{"id":"73cf1e2d-7cfb-4d29-949b-b739950a89ef","arxiv_id":"2506.18734","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Quantum steering harvested from the vacuum near a reflecting wall becomes directionally asymmetric when the detectors are oriented perpendicular to the wall, and the preferred direction can be tuned by the detectors' energy gaps.","lead":"Two small detectors placed near a reflecting wall can harvest quantum steering, a one-way form of quantum correlation, from the vacuum. The paper shows that rotating the detector pair from parallel to perpendicular to the wall changes which detector can steer the other, suggesting geometry can channel quantum information.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The directional claims rest on Eqs. (11)-(12), an entanglement-based sufficient steering witness, whose magnitude is not shown to be a faithful steering quantifier; if loose for these X states, the reported enhancement/suppression may be an artifact.","rationale":"The paper is a clean derivation of a detector-field state near a reflecting boundary, and the algebra from Eq. (16) to Eqs. (20)-(21) is internally consistent. The central physical claim, however, is quantitative: it asserts directional enhancement and suppression of harvested steering. That claim is supported only through Eqs. (11)-(12), which are violations of an entanglement-based sufficient steering criterion. The amount of violation of a sufficient condition is not, without further argument, a steering monotone, so the ordering of S_A→B and S_B→A and the differences between parallel and orthogonal configurations are not established as genuine resource-theoretic comparisons. The reader's weakest assumption correctly identifies this. Since no independent steering measure, such as steering robustness, is provided, the conditional verdict is appropriate rather than full acceptance, but the concern is a gap in evidence, not a demonstrated internal contradiction; thus the verdict need not be changed to reject.","tokens_in":15311,"tokens_out":5405,"duration_ms":61430,"concrete_test":"Recompute the central comparisons of Figs. 4, 5, and 7 with a genuine steering quantifier: for the X states of Eq. (16) at the plotted parameters (Ω_A σ = 0.10, Ω_B σ ∈ {0.10, 0.12, 1.00}, L/σ ∈ [0, 0.2], Δz/σ ∈ [0, 1.5]), evaluate the steering robustness or steering weight via the standard SDP for two-qubit steerability with projective measurements. Then check whether (i) orthogonal > parallel for B→A is preserved, (ii) parallel > orthogonal for A→B is preserved, and (iii) the sign of S_AB^Δ at each plotted point matches the sign of the robust asymmetry. A sign flip or vanishing of these comparisons would show that the headline claim is an artifact of the sufficient witness.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section II defines S_B→A and S_A→B as the largest violation of two steering-detection inequalities obtained by requiring entanglement of τ_AB and τ_BA (Eqs. (3)-(12)). These are sufficient steering witnesses, not established steering measures: a positive value certifies steerability in the corresponding direction, but the amount of violation is not proved to be monotone under one-way LOCC, tight for the X states of Eq. (16), or normalized as a resource measure. The abstract and Figs. 4-7 compare these values across configurations and draw the headline conclusion that orthogonal alignment enhances B→A and suppresses A→B. Every quantitative comparison, including the sign of S_AB^Δ in Fig. 4, the peak structure in Fig. 5, and the difference plots in Fig. 7, depends on the unproved faithfulness of this criterion. A zero in these measures means no violation of a sufficient condition, not no steering; a larger violation does not automatically mean more steerability. The detector-state derivation in Eqs. (17)-(31) is internally consistent, but it is expressed through the same witness, so algebraic consistency does not transfer to the physical directional claim. No independent steering monotone is computed, and no error bound or tightness analysis is given.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies quantum steering harvesting by two static Unruh-DeWitt detectors coupled to a massless scalar vacuum in the presence of a perfectly reflecting boundary, comparing detectors aligned parallel or orthogonal to the boundary. Using the method-of-images Wightman function, the authors derive the leading-order detector density matrix and construct two directional steering witnesses, S_{B→A} and S_{A→B}, from entanglement-based detection inequalities of Refs. [68,69,72]. They then analyze how these quantities depend on detector separation, distance from the boundary, and energy-gap difference, concluding that the boundary can suppress steering in one direction while enhancing it in the other, and that boundary configuration can thus be used to optimize directional steering harvesting.","tokens_in":15558,"tokens_out":6680,"duration_ms":68409,"significance":"The closed-form calculations are a solid technical contribution: the correlation integrals reduce to explicit special functions, no free parameters are introduced, and the numerical results are internally consistent with the analytic expressions. The physical question—whether a boundary can act as a directional control knob for harvested steering, in contrast to boundary-influenced entanglement harvesting—is interesting and well matched to the relativistic quantum information literature. The main caveat is that the directional claims are currently quantitative statements about a sufficient detection witness rather than about a validated steering measure; unless that issue is resolved or the claims are explicitly reframed, the results are significant mainly at the level of steering detection rather than quantitative steering resource.","major_comments":[{"comment":"The quantities S_{B→A} and S_{A→B} are the largest violations of sufficient entanglement-based steering-detection inequalities, not established steering measures. A positive value certifies steerability in one direction, but the magnitude of the violation is not shown to be monotone under one-way LOCC, tight for the X-state family of Eq. (16), or otherwise proportional to a steering resource. The central conclusions—'orthogonal alignment enhances B→A and suppresses A→B', the ordering in Fig. 4(a), and the difference plots in Fig. 7—compare these witness magnitudes as if they were faithful quantitative measures. If the criterion is loose for these states, the reported enhancement/suppression and even the sign of S^Δ_{AB} could be artifacts of the detection criterion. I request an explicit steering measure for the harvested X states (e.g., an SDP-based steering robustness or a tight steering criterion), or a reformulation of all quantitative claims as statements about this witness rather than about steering strength.","section":"II, Eqs. (11)–(12); §IV, Figs. 4–7"},{"comment":"The text interprets vanishing of the same witness as a 'sudden death of quantum steering' and as transitions from two-way to one-way and then to no-way steering. A zero of a sufficient detection inequality only means that the witness does not certify steerability; it does not establish that the state is not steerable. These existence claims require an exact steering criterion for the states under study. Please either prove the absence of steering in the relevant regions (for example with an exact one-way-steering characterization of X states) or reword the discussion in terms of 'no steering detected by the criterion'.","section":"IV.A, Fig. 2 and Conclusion (iii)"},{"comment":"The abstract's claim that 'across most of the parameter space' the orthogonal alignment enhances B→A and suppresses A→B is not supported by the numerical evidence presented. The parameter scan in Figs. 4–7 is limited to ΩA σ = 0.10, with a small set of values for ΩB σ, L/σ and Δz/σ (notably Δz/σ = 1.00 in Figs. 4, 6 and 7, and L/σ = 0.05 in Fig. 5). The phrase 'most of the parameter space' overstates the demonstrated domain. I ask that the claim be qualified to the explored parameter region or supported by a systematic scan.","section":"Abstract and §IV.B"}],"minor_comments":[{"comment":"Several captions say 'for various values of ΩB σ' but show only two or three selected values; please list the exact displayed values in each caption.","section":"Figure captions, Figs. 2, 4 and 5"},{"comment":"The value of the coupling λ used to produce the plots is not stated, so the plotted steering values are meaningful only up to an overall factor λ²; please state λ or indicate that all plotted values are in units of λ².","section":"Figures 2–8"},{"comment":"A brief statement of the perturbative regime (λ ≪ 1) and of the size of the omitted O(λ^4) terms relative to the plotted steering values would make the quantitative discussion more precise.","section":"Eq. (16) and §III"},{"comment":"The notation for the orthogonal configuration alternates between S_V, S^v and S in the text and figure labels; please standardize the symbol.","section":"§IV.B and figure labels"}],"recommendation":"major_revision","confidential_remarks":"The referee's recommendation is driven by the gap between the paper's quantitative language and the sufficient witnesses on which it relies. The derivations appear internally sound; if the authors reframe the results as detection-level statements or add an exact steering measure, the paper would be suitable for publication. I do not see citation or novelty concerns."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a real new result in steering harvesting, with one load-bearing caveat. The new physics is that a reflecting boundary, plus detector orientation and energy-gap difference, can make one-way steering harvesting asymmetric in a way entanglement harvesting cannot. For identical detectors aligned parallel, S_A→B = S_B→A; orthogonal alignment breaks that because one detector sits closer to the boundary. That is a clean, specific statement I don't think is in the earlier literature.\n\nWhat the paper does well: it sets up the standard UDW harvesting calculation carefully. Method-of-images Wightman function, perturbative density matrix to O(λ²), X-state form, and the closed-form expressions for P_D, C, X. I checked the structure of Eqs. (20)-(21) and they follow from the steering inequalities. No fitted parameters, no invented entities. The citation pattern is normal; self-citations are context, not load-bearing.\n\nSoft spot, and it is real: Eqs. (11)-(12) define S_B→A and S_A→B as the amount by which an entanglement-based sufficient steering condition is violated. That certifies steerability when positive; it is not shown to be a faithful steering measure. The magnitude is not proved monotone under one-way LOCC, and the X-state family here is not tested for tightness. The abstract's 'enhance/suppress' language, and Figs. 4-7, compare magnitudes of these witnesses. If the witness is loose, part of the reported asymmetry could be an artifact of the detection criterion rather than a genuine difference in steerability. Also S=0 means 'no violation detected', not necessarily 'no steering'. That should be fixed by either proving tightness for these states, checking a known steering monotone on a subset of parameters, or carefully rewording to 'detected steering' and restricting conclusions to certification.\n\nMinor: numerical exploration is narrow (one gap value, one boundary distance for most figures), so 'across most of parameter space' overshoots. The analytic formulas are general, so a wider scan is cheap. Still, the central steerability-asymmetry observation is plausible and worth taking seriously.\n\nWho benefits: people working on vacuum entanglement/steering harvesting and relativistic quantum information. I'd bring it to a reading group as a good example of how a sufficient steering witness can drive conclusions. It deserves a serious referee; my recommendation is to send it to review, but with a referee who knows the steering-witness literature and demands the tightness question be answered or the claims scaled back.","headline":"Worth a serious referee: the directional steering asymmetry is new and the derivation is clean, but the headline 'enhance/suppress' claims rest on an unproven steering witness magnitude, so the paper needs a tightness check or softened claims.","tokens_in":16041,"tokens_out":2882,"would_cite":false,"duration_ms":31761,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["04.70.Dy","03.65.Ud","04.62.+v"],"model":"deepseek-v4-flash","headline":"The paper claims that a perfectly reflecting boundary can act as a tunable directional control for quantum steering harvested from the vacuum: across most of the parameter space, orthogonal alignment enhances Bob-to-Alice steering and…","keywords":["quantum steering harvesting","Unruh-DeWitt detectors","reflecting boundary","directional steering asymmetry","vacuum correlations","X-state steering witness","one-way steering"],"falsifier":"Compute a certified steering monotone, for example the steering robustness obtained by semidefinite programming, for the same X states used in the paper's plots, and check whether the orthogonal configuration still gives $S_{B\\to A} > S_{A\\to B}$ in the parameter regions where the paper's witness says it does. A reversal or vanishing of the ordering under the certified measure would show that the directional boundary effect is an artifact of the concurrence-based witness; preservation would independently confirm the paper's central claim.","tokens_in":15141,"feed_emoji":"🪞","tokens_out":8797,"duration_ms":85733,"temperature":0.7,"pith_summary":"This paper sets out to show that the direction in which quantum steering can be harvested from the vacuum is not fixed by the two detectors' energy gaps alone: the geometry of a reflecting boundary can tilt the balance one way or the other. Working with two Unruh-DeWitt detectors coupled to a massless scalar field near an infinite perfectly reflecting plane, the authors claim that aligning the detectors orthogonal to the boundary enhances steering from Bob to Alice while suppressing steering from Alice to Bob relative to the parallel alignment, across most of the parameter space they scan. They also report that identical detectors harvest symmetric steering when placed parallel but asymmetric steering when placed orthogonal, because the orthogonal placement puts the two detectors at unequal distances from the boundary. If these claims hold, boundary configuration becomes a practical control parameter for extracting directional quantum correlations, a capability that entanglement harvesting does not offer.","feed_headline":"Mirror geometry picks the direction of vacuum steering","feed_subtitle":"Orthogonal alignment boosts Bob-to-Alice steering and suppresses Alice-to-Bob, unlike symmetric entanglement","key_machinery":"The load-bearing machinery is the X-state density matrix produced at leading order in the detector-field coupling, together with a steering quantifier built from entanglement. The authors take the usual perturbative Unruh-DeWitt result for the two-detector state and record its matrix elements $P_A$, $P_B$ (excitation probabilities) and $C$, $X$ (correlation amplitudes). They then use the fact that steering from Bob to Alice can be witnessed by entanglement of a locally depolarised version of that state, and similarly for Alice to Bob, to write explicit quantities $S_{B\\to A}$ and $S_{A\\to B}$ (Eqs. (11)-(12)). The reflecting boundary enters through the method-of-images Wightman function, producing correlation terms $f(L)-f(\\sqrt{L^2+4\\Delta z^2})$ in the parallel case and $f(L)-f(L+2\\Delta z)$ in the orthogonal case, where $f$ and $g$ are auxiliary functions encoding the Gaussian switching and energy gaps. Because the orthogonal case gives the two detectors different distances to the boundary, $P_A$ and $P_B$ differ even for identical energy gaps, and that difference is what breaks the symmetry.","core_discovery":"On the paper's own terms, the central discovery is a directional asymmetry in vacuum steering harvesting that is controlled by detector orientation. For two detectors with Bob's energy gap at least as large as Alice's, the parallel configuration generally gives stronger steering from Alice to Bob than from Bob to Alice, while the orthogonal configuration tends to reverse or weaken that ordering: as the detector-boundary distance grows, the boundary suppresses steering in one direction and enhances it in the other. Identical detectors orthogonal to the boundary already show one-way bias because the closer detector experiences a different boundary-modified vacuum response than the farther one. The paper interprets this as evidence that steering, unlike entanglement, is a directional probe of spatial structure in the vacuum, and that optimal extraction of steering depends on which direction is wanted: parallel placement favors Alice-to-Bob steering and orthogonal placement favors Bob-to-Alice steering.","pith_inferences":["If the witness is faithful, the same asymmetry should be visible under any exact steering monotone; a numerical check with steering robustness would turn the paper's directional claims into a quantitative prediction rather than a witness-dependent one.","By continuity, detector orientations between parallel and orthogonal should interpolate between the two preferred directions, and the optimal angle may itself depend on separation, boundary distance, and energy-gap difference; scanning intermediate angles could reveal that the extremes are not always optimal.","The method-of-images structure suggests that more complex boundaries, such as two mirrors, a cavity, or curved mirrors, should create spatially varying preferred steering directions, so the effect is likely not limited to the single-plane geometry studied here.","A laboratory analogue with superconducting qubits coupled to an engineered vacuum could test the predicted reversal of one-way steering by state tomography, connecting this relativistic-vacuum result to tabletop quantum-information experiments."],"forward_implications":["To maximize steering from the detector closer to the boundary (Bob to Alice), place the detector pair orthogonal to the reflecting plane; to maximize the reverse direction (Alice to Bob), place them parallel.","Steering harvesting can serve as a directional probe of vacuum structure: the sign and size of the asymmetry reveal which detector is nearer the boundary and how strongly the boundary modifies the field.","Detuning the energy gaps widens the separation range over which Alice-to-Bob steering survives and shrinks the range for Bob-to-Alice steering, so gap difference and geometry can be co-tuned to engineer one-way steering.","The predicted two-way to one-way to no-way transitions as separation grows give an operational way to certify one-way steering from the vacuum in a single experimental setup.","Because entanglement harvesting shows no such directional bias, steering-based protocols gain an extra degree of freedom, detector orientation, that entanglement-based protocols do not have."],"supporting_citations":[{"why":"Supplies the entanglement-based steering detection method, constructing locally noisy states whose entanglement witnesses steering, from which the paper defines the two directional steering quantities.","marker":"[69]"},{"why":"Gives the analytical concurrence formula for X states that converts the steering-detection witness into the explicit inequalities in Eqs. (6)-(10).","marker":"[71]"},{"why":"Provides the perturbative two-detector density matrix (Eq. (16)) used as the starting point for all steering calculations.","marker":"[73]"},{"why":"Supplies the vacuum Wightman function and image method used to incorporate the perfectly reflecting boundary.","marker":"[74]"},{"why":"Derives the boundary-modified correlation functions f(L) and g(L) that appear explicitly in the parallel and orthogonal harvesting expressions.","marker":"[75]"},{"why":"Documents the earlier finding that harvested entanglement remains symmetric under boundary placement, which the paper uses as the contrast showing steering's directional sensitivity.","marker":"[56]"}],"fun_headline_variants":["Boundary orientation flips vacuum steering direction","Mirror angle steers vacuum steering asymmetry","Orientation tunes which detector gains vacuum steering","Parallel versus orthogonal: vacuum steering shows bias","Vacuum steering direction is set by detector alignment"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The directional comparison rests on treating the entanglement-based expressions $S_{B\\to A}$ and $S_{A\\to B}$ as faithful quantitative measures of steerability; if those witnesses overestimate or underestimate one direction for these X states, the claimed boundary-induced enhancement could be an artifact of the detection criterion rather than a property of the harvested vacuum correlations.","fun_headline_variants_meta":{"raw":{"variants":["Boundary orientation flips vacuum steering direction","Mirror angle steers vacuum steering asymmetry","Orientation tunes which detector gains vacuum steering","Parallel versus orthogonal: vacuum steering shows bias","Vacuum steering direction is set by detector alignment"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000515,"raw_usage":{"total_tokens":2514,"prompt_tokens":971,"completion_tokens":1543,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":587,"completion_tokens_details":{"reasoning_tokens":1476}},"tokens_in":587,"tokens_out":1543,"duration_ms":11394,"temperature":1.0,"reasoning_tokens":1476,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T18:44:31.157401+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute a certified steering monotone, for example the steering robustness obtained by semidefinite programming, for the same X states used in the paper's plots, and check whether the orthogonal configuration still gives $S_{B\\to A} > S_{A\\to B}$ in the parameter regions where the paper's witness says it does. A reversal or vanishing of the ordering under the certified measure would show that the directional boundary effect is an artifact of the concurrence-based witness; preservation would independently confirm the paper's central claim.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the entanglement-based steering detection method, constructing locally noisy states whose entanglement witnesses steering, from which the paper defines the two directional steering quantities."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the analytical concurrence formula for X states that converts the steering-detection witness into the explicit inequalities in Eqs. (6)-(10)."},{"cited_title":"Zhang, J","cited_arxiv_id":null,"evidence_quote":"Provides the perturbative two-detector density matrix (Eq. (16)) used as the starting point for all steering calculations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Derives the boundary-modified correlation functions f(L) and g(L) that appear explicitly in the parallel and orthogonal harvesting expressions."}],"review_version":2}