{"id":"4b03b1fc-d257-4919-b1d6-3311f89e42bd","arxiv_id":"2511.21079","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Average fidelity and fidelity deviation form two parallel witness lines that separate teleportation advantage from Bell-inequality violation in any dimension.","lead":"This paper derives a universal relation between the average fidelity and the input-to-input fidelity deviation of d-dimensional quantum teleportation, and uses it to draw two parallel witness lines in the (F,D) plane—one for beating classical teleportation, one for Bell-nonlocality violation. A measured (F,D) point falling above either line certifies a stronger resource than the classical or local bound.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (43) is false for d=3: a single valid wiring X_alpha = diag(1, e^{i pi/3}, e^{-i pi/3}) violates the universal deviation bound, so the witness wedges can certify CGLMP violation for p < p_BV.","rationale":"The reader identified Eq. (43) as the weakest assumption, but treated it as an unproved gap ('by convexity') that might be filled. The stress test shows the gap is not merely a missing proof: the inequality is actually false. The counterexample uses only the paper's own closed-form expressions: with d=3 and all X_alpha equal to a fixed traceless? no, trace-2 unitary, the ratio D/(Fmax-F) is approximately 0.3899, exceeding s_3 = 1/3 for all p>0. This directly violates the universal bound that Theorems 1 and 2 substitute into to define Wlocal and Wcl. Moreover, the violation is not harmless: for p in (0.629, 0.696), the resulting (F,D) point lies outside Wlocal even though p < p_BV(3) ≈ 0.696, so the proposed witness would certify CGLMP violation for a resource that cannot violate CGLMP. Since the central diagnostic map and the claimed separation between teleportation advantage and Bell-nonlocality are derived from this bound, the main conclusions are unsupported and, in the demonstrated regime, incorrect. The reader's conditional acceptance should be replaced by rejection unless the bound is repaired or the witnesses are recalibrated. The counterexample is simple enough to be checked by direct numerical integration or by a symbolic recalculation of Eq. (34), so the issue is decisive rather than stylistic.","tokens_in":19466,"tokens_out":21379,"duration_ms":200778,"concrete_test":"Recompute F and D for the d=3 protocol with X_alpha = diag(1, e^{i pi/3}, e^{-i pi/3}) for all alpha and p = 0.65, using Eqs. (14), (15), (34), and (39). Expected result: F = 0.4958, D = 0.1056, and the point lies outside Wlocal (allowed D ≈ 0.1005). If confirmed, Theorem 1 is falsified; if not, the error is in Eq. (34) or the D reduction, and the bound/proof must be re-derived.","verdict_should_be":"REJECT","load_bearing_attack":"The load-bearing inequality is Eq. (43): D <= s_d (Fmax(p)-F), with s_d = sqrt(2/[d(d+3)]). It is not merely unproved; it is false. Take d=3 and X_alpha = X = diag(1, e^{i pi/3}, e^{-i pi/3}) for all alpha (valid, since V_alpha = X U_alpha). Using the paper's own Eq. (34), E|⟨phi|X|phi⟩|^2 = (|Tr X|^2 + d)/(d(d+1)) = (4+3)/12 = 7/12, and E|⟨phi|X|phi⟩|^4 = 132/360 = 11/30. Hence for the isotropic protocol, F = p(7/12) + (1-p)/3 and D = p sqrt(11/30 - (7/12)^2) = p sqrt(19/720). Since Fmax(p) = p + (1-p)/3, Fmax - F = 5p/12. Thus D/(Fmax-F) = (12/5) sqrt(19/720) ≈ 0.3899 > 1/3 = s_3. So Eq. (43) fails for every p > 0. Concretely, with p = 0.65, (F,D) = (0.4958, 0.1056). For p_BV(3) ≈ 0.696, Fmax_BV ≈ 0.7973, so the Wlocal line permits D ≤ (0.7973 - 0.4958)/3 ≈ 0.1005; the point lies outside Wlocal although p < p_BV. Theorems 1 and 2 and the entire witness-wedge certification therefore rest on a false premise.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a Schur-Weyl/Weingarten calculus for the Haar averages of the teleportation fidelity f(phi) and uses it to compute the average fidelity F and fidelity deviation D for arbitrary dimension d with an isotropic resource of visibility p. It then claims a universal bound D <= s_d (Fmax(p) - F), with s_d = sqrt(2/[d(d+3)]), and uses this bound to draw two witness wedges in the (F,D) plane: one certifying p > p_BV(d) (CGLMP violation) and one certifying p > p_c (teleportation advantage). The main theorems assert that all p <= p_BV resources lie in W_local and all p <= p_c resources lie in W_cl, so points outside these wedges certify the corresponding resource strength.","tokens_in":19947,"tokens_out":11414,"duration_ms":102283,"significance":"If the central bound Eq. (43) were true, the paper would provide a clean and experimentally useful diagnostic: a single measured (F,D) point would yield a visibility certificate and separate Bell-nonlocal from entangled-but-local isotropic resources using only fidelity statistics. The derivation of the closed forms for F (Eq. 35) and the exact covariance expression for D (Eqs. 39-41) is explicit, and the statement of the fourth-moment contraction (Eq. 34) is a useful resource. However, the whole witness-wedge construction rests on Eq. (43), and that inequality is false. The claimed certification theorems therefore do not hold, and the central message of the paper is invalid as stated.","major_comments":[{"comment":"The load-bearing inequality D <= s_d(Fmax(p)-F) is false. Take d=3 and the constant wiring X_alpha = X = diag(1, e^{i pi/3}, e^{-i pi/3}) for every alpha, which is admissible because V_alpha = X U_alpha is unitary. Then f(phi)=p |<phi|X|phi>|^2+(1-p)/3. Using the paper's own Eq. (30) and Eq. (34), E|<phi|X|phi>|^2 = 7/12 and E|<phi|X|phi>|^4 = 11/30. Hence F = p(7/12)+(1-p)/3, Fmax = p+(1-p)/3, and D = p sqrt(11/30-(7/12)^2) = p sqrt(19/720). Therefore D/(Fmax-F) = (12/5) sqrt(19/720) = sqrt(19)/(5 sqrt(5)) ≈ 0.3899 > 1/3 = s_3. Eq. (43) is thus violated for every p>0. The one-line justification 'By convexity (linear interpolation in the trace data entering Eq. (35))' is not a proof, and the counterexample shows it cannot be repaired by a different argument along those lines: D is a square-root of fourth-moment covariance data, not an affine function of the second-moment trace budget.","section":"Sec. IV.B, Eq. (43)"},{"comment":"Because Theorems 1 and 2 are derived by substituting p_BV and p_c into the false inequality Eq. (43), the certification conclusions are invalid. Concretely, with the d=3 wiring above and p=0.65, one has (F,D) = (0.4958, 0.1056). Since p_BV(3) ≈ 0.696, Fmax_BV ≈ 0.7973 and the W_local boundary at this F allows D <= (0.7973-0.4958)/3 ≈ 0.1005. The point therefore lies outside W_local even though p < p_BV, and it lies inside the entangled-but-local window p_c < p < p_BV. This directly falsifies the claimed 'any (F,D) outside W_local certifies CGLMP violation'. The same failure propagates to the visibility-estimation narrative and to the two-door interpretation in Sec. V.B. Eq. (42), the asserted extremal evaluation, is likewise presented without an optimization proof; the counterexample shows that the resulting bound is not a universal property of all wirings.","section":"Sec. V.B, Theorems 1 and 2"}],"minor_comments":[{"comment":"The fourth-moment formula (34) is quoted as 'which can be directly used [18]' and relies on a companion preprint. Since this formula is essential for D, it would be helpful to include a full derivation in an appendix rather than citing the companion work for the explicit result.","section":"Eq. (34) and Appendix A"},{"comment":"The text contains several typos and stylistic errors: 'Equavalently' in Theorem 2, 'casted' for 'cast', and the abstract's 'to analyzed-dimensional quantum teleportation' should be 'to analyze d-dimensional quantum teleportation'.","section":"Throughout"},{"comment":"The numerical CGLMP thresholds p_BV(d) are stated without derivation or a precise citation beyond the original CGLMP paper. A short note on how Q_d is obtained would improve reproducibility.","section":"Sec. V.A"}],"recommendation":"reject","confidential_remarks":"The decisive problem is internal and load-bearing: Eq. (43) is false, and the counterexample in the report is a valid wiring within the paper's own definitions. Since Theorems 1 and 2 and the entire witness-wedge certification are built on this inequality, the main claims of the manuscript fail as stated. This is not a matter of presentation or of disagreement with consensus; it is a concrete mathematical error within the paper's framework."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe main result is invalid. The load-bearing inequality Eq. (43) is false, and the counterexample is simple: for d=3 take X_alpha = diag(1, e^{i pi/3}, e^{-i pi/3}) for every outcome alpha. Using the paper's own Eq. (34), the second and fourth moments give F = p(7/12) + (1-p)/3 and D = p sqrt(19/720). Since Fmax(p) = p + (1-p)/3, the ratio D/(Fmax - F) = (12/5) sqrt(19/720) ≈ 0.390, which exceeds s_3 = 1/3. So Eq. (43) fails for every p > 0. This is an ordinary valid wiring, not an exotic edge case. Consequently Theorems 1 and 2, and the entire witness-wedge certification, do not hold as stated.\n\nWhat is good: the Schur-Weyl/Weingarten machinery is competently executed. I checked the closed forms for F (Eq. 35) and D (Eqs. 39-41) against known qubit cases and they match. The geometric idea—two parallel witnesses separating teleportation advantage from Bell nonlocality—is appealing and would be useful if a correct slope bound existed. The authors are honest about leaning on companion preprints for the core reduction formulas; that is not a problem per se.\n\nThe reader's conditional acceptance is too generous. The 'by convexity' step is not just missing a proof; it is wrong. The bound is not a minor gap; it is the engine of the paper. A minor issue: the numerical CGLMP thresholds for d>2 are quoted without derivation, but that is acceptable.\n\nWho should read this: people working on (F,D) diagnostics should wait for a corrected version. The technique section could be useful to readers interested in Haar-moment evaluations.\n\nRecommendation: not acceptable in current form. I would still send it to a serious referee—the error is instructive and the framework is worth fixing—but the authors need to replace Eq. (43) with a true bound or clearly restrict the wiring class.","headline":"The central witness-wedge result is invalid: Eq. (43) is false for d=3, so the main certification theorems do not hold.","tokens_in":20413,"tokens_out":8533,"would_cite":false,"duration_ms":79772,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In any dimension d, measuring a teleportation protocol's average fidelity and its spread across inputs certifies whether the shared resource beats classical teleportation or violates a Bell inequality.","keywords":["quantum teleportation","fidelity deviation","Schur-Weyl duality","Haar averages","CGLMP inequality","Bell nonlocality","isotropic channels","witness wedges"],"falsifier":"Take d=3, set the isotropic visibility to the CGLMP threshold p_BV ≈ 0.696, and numerically optimize over unitary correction unitaries {X_alpha} to see whether any (F,D) pair lands above the line D = (1/3)(F_max(p_BV) − F); even one such pair would disprove Eq. (43) and destroy both wedges.","tokens_in":19335,"feed_emoji":"⚛️","tokens_out":6229,"duration_ms":65270,"temperature":0.7,"pith_summary":"This paper argues that two numbers from a quantum teleportation experiment — the average fidelity F across all input states and the fidelity deviation D, its spread — together form a calibrated diagnostic chart for the strength of the shared entangled resource in any dimension d. For a resource that is a maximally entangled state mixed with white noise at visibility p, the paper derives a universal trade-off: D can never exceed s_d (F_max(p) − F), where s_d = sqrt(2/[d(d+3)]) and F_max(p) is the best average fidelity at that visibility. This inequality converts any measured (F,D) into a lower bound on p, so a point that falls outside the wedge W_cl — F ≤ 2/(d+1), D ≤ s_d(2/(d+1) − F) — certifies that the protocol beats the classical teleportation benchmark, and a point outside the higher wedge W_local certifies that the resource violates the CGLMP Bell inequality. The two witness lines share the same slope but sit at different heights, so the strip between them is exactly the regime where a resource is entangled enough for teleportation advantage but too weak to be Bell nonlocal. The decisive point is that a large deviation D can certify nonlocality even when the average fidelity F is still inside the Bell-local strip.","feed_headline":"Fidelity spread exposes nonlocality that averages hide","feed_subtitle":"One (fidelity, deviation) reading separates teleportation advantage from real Bell nonlocality","key_machinery":"The load-bearing object is the universal deviation bound D ≤ s_d(F_max(p) − F), with s_d = sqrt(2/[d(d+3)]), derived by evaluating fourth-moment Haar averages over pure input states through Schur-Weyl duality and cycle decomposition of permutation traces (Eqs. 28–43). This bound is what maps measured (F,D) data into a visibility certificate and gives both witnesses their identical slope; without it, the wedges would not be straight lines.","core_discovery":"The central claim is that the (F,D) plane is a calibrated visibility map for isotropic teleportation resources. Using Schur-Weyl duality and permutation-symmetry calculus, the authors reduce the Haar averages defining F and D to finite trace invariants of the composed correction unitaries X_alpha = V_alpha U_alpha^dagger, yielding closed forms valid for all dimensions d. From these they obtain the tight dimension-dependent bound D ≤ s_d(F_max(p) − F). Substituting the two physical thresholds — the CGLMP Bell-visibility threshold p_BV and the separability threshold p_c — produces two parallel witness lines of slope −s_d, anchored respectively at F_max(p_BV) and 2/(d+1). Any measured (F,D) out","pith_inferences":["The wedge logic does not depend on the specific Bell inequality chosen: any local-realist threshold on visibility p* would produce a parallel wedge, so the same chart could be repurposed for steering or one-sided device-independent certification by substituting the relevant p*.","Because D is the empirical variance of single-shot fidelities, the witness is immediately testable with Haar-random or unitary-2-design input sampling; finite-sample estimation of D sets the statistical resolution of the wedge boundary, which the paper does not quantify.","If the noise on the shared resource is not isotropic, the bound D ≤ s_d(F_max(p) − F) may fail; an interesting testable extension is to check whether dephasing or colored noise produces (F,D) points outside the isotropic wedges without actually possessing the certified resource strength."],"forward_implications":["Outside the classical wedge W_cl, (F,D) certifies that the resource visibility p exceeds 1/(d+1), i.e., the protocol outperforms the best measure-and-prepare scheme.","Outside the Bell wedge W_local, (F,D) certifies p > p_BV and therefore a CGLMP inequality violation, even when F alone is below the Bell-local threshold.","The parallel lines with equal slope but different intercepts make the gap between entangled-but-local and genuinely nonlocal resources quantitatively visible as a vertical strip in the (F,D) plane.","The closed-form expressions and bound hold for arbitrary Hilbert-space dimension d, so the same chart applies to qudit teleportation; for d=2 it reduces to the previously known qubit relation D ≤ (F_max − F)/√5.","Since F depends only on the diagonal trace budget of the X_alpha while D depends on pairwise invariants, protocols indistinguishable by F alone can be separated by D at fixed F."],"fun_headline_variants":["(F,D) plane: two parallel lines split teleportation and nonlocality","Fidelity deviation map reveals entangled-but-local gap","Same slope, different intercepts: (F,D) plane separates quantum resources","One (F,D) reading separates teleportation advantage from Bell violation","Deviation vs fidelity: a calibrated diagnostic for nonlocality"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole certification argument hinges on the unproved universal bound D ≤ s_d(F_max(p) − F) holding for every set of correction unitaries in every dimension; the paper justifies it by convexity but supplies neither the interpolation argument nor the extremal optimization that fixes the slope.","fun_headline_variants_meta":{"raw":{"variants":["(F,D) plane: two parallel lines split teleportation and nonlocality","Fidelity deviation map reveals entangled-but-local gap","Same slope, different intercepts: (F,D) plane separates quantum resources","One (F,D) reading separates teleportation advantage from Bell violation","Deviation vs fidelity: a calibrated diagnostic for nonlocality"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000832,"raw_usage":{"total_tokens":3491,"prompt_tokens":788,"completion_tokens":2703,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":532,"completion_tokens_details":{"reasoning_tokens":2611}},"tokens_in":532,"tokens_out":2703,"duration_ms":18876,"temperature":1.0,"reasoning_tokens":2611,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T20:05:40.235194+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take d=3, set the isotropic visibility to the CGLMP threshold p_BV ≈ 0.696, and numerically optimize over unitary correction unitaries {X_alpha} to see whether any (F,D) pair lands above the line D = (1/3)(F_max(p_BV) − F); even one such pair would disprove Eq. (43) and destroy both wedges.","supporting_citations":[],"review_version":1}