{"id":"35bb876c-d34b-449b-804d-4a700b9269e9","arxiv_id":"2411.19619","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Local state discrimination success bounds CHSH violation, maximally entangled-state fidelity, and global energy for ensembles of pure bipartite states.","lead":"The authors show that how well two parties can locally tell apart quantum state preparations limits how nonlocal, entangled, or energetic the average of those preparations can be. This gives a new way to constrain an adversary's global quantum behavior using only local measurements, with potential use in quantum cryptography.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The N≥3 bounds hinge on an unproven monotonicity/rigidity claim (Eq14-15) that inverts observed local success probability into a uniform bound on all global pairwise overlaps; without it, the fidelity and energy bounds are unsupported.","rationale":"The reader's weakest_assumption identifies exactly the same load-bearing gap: Eq14-15 are asserted without proof and are used to convert p_L^s into a bound on the global overlap δ. My stress-test agrees that this is the hinge of the paper. The N=2 CHSH trade-off has a self-contained proof in Appendix A and appears sound. The fidelity and energy bounds, however, are derived by first inverting p_L^s through p_N^s(δ) and then applying the fidelity lemma; without a proof of Eq14-15, the derivation is incomplete. The paper would be conditionally acceptable if the authors supply a rigorous proof of the monotonicity/rigidity claim, or a citation to an existing theorem, and release the SDP code or numerical data used in §VI. A concrete SDP test can settle whether the claim is true; if it fails, the N≥3 claims would need to be substantially revised. The mixed-state limitation is real but secondary, since the paper explicitly restricts to pure preparations. I therefore recommend keeping the conditional verdict rather than accepting or rejecting outright.","tokens_in":15603,"tokens_out":45787,"duration_ms":370510,"concrete_test":"For N=3,4 and δ ∈ {0.2,0.5,0.8}, solve the SDP (or run a random Gram-matrix search) that maximizes the minimum-error discrimination success probability over N pure states whose pairwise inner products are all ≥ δ, and compare the optimum with p_N^s(δ) in Eq13. Eq14 is confirmed only if the optimum equals p_N^s(δ) in every tested case; a single exceedance falsifies it. Also test Eq15 by searching for Gram matrices with at least one off-diagonal entry > δ that still achieve the same optimal success probability; if any is found, the inversion used for Eq17-19 is invalid.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central inversion is: observing p_L^s = p_N^s(δ) is taken to imply that every pairwise global overlap is ≤ δ. This relies on Eq14-15, which assert that among ensembles with all overlaps ≥ δ, the equidistant ensemble is optimally distinguishable and that attaining p_N^s(δ) forces all overlaps ≤ δ. Neither direction is proved or cited. The N=2 case can be justified via reduced-state fidelity and Fuchs-van de Graaf, but for N≥3 the analogous statement is a nontrivial optimization over Gram matrices with pairwise overlap constraints. Eq17 and Eq19 are obtained by feeding this inferred δ into the fidelity bound, so they inherit all the weight of Eq14-15. If Eq14-15 fail, an adversary could have large global overlaps (hence high fidelity/energy restrictions) while still permitting the observed local success probability. The N=3,4 CHSH bounds in Appendix A also assume the same kind of reduction. The numerical SDP in §VI is described but no code or data is provided, so it does not independently verify the monotonicity. This is not a claim that the bounds are false; it is a claim that the proof is incomplete at its load-bearing point.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates a prepare-and-measure scenario in which Charlie prepares N bipartite pure states, and Alice and Bob use fixed local measurements to discriminate them. The authors claim that the optimal local success probability p_L^s restricts global properties of the ensemble: an upper bound on the CHSH winning probability (Eq. (10) for N=2 and extensions for N=3,4), an upper bound on the fidelity with any maximally entangled state (Eq. (17)), and a lower bound on the energy expectation (Eq. (19)). The argument for N=2 combines the Helstrom bound with a spectrum-based CHSH optimization in Appendix A. For general N, the paper constructs an axisymmetric ensemble (Eq. (12)) whose local reduced states are optimally distinguishable, and then invokes a monotonicity statement (Eqs. (14)-(15)) to invert an observed p_L^s into a uniform upper bound on all pairwise global overlaps. A final section compares local versus global discrimination with inconclusive events, using an SDP hierarchy.","tokens_in":15885,"tokens_out":8403,"duration_ms":73292,"significance":"If the main claims are correct, the paper offers a conceptually appealing and potentially useful tool: from a purely local distinguishability measurement one can certify upper bounds on Bell violations and entanglement fidelity, and lower bounds on energy, of an uncharacterized shared ensemble. This is a fresh angle on semi-device-independent certification. The N=2 derivation is rigorous and built on standard tools (Helstrom bound, Verstraete-Wolf CHSH optimization), and Appendix C contains a clean and correct fidelity bound. The numerical SDP in Section VI supports the N=2 trade-off. However, the generalization to N>=3 rests on an unproved monotonicity/rigidity claim that is load-bearing for Eqs. (17) and (19), and the N=3,4 CHSH extension is only computed for the symmetric ensemble. The paper therefore has a valuable core but is not yet complete at its central proof step.","major_comments":[{"comment":"The monotonicity claim in Eq. (14) is asserted without proof or citation: it states that among all ensembles of N pure states with pairwise overlaps at least δ, the equidistant (axisymmetric) ensemble is the most distinguishable. For N>=3 this is a nontrivial optimization over Gram matrices with pairwise overlap constraints, and it is not obvious that the maximum of the minimal-error success probability is attained by the symmetric ensemble. This claim is load-bearing: it is the only mechanism that converts an observed local success probability p_L^s into a bound on all global overlaps, which is then used to derive Eq. (17) and Eq. (19). Until Eq. (14) is proved (or supported by a citable theorem), the N>=3 fidelity and energy bounds are unsupported.","section":"Sec. IV, Eqs. (14)-(15)"},{"comment":"The inversion step in Eq. (15) is not logically established. The argument says that if some overlaps are reduced, distinguishability increases, so an ensemble with some overlaps smaller than δ can still reach p_N^s(δ). This does not imply that every ensemble attaining p_N^s(δ) has all overlaps at most δ: an ensemble with one overlap slightly above δ and another overlap well below δ could conceivably have the same overall success probability. The statement 'p_φ^s = p_N^s(δ) ⇒ 〈φ_z|φ_z'⟩ ≤ δ for all z,z' requires proof. This inversion is exactly what allows the authors to feed a single inferred δ into the fidelity bound, so it is a second load-bearing gap in the N>=3 argument.","section":"Sec. IV, Eq. (15)"},{"comment":"The CHSH bound for N=3,4 is derived only for the equidistant ensemble, i.e., under the assumption that all pairwise overlaps equal δ. The text claims that 'this construction works in the general case', but no proof is given that among all ensembles with pairwise overlaps bounded by δ, the symmetric ensemble maximizes the CHSH violation. The eigenvalues in Eq. (A10) are specific to the equidistant Gram matrix. Without an optimality argument over non-symmetric ensembles, Fig. 4 and the associated N=3,4 trade-off statements are not established for arbitrary ensembles with bounded overlaps.","section":"Appendix A, Eqs. (A8)-(A11)"}],"minor_comments":[{"comment":"The abstract contains a typo: 'maximally entangled sate fidelity' should read 'maximally entangled state fidelity'.","section":"Abstract"},{"comment":"The sentence 'coinciding with Eq. (9)' appears to be a cross-reference error: Eq. (A7) is the CHSH winning-probability bound and should be compared with Eq. (7) of the main text, not with the Helstrom bound in Eq. (9).","section":"Appendix A, after Eq. (A7)"},{"comment":"The comparison between local and global discrimination strategies is presented as a numerical observation, but no code or data are provided. Since the Gram-matrix SDP is an outer approximation, the claim that the local and global feasible regions coincide needs either a formal argument or reproducible numerical evidence.","section":"Sec. VI"},{"comment":"The scope is restricted to pure state preparations, but this is not made explicit in the abstract or introduction. The phrase 'N bi-partite pure state preparations' appears only in Section IV; stating the purity assumption earlier would help readers assess the applicability of the bounds.","section":"Sec. IV"},{"comment":"The two panels of Fig. 3 are labelled 'Free states' and 'Entangled states', but the caption does not explain how these labels correspond to the overlap parameter δ and to the ensemble in Eq. (5). Clarifying the distinction between the two panels would improve readability.","section":"Sec. VI, Fig. 3"}],"recommendation":"major_revision","confidential_remarks":"The paper's central idea is interesting and the N=2 portion is solid, but the N>=3 results depend on an unproved monotonicity/rigidity statement (Eqs. (14)-(15)). I would be willing to reconsider after the authors either supply a complete proof of this statement or clearly restrict the claims to cases where the inversion can be justified. The lack of code for the numerical SDP is a secondary concern but should be addressed in revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nHere is my read on Roch i Carceller and Bernal. The paper's core idea is genuinely nice: use the optimal local state discrimination success probability to bound global properties of a bipartite ensemble—CHSH violation for two pure states, fidelity with a maximally entangled state, and global energy. The N=2 CHSH trade-off is solid: combining the Helstrom bound with the Verstraete-Wolf spectrum-based CHSH bound gives Eq. (10) cleanly, and Appendix A proves it for arbitrary prior probabilities. That part is a real result, though it follows from known ingredients.\n\nWhat is new is the N-state picture: the fidelity bound (Eq. 17) and energy bound (Eqs. 18-19) are not in the prior literature, and the local-vs-global measurement comparison in Section VI is a useful observation. The axisymmetric family construction in Appendix B is careful, and Appendix C gives a legitimate proof of the fidelity bound conditional on the overlaps.\n\nThe soft spot is exactly where the stress-test lands. The whole N≥3 inversion—observing p_L^s and concluding every pairwise overlap is at most δ—rests on the monotonicity/rigidity claim in Eqs. (14)-(15), which is asserted with a heuristic paragraph, not proven. No citation is supplied, and the supplementary material referenced as [28] does not appear to contain it either. For N=2 the claim follows from Fuchs-van de Graaf and reduced state fidelity, so the two-state results stand. But for N≥3, Eqs. (17)-(19) and the N=3,4 CHSH curves are unsupported at their load-bearing point. It is plausible the claim is true—the equidistant ensemble is the natural extremal case—but it is a nontrivial optimization over Gram matrices with pairwise overlap constraints, and the paper gives no proof.\n\nMinor issues: the numerical SDP section has no code or data, and the N=3,4 CHSH generalization is only computed for the symmetric ensemble, so it is really a conjecture unless the monotonicity is proven. The text gestures at these limitations but does not flag the missing proof explicitly.\n\nOverall, this paper is worth engaging. The two-state result is proven, the framework is interesting, and the gap is a missing proof rather than a demonstrated counterexample. A serious referee could push for a proof of Eqs. (14)-(15) or, failing that, a clear conjecture label plus numerical evidence. I would send it to peer review, with the expectation of a major revision.\n\nRecommendation: accept for serious peer review, conditional on the authors proving or properly flagging the monotonicity claim. I would not cite the N≥3 bounds in my own work until the inversion is settled.","headline":"A useful extension of local discrimination bounds to global properties, but the N>=3 results rest on an unproved monotonicity claim that needs a proof or a conjecture label.","tokens_in":16353,"tokens_out":3842,"would_cite":false,"duration_ms":28743,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P45","81P15"],"pacs":["03.67.-a","03.65.Ud"],"model":"deepseek-v4-flash","headline":"The optimal local discrimination probability of a bipartite ensemble bounds its CHSH violation, maximally entangled fidelity, and energy.","keywords":["local state discrimination","CHSH inequality","Bell nonlocality","maximally entangled state fidelity","energy bound","semi-device-independent certification","axisymmetric states","quantum state discrimination"],"falsifier":"Find a set of N pure bipartite states with pairwise overlaps no smaller than δ whose optimal local discrimination success probability exceeds p_N^s(δ), or whose maximal CHSH winning probability exceeds the bound of Eq. (10) for the observed p_L^s; an SDP search over qubit ensembles for N = 3 or 4 would settle it.","tokens_in":15412,"feed_emoji":"⚛️","tokens_out":5153,"duration_ms":40044,"temperature":0.7,"pith_summary":"This paper establishes that a single number—the optimal probability with which two separated parties can tell apart the reduced states in a prepared ensemble—places hard limits on global properties of the shared states. For two pure preparations with overlap δ, the local success probability satisfies p_L^s ≤ (1/2)(1+√(1−δ²)) (Eq. 9), so observing p_L^s fixes a maximal overlap δ; from that, the CHSH winning probability is bounded by Eq. (10). The argument extends to N preparations: if Alice and Bob locally distinguish their shares with probability p_L^s, the ensemble's fidelity with any maximally entangled state is at most Eq. (17), and the expectation value of a global observable such as the energy is bounded below by Eq. (18)–(19). The interest is that global behaviours—non-locality, entanglement fidelity, energy—can be certified from a local distinguishability task alone, which matters for bounding what an entangled adversary can do in quantum cryptography.","feed_headline":"Local state discrimination caps CHSH, fidelity, and energy","feed_subtitle":"A single local success probability bounds non-locality, entanglement fidelity, and energy of a shared ensemble.","key_machinery":"The load-bearing object is the N-state discrimination function p_N^s(δ) = (1/N²)(√(1+(N−1)δ)+(N−1)√(1−δ))² (Eq. 13), together with its inversion: a measured p_L^s fixes a maximum average overlap δ. The paper's axisymmetric family (Eq. 12) saturates this bound through local projections onto the largest eigenvalue of each reduced state, giving the tightest link between local distinguishability and global overlaps. The extremality claim (Eq. 14–15)—equidistant states are the hardest to distinguish among all ensembles with the same minimal overlap—is what turns a concrete family into a universal bound.","core_discovery":"The paper's central claim is that local state discrimination acts as a universal constraint on global features of a bipartite ensemble. Given an observed local success probability p_L^s, one maps it to a pairwise overlap δ through the optimal discrimination bound for N equidistant pure states (Eq. 13), and then δ bounds the CHSH winning probability (Eq. 10), the maximal fidelity with a maximally entangled state (Eq. 17), and the minimum energy (Eq. 18). The construction relies on the axisymmetric ensemble of Eq. (12), whose partial traces are diagonal and optimally distinguishable locally, and on the assertion (Eq. 14–15) that this equidistant ensemble is extremal: any ensemble with pairwise overlaps at least δ is no more distinguishable. If correct, the direction of reasoning runs purely from local measurements to global restrictions.","pith_inferences":["The extremality claim of Eqs. (14)–(15) is where the general N bounds hinge; a reader should expect a proof there rather than a heuristic, since a counterexample would shrink the fidelity and energy bounds to the axisymmetric family only.","For mixed-state ensembles the mapping from p_L^s to δ breaks because reduced states need not be pure; a natural extension would replace pure-state overlaps by some mixed-state distinguishability measure and yield looser but still meaningful bounds.","The energy bound suggests a concrete experimental test: in an optical prepare-and-measure setup, measuring local discrimination success of coherent-state encoding should give a lower bound on average photon number, which can be checked independently."],"forward_implications":["Two parties who only measure how well they can discriminate their local shares can place an upper bound on how much they violate CHSH, without ever running the CHSH game.","In a prepare-and-measure protocol, a bounded local discrimination success probability certifies that the shared ensemble cannot have high fidelity with any maximally entangled state in dimension N².","The same bound translates into a lower bound on the expectation of a global observable such as the vacuum-projector energy, so a power-meter-style estimate follows from a local discrimination measurement.","For N > 2 qubit preparations, the results show a gap between CHSH violation and full distinguishability: non-locality is impossible as soon as preparations are not equivalent.","When only overlaps are constrained, local and global measurement strategies achieve the same optimal success/error region; a gap appears only when the preparations are entangled."],"supporting_citations":[{"why":"Supplies the two-state discrimination bound used to convert a local success probability into an overlap δ.","marker":"[5]"},{"why":"Supplies the quantum upper bound on the CHSH winning probability, used in the non-locality trade-off.","marker":"[27]"},{"why":"Supplementary material carrying the general derivations for the CHSH, fidelity, and energy bounds.","marker":"[28]"},{"why":"Defines the success probability p_N^s(δ) for N equidistant pure states, central to inverting p_L^s.","marker":"[34]"},{"why":"Provides the reduction to Bell-diagonal states that gives the optimal CHSH violation for a fixed spectrum.","marker":"[39]"},{"why":"Supplies the trace bound on the largest eigenvalue used to derive the maximal-fidelity bound.","marker":"[40]"}],"fun_headline_variants":["One local number caps CHSH, fidelity, and energy","Local success odds bound CHSH, fidelity, and energy","Local measurement limits three global state attributes","From local tests to global quantum bounds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The bounds assume that an equidistant ensemble with pairwise overlap δ is the hardest to distinguish among all pure-state ensembles with overlaps at least δ; this monotonicity, stated in Eqs. (14)–(15), is asserted without proof and carries the fidelity and energy bounds.","fun_headline_variants_meta":{"raw":{"variants":["One local number caps CHSH, fidelity, and energy","Local success odds bound CHSH, fidelity, and energy","Local measurement limits three global state attributes","From local tests to global quantum bounds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001859,"raw_usage":{"total_tokens":7220,"prompt_tokens":787,"completion_tokens":6433,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":403,"completion_tokens_details":{"reasoning_tokens":6374}},"tokens_in":403,"tokens_out":6433,"duration_ms":42381,"temperature":1.0,"reasoning_tokens":6374,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T06:00:34.016941+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a set of N pure bipartite states with pairwise overlaps no smaller than δ whose optimal local discrimination success probability exceeds p_N^s(δ), or whose maximal CHSH winning probability exceeds the bound of Eq. (10) for the observed p_L^s; an SDP search over qubit ensembles for N = 3 or 4 would settle it.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the quantum upper bound on the CHSH winning probability, used in the non-locality trade-off."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplementary material carrying the general derivations for the CHSH, fidelity, and energy bounds."},{"cited_title":"Information capacity of quantum communication under natural physical assumptions","cited_arxiv_id":"2405.07231","evidence_quote":"Defines the success probability p_N^s(δ) for N equidistant pure states, central to inverting p_L^s."},{"cited_title":"Verstraete and M","cited_arxiv_id":null,"evidence_quote":"Provides the reduction to Bell-diagonal states that gives the optimal CHSH violation for a fixed spectrum."},{"cited_title":"Wolkowicz and G","cited_arxiv_id":null,"evidence_quote":"Supplies the trace bound on the largest eigenvalue used to derive the maximal-fidelity bound."}],"review_version":1}