overestimates the true concurrence.","fun_headline_variants_meta":{"raw":{"variants":["Thermal spin entanglement solved via symmetry","Linear program unlocks thermal entanglement","Symmetry makes mixed-state entanglement computable","Symmetry turns thermal spin states into linear programs","Exact concurrence from symmetry for thermal spin systems"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000306,"raw_usage":{"total_tokens":1734,"prompt_tokens":909,"completion_tokens":825,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":525,"completion_tokens_details":{"reasoning_tokens":760}},"tokens_in":525,"tokens_out":825,"duration_ms":9489,"temperature":1.0,"reasoning_tokens":760,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T12:15:10.462352+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a numerical search over decompositions, such as a semidefinite relaxation of the convex roof, for the three-spin triangle at parameters where $C_s>0$; if any valid decomposition gives a strictly lower concurrence than $C_s$, the separable list is incomplete. A direct enumeration of the extreme points of the symmetric separable polytope would settle the same question.","supporting_citations":[{"cited_title":"Carvalho, Marek Kuś, and Andreas Buchleitner","cited_arxiv_id":null,"evidence_quote":"Supplies the convex-roof definition of mixed-state entanglement that the paper aims to evaluate."},{"cited_title":"Fidelity and concurrence of conjugated states.Phys","cited_arxiv_id":null,"evidence_quote":"Provides the fidelity and concurrence formulation for conjugated states that underlies the convex roof."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the multipartite concurrence measure of Eq. (4) used for pure states throughout the paper."},{"cited_title":"Hughston, Richard Jozsa, and William K","cited_arxiv_id":null,"evidence_quote":"Classifies all ensembles with a given density matrix, justifying the left-unitary re-decomposition step."},{"cited_title":"Hill and William K","cited_arxiv_id":null,"evidence_quote":"Supplies the exact two-qubit concurrence formula that the new method must reproduce."},{"cited_title":"Wootters","cited_arxiv_id":null,"evidence_quote":"Gives the entanglement-of-formation benchmark for two qubits used as a check in the two-spin examples."},{"cited_title":"Ross, Lucile Savary, Bruce D","cited_arxiv_id":null,"evidence_quote":"Motivates the finite-temperature spin-ice problem that the method is designed to address."},{"cited_title":"Wei and S","cited_arxiv_id":null,"evidence_quote":"Provides the prior 16-site pyrochlore concurrence study that the present method aims to generalize."}],"review_version":1}