{"id":"3d935d56-f991-4374-b805-f10cf922c3be","arxiv_id":"2510.19714","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In quantum field theory, a classical gravitational potential can entangle two superposed masses through virtual matter exchange, so gravitationally induced entanglement is not by itself proof of quantum gravity.","lead":"This paper argues that a classical, non-quantized gravitational interaction can generate entanglement between two masses in the Feynman/Bose-Marletto-Vedral experiment, through virtual matter exchange rather than graviton exchange. Its fourth-order calculation shows the classical effect scales differently with mass and time, so observing entanglement alone would not be unambiguous evidence for quantum gravity.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Unquantified number-transfer sectors may invalidate the spin-entanglement witness used to demonstrate the central claim.","rationale":"The reader's weakest assumption identifies the same point I would stress. The paper's only bridge from the formal β^(4)_RL amplitude to 'classical gravity produces entanglement' is the spin witness after recombination; that witness is computed from the four α_ij amplitudes in the |N,N⟩ sector. The interaction also drives transitions out of that sector, and footnote 39 explicitly asserts they are unobserved rather than proving they are irrelevant. A charitable reading is that the experiment postselects on the |N,N⟩ port, and since a local projection cannot create entanglement from a separable state, a nonseparable conditional spin state would still imply the pre-measurement state was entangled. But the paper does not give the conditional probabilities, the k≠0 amplitudes, or the spatial separation of the ports, so the claimed quantitative result ϑ/φ and the 'entanglement alone cannot evidence quantum gravity' conclusion are not yet established. This supports the reader's CONDITIONAL verdict; it is not a rejection because the mechanism may well survive once the sector analysis is done.","tokens_in":48932,"tokens_out":28150,"duration_ms":270374,"concrete_test":"Compute the k=±1 number-transfer amplitudes at fourth order using the same nonrelativistic contractions as Eq (54), and evaluate the overlap of the resulting |N±1,N∓1⟩ spatial components with the recombined |N,N⟩ port after the proposed Stern-Gerlach reversal. Then construct the full reduced spin density matrix (tracing all number sectors) and recompute the entanglement witness. If the non-N sectors are non-negligible or not orthogonal to the witness port, the Section 5.2 conclusion fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is demonstrated by computing the four survival amplitudes α_ij (Eq 19) and asserting that a branch-dependent fourth-order amplitude β^(4)_RL entangles the spin state (Eq 18). However, the same interaction Hamiltonian (43) at fourth order also produces states |N±k⟩1i|N∓k⟩2j with changed particle numbers. Footnote 39 dismisses these as 'not seen in the experiment' because of different Stern-Gerlach deflections, but no amplitude or overlap estimate is supplied. The spin witness is therefore conditioned on an uncharacterized postselection: if the |N±k,N∓k⟩ components are not cleanly separated from the |N,N⟩ port—or if the measurement is not a simple projection onto the |N,N⟩ sector—the reduced spin state is not the pure state in Eq (18), and the computed entanglement can disappear. This is load-bearing because it is the only place where the paper turns the formal β amplitude into an observable prediction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper challenges the consensus that a classical gravitational interaction cannot entangle two quantum matter systems. Its central claim is that when matter is treated at QFT level, a classical (unquantized) gravitational interaction permits quantum communication through virtual matter, and this can generate entanglement in a Feynman-type experiment. The authors compute the fourth-order Dyson amplitude for two N00N-state mass distributions under semiclassical Einstein gravity and obtain β^(4)_RL ≈ [(6/25) G² m² M³ R t/(ℏ³ d_RL)]², concluding that for Planck-scale masses the observation of entanglement alone cannot be taken as evidence of quantum gravity. The paper includes extensive appendices on relativistic, momentum-space, stochastic, and boundary treatments.","tokens_in":49152,"tokens_out":7521,"duration_ms":75493,"significance":"If the technical concerns below are resolved, this is an important result. It provides a concrete, quantitative mechanism by which a local classical gravity model — semiclassical Einstein gravity — can produce gravitationally mediated entanglement, thereby sharpening the conditions under which entanglement-based quantum-gravity witnesses are valid. The strength of the paper is its explicit QFT calculation: the Dyson-series derivation is detailed and cross-checked in momentum space (Appendix B.3), a quantum-gravity counterpart is computed (Appendix B.6), and the final amplitude gives a falsifiable mass/time/distance scaling. The main weaknesses are the inconsistent sign convention in the central amplitude and an unquantified postselection assumption connecting the computed amplitude to the observable spin entanglement.","major_comments":[{"comment":"There is an internal inconsistency in the definition of ϑ. Eq. (79) states β^(4)_ij = ( (6/25) i G² m² M³ R t / (ℏ³ d_ij) )². Since i² = −1, this is a negative real number, not a pure phase or positive parameter. Eq. (80) then sets β^(4)_RL =: ϑ and writes √ϑ = (6/25) G² m² M³ R t/(ℏ³ d_RL), treating ϑ as a positive real number. This is not a harmless notational slip: the sign and imaginary unit determine whether the branch amplitude α_RL is 1 + ϑ, 1 − ϑ, or 1 + i times a real quantity, which affects the entanglement witness and the ϑ/φ comparison in Fig. 6. The authors should either define ϑ := |β^(4)_RL| and adjust the text, or restore the i consistently and explain how a negative real branch amplitude enters the final state.","section":"Section 5.2, Eqs. (79)–(80), Fig. 6"},{"comment":"The conversion of the computed amplitude β into an observable spin-entanglement witness assumes that the post-measurement state is the pure state (18) with support only on |N⟩_{1i}|N⟩_{2j}. However, the same interaction Hamiltonian (43) at fourth order also produces number-transfer sectors |N+k⟩_{1i}|N−k⟩_{2j} and |N−k⟩_{1i}|N+k⟩_{2j}. Footnote 39 dismisses these as 'not seen in the experiment' because of different Stern-Gerlach deflections, but no amplitude, overlap, or postselection error estimate is supplied. If these components are not cleanly separated from the |N,N⟩ port, or if they are traced out, the reduced spin state is not generically the pure entangled state used in the witness, and the claimed entanglement can disappear. This is load-bearing: it is the only step that turns the formal fourth-order amplitude into a concrete experimental prediction. Please quantify the discarde","section":"Footnote 39; Section 5.2, Eqs. (17)–(18)"},{"comment":"The title and abstract state that 'classical theories of gravity' produce entanglement, but the detailed calculation in Section 5.2 is performed for one specific sourcing prescription: semiclassical Einstein gravity with the Newtonian potential given by the expectation value (63). Appendix B.10 discusses stochastic variants, and Appendix D reviews consistency issues, but the universal claim is broader than the demonstrated case. If the paper is intended as a counterexample to the claim that no local classical gravity can entangle, this should be stated explicitly; if it is intended as a general theorem, additional argument is needed to show that the virtual-matter mechanism survives across the full class of classical theories.","section":"Abstract and Section 6"}],"minor_comments":[{"comment":"The caption reads 'The ratio ϑ/φ ... is identical to ϑ/φ.' The second occurrence is presumably meant to be something else (e.g., the ratio of quantum and classical effects).","section":"Fig. 6 caption"},{"comment":"The notation 'κj≠λj' for non-overlapping spheres is confusing; it should be 'κ≠λ' or 'i≠j' depending on which indices label the objects and branches.","section":"Around Eq. (16)"},{"comment":"The text says 'Note that this preprint has not undergone peer review. The Version of Record of this article is published in Nature'. These two statements are inconsistent in a submission context. If this is a reprint of a published article, that should be disclosed; if it is a fresh submission, the preprint boilerplate should be removed.","section":"Acknowledgements"}],"recommendation":"major_revision","confidential_remarks":"The manuscript appears to be the published Nature article or a postprint of it; the acknowledgements state that the Version of Record is in Nature. If this is being considered as a new submission, the editor should clarify the status and ensure the authors disclose prior publication. The main technical concerns are the sign inconsistency in Eq. (79) and the unquantified number-transfer postselection in footnote 39; both are fixable but are currently load-bearing for the central claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is the most credible version I have seen of the claim that a local, classical gravitational interaction can entangle two quantum masses. The mechanism is distinct from earlier proposals: instead of non-local sourcing or stochastic fluctuations, the entangling channel is a virtual matter propagator that appears once matter is treated as a QFT and gravity is treated as a classical background. The fourth-order Dyson calculation is carried out in detail, with appendices covering the relativistic form, momentum-space derivation, and a check that turning off gravity kills the effect. That is a real piece of work, and it earns attention.\n\nThe scaling result is the useful part. If this calculation holds, the classical effect grows differently in mass and time than the standard quantum-gravity phase, so a BMV-style experiment can be designed to stay below the ϑ threshold. Experimentalists planning those runs should read this.\n\nNow the soft spots. First, the sign/imaginary-unit inconsistency around Eq. (79) is real: the amplitude is written as (iA)², which is negative real, but then it is set equal to a positive ϑ whose square root is taken. This is likely a notational slip, but it matters because the whole entanglement argument hinges on the relative value of the branch amplitudes. A referee should make the authors fix it before citing the result.\n\nSecond, and more importantly, the paper only computes survival amplitudes back into the |N,N⟩ sector. It dismisses the |N−k,N+k⟩ and |N+k,N−k⟩ components in footnote 39 with a one-sentence hand-wave about Stern-Gerlach deflection. The stress-test on this is fair: those components are produced by the same interaction at the same order, and if they carry branch information, the reduced spin state after postselecting on |N,N⟩ is not the pure state in Eq. (18). The paper gives no estimate of how suppressed they are. This is not a fatal flaw in the central idea, but it is a genuine gap between the formal amplitude and the observable prediction. It needs either a quantitative estimate or a redesigned measurement that tracks the number sectors.\n\nThird, the title overclaims. What is shown is that one family of classical-gravity models—semiclassical sourcing—generates this effect. The paper argues the mechanism is generic, and the stochastic case is treated qualitatively, but the generality is asserted rather than proven. For a paper with this title, that should be stated as a conjecture, not a conclusion.\n\nOverall: the virtual-matter channel is new, the calculation is substantial, and the scaling discriminator is useful. The two technical issues are reparable, but they are load-bearing for the experimental claim. I would not desk-reject this. It deserves a serious referee and probably a revised version.\n\nWho is this for? Quantum foundations and quantum-optics experimentalists working on gravity-mediated entanglement. I would bring it to a reading group and cite it if I were writing about BMV constraints.","headline":"A plausible but not airtight loophole in the BMV entanglement witness: the virtual-matter calculation is serious, but two unaddressed technical points keep the central claim conditional.","tokens_in":49602,"tokens_out":5308,"would_cite":true,"duration_ms":53970,"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":"A classical gravitational interaction can entangle two quantum masses by exchanging virtual matter particles, so at Planck-scale masses entanglement alone no longer proves gravity is quantum.","keywords":["classical gravity","quantum entanglement","virtual particles","quantum communication","gravitationally induced entanglement","semiclassical gravity","LOCC","quantum field theory"],"falsifier":"Recompute the final state without truncating to N-particle sectors, or measure spin negativity in a Planck-mass configuration where the predicted classical amplitude is about 0.1: if the number-transfer components destroy the spin witness, or no entanglement appears despite the predicted amplitude, the central claim is falsified.","tokens_in":48795,"feed_emoji":"⚛️","tokens_out":7042,"duration_ms":70256,"temperature":0.7,"pith_summary":"The paper argues that the standard no-go result—classical gravity can only act through local operations and classical communication, hence cannot entangle—misses a process available when matter is treated as a quantum field. Even with a purely classical gravitational field that is identical in every branch of the superposition, the interaction Hamiltonian contains vertices that allow virtual matter particles to travel between the two masses. At fourth order in perturbation theory these virtual-matter exchanges yield a branch-dependent amplitude that creates spin entanglement. The effect scales differently from the quantum-gravity phase, so for low masses near-future experiments are unaffected, but for Planck-scale masses entanglement can no longer be treated as unambiguous evidence of quantum gravity.","feed_headline":"Classical gravity can entangle masses via virtual particles","feed_subtitle":"At Planck-scale masses, seeing entanglement would no longer prove gravity is quantum.","key_machinery":"The central object is the fourth-order perturbative amplitude for virtual-matter exchange under a classical gravitational interaction, built from: (i) N-particle non-relativistic wavepacket states localized in spheres of radius R; (ii) a classical potential sourced by the expectation value of the matter energy-momentum tensor, hence identical in all superposition branches; and (iii) Feynman propagators of a massive complex scalar field connecting the two objects. The key identity is β^(4)_ij proportional to [∫ dx dy Φ(x)Φ(y) θ_{1i}(x)θ_{2j}(y)/|x−y|]^2, whose d_ij^{-2} branch dependence generates the entanglement. It does this by letting virtual matter particles traverse different distances","core_discovery":"In a version of the two-superposed-masses experiment, with each mass described by an N-particle quantum-field-theory wavepacket and gravity by a classical gravitational potential that is the same in every branch, the leading entangling process is not the classical analogue of the graviton diagram but a fourth-order diagram in which the two objects exchange two virtual matter particles. The amplitude for branch (i,j) is β^(4)_ij ≈ (6/25 i G^2 m^2 M^3 R t / (ħ^3 d_ij))^2. Because the virtual matter propagator involves |x−y|, the distance between the spherical wavepackets depends on the branch, so β differs across branches even though the classical potential does not; this branch-dependence is","pith_inferences":["The same fourth-order virtual-matter mechanism should operate for any classical field coupled to quantum matter—for instance a classical electromagnetic potential—so the ability of a classical field to entangle is not peculiar to gravity; tabletop experiments may need to control classical electromagnetic backgrounds at the same perturbative order.","The material dependence of the classical amplitude (it scales with the constituent particle mass and radius) versus the quantum-gravity phase's independence of those parameters offers a direct experimental discriminator: changing the material at fixed total mass should change the classical contribution but not the quantum one.","The calculation truncates to fixed-N sectors; including the |N±k⟩|N∓k⟩ number-transfer channels could either introduce decoherence that suppresses the spin entanglement or open additional witnesses, so a full calculation is needed before Planck-scale experiments are designed.","If the effect is real, the experimental question shifts from 'does entanglement occur?' to 'does the entanglement amplitude scale as predicted?', making classical-sourcing models directly testable through the predicted dependence on mass, radius, time, and separation."],"forward_implications":["For the near-term low-mass parameter regime (M≈10^-14 kg, t≈2 s), the classical-gravity entanglement amplitude is far below the quantum-gravity phase, so observing entanglement there retains its status as evidence for quantum effects in gravity.","At masses approaching the Planck mass and above, the classical amplitude reaches order 0.1 even at short times, so entanglement observation in that regime would not discriminate classical from quantum gravity.","The two amplitudes scale differently with mass, radius, and distance, so scanning these parameters can separate the classical-gravity contribution from the quantum-gravity contribution in a single experiment.","The LOCC-based theorems are not generic: classical gravity models with quantum-field-theoretic matter are not restricted to classical communication, because the matter sector itself provides a quantum channel.","In stochastic classical-gravity models with fundamental decoherence, the virtual-matter entanglement can still dominate when its generation rate exceeds the gravitational decoherence rate, so entanglement is not unique to semiclassical sourcing."],"fun_headline_variants":["Classical gravity can entangle via virtual matter swap","Entanglement from classical gravity: virtual particles route","Gravity's classical field still yields entanglement","Virtual particles let classical gravity entangle masses","Classical gravity entangles—no quantum gravity needed"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The paper's spin-entanglement witness computes amplitudes from states that keep the particle number N fixed in each object; if the states |N±k⟩|N∓k⟩ that the same interaction creates are not truly negligible or are not traced out, they could carry which-branch information and spoil the predicted spin entanglement.","fun_headline_variants_meta":{"raw":{"variants":["Classical gravity can entangle via virtual matter swap","Entanglement from classical gravity: virtual particles route","Gravity's classical field still yields entanglement","Virtual particles let classical gravity entangle masses","Classical gravity entangles—no quantum gravity needed"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000135,"raw_usage":{"total_tokens":954,"prompt_tokens":690,"completion_tokens":264,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":434,"completion_tokens_details":{"reasoning_tokens":203}},"tokens_in":434,"tokens_out":264,"duration_ms":3194,"temperature":1.0,"reasoning_tokens":203,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T08:38:02.214945+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the final state without truncating to N-particle sectors, or measure spin negativity in a Planck-mass configuration where the predicted classical amplitude is about 0.1: if the number-transfer components destroy the spin witness, or no entanglement appears despite the predicted amplitude, the central claim is falsified.","supporting_citations":[],"review_version":1}