{"id":"b0c3e0a2-5fca-460c-98f9-05c4725ce4bb","arxiv_id":"2411.14234","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Quantum steering in three-flavor neutrino oscillations deviates by up to 21 percent when non-standard interactions are included, most visibly for the DUNE baseline.","lead":"This paper calculates how quantum steering, a correlation between two distant quantum systems, changes when neutrinos interact with matter in a non-standard way, for the NOvA and DUNE experiments. It reports up to 21 percent deviations between standard and non-standard steering values in DUNE's energy range, proposing quantum correlations as a new probe of new physics.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Eq. (19) is asserted as an if-and-only-if two-setting steering criterion, but Tr(T^T T)>1 is only a witness; for the paper's own X-shaped reduced states it can flag steerability when the Cavalcanti two-setting inequality is not violated, so the 'steering' interpretation of the 21% claim is not…","rationale":"I read the paper as a phenomenological application of the steering witness S_max, defined from Eq. (19), to neutrino oscillation probabilities in the presence of NSI. The algebraic core—expressing S_AB, S_AC, S_BC through Eqs. (39)-(44)—is internally consistent: I checked that these are the correct traces T^T T for the X-shaped reduced density matrices, and the reduction to the two-flavor limit (46) is correct. The reader's weakest assumption is exactly where I also locate the main risk: the paper cites Ref. [35] for an if-and-only-if statement that is not generally true for two-qubit states. The trace Tr(T^T T)>1 is a known steering witness in some of the cited literature, but it is not equivalent to violation of the two-setting Cavalcanti inequality, whose optimal form depends on the largest singular values of T. Because the paper's quantitative headline is the change in S_max under NSI, and S_max is only meaningful if it is a faithful steering quantifier or at least a valid witness, this is load-bearing. The counterexample I construct—a valid reduced state of the paper's own family with P_τ=0.30, P_μ=P_e=0.35—illustrates the gap: S_AB>1 yet the maximized linear two-setting expression does not exceed 1. Whether exact two-setting steerability holds for this state must be settled by the LHS SDP, which is why I propose that as the decisive test. I do not treat disagreement with consensus as a flaw; rather, this is a correctness risk internal to the paper's own stated criterion. The missing exact NSI parameter point is a reproducibility concern, but the steering-witness issue is more fundamental. I therefore keep the reader's CONDITIONAL verdict: the paper's idea is plausible and its probability algebra is sound, but the interpretation of the numerical results requires either a proof/clear citation for the special-case iff, or a switch to a verified steering criterion. If the SDP test shows false positives in the DUNE energy range, the 21%/15% claims would need to be re-evaluated, possibly moving toward rejection.","tokens_in":12607,"tokens_out":31765,"duration_ms":307209,"concrete_test":"Take the three-flavor oscillation probabilities from the DUNE NO curve in Fig. 1 at, say, E=2.5 GeV, and form the reduced states (30)-(32). Compute the singular values of each T and test the exact two-setting steering condition by solving the LHS semidefinite program (or, if a closed criterion is available, the condition s1^2+s2^2>1 for the two largest singular values). Compare this against S_AB>1 of Eq. (19) bin by bin in the energy range [2.3,3.1] GeV. If there is any bin where Eq. (19) gives S>1 but the exact two-setting SDP says unsteerable, Eq. (19) is false as an iff and the 21% claim must be recomputed with a valid steering witness.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that S_max, built from Eq. (19), is a steering measure and that its 21%/15% shifts under NSI are physically meaningful. Equation (19) states that a two-qubit state is steerable with two measurement settings iff S_AB = Tr(T^T_AB T_AB) > 1, citing Ref. [35]. But the two-setting steering inequalities of Cavalcanti et al. [44], from which Ref. [35] derives its witness, involve the largest singular values of T, not the sum of all squared singular values. For the paper's reduced states (30)-(32), T has singular values s1=s2=2sqrt(P_μ P_e) and s3=|P_τ-P_μ-P_e| (where P_τ, P_μ, P_e are the probabilities associated with the |00>, |01>, |10> weights). A concrete member of this family, P_τ=0.30, P_μ=P_e=0.35, gives S_AB=1.14>1, so Eq. (19) labels it steerable, while the maximized two-setting linear expression is F_2=(s1+s2)/sqrt(2)=0.99<1. Unless Ref. [35] established a special equivalence for this class of single-excitation/rank-deficient states, which the paper does not demonstrate, the trace criterion is at best a sufficient steering witness and not an iff. Since every subsequent statement—'S_max exceeds 1 in all cases' and the DUNE 21%/15% deviations—interprets S_max as steering, the load-bearing link between the computed quantity and quantum steering is unsupported. A secondary reproducibility issue is that Table II gives NSI ranges but the figures use unspecified points within those ranges, so the numerical headline cannot be independently checked.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies quantum steering in three-flavor neutrino oscillations in the presence of non-standard interactions (NSI). It derives formulas for the quantity S_max, defined as the maximum of Tr(T^T T) over the three reduced two-qubit states of a single-excitation three-flavor state, and expresses S_max directly in terms of oscillation probabilities (Eqs. 42-45). Using standard oscillation parameters and NSI ranges from external fits, it computes S_max for NOvA and DUNE under both mass orderings and reports characteristic deviations of the NSI case from the Standard Model, notably about 21% for DUNE normal ordering and about 15% for inverted ordering in the multi-GeV energy range. The paper also compares steering with Bell-CHSH nonlocality and concurrence, and gives relations between S_AB and reduced concurrences.","tokens_in":13030,"tokens_out":14620,"duration_ms":150760,"significance":"If the steering interpretation were valid, the paper would provide an interesting bridge between neutrino oscillation phenomenology and quantum information: explicit probability-only formulas for a correlation quantifier, a concrete comparison of nonlocality, steering, and entanglement in long-baseline experiments, and a falsifiable prediction that NSI shifts the quantifier by tens of percent in DUNE. The analytical content is explicit and checkable, and the setup uses externally fitted oscillation and NSI parameters rather than fitting to the target result, so there is no obvious circularity. However, the central physical claim hinges on the steering criterion in Eq. (19), and that criterion is not supported by the paper's own stated framework. The numerical conclusions are also presented for a single, incompletely specified NSI benchmark, so the headline percentages are not independently reproducible as stated.","major_comments":[{"comment":"The criterion S_AB = Tr(T^T T) > 1 is asserted as an if-and-only-if condition for two-setting steering, citing Ref. [35], but it is inconsistent with the Cavalcanti inequality in Eq. (17). For n=2, the maximum of Eq. (17) is F_2 = (1/sqrt(2)) sqrt(lambda_1+lambda_2), where lambda_i are the eigenvalues of T^T T, so two-setting steerability requires lambda_1+lambda_2 > 2, not Tr(T^T T) > 1. For the reduced states in Eqs. (30)-(32), the correlation matrix has singular values s1=s2=2 sqrt(P_mu P_e) and s3=|P_tau - P_mu - P_e|, giving F_2 = 2 sqrt(P_mu P_e), which is at most 1 for all physical probabilities since P_mu+P_e <= 1. Thus no state of the single-excitation form studied in this paper ever violates the two-setting steering inequality. Concretely, taking P_tau=0.30, P_mu=P_e=0.35 gives S_AB=1.14 by Eq. (42), but F_2=0.7<1. Therefore the interpretation of S_max>1 as steering, and the derived 21% and 15% claims in the abstract and Section V, are unsupported without a proof that Ref. [35] establishes an equivalence for this specific class of states. The trace condition is at best a necessary condition for two-setting steerability, not a sufficient one.","section":"Section III, Eq. (19); Section IV, Eqs. (42)-(45); Section V"},{"comment":"The numerical NSI curves are said to use the 2 sigma upper limits of Table II, but Table II lists ranges, several of which are unions of disjoint intervals, and the text does not specify which point in each range is used or whether all parameters are set simultaneously. The 21% and 15% deviation numbers therefore cannot be independently reproduced from the information given. Please state the exact benchmark values and show how the results vary across the allowed NSI ranges, for instance with uncertainty bands.","section":"Section V and Table II"},{"comment":"The text states that in the SM the NOvA steering maximum occurs at E=0.8 GeV for NO and E=1.26 GeV for IO, but Figure 1 shows an energy axis from 2 to 10 GeV, so the cited energies are outside the plotted range. This mismatch makes the NOvA comparison, including the claimed enhancements at E approximately 1.26 GeV, unverifiable as presented. The figure or the text should be corrected so that the quantitative statements refer to the same energy range.","section":"Section V, paragraph on NOvA; Figure 1"}],"minor_comments":[{"comment":"The horizontal axis label reads \"Energy(Gev)\" and the axis appears duplicated in the extracted figure; please correct the typo and the layout.","section":"Figure 1"},{"comment":"For parameters such as epsilon_ee and epsilon_mu_mu, the quoted 2 sigma ranges are unions of two disjoint intervals; the phrase \"upper limit\" should define whether the largest positive endpoint or the boundary of the full interval is used.","section":"Table II"},{"comment":"Equation (45) is stated to be valid for vacuum oscillations and for NSI-modified probabilities, but the derivation assumes a single-excitation three-qubit pure state; please state explicitly that NSI enters only through the replacement of the probabilities P_alpha beta by their NSI-modified values.","section":"Section IV, Eq. (45)"}],"recommendation":"reject","confidential_remarks":"The main reason for rejection is the unsupported steering criterion in Eq. (19), which is load-bearing for every physical conclusion in the paper. If a correct two-setting steering condition is used, the specific reduced states studied here never violate it, so the central claim cannot be repaired by a local correction. The authors would need to reformulate the observable, recompute the numerical results under a valid steering inequality or a different recognized steering measure, and then reassess whether any NSI-induced steering signature survives."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this before reading: the headline 21%/15% numbers are about a quantity that is probably not quantum steering. Eq. (19) asserts the iff criterion S_AB = Tr(T^T T) > 1 for two-setting steerability, citing [35]. The two-setting linear steering inequality of Cavalcanti et al. is violated when the sum of the two largest singular values of T exceeds sqrt(2), not when the Frobenius norm squared exceeds 1. The stress-test note gives a concrete counterexample from the paper's own state family: with P_tau=0.30, P_mu=P_e=0.35, S_AB=1.14>1 while the maximized two-setting expression is (s1+s2)/sqrt(2)=0.99<1. So the trace condition is at best a sufficient witness, not an iff, and the claim that Smax>1 means steerable does not follow. Unless [35] proved a special equivalence for this rank-deficient class—which the paper does not demonstrate—the word 'steering' in the abstract and conclusions is doing work it cannot support.\n\nWhat is actually new and useful: the paper gives a clean three-flavor expression for the trace measure in terms of oscillation probabilities (Eqs. 42-45), checks the two-flavor limit, and connects it to concurrence. That algebra is correct. The numerical study of how NSI shifts this trace measure in NOvA and DUNE is a legitimate scenario exercise, even if the name 'steering' is wrong.\n\nThe other soft spots are real but minor by comparison. The NSI plots use a single point inside the 2-sigma ranges of Table II, with no error bands, so the 21%/15% numbers cannot be independently reproduced. There is also a text-figure mismatch for NOvA energy values that the figures do not resolve.\n\nBottom line: this is a salvageable paper, but the central interpretational claim is currently unsupported. The authors should either prove that their trace measure is equivalent to a genuine steering inequality for this state class, or recompute with the standard two-setting measure—and then the 21% claim may change. As it stands, a careful reader cannot trust the 'steering' part. Worth sending to referees, because the issue is subtle and fixable, but I would not cite it in its current form.","headline":"The paper's central 'steering' numbers are driven by a trace criterion that is not equivalent to any standard steering inequality; the oscillation-probability algebra is clean, but the main interpretation is unsupported.","tokens_in":13555,"tokens_out":8746,"would_cite":false,"duration_ms":79363,"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":"This paper argues that quantum steering in neutrino oscillations is measurably altered by non-standard interactions, with DUNE showing a 21% deviation from the Standard Model at its peak-flux energies.","keywords":["quantum steering","neutrino oscillations","non-standard interaction","DUNE","NOvA","three-flavor mixing","Bell nonlocality","concurrence"],"falsifier":"Compute the same DUNE curves using the general two-setting steering parameter—the sum of the two largest singular values of each correlation matrix—in place of $\\mathrm{Tr}(T^T T)$; if the $S_{\\max} > 1$ windows or the reported 21% and 15% NSI deviations shift materially, the paper's central numerical claim fails.","tokens_in":12402,"feed_emoji":"⚛️","tokens_out":9003,"duration_ms":80439,"temperature":0.7,"pith_summary":"This paper argues that quantum steering—the ability of one neutrino's flavor measurement to change what can be predicted about a partner neutrino's state—is a practical probe of new physics in neutrino oscillations. The authors show that the maximum steering value of a three-flavor neutrino state, $S_{\\max}$, is a simple function of the three oscillation probabilities $P_{\\alpha e}, P_{\\alpha \\mu}, P_{\\alpha \\tau}$, so it can be computed directly from measured oscillation data. Applying this to the NOvA and DUNE experiment geometries with a Standard Model plus non-standard interaction (NSI) Hamiltonian, they find that DUNE's longer baseline makes steering noticeably more NSI-sensitive: around DUNE's peak-flux energies, $S_{\\max}$ deviates by about 21% from the Standard Model prediction for normal mass ordering and rises about 15% for inverted ordering. If correct, this gives a new, oscillation-probability-based route to spotting NSI effects and to comparing entanglement, steering, and Bell nonlocality in the same neutrino system.","feed_headline":"Neutrino steering shifts 21% under non-standard interactions","feed_subtitle":"Longer baselines make DUNE's quantum-steering signal more NSI-sensitive than NOvA's.","key_machinery":"The load-bearing object is quantum steering—Alice's ability to update Bob's state by a local measurement—quantified by the steering measure $S_{AB} = \\mathrm{Tr}(T_{AB}^{T} T_{AB})$ built from the correlation matrix $T_{AB}$ of a two-qubit reduced state. For a three-flavor neutrino in the flavor-mode encoding $|\\nu_e\\rangle = |100\\rangle$, $|\\nu_\\mu\\rangle = |010\\rangle$, $|\\nu_\\tau\\rangle = |001\\rangle$, the three bipartite reduced states yield $S_{AB}$, $S_{AC}$, $S_{BC}$, and their maximum $S_{\\max} = \\max\\{S_{AB}, S_{AC}, S_{BC}\\}$. The key identity is that each $S$ collapses to a polynomial in the oscillation probabilities, for example $S_{AB} = 8P_{\\alpha e}P_{\\alpha \\mu} + [P_{\\alpha \\tau} - P_{\\alpha e} - P_{\\alpha \\mu}]^2$, so the whole steering analysis reduces to computing those probabilities. The probabilities themselves come from the evolution operator $U_f(L)$ of the three-flavor Hamiltonian with matter potential plus NSI parameters $\\epsilon_{\\alpha\\beta}$, and the experimental numbers use the NOvA and DUNE baselines and flux ranges. This machinery lets the authors translate any modification of oscillation probabilities—including NSI-induced ones—directly into a modification of the steering witness.","core_discovery":"On the paper's own terms, the central discovery is a set of closed-form expressions—Eqs. (42)–(45)—that write the steering measures $S_{AB}$, $S_{AC}$, $S_{BC}$, and their maximum $S_{\\max}$ entirely in terms of three-flavor neutrino oscillation probabilities; this formula is valid for vacuum oscillations and, because it uses probabilities, also for NSI-modified oscillations. From those expressions, the paper numerically finds that in the DUNE setup, within the energy window $[2.3, 3.1]$ GeV where DUNE's flux peaks, including NSI changes $S_{\\max}$ by about 21% below the Standard Model value for normal ordering and by about 15% above it for inverted ordering. The same comparison for NOvA produces smaller or energy-dependent shifts. Throughout the studied energies $S_{\\max}$ stays above 1, which the paper interprets as the neutrino states being steerable, and the DUNE comparison of Bell-CHSH, steering, and concurrence exhibits the expected hierarchy: Bell nonlocality implies steering implies entanglement, but not conversely. The paper also derives exact relations between the pairwise steering measures and the pairwise concurrences of the three-qubit flavor state.","pith_inferences":["Beyond the paper's explicit claims: the probability-only form of $S_{\\max}$ means DUNE could report an $S_{\\max}$-versus-energy curve from reconstructed oscillation probabilities, and persistence of the 21%/15% NSI deviations in such a curve would be a new, largely systematic-independent NSI diagnostic.","The trace criterion $\\mathrm{Tr}(T^T T) > 1$ is not the general two-setting steering parameter; checking whether the general sum-of-two-largest-singular-values condition changes the steerable regions would settle whether the reported windows and percentages are artifacts of the adopted criterion.","The same probability-polynomial construction applies to any three-flavor unitary evolution, so with appropriate probabilities it could also probe sterile mixing, non-unitarity, or modified matter potentials in the same steering language.","The concurrence relations suggest a possible entanglement-based lower bound on steering in mixed flavor states; whether the pure-state formulas survive partial tracing in realistic detector scenarios is a direct testable extension."],"forward_implications":["Because $S_{\\max}$ is a function only of $P_{\\alpha e}$, $P_{\\alpha \\mu}$, and $P_{\\alpha \\tau}$ (Eq. (45)), any experiment that measures three-flavor oscillation probabilities can compute the steering witness directly, without additional quantum measurements.","In DUNE's peak-flux window $[2.3, 3.1]$ GeV, NSI shifts $S_{\\max}$ by about 21% for normal ordering and about 15% for inverted ordering, making steering a more sensitive NSI indicator there than the shorter-baseline NOvA.","When $P_{\\alpha \\tau} = 0$, the three-flavor formula reduces to the two-flavor result $S_{AB} = (P_{\\alpha e} + P_{\\alpha \\mu})^2 + 8P_{\\alpha e}P_{\\alpha \\mu}$, linking the new expressions to the existing two-flavor literature.","The relations $S_{AB} = 1 + 2C_{AB}^2 - C_{AC}^2 - C_{BC}^2$ (and cyclic versions) tie steering violations directly to pairwise concurrences, so pairwise entanglement measurements bound the steering inequality.","The DUNE comparison of Bell-CHSH, steering, and concurrence shows the expected nesting of nonlocality, steering, and entanglement, with steering present at all energies studied and the hierarchy respected."],"supporting_citations":[{"why":"Supplies the two-setting steering criterion $S_{AB} = \\mathrm{Tr}(T_{AB}^T T_{AB}) > 1$ on which all steerability claims rest.","marker":"[35]"},{"why":"Defines the Cavalcanti linear steering inequalities that ground the notion of steerability used here.","marker":"[44]"},{"why":"Gives the three-flavor evolution operator in matter with NSI, used to compute the oscillation probabilities.","marker":"[39]"},{"why":"Provides the best-fit neutrino oscillation parameters (Table I) for the numerical SM and NSI curves.","marker":"[55]"},{"why":"Provides the NSI parameter ranges (Table II) whose 2σ upper limits generate the NSI plots.","marker":"[56]"},{"why":"Defines the NOvA experiment baseline and setup that anchors the shorter-baseline comparison.","marker":"[12]"},{"why":"Defines the DUNE experiment baseline and setup that anchors the main 21%/15% result.","marker":"[13]"},{"why":"Establishes the two-flavor steering-in-probabilities expression and pure-state steering-concurrence relation that the three-flavor result extends and reduces to.","marker":"[51]"}],"fun_headline_variants":["NSI alters neutrino steering 21% at DUNE","DUNE's steering signal shifts 21% under NSI","Closed-form steering shows NSI changes DUNE by 21%","Neutrino steering: NSI effect hits 21% at DUNE"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The results assume that the simple correlation-matrix condition 'trace of $T^T T$ exceeds 1' exactly matches the accepted definition of two-setting steerability for the neutrino states considered; if it only approximates it, the regions labeled steerable, and the 21% and 15% deviation numbers, could change.","fun_headline_variants_meta":{"raw":{"variants":["NSI alters neutrino steering 21% at DUNE","DUNE's steering signal shifts 21% under NSI","Closed-form steering shows NSI changes DUNE by 21%","Neutrino steering: NSI effect hits 21% at DUNE"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000254,"raw_usage":{"total_tokens":1565,"prompt_tokens":940,"completion_tokens":625,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":556,"completion_tokens_details":{"reasoning_tokens":551}},"tokens_in":556,"tokens_out":625,"duration_ms":6569,"temperature":1.0,"reasoning_tokens":551,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:27:18.745746+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the same DUNE curves using the general two-setting steering parameter—the sum of the two largest singular values of each correlation matrix—in place of $\\mathrm{Tr}(T^T T)$; if the $S_{\\max} > 1$ windows or the reported 21% and 15% NSI deviations shift materially, the paper's central numerical claim fails.","supporting_citations":[{"cited_title":"Yang et al., Phys","cited_arxiv_id":null,"evidence_quote":"Defines the Cavalcanti linear steering inequalities that ground the notion of steerability used here."},{"cited_title":"Grossman, Phys","cited_arxiv_id":null,"evidence_quote":"Gives the three-flavor evolution operator in matter with NSI, used to compute the oscillation probabilities."},{"cited_title":"Bittencourt, M","cited_arxiv_id":null,"evidence_quote":"Provides the best-fit neutrino oscillation parameters (Table I) for the numerical SM and NSI curves."},{"cited_title":"Adamson et al","cited_arxiv_id":null,"evidence_quote":"Defines the NOvA experiment baseline and setup that anchors the shorter-baseline comparison."},{"cited_title":"Qiu, Quantum Inf","cited_arxiv_id":null,"evidence_quote":"Establishes the two-flavor steering-in-probabilities expression and pure-state steering-concurrence relation that the three-flavor result extends and reduces to."}],"review_version":1}