{"id":"9d8441be-f15b-4e0f-800e-cee7dff1ff34","arxiv_id":"2504.18111","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A teleportation-based interferometric scheme turns a position meter into a speed meter, beating the standard quantum limit in both real-time and post-processed implementations.","lead":"Researchers propose using quantum teleportation to build a speed meter, a device that measures how fast a mirror moves while canceling the push of the measuring laser. This could improve gravitational wave detectors, which today are limited by exactly this back-action noise.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The lossy equivalence of the offline approach is asserted without derivation; if it fails, the claim that both implementations beat the SQL with losses is unsupported.","rationale":"The paper's central claim is conditional on the offline and online implementations remaining equivalent after losses are introduced. The lossless equivalence is demonstrated (Eqs. 27-34), but the lossy case is exactly where equivalence is nontrivial because loss is a dissipative beamsplitter that does not commute with the feedforward displacement: online injects Bell-measurement noise into the interferometer, offline does not, yet the final estimators are asserted to coincide. Since both the abstract and Discussion promise 'both implementations ... even in the presence of losses', and since the offline path is the practically preferred one (online displacement may add loss, per Discussion), the missing derivation is load-bearing. I do not see an internal inconsistency in the lossless derivation; the ideal speed-meter response seems correctly reproduced. The concern is not that the result disagrees with consensus, but that a needed calculation is absent. The proposed test, recomputing the offline lossy sensitivity with optimized Wiener filters, would settle the question directly. If the test reproduces the online lossy curve for the offline case, the concern is resolved and the manuscript's claim stands; if not, the claim must be restricted to the online case or adjusted. This does not change the reader's CONDITIONAL verdict, because the reader already required exactly this check before full acceptance.","tokens_in":11806,"tokens_out":11885,"duration_ms":135995,"concrete_test":"Derive the offline lossy sensitivity from first principles: include arm loss γ2 in Eqs. (38), input/output losses ϵin and ϵout in Eqs. (39)-(41), finite EPR squeezing r, and the lossy Bell measurement; express b2,out and the Bell outcomes as linear functions of all vacuum inputs; optimize the Wiener filters g1(Ω) and g2(Ω) to minimize S_x; compare the resulting offline curve with the online lossy curve in Fig. 6 at 8 Hz and 50 Hz. If the optimized offline S_x differs from the online value by more than a small numerical tolerance, or crosses the SQL where the online curve does not, the asserted equivalence fails. A simpler first check is to apply the same loss matrices to Eq. (27) and verify whether the online-optimal filter form in Eqs. (33)-(34) still achieves the claimed offline sensitivity.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is in 'Loss analysis': after introducing losses, the paper states 'we focus on the online approach, but we confirmed that the results are equivalent for the offline case' without showing the algebra. This matters because the central claim explicitly covers both implementations ('both ... even in the presence of losses'), and the Discussion motivates the offline method as more practical because the online displacement operation may add losses. In the lossless case, equivalence is established by choosing Wiener filters (Eqs. 33-34). With losses this does not follow automatically: in the online scheme the Bell-measurement outcomes are coherently added to Bob's input before the loss ports of Eqs. (38)-(41), so teleportation noise is injected into the interferometer; in the offline scheme the same outcomes are used only as classical post-processing weights. Loss is a beamsplitter operation and does not commute with this feedforward, so the optimal filters and the resulting S_x(Ω) need not coincide. Without the missing derivation, the SQL-beating claim for the offline path, and hence the 'both implementations' claim, is unverified. The Discussion's own caveat that the simulation omits displacement-operation losses further weakens the quantitative loss analysis for the online path.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a continuous-variable teleportation protocol as a way to realize a speed meter for interferometric displacement sensing. It models a non-reciprocal two-mode optomechanical system in which the probe field interacts with the test mass twice with opposite radiation-pressure signs, and it shows that this can be implemented either online, by physically displacing Bob's field using the Bell-measurement outcomes, or offline, by post-processing the homodyne output with Wiener filters. Using standard input-output theory, the authors derive transfer matrices, the displacement noise spectral density, the optimal homodyne angle, and the offline filtering conditions, and they show that in the lossless case the sensitivity agrees with the known speed-meter result of Ref. [25]. They then add input, output, and arm-cavity losses, present numerical sensitivity curves with 1% losses and 15 dB of generated squeezing, and sketch an application to a LIGO-type gravitational-wave detector.","tokens_in":12114,"tokens_out":13686,"duration_ms":144550,"significance":"If the central claim holds, the scheme provides a conceptually new route to back-action evasion in interferometers: a teleportation-based speed meter that does not require changing the core optics or mirror coatings. The offline version is particularly attractive because it moves the displacement operation into post-processing. The paper is built on standard input-output methods, gives explicit transfer matrices, and makes a clear, falsifiable prediction of sub-SQL sensitivity in a specified loss and squeezing parameter regime. The agreement with the earlier speed-meter sensitivity of Ref. [25] in the lossless limit is a useful consistency check. The main weakness is that the lossy off-line equivalence, which is needed for the paper's advertised 'both implementations' claim, is asserted rather than demonstrated.","major_comments":[{"comment":"After Eq. (41), the statement 'we focus on the online approach, but we confirmed that the results are equivalent for the offline case' is unsupported. This is a load-bearing point because the abstract and the Discussion claim that both implementations beat the SQL even in the presence of losses. In the online scheme, the Bell-measurement outcomes are inserted through the displacement operation before the loss ports of Eqs. (38)-(41), whereas in the offline scheme those same outcomes are used only as classical post-processing weights on the final homodyne output. Losses are beamsplitter operations and do not trivially commute with this feedforward, so the optimal Wiener filters and the resulting S_x(Omega) need not coincide. Please provide the missing derivation for the lossy offline case, including the noise covariance of the Bell-measurement data, or explicitly restrict the lossy sub-SQL claim to the online implementation.","section":"Loss analysis"},{"comment":"The Discussion admits that the simulation 'omits some experimental details of the teleportation procedure' and that the online displacement operation 'may introduce additional losses.' This displacement operation is precisely the element that distinguishes the online approach, so the loss model in Eqs. (38)-(41) does not fully cover the online implementation as described. The quantitative statement that the online approach surpasses the SQL with losses is therefore stronger than what the simulation demonstrates. Please include a quantitative model of the displacement-operation loss, or state more cautiously that the sub-SQL result applies to the modeled input, arm, and output losses only.","section":"Discussion"},{"comment":"Eqs. (10)-(12) contain an apparent sign inconsistency in the optomechanical coupling terms. Equation (10) gives opposite signs for the A and B coupling terms, while Eq. (12) gives the same sign for the displacement terms in the A2 and B2 equations. Since the cancellation of radiation-pressure back action is the central mechanism of the speed meter, please reconcile the Hamiltonian, the Heisenberg equations, and the quadrature equations, or state explicitly the convention that makes the printed signs consistent.","section":"State Preparation and System Dynamics"}],"minor_comments":[{"comment":"In the fourth line of Eq. (38), the loss-induced vacuum term should be b'_2, not b'_1, which appears to be a typographical error in the B2 quadrature equation.","section":"Loss analysis, Eq. (38)"},{"comment":"The text refers to 'Fig. 1b' and to panels (b) and (c) for the online and offline approaches, but the caption lists only panels (a), (c), and (d), with online and offline described in (c) and (d). Please make the panel labels consistent between the text and the figure.","section":"Figure 1"},{"comment":"There is a typo in the speed-meter description: 'the probe laser is coupled the the object mass twice' should read 'coupled to the object mass twice.'","section":"Introduction"},{"comment":"The paper references supplementary information I and II for derivations of the position-meter noise and the loss calculation, but the supplementary material is not included in the arXiv version. Please ensure it is part of the submitted manuscript and is clearly linked from the text.","section":"General"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe genuinely new thing here is the offline teleportation-based speed meter: you can cancel back-action in data analysis instead of with a physical frequency converter. That is real, and it is the reason to read the paper. The ideal (lossless) derivation is straightforward input-output theory, and the recovered speed-meter sensitivity agrees with earlier results, which is a good consistency check. The GW application is also sensible: dark-port modification only, no new mirror coatings.\n\nSoft spots, in order of importance. The paper claims both online and offline implementations beat the SQL with losses, but the loss analysis only shows the online case. The sentence \"we confirmed that the results are equivalent for the offline case\" is not a derivation. In the online case the Bell outcomes are coherently added to Bob's input before the loss ports; in the offline case they are used as post-processing weights. Losses are beamsplitters, and there is no reason the optimal filters and noise spectrum must coincide. The stress-test note makes this point correctly. If the equivalence fails, the \"both implementations\" claim is unsupported, though the offline scheme might actually be more robust because it avoids injecting teleportation noise. This needs to be proven or the claim softened.\n\nThe second soft spot is the jump from teleportation to the non-reciprocal Hamiltonian in Eq. (10). The paper argues it by saying \"with perfect fidelity this is equivalent,\" which is plausible but not shown from the teleportation protocol. In the lossless limit the online approach with infinite squeezing effectively realizes it, so I am not too worried, but a referee should ask for a derivation or a reference.\n\nMinor issues: the figure callouts are inconsistent (text refers to panels (b) and (c), caption has (c) and (d)); Eq. (38) appears to have a typo in the last line (b'_1 should be b'_2). None of these affect the physics.\n\nWho is this for? People working on speed meters, quantum back-action evasion, and GW detector configurations. It is a solid methods paper with a clearly stated novel element. The central ideal-result holds up; the lossy claim for offline needs work. I would send it to peer review with a request for a full loss derivation for the offline case.","headline":"The offline teleportation speed meter is a genuinely new idea, but the paper's claim that both implementations beat the SQL with losses rests on an unproven equivalence; the ideal result is sound, so it deserves a careful referee rather than a desk reject.","tokens_in":12541,"tokens_out":2018,"would_cite":true,"duration_ms":20638,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["03.67.-a","04.80.Nn","42.50.Lc"],"model":"deepseek-v4-flash","headline":"A teleportation-based speed meter can beat the standard quantum limit in interferometric displacement sensing.","keywords":["quantum teleportation","speed meter","standard quantum limit","radiation pressure noise","back-action evasion","gravitational-wave detector","EPR entanglement","optomechanics"],"falsifier":"Run the offline scheme on a tabletop optomechanical setup with the optimal fixed homodyne angle, 15 dB of generated two-mode squeezing, and 1% input and output losses, and compare the measured displacement-noise spectrum to Eq. (26); a curve that fails to cross the standard quantum limit in the band where $K_{\\rm sm}$ is flat would disprove the central claim. A simpler calculation would derive the full input-output map of the teleportation chain including Bell-measurement noise and check whether the back-action terms cancel before replacing the chain by Eq. (10).","tokens_in":11656,"feed_emoji":"🔭","tokens_out":10098,"duration_ms":92426,"temperature":0.7,"pith_summary":"This paper proposes turning a conventional interferometric position meter into a quantum speed meter with continuous-variable teleportation. The probe light that first reflects from a test mass is teleported back to strike it from the opposite side, using an entangled pair and a Bell measurement, so the two radiation-pressure kicks cancel and the device effectively reads the mass's speed rather than its position. Two versions are analyzed: an online one that displaces the returning field in real time, and an offline one that performs the same cancellation by filtering the recorded data afterward. The authors claim both reach sensitivities below the standard quantum limit, the floor set by the trade-off between measurement precision and back-action noise, and that the offline version keeps this advantage with 1% input/output losses, a 30 ppm arm-cavity loss, and 15 dB of generated squeezing. Because the core interferometer optics are untouched, the scheme offers gravitational-wave detectors a practical route to back-action evasion.","feed_headline":"Teleportation-based speed meter beats the standard quantum limit","feed_subtitle":"Both live and post-processed versions cut radiation-pressure noise below the SQL without changing core optics.","key_machinery":"The load-bearing object is the nonreciprocal Hamiltonian of Eq. (10), in which two cavity modes $\\hat{A}$ and $\\hat{B}$ both couple to $\\hat{x}$ but with opposite signs, so their radiation-pressure forces subtract instead of adding. The teleportation chain implements this nonreciprocity: an entangled pair of optical modes, a Bell measurement on the reflected probe and one partner mode, and a displacement of the other partner mode by the measurement outcomes $\\{x_-,p_+\\}$ turns the second interaction into a back-action-canceling one. In the offline version the displacement is replaced by optimal filtering of the recorded quadratures, using the relation $|K_z|^2=K_{\\rm sm}K_{\\rm pm}$. The key performance parameter is $K_{\\rm sm}=16\\hbar\\omega_a^2\\gamma/[m(\\gamma^2+\\Omega^2)^2]$, which is independent of frequency for $\\Omega\\ll\\gamma$; this flatness is what lets a fixed homodyne angle $\\phi_{\\rm opt}=\\mathrm{arccot}(K_{\\rm sm}(0))$ beat the SQL without extra filter cavities.","core_discovery":"The central discovery is that teleportation supplies the non-reciprocal coupling a speed meter requires: two cavity modes $\\hat{A}$ and $\\hat{B}$ couple to the same mechanical position $\\hat{x}$ with opposite signs in the Hamiltonian, so the second light–mirror interaction cancels the radiation pressure of the first and the measured phase becomes proportional to $\\hat{x}(t+\\tau)-\\hat{x}(t)\\sim\\tau\\bar{v}$. With perfect teleportation the process is exactly the textbook speed meter; with finite squeezing the residual noise enters through a factor $e^{-2r}$, and the optimum homodyne sensitivity takes the form $S^{\\rm sm}_{x,\\pi/2}=(x_{\\rm SQL}^2/2)(1/K_{\\rm sm}+K_{\\rm sm})$, which lies below the standard quantum limit because $K_{\\rm sm}$ is flat at low frequencies while the position-meter coupling $K_{\\rm pm}\\propto\\Omega^{-2}$ grows. The offline implementation recovers the same sensitivity by combining the recorded quadrature with the Bell-measurement outcomes through the optimal filters $g_1=e^{2i\\beta}$ and $g_2=-K_z^{*}e^{2i\\beta}$. Losses are modeled as extra vacuum modes entering the cavity and beamsplitter losses at the input and output, and the paper claims the sub-SQL enhancement survives those losses for a long-baseline gravitational-wave detector with 15 dB of generated squeezing.","pith_inferences":["A direct implication the authors do not develop is that finite teleportation fidelity should be folded into the equivalence itself; adding realistic Bell-measurement noise to the derivation of Eq. (10) would produce an explicit squeezing-dependent penalty for the SQL beating.","The same online/offline teleportation split could in principle be borrowed for other non-reciprocal sensing tasks, such as quantum nondemolition monitoring of momentum in cavity optomechanics, where a physical optical circulator is difficult; the paper restricts itself to displacement sensing, so this is an extrapolation.","A tabletop experiment comparing the online and offline outputs on the same mechanical oscillator would isolate the entanglement contribution: if the filtered offline curve tracks the online curve, the tripartite correlation is doing the back-action cancellation."],"forward_implications":["A long-baseline gravitational-wave detector could gain roughly an order of magnitude in displacement sensitivity near 8 Hz without replacing mirrors or coatings, by adding the dark-port teleportation hardware and a second pump.","The offline variant means the back-action force is not canceled during the measurement; it is erased afterward in the data, shifting the experimental burden from real-time feedforward to stable readout and post-processing.","Two-color pumping with a detuning of order MHz is sufficient, and the required entanglement level is 15 dB of generated two-mode squeezing, so retrofitting an existing detector would be a matter of dark-port hardware plus post-processing.","Because $K_{\\rm sm}$ is flat at low frequencies, fixed-angle homodyne readout suffices; the frequency-dependent filter cavities used in variational or frequency-dependent readout are unnecessary.","Even with realistic losses, the low-frequency sensitivity scales only as $\\Omega^{-1}$ and remains below the SQL at the optimized readout angle, so the speed-meter advantage is not confined to the lossless ideal."],"supporting_citations":[{"why":"Provides the canonical two-interaction speed-meter model and the argument that the second pulse cancels the first pulse's back-action, which the teleportation scheme reproduces.","marker":"[22]"},{"why":"Supplies the earlier speed-meter interferometer whose sensitivity formula Eq. (26) must match, and the system parameters used in the gravitational-wave application.","marker":"[25]"},{"why":"Defines the continuous-variable teleportation protocol, including the Bell measurement and displacement operation, from which both implementations are built.","marker":"[41]"},{"why":"Establishes the two-photon quadrature formalism used for all field operators and input-output calculations.","marker":"[43]"},{"why":"Gives the spectral density of the two-mode squeezed entangled state that quantifies how finite squeezing $r$ enters the teleported field as $e^{-2r}$ noise.","marker":"[44]"},{"why":"Earlier EPR-based speed-meter scheme that this work extends by using teleportation, and one of the polarization-circulator alternatives against which the nonreciprocal coupling is framed.","marker":"[26]"}],"fun_headline_variants":["Teleportation speed meter cancels radiation-pressure noise","Quantum teleportation yields sub-SQL displacement sensing","Speed meter via teleportation beats standard quantum limit","Offline teleportation speed meter surpasses SQL","Teleportation enables speed meter without core optic changes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a teleportation step performed with finite squeezing and losses remains exactly equivalent to routing the light to the mirror's back side, so the radiation-pressure kick from the first interaction cancels the kick from the second; if realistic teleportation noise breaks that equivalence, the predicted sub-SQL sensitivity does not survive.","fun_headline_variants_meta":{"raw":{"variants":["Teleportation speed meter cancels radiation-pressure noise","Quantum teleportation yields sub-SQL displacement sensing","Speed meter via teleportation beats standard quantum limit","Offline teleportation speed meter surpasses SQL","Teleportation enables speed meter without core optic changes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000215,"raw_usage":{"total_tokens":1414,"prompt_tokens":918,"completion_tokens":496,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":534,"completion_tokens_details":{"reasoning_tokens":420}},"tokens_in":534,"tokens_out":496,"duration_ms":4710,"temperature":1.0,"reasoning_tokens":420,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T10:23:37.139705+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the offline scheme on a tabletop optomechanical setup with the optimal fixed homodyne angle, 15 dB of generated two-mode squeezing, and 1% input and output losses, and compare the measured displacement-noise spectrum to Eq. (26); a curve that fails to cross the standard quantum limit in the band where $K_{\\rm sm}$ is flat would disprove the central claim. A simpler calculation would derive the full input-output map of the teleportation chain including Bell-measurement noise and check whether the back-action terms cancel before replacing the chain by Eq. (10).","supporting_citations":[],"review_version":1}