{"id":"b984d016-0e22-4b45-b3fc-d66b8653916f","arxiv_id":"2607.25029","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":3,"one_line_summary":"CP-odd SMEFT top interactions appear as ΔB and antisymmetric C_A in the tt̄ production density matrix; direct markers beat most QI measures for CP sensitivity at LHC and FCC-ee.","lead":"CP violation in top-quark pair production leaves two clear fingerprints in the pair’s spin quantum state: unequal top/antitop polarisations and an antisymmetric spin-correlation matrix. The paper maps those fingerprints across collider channels and shows which quantum observables can actually see them with present and future data.","discovery_kind":"extension","skeptic_critique":{"model":"moonshotai/kimi-k3","headline":"The structural CP result (ΔB, C_A) is sound; the load-bearing weakness is the quantitative claim that a_k beats existing Im(C_tG) bounds, which rests on an uncorrelated χ² over CMS bins and LO EFT shapes grafted onto NNLO SM central values.","rationale":"My read converges with the reader's: the production-level framework (CP-odd structures ΔB and C_A, channel taxonomy, analytic Fano–Bloch coefficients cross-checked against Refs. [23, 42, 59, 122]) is the solid core, and the sensitivity numbers are explicitly framed by the authors as illustrations under LO, one-operator, uncorrelated-error, fixed-5% assumptions. The reader's weakest_assumption already identifies exactly the locus I would attack — the uncorrelated χ² and truncation choices underlying the a_k superiority claim — so I agree rather than adding a new concern. I considered two alternative attacks and set them aside: (1) that C_A could be faked by absorptive phases — it cannot, since the CP relations in App. B are symmetry statements about the density matrix for CP-conjugate configurations, not about the absence of phases; the paper's caveats (unpolarized initial states, common spin basis, CP-related kinematics) are the correct and sufficient ones; (2) that retaining O(Λ⁻⁴) while dropping dim-8 is inconsistent — it is, formally, but the authors state this plainly and use it only to keep ρ positive, so it is a flagged limitation rather than a hidden flaw. Neither rises to load-bearing. The one genuinely load-bearing point is whether the single quantitative headline — a_k beating representative Im(C_tG) intervals — survives inclusion of the published CMS covariance and a truncation check; the proposed test settles it directly. This does not overturn the paper's conditional value: the framework and the ranking of direct CP markers over global QI observables stand regardless. Hence UNCHANGED: CONDITIONAL remains the right verdict, with the condition being precisely that the sensitivity numbers be treated as indicative until the correlated fit is done.","tokens_in":64484,"tokens_out":3448,"duration_ms":107526,"concrete_test":"Rebuild the a_k fit of §6.1.2 using the full CMS covariance matrix (statistical + systematic, all 12 m_tt × |cosθ| bins) from the HEPData record of Phys. Rev. D110 (2024) 112016, and repeat the χ² scan of Im(C_tG) (a) with the quadratic terms as in the paper and (b) truncated at O(Λ⁻²). If the correlated-fit 2σ interval on Im(C_tG) widens past the Table 3 reference interval, or if the linear/quadratic truncation moves the interval by a comparable amount, the \"a_k improves on existing bounds\" claim should be downgraded to illustrative; if both checks are stable, the claim stands.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The strongest claim has two parts. The structural part — CP violation in production is encoded in ΔB and C_A once spins are expressed in a common basis and CP-related kinematics are compared — is essentially a symmetry statement (App. B: CP invariance ⟺ B=B̄, C=C^T for unpolarized, CP-self-conjugate initial configurations). It is basis-aware, correctly hedged, and does not depend on the EFT truncation. I find no soft spot there; even absorptive phases cannot fake C_A for a genuinely CP-invariant theory, and the paper is explicit about the unpolarized-initial-state and common-basis qualifications.\n\nThe load-bearing part is therefore the phenomenological sentence: \"the component a_k of the antisymmetric correlation vector currently gives the strongest individual constraint on Im(C_tG) from CMS data, improving on representative existing intervals.\" This rests on three stacked approximations in §6.1: (i) the χ² treats the CMS [15] bin-to-bin uncertainties as uncorrelated, whereas spin-correlation coefficients in neighboring m_tt–|cosθ| bins share unfolding, luminosity, and modeling systematics and are published with covariance information; (ii) the SM central values are taken from MiNNLO+P8 (NNLO) while the entire C_tG dependence — the shape that the fit actually uses — is LO, including quadratic O(Λ⁻⁴) terms retained for positivity but formally incomplete against dim-8; (iii) the \"improvement\" is judged against Table 3 intervals obtained with different normalizations (e.g., Im(C_tG)/y_t g_s from [12]), different observables, and different datasets. If correlations widen the a_k interval by more than the margin over the reference interval, or if the linear-only truncation shifts the best-fit region, the \"strongest individual constraint, improving on existing intervals\" statement — the paper's only claim of new quantitative reach — does not survive, even though the framework claim does.","agreement_with_reader":"agree"},"referee_report":{"model":"moonshotai/kimi-k3","summary":"The paper develops the production-side framework for diagnosing CP violation in top-quark interactions through the two-qubit spin density matrix of the tt̄ pair. Working in a common spin basis, the authors show (Sec. 4.1, App. B) that for unpolarised, CP-self-conjugate production configurations, CP invariance is equivalent to B = B̄ and C = C^T, so that CP violation is encoded in exactly two structures: the polarisation difference ΔB and the antisymmetric correlation matrix C_A (equivalently the vector a). They derive analytic production density matrices via the Bouchiat–Michel formalism for five benchmark channels (scalar decay, e⁺e⁻, γγ, qq̄, gg), retaining linear and quadratic dimension-six SMEFT contributions with complex dipole coefficients, and study the response of discord, concurrence, magic, and a CP-sensitive trace distance. A phenomenological section (Sec. 6) compares against CMS spin-correlation measurements and FCC-ee projections, concluding that the component a_k currently provides the strongest individual constraint on Im(C_tG), and that a_k also leads the projected sensitivity to Im(C_tB) at a 365 GeV lepton collider. A companion paper will treat tomographic reconstruction.","tokens_in":64854,"tokens_out":3541,"duration_ms":139649,"significance":"If the results hold, this is a useful and well-organised contribution to the growing quantum-information program in top physics. Particular strengths: (i) the CP classification is derived from the action of CP on the density matrix itself (App. B) with explicit statements of its basis dependence and its restriction to unpolarised initial states and CP-related kinematics — it is not fitted or assumed; (ii) complete analytic Fano–Bloch coefficients are given for five production channels (App. C), including CP-odd structures at quadratic order, cross-checked against Refs. [23, 42, 59, 122] and verified numerically with the code of Ref. [104]; (iii) the scalar-decay example (Sec. 5.1) cleanly demonstrates that maximal entanglement and maximal CP violation are logically independent, a genuinely instructive result; (iv) the EFT-truncation limitations (quadratic dim-6 retained for positivity without dim-8; LO QCD in the LHC fit; uncorrelated uncertainties) are stated openly in Secs. 6–7 rather than hidden. The main quantitative claim — that a_k beats existing Im(C_tG) intervals — is falsifiable and, if it survives a proper covariance treatment, would be of direct interest to the LHC EFT社区","major_comments":[{"comment":"The headline phenomenological claim (Sec. 6.1.2, Fig. 21; repeated in the conclusions, Sec. 7) that a_k 'improves on the representative existing interval' for Im(C_tG) rests on a χ² over the CMS [15] bins in which all bin-to-bin uncertainties are treated as uncorrelated (stated explicitly at the end of Sec. 6.1). Spin-correlation coefficients in neighbouring m_tt–|cos θ| bins share unfolding, luminosity, and modelling systematics, and the CMS analysis provides covariance information via HEPData [106]. Neglecting these correlations generically overstates the constraining power and can also shift the preferred region. Since the 'strongest individual constraint' sentence is load-bearing for the paper's phenomenological punchline, the fit should be redone (or at least bracketed) with the published covariance matrix, or the claim should be downgraded to an illustration under diagonal uncertai","section":"§6.1.2, Fig. 21"},{"comment":"The comparison underlying the improvement claim mixes normalisations without showing the conversion. Table 3 quotes Im(C_tG)/(y_t g_s) ∈ [−0.33, 0.20] from [12] and Re(C_tG)/g_s from [76], while the fits in Figs. 21–22 bound Im(C_tG) directly in the dim6top_LO convention of [62]. The manuscript never states the explicit mapping used to translate the Table 3 intervals into the (Re C_tG, Im C_tG) plane of Figs. 21–22. A one-line conversion (with the values of y_t, g_s used) is needed in the caption of Fig. 21 or in Sec. 6.1.2; without it, the statement that the a_k contour is 'smaller compared to the experimental bound' is not verifiable from the manuscript.","section":"§6.1.2, Fig. 21 and Table 3"},{"comment":"In the FCC-ee projections, the CP-sensitive observables ΔB_n and a_k lose all sensitivity to Im(C_tB) near Re(C_tB) ≃ −0.35 and ≃ −1.5 respectively. The text correctly traces this to a cancellation between the O(Λ⁻²) and O(Λ⁻⁴) terms in the numerator (footnote 6). However, the O(Λ⁻⁴) terms are retained only as dim-6-squared contributions, without the dim-8 interference terms of the same formal order — a truncation the authors themselves flag as incomplete (Sec. 5.2, Sec. 7). The blind spots are therefore artifacts of an inconsistent-order numerator, yet they visibly deform the allowed contours in Fig. 23 and propagate into the conclusion that a_k is 'the strongest projected sensitivity' in the range Re(C_tB) ∈ [−1,1]. The contours should either be recomputed at consistent O(Λ⁻²) (accepting possible non-positivity of ρ and restricting observables accordingly) or the regions near the zeros","section":"§6.2, Figs. 23–24, footnote 6"},{"comment":"The comparison with Refs. [23] and [59] reports that the spin-correlation coefficient C̃_rk 'differs by an overall sign' — and this same unresolved discrepancy is stated identically in App. C.3 (qq̄) and App. C.4 (gg). No origin is identified (convention vs. genuine disagreement). Since C_rk enters the symmetric correlation structure used in the Sec. 6 fits (via the CMS-basis conversion, Eq. 6.5, and the a_n component (C_rk − C_kr)/2), an unresolved sign discrepancy with two independent published computations is a correctness risk that should be settled before publication — e.g., by tracing it to a specific basis or ε-tensor convention, or by a numerical cross-check at a fixed phase-space point against one of the two references.","section":"App. C.3 (Eq. C.22) and App. C.4 (Eq. C.25)"}],"minor_comments":[{"comment":"In Eq. (C.22), the Λ⁻² term of C̃^A_kn contains the factor 'βeγgsmt...'; 'eγ' appears to be a typo for ˜γ (cf. the Λ⁻⁴ term of the same coefficient, which correctly carries β˜γ). Please check and correct.","section":"App. C.3, Eq. (C.22)"},{"comment":"The caption of Fig. 12 (and similarly Fig. 13) describes the upper panels as showing ∥a∥ and δM₂, but the figure layout places ∥a∥ and δM₂ in the top row and δD, δC in the bottom row; the wording 'upper panels ... and the lower panels' is ambiguous given the 2×2 arrangement. Please rephrase to 'top row / bottom row'.","section":"Figs. 12–13 captions"},{"comment":"The companion-paper reference [73] is a placeholder ('2607.XXXXX') and the fourth author's name is misspelled ('Vrynidou' for 'Vryonidou'). Please update at revision.","section":"Reference [73]"},{"comment":"Eq. (4.39): the mixed-state SRE₂ formula is used as a 'diagnostic of non-stabilizerness rather than a fully faithful magic monotone' — this caveat is welcome, but it would help the reader to state explicitly whether M₂ can be nonzero for stabilizer mixtures (false positives) or only fail as a monotone, since the phenomenological projections in Sec. 6 treat M₂ as a measurable discriminant.","section":"§4.2.4"},{"comment":"Table 3: the caption notes that different normalisations are employed, but does not flag that the two C_tG rows from [12] are quoted per (y_t g_s) while the [76] row is per g_s. A footnote giving the numerical conversion factor used elsewhere in the paper would prevent misreading.","section":"Table 3"},{"comment":"Sec. 6.1.1: the basis conversion in Eq. (6.5) introduces sgn(cos θ) factors on n̂ and r̂ to match the CMS convention. Since this redefinition is discontinuous at cos θ = 0 and the binning includes |cos θ| ∈ [0, 0.4], one sentence clarifying that the discontinuity lies inside a single bin and does not mix CP-even and CP-odd components would be useful.","section":"§6.1.1, Eq. (6.5)"}],"recommendation":"major_revision","confidential_remarks":"The structural part of the paper (Secs. 3–5, Apps. A–C) is careful and, in my reading, correct in its symmetry logic; the authors are commendably explicit about basis dependence and EFT-truncation limits. The revision is driven entirely by the phenomenology section: the 'strongest individual constraint on Im(C_tG)' sentence is the paper's quotable result, and as it stands it is produced by a diagonal χ² on data whose covariance is publicly available, compared against a Table 3 interval in a different normalisation, with an unresolved C_rk sign discrepancy in the underlying coefficients. All of these are fixable within the manuscript's scope — none requires new physics or new formalism — but they are load-bearing for the advertised quantitative conclusion, hence major rather than minor revision. The manuscript is a good fit for the journal's scope in collider phenomenology/EFT."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The load-bearing result here is structural and clean. For the unpolarised channels they consider, once spins sit in a common basis, CP violation in production is exactly the pair (ΔB, C_A). That follows from the CP action on the two-qubit density matrix (App. B) and does not depend on the EFT truncation. They then deliver analytic production matrices, with complex electroweak and QCD dipoles, across S→tt̄, e⁺e⁻, γγ, qq̄ and gg, organised so the CP-even/odd pieces are visible at a glance. That taxonomy, plus the explicit comparison of direct markers versus discord/concurrence/magic, is the real addition to the existing spin-correlation and collider-QI literature.\n\nThe analytic work looks careful. Bouchiat–Michel, spin-basis conversion, and cross-checks against prior SM/EFT results are all there; they flag the C_rk sign difference themselves. The channel-by-channel plots make the process dependence obvious: entanglement is blind in the scalar case, polarisations appear only in e⁺e⁻, and photonic/QCD CP-odd information sits in C_A. Magic and discord move under both CP-even and CP-odd deformations, so they are not null tests—exactly the point the paper wants to make.\n\nThe soft spot is the phenomenology in §6, and it is the one the stress-test flags. The claim that a_k currently gives the strongest individual constraint on Im(C_tG) and improves on Table 3 rests on an uncorrelated χ² over CMS bins, LO EFT shapes grafted onto NNLO SM centrals, one-operator scans, and a fixed 5% error for projected QI observables. They state the limitations, and they treat the numbers as illustrations, but that sentence is still the only claim of new quantitative reach. If the published covariances widen the interval or the LO shape is off, the “improves on existing” part does not hold; the framework claim does.\n\nThis is for people already working on top spin, SMEFT dipoles, or collider QI who need a unified production target list before the tomography companion. The math and citation pattern are solid; the free parameters are the usual projection knobs. I would send it to referees. Engage with the structural results and the observable ranking; treat the numerical bounds as indicative until covariances and NLO are in.","headline":"Solid production-level framework: CP-odd effects live cleanly in ΔB and C_A; the a_k “beats existing bounds” claim is the soft quantitative spot, not the structure.","tokens_in":63439,"tokens_out":590,"would_cite":true,"duration_ms":13878,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"CP violation in top-pair production is stored in two definite spin-density structures that current LHC data already constrain.","keywords":["top-quark pair production","CP violation","SMEFT","spin density matrix","Fano-Bloch decomposition","quantum information observables","chromoelectric dipole moment","collider phenomenology"],"falsifier":"A statistically consistent combination of the full set of measured Fano–Bloch coefficients (including a_k) that either tightens or fails to improve the present interval on Im(C_tG), or a future measurement of concurrence at the few-percent level that does not follow the projected exclusion contours.","tokens_in":63061,"feed_emoji":"⚛️","tokens_out":1021,"duration_ms":18251,"temperature":0.7,"pith_summary":"This paper shows that new sources of CP violation in top-quark interactions leave a clear fingerprint on the quantum spin state of a produced top-antitop pair. Working in a common spin basis, CP invariance forces the top and antitop polarisation vectors to be equal and the spin-correlation matrix to be symmetric. Any violation therefore appears as a polarisation difference ΔB and an antisymmetric correlation piece C_A. The authors derive the full production density matrix analytically for scalar decay, electron-positron annihilation, photon fusion, and the quark- and gluon-initiated channels at hadron colliders, mapping SMEFT operators onto these CP-odd structures. They then build direct probes (the norms of ΔB and of the antisymmetric vector a, plus the trace distance to the CP-transformed state) and compare them with quantum-information measures such as discord, concurrence and magic. Using existing CMS spin-correlation measurements and projections for the LHC and a future lepton collider, they find that the antisymmetric component a_k already supplies the strongest individual bound on the imaginary part of the chromoelectric dipole coefficient, while concurrence offers promising projected reach. The work supplies the production-level target whose experimental reconstruction is treated in a companion paper.","feed_headline":"Top-pair CP violation lives in two spin structures","feed_subtitle":"LHC data already bound the antisymmetric correlation; concurrence could tighten the reach further","key_machinery":"The Fano–Bloch decomposition of the two-qubit production density matrix in a common spin basis, which converts CP invariance into the elementary conditions B = B̄ and C = C^T and thereby isolates the two CP-odd markers ΔB and C_A.","core_discovery":"For the unpolarised production processes considered, after expressing the top and antitop spins in a common basis, CP violation in production is encoded exactly in two independent Fano–Bloch structures: the polarisation difference ΔB = (B − B̄)/2 and the antisymmetric part of the spin-correlation matrix C_A = (C − C^T)/2. Direct observables built from these structures, especially the component a_k, currently give the strongest individual constraint on Im(C_tG) from CMS data.","pith_inferences":["Once the companion tomography paper is available, production-side and decay-side CP-odd effects can be separated experimentally, turning the two-paper programme into a complete CP diagnostic for top pairs.","Because magic and concurrence respond differently to real versus imaginary dipole coefficients, a joint measurement of both could help discriminate CP-even from CP-odd new physics even when the direct markers are statistically limited.","The clean separation of CP-odd entries in the photon-fusion and gluon-fusion correlation matrices suggests that a high-energy photon collider would offer an especially transparent laboratory for top electric-dipole moments."],"forward_implications":["Direct CP markers built from ΔB and C_A become standard null tests for new top CP violation at the LHC and future colliders.","The component a_k of the antisymmetric correlation vector can already improve existing bounds on the imaginary chromoelectric dipole coefficient.","Projected concurrence measurements with few-percent precision would add competitive sensitivity to both CP-even and CP-odd dipole operators.","Electron-positron colliders gain an extra CP-odd handle (the polarisation difference ΔB) that is absent in leading-order hadronic production.","The same production-density-matrix framework extends immediately to tau-pair channels once the corresponding dipole operators are inserted."],"fun_headline_variants":["CP violation in top pairs sits in ΔB and antisymmetric C_A","Two Fano-Bloch structures encode top-pair CP violation","Polarisation difference and C_A capture production CP oddness","a_k observable leads CMS bounds on Im(C_tG) in ttbar","Common spin basis isolates exact CP-odd top-pair signatures"],"cache_read_input_tokens":49280,"weakest_assumption_plain":"The sensitivity claims rest on leading-order production matrices, one-operator scenarios, uncorrelated experimental errors, and a fixed 5 percent uncertainty assigned to projected quantum observables, while quadratic dimension-six terms are kept without the matching dimension-eight operators.","fun_headline_variants_meta":{"raw":{"variants":["CP violation in top pairs sits in ΔB and antisymmetric C_A","Two Fano-Bloch structures encode top-pair CP violation","Polarisation difference and C_A capture production CP oddness","a_k observable leads CMS bounds on Im(C_tG) in ttbar","Common spin basis isolates exact CP-odd top-pair signatures"]},"model":"grok-4.5","effort":"low","cost_usd":0.003447,"raw_usage":{"total_tokens":1137,"prompt_tokens":735,"num_sources_used":0,"completion_tokens":97,"cost_in_usd_ticks":34468000,"prompt_tokens_details":{"text_tokens":735,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":305,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":735,"tokens_out":97,"duration_ms":5690,"temperature":1.0,"reasoning_tokens":305,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-31T03:02:31.417002+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"A statistically consistent combination of the full set of measured Fano–Bloch coefficients (including a_k) that either tightens or fails to improve the present interval on Im(C_tG), or a future measurement of concurrence at the few-percent level that does not follow the projected exclusion contours.","supporting_citations":[],"review_version":1}