{"id":"abef82ea-4875-4011-8573-dd867dc4493d","arxiv_id":"2412.18331","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Two copies of certain biseparable three-qubit states are genuinely multipartite entangled, yet every local reduction to one copy is biseparable; this incompressible entanglement is constructed explicitly.","lead":"This paper shows how to turn quantum states that have no genuine multipartite entanglement on their own into states that do, by taking two copies and applying local maps. The authors also construct states whose two-copy entanglement can never be compressed back to a single copy, a new resource phenomenon called incompressible entanglement.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"ICE proof depends on unshipped numerical biseparability certificates in Appendix G; exact decompositions, tolerances, and code are not provided.","rationale":"I read the argument in Appendix G in good faith and checked the logical reduction from arbitrary local Kraus maps to the rank-3 projection P = 1 - |00><00|. The product-vector argument for two-dimensional subspaces, the use of local unitary invariance under U and V independently on the two copies, and the step from a rank-3 projection to a rank-2 projection are all sound. The single load-bearing weak point is the assertion that the projected 3x4x4 states are biseparable via the algorithms of Refs. [70-72]. Because the paper reports only parameter intervals and not the promised explicit decompositions, and because the GME and biseparability intervals nearly touch, the existence proof for ICE is not reproducible as written. This is exactly the concern identified by the reader, so no change to the CONDITIONAL verdict is needed; the condition should be to release the certificates or code.","tokens_in":30267,"tokens_out":16502,"duration_ms":157814,"concrete_test":"Recompute the 'Projection BS' rows of Table I at p = 0.781, 0.782, 0.783, and 0.784 for rho^(s) (and at the analogous endpoints for rho^(u)) using the algorithms of Refs. [70-72], then verify each returned biseparable decomposition in exact rational arithmetic or with interval-arithmetic error bounds. If no certified decomposition exists at any p >= 0.781, the ICE interval in Table I is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that incompressible entanglement exists is proved only in Appendix G, and the proof reduces to a numerical assertion: for the projected 3x4x4 states, the algorithms of Refs. [70-72] return explicit biseparable decompositions over the parameter windows in Table I, e.g., 0.781 <= p <= 0.784 for rho^(s). These decompositions are not shown, no solver tolerances are given, and no code or certificate is shipped. The PPT-mixture SDP alone does not certify biseparability: it only shows the projected state is a PPT mixture (p <= 0.800), which is a necessary condition, not a sufficient one. The claimed ICE interval is therefore supported by an unverifiable black-box computation, and the interval is only 0.003 wide. If the true biseparable threshold lies below 0.781, or if the decomposition is only approximately valid, the claimed window for ICE disappears. The GHZ-diagonal ICE evidence in the main text is correctly labeled as numerical evidence, but Appendix G is presented as a proof and inherits this unverified computational step.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper develops a framework for studying superactivation of genuine multipartite entanglement (GME) in the multi-copy scenario. The authors reformulate the detection of two-copy GME via local projections as an optimization over product vectors, combine this with PPT-mixture semidefinite programs and graph-state symmetries, and derive quantitative results such as white-noise robustness and achievable GHZ fidelity after projection. They also derive a k-copy generalization of the Gabriel-Hiesmayr-Huber criterion, construct nonlinear witnesses with a statistical model for experiments, and analyze GHZ-symmetric states in relation to SLOCC classes. The central new claim is the existence of incompressible entanglement (ICE): a biseparable tripartite state ρ such that ρ⊗2 is GME while every local projection to a single copy is biseparable. The proof of this claim is given in Appendix G, based on numerical biseparability certificates for the projected states.","tokens_in":30476,"tokens_out":22957,"duration_ms":200604,"significance":"If the ICE existence claim can be made fully rigorous, the paper constitutes a significant advance: it introduces a new phenomenon analogous to bound entanglement in the multipartite setting, provides a quantitative theory of GME superactivation, and delivers analytic criteria (e.g., the k-copy GHH inequality) that are directly usable. The analytic derivations in Appendices E and F are clear and appear correct, and the numerical evidence for GHZ-diagonal states is properly labeled. The paper is broad and well organized, and the proposed nonlinear witnesses with a concrete statistical analysis are a strength. However, the central proof of ICE currently rests on (i) an unjustified symmetry reduction and (ii) unverifiable numerical certificates; both need to be addressed before the main claim can be accepted.","major_comments":[{"comment":"The proof that the projected states are biseparable rests entirely on the algorithms of Refs. [70–72], but no explicit biseparable decompositions, solver tolerances, or code are provided. The PPT-mixture bounds (e.g., p ≤ 0.800 for ϱ(s)) are only necessary conditions for biseparability, and the claimed ICE window 0.781 ≤ p ≤ 0.784 is narrow; a small numerical error would erase it. Please supply machine-checkable certificates (e.g., explicit decompositions as ancillary data) or replace this step with an analytic argument. As written, the central existence claim is not independently verifiable.","section":"Appendix G, Table I"},{"comment":"The reduction to the single projection P = 1 − |00><00| is not justified. The families ϱ(s) and ϱ(u) are invariant under local unitaries U⊗U⊗U with the same U on the three parties; on Alice's two-qubit subsystem the induced action is U⊗U, whose orbit through |00> consists only of product states |u>|u>. A general product vector |ab> in Eq. (G6) with |a> ≠ |b> (up to phase) cannot be brought to |00> by such a unitary, and not every two-dimensional subspace contains a vector of the form |u>|u>. Hence the numerical certificate for P = 1 − |00><00| does not cover all local projections, and the assertion that 'it suffices to consider' this projection is unsupported. Please either prove a larger symmetry, or certify biseparability for representatives of all orbits of product vectors (e.g., |a> = |0>, |b> = cosθ|0> + sinθ|1> for θ ∈ [0, π/2]).","section":"Appendix G, Eq. (G8)"}],"minor_comments":[{"comment":"The sentence 'Both methods proved the existence of ICE' is not consistent with the table, where the Gilbert algorithm row is empty/inconclusive for ϱ(s); please adjust the wording.","section":"Appendix G, Table I"},{"comment":"The display equations (e.g., Eq. (A1) and Eq. (D5)) contain garbled symbols in the provided version and should be typeset cleanly.","section":"Appendix A and Appendix D"},{"comment":"Please state the solver versions and numerical tolerances used for the SDP/LP computations, and consider making the code publicly available, since several quantitative claims depend on these values.","section":"Appendix B"}],"recommendation":"major_revision","confidential_remarks":"The manuscript has strong analytic components and a potentially important central claim. My main concern is that the ICE proof is not yet verifiable: the symmetry reduction in Eq. (G8) appears mathematically wrong, and the numerical certificates are not shipped. I recommend major revision rather than rejection, because both issues are potentially fixable within the paper's scope, but the authors should be asked to either provide a valid proof of the reduction or extend the certificates to the full one-parameter family of product-vector orbits, and to make the computational certificates publicly available."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Key thing to know: this paper actually proves a new phenomenon—incompressible entanglement—and does a serious quantitative study of superactivation of genuine multipartite entanglement. The main caveat is the one flagged in the stress test: the central existence proof in Appendix G rests on numerical biseparability certificates that are not shipped, so the headline claim is not independently verifiable as written.\n\nWhat's genuinely new: the explicit construction of ICE states—biseparable tripartite states whose two-copy state is GME but every local projection to a single copy is biseparable—plus the k-copy GHH criterion, the extension of Hadamard-type maps to two-colorable graph states, and the noise-robustness bounds. The paper is careful to separate proof from numerical evidence: the GHZ-diagonal results are labeled as strong numerical evidence, and the ICE proof is confined to Appendix G. The derivations of criteria like Eq. (D12) and the k-copy GHH criterion are clean and directly usable. The experimental section with concrete witnesses and a statistical model is a real plus.\n\nThe soft spot is the ICE proof. The argument reduces to a numerical assertion: for the projected 3x4x4 states, two biseparability algorithms return explicit decompositions over narrow parameter windows (e.g., 0.781 ≤ p ≤ 0.784 for one family). Those decompositions, the solver tolerances, and the code are not included. The PPT-mixture bound alone does not certify biseparability—it is necessary, not sufficient. The windows are only 0.003–0.014 wide, so if the true biseparable threshold is even slightly lower, the claimed ICE intervals vanish. This is a reproducibility gap, not an evident mathematical error; the authors say the tools deliver explicit decompositions with many terms, and two independent algorithms agree. But for a central claim, that is not enough to verify as written. A minor point: the seesaw algorithm for finding projections is heuristic, but the paper acknowledges this and uses SDP relaxations to certify some no-projection cases. That is handled honestly.\n\nBottom line: this paper is for people working on multipartite entanglement, distillation, and resource theories. It deserves serious refereeing. The referee should ask for the numerical certificates or code for Appendix G and for explicit statements of numerical precision. If those check out, the result stands; if not, the ICE claim should be downgraded to a conjecture supported by numerical evidence. Either way, the rest of the paper—the criteria, the quantification, the experimental protocols—is solid and useful.","headline":"Proves a new phenomenon—incompressible entanglement—with clean criteria and honest numerical labeling; the central ICE proof needs shipped certificates to be fully verifiable.","tokens_in":31008,"tokens_out":3472,"would_cite":true,"duration_ms":28679,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["03.67.Mn","03.67.-a"],"model":"deepseek-v4-flash","headline":"The paper proves that genuine multipartite entanglement can be superactivated in two copies yet be incompressible: no local projection back to a single copy preserves the activated entanglement.","keywords":["genuine multipartite entanglement","superactivation","incompressible entanglement","biseparability","GHZ-diagonal states","PPT mixtures","local projection maps","entanglement distillation"],"falsifier":"Take $\\rho^{(s)}$ with $p = 0.782$ and search numerically or analytically for any local Kraus operators $F_A$, $F_B$, $F_C$ such that the renormalized projected state $\\tau_2$ is genuinely multipartite entangled; finding one such operator would disprove the incompressibility claim for that state.","tokens_in":30072,"feed_emoji":"🔗","tokens_out":9274,"duration_ms":78353,"temperature":0.7,"pith_summary":"This paper asks when correlations that are absent in a single copy of a quantum state appear once several copies are combined, and how much of that newly created correlation can be brought back to one copy. It develops a systematic toolbox for superactivation of genuine multipartite entanglement (GME), where two copies of a state that is only biseparable become entangled across all parties. The main proof is that this phenomenon can be incompressible: there are tripartite states whose two-copy versions are GME, yet every local projection back to a single copy yields a biseparable state again. If correct, this establishes a multipartite analogue of bound entanglement, and the quantitative criteria provided here make the effect measurable in experiments.","feed_headline":"Entanglement that only appears in two copies is provably incompressible","feed_subtitle":"Two copies are genuinely entangled; every local projection to one copy is biseparable.","key_machinery":"The load-bearing object is the local projection map from the $k$-copy Hilbert space to a single-copy space, $\\tau_k = \\frac{1}{N}[F_A\\otimes F_B\\otimes F_C]\\,\\rho^{\\otimes k}\\,[F_A^\\dagger\\otimes F_B^\\dagger\\otimes F_C^\\dagger]$, with the Hadamard map $E_X = \\lvert 0\\rangle\\langle 00\\rvert + \\lvert 1\\rangle\\langle 11\\rvert$ the special case already used in the literature. Incompressible entanglement is defined by the absence of any such projection producing a GME single-copy state. The proof machinery combines the reformulation of projection existence as a product-vector minimization $\\langle abc\\rvert \\rho^{\\otimes 2}\\otimes W \\lvert abc\\rangle < 0$, the PPT-mixture semidefinite relaxation for detecting GME, and partial projections that reduce the problem to checking biseparability of a $3\\times 4\\times 4$ state; biseparability decompositions from numerical algorithms then close the proof.","core_discovery":"On its own terms, the paper establishes the existence of incompressible entanglement (ICE) at the two-copy level. For the symmetric family $\\rho^{(s)} = \\frac{1}{3}(\\sigma_{AB}\\otimes \\mathbb{1}_C/2 + \\mathbb{1}_A/2\\otimes \\sigma_{BC} + \\sigma_{AC}\\otimes \\mathbb{1}_B/2)$ built from noisy singlet states, the two-copy state $(\\rho^{(s)})^{\\otimes 2}$ is GME for $p \\geq 0.781$; for the asymmetric family $\\rho^{(u)} = \\frac{1}{2}(\\sigma_{AB}\\otimes \\mathbb{1}_C/2 + \\mathbb{1}_A/2\\otimes \\sigma_{BC})$, two copies are GME for $p \\geq 0.708$. Yet after any local Kraus-map reduction $\\tau_2 = \\frac{1}{N}[F_A\\otimes F_B\\otimes F_C](\\rho)^{\\otimes 2}[F_A^\\dagger\\otimes F_B^\\dagger\\otimes F_C^\\dagger]$ to a single copy, the result is biseparable, with explicit windows such as $0.781 \\leq p \\leq 0.784$ for $\\rho^{(s)}$. The proof uses the permutation symmetry of the states to reduce the projection problem to a single partial projection, and then certifies biseparability of the projected state by explicit decompositions. This is the promised multipartite analogue of bound entanglement: the superactivated GME cannot be compressed onto three qubits.","pith_inferences":["If ICE exists at the two-copy level, analogous families likely exist for $k > 2$ copies, giving a hierarchy of states whose GME appears only at some copy number and can never be reduced; this would make 'copy depth' a resource parameter.","The gap in the numerical results between states certified GME by PPT mixtures and states where all projections are excluded suggests there may be two-copy states that are 'bound' in a stronger sense, which would sharpen the analogy to NPT bound entanglement.","The optimized projections that turn $\\chi(1,1,0.05,0)$ into a high-fidelity GHZ-like single copy suggest practical distillation protocols; a natural extension is to test whether such projections can be chained across more copies to purify GHZ states."],"forward_implications":["Superactivation is quantifiable: white-noise robustness of two-copy GME can be computed by semidefinite programming, and for the most robust GHZ-diagonal state $\\chi(1,1/3,1/3,1/3)$ the robustness is at least $0.5294$.","Two copies of a biseparable GHZ-diagonal state can be mapped to a single copy with GHZ fidelity $0.97561$, which is high enough to certify GHZ-class entanglement rather than only W-class entanglement.","The $k$-copy GHH criterion detects $k$-copy activatability using only single-copy expectation values, and for large $k$ its threshold converges to the PPT boundary for partition separability.","Incompressible entanglement is an explicit phenomenon: for example, for $p$ between $0.781$ and $0.784$ the two-copy state of the symmetric family is GME, while every single-copy reduction is biseparable."],"supporting_citations":[{"why":"Defines the Hadamard-map activation scheme and the projection-based detection that the paper generalizes and tests.","marker":"[14]"},{"why":"Proves every biseparable but not partition-separable state becomes GME for some number of copies, the superactivation fact underpinning the whole study.","marker":"[15]"},{"why":"Provides the PPT-mixture semidefinite programming method used to certify two-copy GME throughout.","marker":"[40]"},{"why":"Gives the necessary-and-sufficient biseparability criterion for GHZ-diagonal states used to benchmark witnesses and noise robustness.","marker":"[42]"},{"why":"Supplies the GHH k-separability criterion that the paper lifts to a k-copy GME witness.","marker":"[61]"},{"why":"The iterative and Gilbert-type algorithms that deliver the explicit biseparable decompositions proving incompressibility.","marker":"[70–72]"},{"why":"Frames the distillability problem that incompressible entanglement is positioned as a multipartite analogue of.","marker":"[23]"},{"why":"Experimental activation benchmark on noisy GHZ states used to model statistical signatures for the nonlinear witnesses.","marker":"[17]"}],"fun_headline_variants":["Two-copy entanglement can be provably incompressible","Superactivation forces entanglement to stay multipartite","Proof: two-copy GME cannot be reduced to single copy","Incompressible multipartite entanglement exists in two copies","Entanglement superactivation yields irreversible multipartite states"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The existence proof for incompressible entanglement relies on computer-generated biseparability certificates for the projected states, and those certificates and their solver tolerances are not included in the paper.","fun_headline_variants_meta":{"raw":{"variants":["Two-copy entanglement can be provably incompressible","Superactivation forces entanglement to stay multipartite","Proof: two-copy GME cannot be reduced to single copy","Incompressible multipartite entanglement exists in two copies","Entanglement superactivation yields irreversible multipartite states"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000136,"raw_usage":{"total_tokens":1186,"prompt_tokens":1022,"completion_tokens":164,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":638,"completion_tokens_details":{"reasoning_tokens":84}},"tokens_in":638,"tokens_out":164,"duration_ms":2027,"temperature":1.0,"reasoning_tokens":84,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T04:48:40.406942+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take $\\rho^{(s)}$ with $p = 0.782$ and search numerically or analytically for any local Kraus operators $F_A$, $F_B$, $F_C$ such that the renormalized projected state $\\tau_2$ is genuinely multipartite entangled; finding one such operator would disprove the incompressibility claim for that state.","supporting_citations":[{"cited_title":"Brunner, D","cited_arxiv_id":null,"evidence_quote":"Defines the Hadamard-map activation scheme and the projection-based detection that the paper generalizes and tests."},{"cited_title":"Cavalcanti and P","cited_arxiv_id":null,"evidence_quote":"Proves every biseparable but not partition-separable state becomes GME for some number of copies, the superactivation fact underpinning the whole study."},{"cited_title":"Loulidi and I","cited_arxiv_id":null,"evidence_quote":"Provides the PPT-mixture semidefinite programming method used to certify two-copy GME throughout."},{"cited_title":"Burchardt, G","cited_arxiv_id":null,"evidence_quote":"Gives the necessary-and-sufficient biseparability criterion for GHZ-diagonal states used to benchmark witnesses and noise robustness."},{"cited_title":"Eltschka and J","cited_arxiv_id":null,"evidence_quote":"Supplies the GHH k-separability criterion that the paper lifts to a k-copy GME witness."},{"cited_title":"Steinberg, H","cited_arxiv_id":null,"evidence_quote":"Frames the distillability problem that incompressible entanglement is positioned as a multipartite analogue of."}],"review_version":1}