{"id":"7eccc30b-e84e-4273-9966-7e4709a336cb","arxiv_id":"2412.16288","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Quantum-controlled interactions without field degrees of freedom are shown to be unavoidably retrocausal, but the retrocausal contribution scales as L/T for long interaction times, and in current GME proposals it lies far below experimental resolution.","lead":"The paper shows that a quantum-controlled model of two systems interacting without a mediating quantum field predicts retrocausal effects whenever one system is in the causal future of the other, though the effects shrink relative to the causal signal as interaction time grows. The result is used to argue that current gravity-mediated entanglement experiments cannot rule out a mediator-free description of gravity.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; the Huygens assumption is standard for flat-space linearized gravity and the minor split inconsistency does not affect the central bound.","rationale":"The reader's ACCEPT verdict is well supported. I re-derived the main signalling-estimator integrals for the 3+1 setup and confirmed the ratios C_r/C_c = L/(T-L) and C_r/C_a = L/T for T > 2L. The corresponding non-perturbative phase-shift analysis is consistent: the retrocausal contribution appears as a constant phase lambda^2/(4 pi) in the modulus estimator, with a time shift delta T = L and period 8 pi^2 L/lambda^2. For the GME parameters quoted, both L/T ~ 1e-14 and lambda^2 ~ 1e-14 are far below achievable resolution, so the qc-model remains an experimentally indistinguishable alternative to a fully quantum gravitational field description. The weakest assumption flagged by the reader, the strong Huygens property, is not a live threat for linearized gravity in flat spacetime; linearized Einstein equations in harmonic gauge share the massless scalar retarded Green's function, and corrections from curvature, nonlinearity, or a massive graviton are suppressed far below the relevant scales. The single concrete discrepancy I found is the causal split in Eq. (91), which omits the -L offset present in the perturbative expression Eq. (67). This is a minor algebraic typo rather than a load-bearing flaw: it changes the split but not the relative phase shift or the L/T scaling, and the final GME bound is robust. I therefore see no reason to change the reader's ACCEPT verdict, though confidence remains moderate because the physical interpretation of retrocausality in the qc-model and the gravitational extension rest partly on modelling choices.","tokens_in":25214,"tokens_out":36674,"duration_ms":338821,"concrete_test":"Re-evaluate the degenerate-detector split of Section IV.B by directly integrating Delta(Lambda_a, Lambda_b) with the rectangular windows of Eq. (65). Confirm that Delta_c = lambda^2 (T-L)/(2 pi L), Delta_r = lambda^2/(2 pi), and that the difference between N_a and N_c is |sin(lambda^2 T/(4 pi L)) - sin(lambda^2 (T-L)/(4 pi L))| ~ lambda^2/(4 pi) for T >> L. If the offset instead changes the relative phase by more than order lambda^2/(4 pi), the GME time-resolution bound would need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"No significant objection identified. The central estimate, C_r/C_a = L/T for 3+1 massless fields, follows from the strong Huygens propagator (Eq. 62), and the extension to linearized gravity in Minkowski spacetime is sound: in harmonic gauge the metric perturbation satisfies sourced wave equations with the same delta-supported retarded Green's function, so no tails arise. Tail contributions from background curvature, nonlinearities, or a possible graviton mass are suppressed by tiny dimensionless parameters for the proposed GME scales (L ~ 1e-6 m, T ~ 1 s, lambda^2 ~ 1e-14), and the 1+1 counterexample does not transfer to this regime. The only concrete issue found is a secondary algebraic discrepancy: Eq. (91) gives the causal split as lambda^2 T/(2 pi L), whereas the explicit split using the same windows as Eq. (65) gives lambda^2 (T-L)/(2 pi L). This changes an offset of order L/T ~ 1e-14 and does not alter the retrocausal phase shift lambda^2/(4 pi) in the modulus estimator, the time-shift delta T = L, or the GME conclusion. The perturbative ratios, the non-perturbative phase argument, and the operational definitions of retrocausality are otherwise internally consistent.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper analyzes the quantum-controlled (qc) model, in which two quantum systems interact through a direct, retarded two-body coupling with no local quantum field degrees of freedom, and compares it with a full QFT description. It provides operational definitions of non-retrocausal interactions (Definitions III.1 and III.2) and argues from the symmetric-propagator structure of the qc evolution (Eqs. (8), (16)-(18), (39)) that the qc-model necessarily contains retrocausal contributions whenever one system lies in the causal future of the other, so that any regime in which the qc-model approximates QFT carries some retrocausal signalling. The paper quantifies this in explicit qubit setups: for a massless scalar in (3+1) dimensions with window switching and pointlike smearing, the retrocausal part of the signalling estimator is C_r/C_a = L/T for T > 2L (Eqs. (66)-(70)), while in (1+1) dimensions the retrocausal contribution can dominate, with C_r/C_a tending to 1 as S → ∞ (Eqs. (75)-(80)). A nonperturbative (gapless) analysis gives a retrocausal phase shift of order λ^2/(4π) and a time shift δT = L (Eqs. (91)-(97)). These results are applied to gravity-mediated entanglement (GME) experiments with L ~ 10^-6 m, T ~ 1 s, λ^2 ~ 10^-14, concluding that the retrocausal predictions of the qc-model are below the resolution of current proposals, so those experiments do not yet discriminate between a qc-description and a full quantum-gravitational one.","tokens_in":25427,"tokens_out":29092,"duration_ms":226189,"significance":"If the results hold, the paper delivers a quantitative, checkable bound on the regime of validity of relativistic direct-coupling models and sharpens the GME debate: for the currently proposed parameters, a qc-description without gravitational field degrees of freedom remains experimentally indistinguishable, so those experiments would not by themselves reveal the quantum nature of gravity. The paper's strengths are its explicit operational definitions of retrocausality, the fully analytic perturbative computations of the signalling estimators (Eqs. (66)-(68) and (75)-(79), which I verified), the nonperturbative gapless check, and the honest identification of the strong-Huygens assumption through the (1+1)-dimensional counterexample in which retrocausal signalling is not bounded. The central conclusions are conditional on the absence of tails in the mediating propagator, which is standard for massless fields in (3+1)-dimensional Minkowski spacetime and for linearized gravity in harmonic gauge; the paper itself demonstrates where the assumption fails.","major_comments":[{"comment":"The claimed additive split ∆ab = ∆(c)ab + ∆(r)ab is inconsistent as written. With the same window functions as in Eq. (65), the causal part of the perturbative signalling estimator for T > 2L is C(c)a = (T−L)/(2πL) (from Eq. (67)), so the nonperturbative split should read ∆(c)ab = λ²(T−L)/(2πL) and ∆(r)ab = λ²/(2π); as stated, Eq. (91) gives ∆(c)ab = λ²T/(2πL), and then ∆(c)ab + ∆(r)ab = λ²(T+L)/(2πL) ≠ ∆ab = λ²T/(2πL). Correspondingly, Eq. (93) should use Na = |sin(λ²T/(4πL))| and N(c)a = |sin(λ²(T−L)/(4πL))|. I emphasize that the downstream conclusions survive: the relative phase shift between Na and N(c)a is still λ²/(4π), the time shift is still δT = L, and the tolerance conditions (95)-(97) and the GME estimates are unchanged. The error is local, but Eqs. (91)-(93) should be corrected.","section":"§IV.B, Eqs. (91)-(93)"},{"comment":"The argument for the claim that approximating QFT inescapably implies retrocausal signalling is stated in a way that can mislead. Since the qc-model's reduced state depends only on ∆ab and not on Eab, the non-retrocausality condition ρa(∆ab, Eab) = ρa(∆ab − G̃ba, Eab + G̃ba) fails whenever G̃ba ≠ 0 regardless of whether Eab vanishes; the sentence 'imposing Eab = 0 prevents Eq. (46) from being satisfied' suggests that a nonzero Eab could restore non-retrocausality, which is not the case for the unitary direct-coupling form. The intended point—that a non-retrocausal model would need to depend on the antisymmetric combination in a way the symmetric-propagator structure cannot provide—is correct, but the paragraph should be reformulated for precision.","section":"§III.B, Eq. (46)"}],"minor_comments":[{"comment":"The sentence in the Conclusions stating that retrocausal effects are 'upper bounded by the square of the interaction strength' should be made more precise: the signalling estimator C(r)a = 1/(2π) is O(1) and λ-independent, while it is the phase shift λ²/(4π) that is O(λ²); the bound C(r)a/Ca = L/T is geometric. Please clarify which quantity is being bounded.","section":"§IV.C, paragraph on GME parameters"},{"comment":"The extension of the (3+1) bound to linearized gravity rests on the strong Huygens property (delta-supported retarded propagator, Eq. (62)). The paper makes this assumption at the outset but should state explicitly in Section IV.C that a massive graviton, background curvature, or nonlinearities would reintroduce tails and invalidate the L/T bound, as the (1+1) computation in Eqs. (75)-(80) demonstrates; the current text leaves this as an implicit inference.","section":"§IV.C and §V"},{"comment":"The index structure of the fourth operator term in Eq. (37) appears unbalanced: the term written as ˆjb(a)(x)ˆρab,0ˆja(a)(x′) should presumably involve a b-index on the last current to match the Hermitian structure of the adjacent terms. Please check all index pairings in Eqs. (37)-(38).","section":"§II.B.2, Eq. (37)"},{"comment":"There are several typos and redundancies that should be cleaned up: 'the the setups' (Section I), 'stystem' (Section II.B.3), 'von Newman algebras' (Section II.B.1), 'quantum quantum degrees of freedom' (Section V), 'are are of the order' (Section IV.C), and 'buf for completeness' (footnote 1).","section":"Throughout"}],"recommendation":"minor_revision","confidential_remarks":"The report confirms the reader's assessment: the central integral estimates in Eqs. (66)-(68) and (75)-(79) check out, the nonperturbative phase-shift analysis is consistent at leading order, and the only concrete defect is the local algebraic inconsistency in Eqs. (91)-(93), which does not affect the central bound or the GME conclusion. The paper is a modest but useful extension of the authors' own prior work, Refs. [4,5]; the self-citations are appropriate because the qc-model originates there. The stress-test concern about the strong-Huygens assumption is handled honestly: the paper itself shows via the (1+1) computation that the bound is assumption-dependent, not a general theorem. Scope fits the journal; recommend minor revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read it. This paper makes the retrocausality worry in quantum-controlled models quantitative, and it does so cleanly. The new pieces are the explicit definitions of retrocausality (III.1, III.2), the split of the signalling estimator into causal and retrocausal parts, and the L/T scaling conditions that follow from strong Huygens propagation in 3+1 dimensions. The nonperturbative phase-shift analysis for degenerate detectors is also new and consistent with the perturbative result. It builds on your group's earlier qc-model work, so the novelty is incremental, but it is real: the earlier papers didn't derive a bound on retrocausal effects.\n\nThe derivations hold up. I re-derived the integral estimates in Section IV.A and the phase-shift argument in IV.B; the algebra is consistent. The conclusion that the qc-model always has some retrocausal content when it approximates QFT is derived from Eq. (8) and the symmetric propagator, not assumed. The GME application is the payoff: retrocausal effects in the proposed parameter range are below experimental resolution by a large margin, so the paper reinforces the earlier claim that these experiments won't witness field degrees of freedom.\n\nSoft spots are minor. The stress test flagged an offset in Eq. (91): the causal split should be lambda^2 (T-L)/(2 pi L) rather than lambda^2 T/(2 pi L) if you use the same windows as Eq. (65). This is a genuine slip, but it changes an L/T-relative offset of order 10^-14 and leaves the retrocausal phase lambda^2/4 pi, the time shift delta T = L, and the GME conclusion untouched. Worth a footnote, not a rewrite.\n\nThe load-bearing assumption is strong Huygens propagation for linearized gravity in Minkowski space. That's standard for massless higher-spin fields in flat space, so I don't see it as a flaw, but the paper should say more explicitly that the gravity conclusion lives or dies with that assumption. The 1+1 example shows the bound disappears with tails, so a sentence acknowledging the tail caveat would be honest.\n\nSelf-citation is present but not a problem here; the qc-model is the authors' construction and the GME claim inherits from Ref. [5]. The argument doesn't loop.\n\nWho is this for: people working on GME experiments, effective models of relativistic interactions, and causality in nonlocal models. It deserves a serious referee. I'd accept after minor revision, with the Eq. (91) fix and a more explicit caveat on Huygens.\n\nRecommendation: send to peer review.","headline":"Quantitative retrocausality bounds for the qc-model, internally consistent, with a minor slip that doesn't change the GME conclusion; deserves serious refereeing.","tokens_in":26036,"tokens_out":2821,"would_cite":true,"duration_ms":24790,"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":"Field-free quantum interactions that mimic QFT always carry retrocausal signals, yet today's gravity-entanglement experiments cannot resolve them.","keywords":["quantum-controlled model","retrocausality","signalling estimator","gravity-mediated entanglement","relativistic quantum information","Huygens principle","quantum field theory","Unruh-DeWitt detector"],"falsifier":"Measure the phase of the signalling estimator in a gravity-mediated entanglement experiment with time resolution better than the light-crossing time L between the masses: the qc-model predicts a constant retrocausal phase shift of order \\(\\$lambda^{2}$/(4\\pi)\\), which a local QFT description does not produce.","tokens_in":24931,"feed_emoji":"⏳","tokens_out":9005,"duration_ms":73197,"temperature":0.7,"pith_summary":"The paper analyzes the quantum-controlled (qc) model, an effective description in which two quantum systems interact through a retarded relativistic potential but the mediating field has no quantum degrees of freedom of its own. Comparing this model to a full quantum field theory (QFT) description, the authors show that the qc-model can approximate QFT only in regimes where it also predicts retrocausal signalling: a system can be affected by another system that lies in its causal future. They quantify this effect with a signalling estimator and find that in 3+1 dimensions with a massless field the retrocausal part is a fraction L/T of the total signal, vanishing for long interactions, while in 1+1 dimensions it can dominate. Applied to gravity-mediated entanglement proposals, the numbers yield retrocausal phase shifts far below experimental resolution, so those experiments cannot distinguish the qc-model from QFT and cannot, on these grounds, claim to witness the quantum degrees of freedom of gravity.","feed_headline":"Retrocausal signals always accompany field-free quantum models","feed_subtitle":"In proposed gravity-entanglement tests the effect stays below detection, so quantum gravity remains unwitnessed.","key_machinery":"The machinery is the decomposition of the qc-model's evolution in terms of the symmetric propagator \\(\\$\\Delta$(x, x') = G_R(x, x') + G_A(x, x')\\), the sum of retarded and advanced Green's functions, which replaces the field's quantum degrees of freedom. The signalling estimator \\(C_a = \\$\\Delta$(\\Lambda_a, \\Lambda_b)\\) splits into causal and retrocausal parts, and the strong Huygens principle in 3+1 dimensions — the retarded propagator is a delta-function on the light cone — is what keeps the retrocausal part bounded by \\(L/T\\). The trade-off is encoded in the identity relating the symmetric and causal propagators: the qc-model discards the antisymmetric combination \\(E = G_R - G_A\\), and the paper proves that discarding \\(E\\) while retaining \\(\\$\\Delta$\\) necessarily permits advanced propagation.","core_discovery":"The central claim is that the qc-model inevitably contains retrocausal contributions whenever one system has support in the causal future of the other, and these contributions are inseparable from the model's ability to match QFT: approximating QFT forces the evolution to depend on the symmetric propagator \\(\\$\\Delta$ = G_R + G_A\\), which carries advanced (future-to-past) information. The paper shows quantitatively that for massless fields in 3+1 dimensions, where the retarded propagator is delta-supported on the future light cone, the retrocausal part of the signalling estimator is \\(L/T\\) of the total for interaction times \\(T > 2L\\), and the absolute retrocausal contribution is bounded by the squared coupling. For the proposed gravity-mediated entanglement parameters (\\(L \\sim $10^{{-6}}$\\) m, \\(T \\sim 1\\) s, \\(\\$lambda^{2}$ \\sim $10^{{-14}}$\\)), this is \\($10^{{-14}}$\\) of the signal, requiring a time resolution of order the light-crossing time to observe. The conclusion is that current GME proposals are experimentally indistinguishable from a field-free qc-description, so they do not yet access quantum aspects of the gravitational interaction through retrocausality.","pith_inferences":["A future GME experiment with time resolution near the light-crossing time could convert the retrocausal phase shift into an unambiguous discriminator between a field-free qc-description and a local QFT description, a test the paper's bounds make concrete.","The same trade-off between matching QFT and introducing advanced-propagation artefacts likely applies to any effective model that truncates mediator degrees of freedom, not just the specific qc-construction studied here.","Since the qc-model cannot reproduce the Hadamard (noise) contribution of the quantum field, measuring local decoherence of the sources — rather than their mutual entanglement — may be a cleaner experimental route to detect the field's quantum degrees of freedom in regimes where retrocausality is negligible.","The conclusions about GME experiments rest on gravity behaving like a massless higher-spin field with no tails; computing the same estimators for massive-graviton or non-linear corrections would test whether the retrocausal bound survives more realistic gravitational dynamics."],"forward_implications":["Any qc-model that faithfully approximates QFT during a finite interaction is necessarily retrocausal, so fully causal and fully field-free relativistic models cannot both match QFT.","In 3+1 dimensions, retrocausal effects in qc-interactions are suppressed by the factor \\(L/T\\), so long interaction times make the model effectively causal even though it is not exactly causal.","In settings where the strong Huygens principle fails, such as 1+1-dimensional massless fields or massive mediators, the retrocausal signal is not bounded and can dominate the causal signal.","With the parameters of current gravity-mediated entanglement proposals, the retrocausal contribution is \\(\\lambda^2 \\sim 10^{-14}\\) of the signal, so the qc-model and the full QFT description are experimentally indistinguishable in those experiments.","Detecting the retrocausal signature would require resolving relative state variations of order \\(10^{-14}\\) or time intervals of order the light-crossing time (\\(10^{-14}\\) s for the proposed setups), far beyond current experimental capabilities."],"supporting_citations":[{"why":"Introduces the qc-model and identifies the long-interaction, causal-contact regime where it approximates QFT; the present paper builds on this regime analysis.","marker":"[4]"},{"why":"Shows that the proposed GME experiments can be modelled by a qc-interaction; the present paper applies its retrocausality bounds to that same context.","marker":"[5]"},{"why":"Provides the reference parameters for gravity-mediated entanglement experiments used to estimate the observability of retrocausal effects.","marker":"[6]"},{"why":"The companion GME proposal that the paper's quantitative bounds are checked against.","marker":"[7]"},{"why":"Raises the retrocausality objection to using GME to infer quantum gravity that this paper quantifies and addresses.","marker":"[17]"},{"why":"Supplies the signalling estimator method used to split causal and retrocausal contributions.","marker":"[51]"},{"why":"Establishes the strong Huygens principle for wave propagation that underlies the (3+1)-dimensional bound.","marker":"[49]"},{"why":"Provides the nonperturbative solution for gapless detectors used in the exact analysis of retrocausal phase shifts.","marker":"[18]"}],"fun_headline_variants":["Field-free quantum models always include retrocausality","Gravity-entanglement tests cannot see retrocausal effects","Retrocausality unavoidable in no-mediator quantum models","Quantum gravity stays hidden: retrocausal signal too weak","Model without mediators yields retrocausality, but too small"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The bound on retrocausality in 3+1 dimensions assumes the mediating field obeys the strong Huygens principle, so the retarded propagator is concentrated exactly on the future light cone; if the field has tails, retrocausal effects are no longer bounded, as the 1+1-dimensional case shows.","fun_headline_variants_meta":{"raw":{"variants":["Field-free quantum models always include retrocausality","Gravity-entanglement tests cannot see retrocausal effects","Retrocausality unavoidable in no-mediator quantum models","Quantum gravity stays hidden: retrocausal signal too weak","Model without mediators yields retrocausality, but too small"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000232,"raw_usage":{"total_tokens":1472,"prompt_tokens":912,"completion_tokens":560,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":528,"completion_tokens_details":{"reasoning_tokens":478}},"tokens_in":528,"tokens_out":560,"duration_ms":5243,"temperature":1.0,"reasoning_tokens":478,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T10:46:39.424577+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the phase of the signalling estimator in a gravity-mediated entanglement experiment with time resolution better than the light-crossing time L between the masses: the qc-model predicts a constant retrocausal phase shift of order \\(\\$lambda^{2}$/(4\\pi)\\), which a local QFT description does not produce.","supporting_citations":[{"cited_title":"In analogy to the classical interaction (5), we prescribe the interaction Hamiltonian as ˆHint(t) = λ Z d3x ˆja (a)(x) ˆϕ(a)(x) + ˆjb (a)(x) ˆϕ(a)(x)","cited_arxiv_id":null,"evidence_quote":"Introduces the qc-model and identifies the long-interaction, causal-contact regime where it approximates QFT; the present paper builds on this regime analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows that the proposed GME experiments can be modelled by a qc-interaction; the present paper applies its retrocausality bounds to that same context."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the reference parameters for gravity-mediated entanglement experiments used to estimate the observability of retrocausal effects."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The companion GME proposal that the paper's quantitative bounds are checked against."},{"cited_title":"Mart ´ ın-Mart ´ ınez and T","cited_arxiv_id":null,"evidence_quote":"Raises the retrocausality objection to using GME to infer quantum gravity that this paper quantifies and addresses."}],"review_version":1}