{"id":"82240629-9377-4545-b433-c498b5acb1e8","arxiv_id":"2412.13266","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Every pure multipartite entangled state of qubits can be self-tested in the standard Bell scenario up to complex conjugation, using at most 9·2^{n−2}−4 two-outcome measurements per party.","lead":"The paper proves that every pure multipartite entangled state of qubits can be certified from measurement correlations alone, up to the inherent ambiguity of complex conjugation. This resolves the multipartite qubit case of a long-standing open problem in device-independent quantum information.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Section 3's basis-existence step is not proved: the C.4 stitching requires sin(phi001-phi000) and sin(phi101-phi100) nonzero, not merely unequal phases, and the cited [44] plus the phase/Hadamard sketch does not establish all conditions simultaneously.","rationale":"The reader's weakest_assumption identifies exactly the unproven basis-existence step in Section 3, and I agree that it is the most load-bearing point: the entire tripartite stitching in Appendix C.4 depends on comparing coefficients lambda000 and lambda001, and lambda100 and lambda101, to equate the partial-isometry vectors |xi0>, |xi1>, |xi'0>, |xi'1>. My stress-test sharpens this concern in two ways. First, the equations in C.4 require the sines of the phase differences to be nonzero, not just the phases to be unequal; if the phases differ by pi, the sine vanishes and the argument degenerates. The paper's wording 'phases differ' does not exclude this. Second, the proposed construction (phase gate then Hadamard) is not shown to satisfy the nonvanishing, sine-nonzero, and three-projection-entanglement conditions simultaneously; the citation to [44] covers only a single entangled projection statement. These are concrete, checkable algebraic conditions. Despite this gap, I do not recommend changing the CONDITIONAL verdict to REJECT: the asserted basis existence is almost certainly true. For any non-product three-qubit state, if the pencil of Alice-slices were entirely contained in the rank-one variety, the state would be biseparable, so a generic Alice basis makes both post-measurement states entangled; a generic local unitary on Charlie preserves the three rank-two determinants while making the four amplitudes and the two phase sines nonzero. Thus the missing piece is a proof of a true lemma, not a counterexample to the theorem. The measurement-count and induction arguments in Section 4 are secondary: they inherit any defect from the tripartite base case but otherwise appear coherent. I therefore keep the reader's CONDITIONAL verdict.","tokens_in":50958,"tokens_out":17561,"duration_ms":176913,"concrete_test":"Run a feasibility search over local unitaries U_A x U_B x U_C for a dense sample of GME three-qubit states, including the GHZ-class and W-class canonical families, checking whether the following quantities can be made simultaneously non-zero: |lambda000|, |lambda001|, |lambda100|, |lambda101|, |sin(phi001-phi000)|, |sin(phi101-phi100)|, det[[lambda0jk]], det[[lambda1jk]], and det[[lambdai0k]]. If a GME state is found where the maximum over all local unitaries of the minimum of these quantities is zero, the basis-selection lemma is false and the proof fails; if none is found, the concern is a missing derivation rather than a false central claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 3 is proved by combining three sub-tests through the SWAP isometry. In Appendix C.4, the identification of the flag states |xi1> = |xi''1> and |xi'1> = |xi''1> is obtained by subtracting equations after multiplying by exp(i*phi000) and exp(i*phi001); this division is by sin(phi001 - phi000), and similarly for the (100,101) pair. Thus the proof requires the two phase differences to be non-zero modulo pi. Section 3 only asserts that the phases 'differ' and sketches a construction using a diagonal phase on Charlie followed by a Hadamard. No argument is given that this two-step operation preserves the three projected-state entanglement conditions obtained from [44], keeps lambda000, lambda001, lambda100, lambda101 non-zero, and achieves sin(phi001-phi000) and sin(phi101-phi100) simultaneously non-zero. The cited [44] result concerns a single entangled projection, whereas the proof needs three projections -- after A=0, A=1, and B=0 -- to be entangled simultaneously. If for some GME three-qubit state no local computational basis satisfies all of these algebraic conditions, the derivation of Eq. (34) collapses and Theorem 4, which is built inductively on the tripartite result, does not follow from the presented argument. The gap is likely repairable by a generic Charlie rotation, but as written the load-bearing step is asserted rather than proved.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper claims that every pure multipartite entangled state of n qubits can be self-tested in the standard Bell scenario up to complex conjugation. The protocol is modular: it uses tilted-CHSH self-tests of two-qubit projected states, a measurement lemma for post-hoc certification of observables, and a SWAP isometry. The tripartite case (Theorem 3) is built from three sub-tests that certify the bipartite projections after Alice's and Bob's computational-basis measurements; the general case (Theorem 4) is obtained by induction and is claimed to require at most 9·2^{n−2}−4 two-outcome measurements per party. Detailed proofs of the bipartite self-testing lemma and the measurement lemma are provided in Appendices A and B.","tokens_in":51309,"tokens_out":18171,"duration_ms":182019,"significance":"If the proof is completed, this is a major result: it would settle the open problem of self-testing all pure qubit multipartite entangled states in the standard Bell scenario, with an explicit measurement count. The paper is unusually concrete about the correlations needed for each sub-test, and it re-proves its central bipartite ingredients rather than merely citing them. The up-to-complex-conjugation certification is the right target given that some multipartite states are not locally unitarily equivalent to their conjugates. The modular construction is a natural and potentially influential framework for subsequent device-independent certification results.","major_comments":[{"comment":"The proof of Theorem 3 relies on an asserted but unproved basis-existence step. The derivation of Eq. (34) requires, for every GME three-qubit state, local computational bases in which the three projected states |ψa+>, |ψa−>, and |ψb+> are all entangled, the four coefficients λ000, λ001, λ100, and λ101 are all nonzero, and the phase differences satisfy sin(φ001−φ000) ≠ 0 and sin(φ101−φ100) ≠ 0. Section 3 cites [44] only for the existence of an entangled projection, and the subsequent diagonal-phase-plus-Hadamard sketch is not shown to preserve the simultaneous entanglement of all three projections while enforcing the required phase inequalities. Appendix C.4 divides by these sine factors to identify the flag states, and Appendix D invokes the analogous coefficient conditions for every sub-vector in Eq. (D15) without proof. This is a load-bearing gap: if any GME three-qubit state fails to admit such a basis, the stitching argument after Eq. (C70) and hence Eq. (34) do not follow from the presented argument. The authors should supply a proof or an explicit construction of such a basis, or state and prove a dedicated lemma.","section":"Section 3 and Appendix C.4"},{"comment":"The stitching step that identifies |ξ′0> with |ξ″0> and |ξ′1> with |ξ″1> is also obscured by apparent sign and index errors at the exact point where the third sub-test is combined with the first two. Eq. (C69c) writes the coefficient of |ξ″1> as λ*001, but consistency with line (C60e) requires λ*100. Immediately after Eq. (C69d) the text states that the equations are used 'under the assumption ϕ100 = ϕ101'; the derivation actually requires ϕ100 ≠ ϕ101, since it subtracts the two equations and divides by sin(ϕ101−ϕ100). These are likely typographical errors, but because they occur in the most delicate part of the proof, the corrected equations and the resulting division should be written out explicitly.","section":"Appendix C.4, Eqs. (C69c)–(C69d)"}],"minor_comments":[{"comment":"The sentence 'we select a basis in which the complex phases of λ000 and λ001 differ, as do those of λ100 and λ101' should specify that the phase differences are nonzero modulo π, since the later proof divides by the corresponding sines.","section":"Section 3"},{"comment":"The notation |0*0*> and |0*1*> used in Eqs. (C21), (C40), and (C59) for the complex-conjugated basis kets is never formally defined; a sentence explaining that the star denotes the complex-conjugate basis vectors would improve readability.","section":"Appendix A"},{"comment":"The proof of Theorem 4 states that the local computational bases are chosen so that for every vector a2 the coefficients λ^{a2}_{00} and λ^{a2}_{10} are nonzero and have different phases, but no justification is given for the simultaneous existence of these bases beyond the tripartite discussion; this should at least be linked to the missing lemma flagged in the major comment.","section":"Appendix D"}],"recommendation":"major_revision","confidential_remarks":"This is a strong and important manuscript, and the main gap is likely repairable by a generic-rotation argument or a dedicated lemma on simultaneous entangled projections with prescribed phase differences. I therefore recommend major revision rather than rejection. Please ensure the corrected Appendix C.4 equations are carefully re-checked, since the current text contains both an index typo and a reversed inequality assumption at the load-bearing stitching step."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe headline: this paper genuinely resolves the multipartite qubit self-testing question: for any pure n-qubit state there is a standard Bell scenario protocol that certifies it up to a flag superposition with its complex conjugate. That is new and important.\n\nWhat it does well: the modular structure is clean. It reduces to bipartite self-testing as a black box, uses a measurement lemma to transfer self-tests between parties, and a SWAP isometry to assemble the pieces. The appendices re-prove the bipartite lemma, so the paper is fairly self-contained. The treatment of complex conjugation as an unavoidable ambiguity is careful and correctly reflected in the definition.\n\nThe soft spots: Section 3 asserts there is a local computational basis in which the projected two-qubit states are all entangled and the four coefficients λ000, λ001, λ100, λ101 are nonzero with pairwise different phases. The sketch (diagonal phase on Charlie plus Hadamard) does not prove that these conditions can be imposed simultaneously. This matters: in C.4, the stitching step divides by sin(φ001-φ000) and sin(φ101-φ100), so you need those sines nonzero, not merely unequal phases. The cited [44] only covers the entangled projection, not the simultaneity with the phase conditions. This is a real gap in the written proof, though it looks repairable—a generic rotation of Charlie's local basis should do the job.\n\nThere are also typos in critical equations: (C69c) writes λ*001 where λ*100 is required, and the sentence after (C69d) refers to \"under the assumption φ100 = φ101\"—exactly backwards from what the proof needs. These are fixable but need attention.\n\nBottom line: the central theorem is new and the architecture is credible. The gap is in a technical basis-existence lemma, not in the overall strategy. The paper deserves a serious referee, but not acceptance in the current form. I'd send it to review with a request to complete the proof of the basis conditions and clean up the appendix equations.\n\nRecommendation: peer review, conditional accept after revision.","headline":"New result of real importance; proof has a repairable gap in basis selection, plus fixable typos; deserves review.","tokens_in":51819,"tokens_out":4070,"would_cite":true,"duration_ms":35104,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P40","81P68","81P15"],"pacs":["03.67.-a","03.65.Ud"],"model":"deepseek-v4-flash","headline":"Every pure multipartite state of qubits can be self-tested from correlations alone, up to complex conjugation, using a finite explicit protocol for each n.","keywords":["self-testing","multipartite entanglement","qubits","complex conjugation","Bell nonlocality","device-independent certification","tilted CHSH","SWAP isometry"],"falsifier":"Search for a genuinely multipartite entangled three-qubit state for which no choice of local bases satisfies all of the following at once: the projected two-qubit states are entangled and $\\lambda_{000},\\lambda_{001},\\lambda_{100},\\lambda_{101}$ are all nonzero with $\\phi_{000}\\neq\\phi_{001}$ and $\\phi_{100}\\neq\\phi_{101}$. If such a state exists, the SWAP stitching step in Appendix C.4 has no basis to start from and Theorem 3 would not follow from the presented argument for that state.","tokens_in":50815,"feed_emoji":"⚛️","tokens_out":11143,"duration_ms":86820,"temperature":0.7,"pith_summary":"The paper aims to close the multipartite self-testing question for qubits: it claims that every pure entangled state of any number of qubits can be certified from Bell correlations alone in the standard Bell scenario, with the only residual freedom being complex conjugation of the target state. Complete characterizations existed for two parties and for special multipartite families, but not for all pure multipartite qubit states. The proposed recipe is modular: project all but two parties in self-tested bases, self-test each resulting partially entangled two-qubit state, and stitch the pieces together with a SWAP isometry. If the claim holds, any pure multi-qubit entangled state admits a device-independent certificate with at most $9\\cdot 2^{n-2}-4$ two-outcome measurements per party, and the certificate does not assume purity or projective measurements.","feed_headline":"Every pure multipartite qubit state can be self-tested","feed_subtitle":"Certifies any n-qubit entangled state up to complex conjugation using at most 9·2^(n−2)−4 binary measurements per party.","key_machinery":"The load-bearing object is a modular projection-and-stitching scheme. In each sub-test, all but two parties measure their self-tested diamond observables, projecting the state onto a partially entangled pair; that pair is self-tested with the tilted CHSH inequalities of Lemma 1, which also certify three Pauli observables on one side. Lemma 2 (the measurement lemma) then certifies the projecting observables themselves from their correlations with the already-certified ones. A SWAP isometry built from the certified observables extracts the physical state into auxiliary qubits, and the sub-tests are designed so that the extracted pieces coalesce into exactly two orthogonal vectors, attached to $|\\Psi\\rangle$ and $|\\Psi^{*}\\rangle$.","core_discovery":"The central claim is that self-testing is complete for pure multipartite qubit states. For any $n$-qubit pure state $|\\Psi\\rangle$ there is a correlation, built from at most $9\\cdot 2^{n-2}-4$ two-outcome measurements per party, such that every physical realization reproducing the correlation is, up to local unitaries, a flag superposition $\\sqrt{p}|\\Psi\\rangle|0\\rangle + \\sqrt{1-p}|\\Psi^{*}\\rangle|1\\rangle$ with orthogonal flag states $|0\\rangle,|1\\rangle$. When the target state is locally equivalent to its complex conjugate, self-testing is exact. The proof treats non-genuinely-multipartite-entangled states through known bipartite self-tests and develops the genuinely-multipartite case by a sequence of sub-tests; the final output of the SWAP isometry is forced to the form $|\\xi_0\\rangle|\\Psi\\rangle + |\\xi_1\\rangle|\\Psi^{*}\\rangle$, which is exactly the flag form of Definition 1.","pith_inferences":["The exponential growth in measurements is a property of this universal construction, not a proven lower bound; families with symmetry or algebraic structure plausibly admit polynomial or constant-size protocols, as already happens for W and graph states in the literature the paper cites.","The flag formulation gives a device-independent way to probe the complex-conjugation class of a state: estimating $p$ from the extracted flag states would tell whether the physical state is locally equivalent to $|\\Psi\\rangle$ or to $|\\Psi^{*}\\rangle$.","A numerical search over three-qubit genuinely multipartite entangled states could test the unproved basis-selection premise directly; if violations of the phase and nonvanishing conditions are found, the stitching argument would need a modified extraction step, while the conclusion might still hold."],"forward_implications":["For every pure $n$-qubit state, the protocol certifies either the state itself up to local unitaries when it is locally equivalent to its complex conjugate, or the two-dimensional flag subspace spanned by $|\\Psi\\rangle$ and $|\\Psi^{*}\\rangle$.","The protocol is fully device-independent: it assumes neither purity of the physical state nor projective measurements, and the support-preserving condition is derived from the correlations rather than imposed.","The number of measurement settings per party is explicit and finite for every fixed $n$: at most $9\\cdot 2^{n-2}-4$ binary-outcome settings, with party 1 using $3+9(2^{n-2}-1)$ of them.","Any non-genuinely-multipartite-entangled $n$-qubit state inherits a self-test from the bipartite case, so the characterization covers all pure $n$-qubit states, entangled or not."],"supporting_citations":[{"why":"Supplies Lemma 1, the self-test of a partially entangled two-qubit pair and three Pauli observables via tilted CHSH correlations.","marker":"[41]"},{"why":"Establishes that for any genuinely multipartite entangled state there are local bases in which the projection onto two parties is entangled.","marker":"[44]"},{"why":"Provides the post-hoc measurement self-testing formalism that Lemma 2 and Lemma 5 rely on.","marker":"[22]"},{"why":"Introduces the projection-based modular self-testing method that the present protocol generalizes.","marker":"[26]"},{"why":"Shows that arbitrary bipartite qubit states can be self-tested, which covers the non-genuinely-multipartite case.","marker":"[43]"},{"why":"Establishes that all pure bipartite entangled states can be self-tested, the bipartite completeness result being extended here.","marker":"[24]"},{"why":"Shows three-qubit states that are not locally equivalent to their complex conjugate, motivating the flag-qubit formulation.","marker":"[35]"},{"why":"Characterizes local unitary equivalence of multipartite pure states, including the complex-conjugation obstruction.","marker":"[36]"}],"fun_headline_variants":["All pure multipartite qubit states self-test up to complex conjugation","Self-testing complete for all pure multipartite qubit states","Every pure multipartite qubit state self-tests with finite measurements","Complete self-testing for any pure multipartite qubit state","Multipartite qubit self-testing solved up to complex conjugation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The weakest premise is the asserted existence, for every genuinely multipartite entangled three-qubit state, of local computational bases in which the two-qubit projections used in the sub-tests are entangled and the four amplitudes $\\lambda_{000},\\lambda_{001},\\lambda_{100},\\lambda_{101}$ are nonzero with unequal phases; the paper cites a previous result for the entangled-projection part but does not prove that the nonvanishing and phase conditions can be forced simultaneously.","fun_headline_variants_meta":{"raw":{"variants":["All pure multipartite qubit states self-test up to complex conjugation","Self-testing complete for all pure multipartite qubit states","Every pure multipartite qubit state self-tests with finite measurements","Complete self-testing for any pure multipartite qubit state","Multipartite qubit self-testing solved up to complex conjugation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000718,"raw_usage":{"total_tokens":3207,"prompt_tokens":906,"completion_tokens":2301,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":522,"completion_tokens_details":{"reasoning_tokens":2208}},"tokens_in":522,"tokens_out":2301,"duration_ms":14851,"temperature":1.0,"reasoning_tokens":2208,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T13:18:58.292551+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search for a genuinely multipartite entangled three-qubit state for which no choice of local bases satisfies all of the following at once: the projected two-qubit states are entangled and $\\lambda_{000},\\lambda_{001},\\lambda_{100},\\lambda_{101}$ are all nonzero with $\\phi_{000}\\neq\\phi_{001}$ and $\\phi_{100}\\neq\\phi_{101}$. If such a state exists, the SWAP stitching step in Appendix C.4 has no basis to start from and Theorem 3 would not follow from the presented argument for that state.","supporting_citations":[{"cited_title":"Maximal randomness from partially entangled states","cited_arxiv_id":null,"evidence_quote":"Supplies Lemma 1, the self-test of a partially entangled two-qubit pair and three Pauli observables via tilted CHSH correlations."},{"cited_title":"Device-independent detection of genuine mul- tipartite entanglement for all pure states","cited_arxiv_id":null,"evidence_quote":"Establishes that for any genuinely multipartite entangled state there are local bases in which the projection onto two parties is entangled."},{"cited_title":"All real projective measurements can be self- tested","cited_arxiv_id":null,"evidence_quote":"Provides the post-hoc measurement self-testing formalism that Lemma 2 and Lemma 5 rely on."},{"cited_title":"Self-testing multipartite entangled states through projections onto two systems","cited_arxiv_id":null,"evidence_quote":"Introduces the projection-based modular self-testing method that the present protocol generalizes."},{"cited_title":"Sum-of-squares decompositions for a family of Clauser-Horne- Shimony-Holt-like inequalities and their application to self-testing.Phys","cited_arxiv_id":null,"evidence_quote":"Shows that arbitrary bipartite qubit states can be self-tested, which covers the non-genuinely-multipartite case."},{"cited_title":"Three-qubit pure-state canonical forms.Journal of Physics A: Mathematical and General, 34(35):6725–6739, aug 2001","cited_arxiv_id":null,"evidence_quote":"Shows three-qubit states that are not locally equivalent to their complex conjugate, motivating the flag-qubit formulation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Characterizes local unitary equivalence of multipartite pure states, including the complex-conjugation obstruction."}],"review_version":1}