{"id":"feb25cc1-7f0e-4f73-86f6-5bdce9e82064","arxiv_id":"2607.26122","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In h→τ^-τ^+ Z decays, the spin state is genuinely qubit-qubit-qutrit entangled almost everywhere, violates Bell inequalities throughout, and carries up to 1.95 bits of non-local magic.","lead":"This paper maps out every major quantum correlation in the decay of a Higgs boson into two tau leptons and a Z boson, a system with two qubits and one qutrit. It shows the state is genuinely three-way entangled, violates Bell inequalities everywhere, and carries up to about two bits of 'magic' — with the highest decay rate falling in the most quantum region.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Non-local magic values hinge on an unverified global minimum: local minimization of M_nl gives only an upper bound, so the 'one bit' and log2(27/7) peak could be overestimates.","rationale":"The paper's central Bell and entanglement claims are well supported by analytic limits and closed-form expressions; the Bell violation everywhere is robust because the threshold value is analytic and far above the LHV bound, and the semi-analytic reductions reduce the numerical dimension substantially. The main soft spot is the non-local magic minimization. Unlike Bell maximization, where a missed optimum would only increase the violation (and thus cannot falsify 'violated everywhere'), the minimization in Eq. (5.12) means every local-search result is an upper bound on the true M_nl. The advertised quantitative results—'almost exactly one bit' at the endpoint and the peak log2(27/7)—depend on the claim that the numerical minimum equals the analytic candidate. The manuscript itself flags 'repeated local searches' and 'numerical artefacts', and it does not provide code or a global-convergence argument. This does not undermine the qualitative statement that the state has non-local magic, nor the entanglement and Bell results, but it makes the precise magic numbers conditional. I therefore keep the reader's CONDITIONAL verdict; the concern is essentially the same one identified by the reader, though sharpened to the asymmetry between maximization and minimization in the two numerical optimizations.","tokens_in":906,"tokens_out":870,"duration_ms":136246,"concrete_test":"At the upper endpoint m_AB = m_H − m_V and at the collinear point κ^2 = 1/2, recompute M_nl of Eq. (5.12) with a certified global optimization method on the 14-parameter local-unitary manifold (e.g. branch-and-bound or differential evolution with many restarts plus local refinement, or by solving the stationarity equations ∇M_2(U_A,U_B,U_V)=0 and enumerating stationary points). If any admissible local unitary gives M_2 strictly below Eq. (5.15) (endpoint) or Eq. (5.16) (collinear), the 'one bit' and log2(27/7) values are overestimates; if no lower value is found, the numerical saturation claim is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing numerical assumption is not in the Bell optimization (where a missed higher setting would only strengthen violation, and the threshold value is analytic), but in the non-local magic claims. M_nl is defined in Eq. (5.12) as a minimum over a 14-parameter local-unitary manifold. Any value produced by a local search is an upper bound on the true minimum. The paper asserts that near the upper endpoint the embedded GHZ state has M_nl exactly as in Eq. (5.15), and that along the collinear family M_2(κ)=M_nl(κ) in Eq. (5.16), saying only 'we have verified numerically' / 'we find numerically' (Sec. 5.3). If a better local-unitary basis exists and lowers M_2 below those formulas, the advertised 'almost exactly one bit of non-local magic' and the peak 'log2(27/7) ≈ 1.95' are overestimates. Since M_2 ≥ 0, the qualitative presence of magic is robust, but the quantitative headline values are not. The paper is honest that the monotonicity of SRE for unequal dimensions is unproven and that the residual optimizations use repeated local searches, but no certificate or code is supplied for the crucial saturation claims.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the pure three-partite spin state produced in H -> f fbar V, specializing to h -> tau^- tau^+ Z, and maps its quantum correlations across phase space. Working at tree level with a systematic expansion in the fermion mass, it obtains compact analytic formulas for bipartite and genuine 2⊗2⊗3 entanglement, for the maximized 4×4×2 Bell operators, and for a new extension of the stabilizer Rényi entropy and non-local magic to unequal local dimensions. The central claims are that the state is genuinely 2⊗2⊗3 entangled over almost all phase space, that all three tight Bell inequalities are violated everywhere (reaching near the quantum bound at the upper endpoint), and that the non-local magic is approximately one bit at the endpoint and peaks at log2(27/7) ≈ 1.95 in collinear regions, with the differential decay rate concentrated in the most nonclassical region.","tokens_in":47527,"tokens_out":17411,"duration_ms":148559,"significance":"If the results hold, the paper makes two genuinely useful contributions: a semi-analytic treatment of tight 4×4×2 Bell inequalities for 2⊗2⊗d systems, and a systematic definition of non-local magic for unequal local dimensions, applied to a physically realistic and analytically tractable collider process. The derivation is self-contained and parameter-free: the spin state is computed from first principles, and the compact formulas are checked against the numerical maps in kinematic limits. The identification of the pointer-state structure, the CP-imposed mirror symmetries, and the monogamy-like trade-offs are valuable. The main weakness is numerical certification: the non-local magic is defined through a 14-parameter minimization and the headline values rely on repeated local searches, so the reported numbers are not proven global extrema. This affects the quantitative claims but not the robust qualitative conclusions such as Bell violation and the presence of magic.","major_comments":[{"comment":"The non-local magic is defined as a minimum over a 14-parameter local-unitary manifold and computed with repeated local searches. For a minimization, a local search yields an upper bound on M_nl, not a certified value; a missed local-unitary basis would lower M_nl below the reported values. The saturation claims underlying the two headline results -- the endpoint value 'almost exactly one bit' (Eq. (5.15): 'we have verified numerically') and M_nl(kappa)=M_2(kappa) along the collinear family with peak log2(27/7) (Eqs. (5.16)-(5.17): 'we find numerically') -- are not backed by a certificate, an analytic proof, or released code. Since these values are central to the abstract and conclusions, this is load-bearing. Please provide an analytic proof of the saturation claims, a certified global optimization, or make the numerical code, seeds, and convergence criteria available; otherwise the qua","section":"Sec. 5.2 / Eq. (5.12)"},{"comment":"The Bell-inequality maxima are obtained after analytic reduction to residual searches over 8 (B_442) and 9 (B'_424) parameters, which are then optimized by local searches. This does not invalidate the core violation claim: the reported values exceed the LHV bound 4 everywhere, and the threshold value 4*sqrt(1+C_AB^2) ≈ 5.6 is analytic, so even an underestimated maximum would leave the violation intact. However, the quantitative statements 'within a few per cent of the quantum bound' and the 0.01-level differences shown in Fig. 10 depend on convergence of these local searches. The manuscript should specify the number of restarts, stopping criteria, and ideally provide the code or data for the residual optimizations so that the maps are reproducible.","section":"Sec. 4.1/4.2, Eqs. (4.8), (4.19)"}],"minor_comments":[{"comment":"The caption admits 'numerical artefacts' in the pairwise Bell differences. With a claimed semi-analytic optimization, such artefacts should be eliminated or explained; otherwise they undermine confidence in the 0.01-level quantitative comparisons.","section":"Fig. 10 caption"},{"comment":"The paper correctly notes that monotonicity of the stabilizer Rényi entropy has not been established for mixed local dimensions. Since the term 'non-local magic' carries resource-theoretic connotations, please state explicitly which properties of M_nl are proven and that monotonicity is not needed for the present physical conclusions.","section":"Sec. 5.1"},{"comment":"The phrase 'most nonclassical region' is ambiguous: the Bell violation is maximal near the endpoint, whereas M_nl peaks in the collinear hot-spot region, which is rate-suppressed. Please specify which resource is meant, or qualify the statement.","section":"Abstract / Sec. 6"},{"comment":"Reference [113] (HL-LHC technical design report) lacks volume/publisher/DOI information; please provide the full citation.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is solid in its analytic core: the spin-state derivation, the kinematic-limit formulas, and the closed-form Bell and SRE expressions are careful and check out. The only blocking issue is the certification of the global minima/maxima for the numerical optimizations, particularly for M_nl where a local search gives only an upper bound. If the authors can supply analytic proofs or reproducible global-optimization code, or suitably downgrade the quantitative claims, I would support publication. The Bell-violation claim itself is robust and should not be held hostage to the residual search concerns."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things. First, this paper is a real step forward for quantum-information-at-colliders: it gives the first complete treatment of a pure 2x2x3 final state from a collider process, and it produces two genuinely new tools — semi-analytic optimization of tight 4x4x2 Bell inequalities for 2x2xd systems, and an extension of stabilizer Rényi entropy and non-local magic to unequal local dimensions. Second, the analytic core is strong, but the quantitative magic claims rest on unverified numerical minimization, and that is the part to scrutinize.\n\nThe paper does much well. The spin state is derived from first principles at tree level, no fitting parameters, and the closed-form limits (threshold, collinear, endpoint) match the numerical maps. The entanglement analysis, including the Miyake classification and the 223-tangle, is careful and the monogamy trade-off is nicely quantified. The Bell-inequality optimization is a genuine generalization of the three-qubit result, and the reduction to a low-dimensional search is elegant. The authors are also honest: they state that SRE monotonicity for unequal dimensions is unproven, that residual optimizations use repeated local searches, and that experimental prospects are estimates, not a sensitivity study. That honesty earns credit.\n\nThe soft spot is exactly what the stress-test note flags. M_nl is defined as a minimum over a 14-parameter local-unitary manifold, and any local search gives only an upper bound. The paper claims saturation of the bound near the endpoint and along the collinear family with 'we have verified numerically', but no code or certificate is provided. If a better local unitary lowers M_2 below those formulas, the advertised 'one bit' and the log2(27/7) peak are overestimates. The qualitative presence of non-local magic is robust — M_nl is non-negative and the analytic lower bounds from the embedded GHZ and collinear states are solid — but the headline quantitative values are not certified. The Bell violations are on much firmer ground, since a missed setting would only strengthen the violation and the minimum analytic value is far above the LHV bound.\n\nThe central physics claims — genuine tripartite entanglement over almost all phase space, Bell violation everywhere, and non-local magic that peaks where the qutrit is most active — hold up. The weakest load-bearing number is the magic peak. That should be addressed, ideally by supplying code or a more rigorous global-optimization argument.\n\nWho is this for? Anyone working on quantum information in particle decays, and anyone interested in multipartite Bell inequalities or magic in asymmetric dimensions. It deserves serious peer review. I'd send it to a competent referee and ask them to focus on the numerical optimization for M_nl. I'd also bring it to the reading group — the techniques will be reused.","headline":"Genuinely new analytic results on 2x2x3 Bell inequalities and asymmetric magic in a collider decay; the numerical optimization for non-local magic is the one quantitative claim that needs support before I'd trust the headline numbers.","tokens_in":48098,"tokens_out":1606,"would_cite":true,"duration_ms":18726,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The pure qubit–qubit–qutrit spin state produced in the three-body Higgs decay h→τ−τ+Z is genuinely tripartite-entangled over almost all of phase space, violates the tight 4×4×2 Bell inequalities everywhere, and carries non-local magic peaki","keywords":["Higgs decay","tripartite entanglement","qubit-qubit-qutrit","Bell inequality","non-local magic","stabiliser Rényi entropy","hyperdeterminant","tau-Z spin correlations"],"falsifier":"Reconstruct the spin density matrix of h→τ−τ+Z in an event sample and evaluate the three optimised Bell observables B_442, B'_424, B'_244; finding any phase-space point with a value below 4 would refute the universality claim, and finding a value above the predicted ≈7.9 near the endpoint would signal a defect in the amplitude. Alternatively, run a certified global optimisation of the qutrit projectors at a fixed phase-space point and check whether the value exceeds the reported maximum.","tokens_in":47132,"feed_emoji":"⚛️","tokens_out":7982,"duration_ms":70930,"temperature":0.7,"pith_summary":"Three-body Higgs decays into a tau pair and a Z boson produce a pure spin state with two qubits and one qutrit. The paper shows that this state is nonclassical everywhere: it is genuinely 2⊗2⊗3 entangled over almost the whole phase space, it violates the tight 4×4×2 Bell inequalities at every point (from ≈5.6 at threshold to ≈7.9 near the di-tau endpoint, against an LHV bound of 4 and quantum bound 8), and it carries non-local magic that peaks at log2(27/7)≈1.95 exactly where the tripartite entanglement is largest. The analysis is made possible by expanding the tree-level amplitude in the small fermion mass, which exposes a pointer-state structure and yields compact closed formulas for every observable. Two results are new: a semi-analytic optimisation of the tight 4×4×2 Bell inequalities for any 2⊗2⊗d system, and an extension of the stabiliser Rényi entropy and non-local magic to systems with unequal local dimensions. If the claims hold, h→τ−τ+Z becomes a practical laboratory for genuine three-party quantum correlations, and the differential decay rate is shown to concentrate precisely in the most nonclassical region.","feed_headline":"Higgs decay spins violate Bell bounds over the entire phase space","feed_subtitle":"A pure qubit-qubit-qutrit spin state stays genuinely tripartite-entangled almost everywhere; its non-local magic peaks at log2(27/7).","key_machinery":"The argument is carried by four pieces. (1) The ε-expansion of the tree-level amplitude around the massless-fermion limit, where the zeroth-order state is a superposition of two pointer states tagged by fermion helicities and the sub-leading corrections occupy the same-helicity subspace; this expansion makes every observable a compact formula in each corner of phase space. (2) The hyperdeterminant M=m1m4−m2m3 of the amplitude matrix together with the local ranks, which separates the SLOCC classes and detects genuine 2⊗2⊗3 entanglement; the leading minors are O(ε), and their zeros define nodal curves governed by the vector-to-axial coupling ratio. (3) For Bell inequalities: a reduction of the","core_discovery":"On the paper's own terms, the central discovery is that the spin state produced in H→f fbar V, specialised to h→τ−τ+Z, is a pure state whose nonclassicality is controlled by a tiny mass parameter ε. At leading order the state is a GHZ-like superposition of two branches in which the fermion helicities are correlated with two vector-boson pointer states; the overlap of the pointer states governs where entanglement lives. The genuinely tripartite 2⊗2⊗3 character is switched on at O(ε), detected by the minors and hyperdeterminant of the 3×4 amplitude matrix, and the state belongs to the generic class over essentially the whole plane except along nodal curves and at the di-tau threshold. The tigh","pith_inferences":["If the predicted non-local magic is confirmed, the 'one bit from dimensional mismatch' phenomenon suggests a general resource: embedding a two-qubit stabiliser GHZ state in a higher-dimensional local space creates magic without changing the entanglement spectrum, which could be exploited in other settings where one party has a natural three-level structure.","The numerical global-extrema assumption is the soft spot; a certified global optimisation or analytic closed form for the qutrit projectors would be a clean follow-up. Even if the quoted maxima shift by a few percent, the minimum analytic bound ≈5.6 keeps the Bell violation robust.","The location of the minor nodal curves is set by c_V/c_A; since radiative corrections shift it only by a few percent, the angular pattern of genuine tripartite entanglement is effectively a tree-level probe of the vector-to-axial coupling ratio, potentially usable as a new physics discriminant.","The authors' rough yield estimates (of order 10^3–10^4 leptonically tagged events at a high-luminosity hadron collider) suggest that a first measurement of three-party correlations in this channel is within reach; a dedicated detector-level study with tau spin-analysing power would be the natural test."],"forward_implications":["Because the Bell violation is everywhere at least ≈5.6, with the endpoint within a few percent of the quantum bound 8, even a modest, non-exhaustive scan of settings certifies tripartite nonlocality in this decay.","The differential decay rate is concentrated in the near-endpoint region, where the state approaches the generalised GHZ form and the non-local magic sits on its plateau; most produced events are therefore the most nonclassical ones.","The maxima of genuine tripartite entanglement and non-local magic occur at the same point (collinear kinematics, m_ττ≈12 GeV), so one observable can be used to locate the other.","The monogamy-like trade-off between the fermion-pair concurrence and the fermion–boson entanglement means that an apparent suppression of di-tau entanglement is not a loss of quantumness but a transfer to fermion–boson pairs.","The new semi-analytic Bell optimisation and the unequal-dimension stabiliser Rényi entropy apply to any 2⊗2⊗d state and to arbitrary collections of prime-dimensional subsystems, not just this decay."],"fun_headline_variants":["Tau-Z spin state stays entangled across full phase space","Higgs decay state violates Bell inequalities everywhere","Qubit-qubit-qutrit magic peaks in Higgs di-tau endpoint","Collinear Higgs spin state carries near-maximal magic","Bell bound violated over entire Higgs decay phase space"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The reported phase-space maxima of the Bell observables and of non-local magic rely on numerical optimisation (an 8-parameter search for the qutrit measurements and a 14-parameter search over local unitaries) rather than a certified global method; if the true extrema differ, the quantitative values could shift, although the Bell violation itself is robust because the minimum analytic value ≈5.6 is far above the bound 4.","fun_headline_variants_meta":{"raw":{"variants":["Tau-Z spin state stays entangled across full phase space","Higgs decay state violates Bell inequalities everywhere","Qubit-qubit-qutrit magic peaks in Higgs di-tau endpoint","Collinear Higgs spin state carries near-maximal magic","Bell bound violated over entire Higgs decay phase space"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000265,"raw_usage":{"total_tokens":1540,"prompt_tokens":938,"completion_tokens":602,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":682,"completion_tokens_details":{"reasoning_tokens":522}},"tokens_in":682,"tokens_out":602,"duration_ms":5088,"temperature":1.0,"reasoning_tokens":522,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T00:41:52.079670+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Reconstruct the spin density matrix of h→τ−τ+Z in an event sample and evaluate the three optimised Bell observables B_442, B'_424, B'_244; finding any phase-space point with a value below 4 would refute the universality claim, and finding a value above the predicted ≈7.9 near the endpoint would signal a defect in the amplitude. Alternatively, run a certified global optimisation of the qutrit projectors at a fixed phase-space point and check whether the value exceeds the reported maximum.","supporting_citations":[],"review_version":1}