{"id":"500105e3-2c17-4810-8c20-b3b29bf6dd01","arxiv_id":"1908.03240","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"The strong-to-weak coupling transition in optomechanical state transfer is interpreted as a passive PT-symmetry breaking transition, with exact Lindblad/non-Hermitian agreement only in the single-excitation subspace.","lead":"This paper shows that a standard optomechanical setup, where light and a vibrating mirror exchange energy, can be described by an effective parity-time symmetric model with unbalanced loss. It compares the full quantum description with the simplified non-Hermitian one and finds that they agree exactly only for states with one quantum of excitation.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Multi-excitation renormalization is not fixed-N postselection: Eqs. (13)-(15) mix loss sectors, so Figs. 2-5 do not cleanly separate Lindblad from non-Hermitian PT dynamics.","rationale":"The core mapping to the passive PT dimer H_PT = g sigma_x - i Gamma sigma_z is correct, and the single-excitation identity is solid. The abstract and Section II B, however, extend this equivalence to multi-excitation states via instantaneous renormalization, and the numerical comparison in Figures 2-5 depends on that extension. Because loss jumps transfer population between excitation sectors, the renormalized Lindblad expectation values are sector-weighted mixtures rather than fixed-N no-jump trajectories. This is exactly the reader's identified weakest assumption, and it is load-bearing for the multi-excitation signature claim and the 'pathway for realizing non-Hermitian Hamiltonians at a quantum level.' The finite-temperature Langevin results provide independent evidence for the strong-to-weak coupling threshold, but they do not repair the multi-excitation postselection equivalence. Therefore the conditional verdict remains appropriate: the authors should either prove or substantially qualify the fixed-N equivalence and temper the pathway claim. No change to the reader's verdict is needed.","tokens_in":9305,"tokens_out":5948,"duration_ms":66681,"concrete_test":"Integrate the zero-temperature Lindblad master equation, Eq. (3), and the non-Hermitian evolution generated by Eq. (8) for N=2 with initial state |2,0>, using the experimental gamma_a, gamma_b, and g values for each phase. Compute Delta_n(t) = [<c^dagger c>/<N>]_Lindblad - [Tr(c^dagger c e^{-iH_eff t}|2,0><2,0|e^{iH_eff^dagger t})/Tr(e^{-iH_eff t}|2,0><2,0|e^{iH_eff^dagger t})]_no-jump, and similarly for g^(1)(t). Also compute the conditional no-jump average from a quantum-jump unraveling, or a small-time expansion to order t^2, to determine which quantity actually corresponds to postselected fixed-N evolution. If Delta(t) is nonzero for t>0, the Section II B equivalence fails for multi-excitation states.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing weak point is the assertion in Section II B after Eqs. (13)-(15) that \"instantaneous renormalization is equivalent to restricting to a fixed excitation number sector\" and that post-selecting to this sector is equivalent to measuring n_a(t), n_b(t), and g^(1)(t). This equivalence is exact only in the single-excitation subspace. For N > 1, the unconditional Lindblad solution is a mixture over excitation sectors: each loss jump removes one excitation but the remaining N-1 excitations continue evolving under the beam-splitter Hamiltonian. Thus <O>(t) = sum_N p_N(t)<O>_N(t), and dividing by <N> weights sectors by their excitation number, not by the no-jump probability. The non-Hermitian trajectory generated by Eq. (8), and the equations of motion (10)-(11), stay within the initial sector N. Consequently, Eqs. (13)-(15) applied to Lindblad dynamics describe a different object from the fixed-N no-jump evolution. The claimed equivalence underpins the interpretation of Figures 2-5 as clean signatures of the PT-phase transition; the observed differences include sector-mixing corrections that are not part of the PT-dimer comparison. The single-excitation identity, Eq. (12), and the finite-temperature Langevin threshold in Figure 6 are unaffected, but they do not justify the multi-excitation comparison. The conclusion also states that multi-excitation Lindblad and non-Hermitian dynamics are \"not identical,\" so the text is internally inconsistent about the status of this equivalence.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies a linearized optomechanical beam-splitter system with unequal mode losses and argues that its dynamics realize a passive PT-symmetric dimer. The Lindblad and Langevin descriptions of the open system are compared with the non-unitary evolution generated by an effective non-Hermitian Hamiltonian H_L. In the single-excitation subspace the two descriptions coincide exactly (Eq. (12)). For multi-excitation states, the authors define instantaneously renormalized occupation numbers and coherences (Eqs. (13)-(15)) and claim that this renormalization is equivalent to postselecting onto a fixed total-excitation sector, which would make the numerical comparisons in Figs. 2-5 direct signatures of the PT transition. They also study finite-temperature thermal initial states and identify the strong-to-weak-coupling crossover with the PT-symmetry-breaking threshold g=|Gamma|. The central claim is that mode hybridization versus damped dynamics in this optomechanical state-transfer protocol is a signature of passive PT symmetry breaking.","tokens_in":9606,"tokens_out":9157,"duration_ms":98025,"significance":"If the mapping is valid, the paper provides a concrete optomechanical route to observing passive PT-symmetric dimer physics, with the exact single-excitation identity in Eq. (12) being a clean and useful result. The finite-temperature Langevin calculations use experimental parameters and yield a specific, in-principle testable threshold for the loss of state transfer at g=|Gamma|. The algebraic reduction of the lossy beam-splitter Hamiltonian to H_PT = g sigma_x - i Gamma sigma_z is correct and parameter-free in its derivation. However, the multi-excitation comparison rests on an invalid equivalence between instantaneous renormalization and fixed-N no-jump postselection, so the interpretation of Figs. 2-5 as clean PT signatures is not supported. In addition, the PT threshold is built into the construction by definition, so the paper's independent predictive content is mostly the single-excitation/no-jump mapping and the finite-temperature state-transfer switch, not a new prediction of the PT transition itself.","major_comments":[{"comment":"The assertion after Eqs. (13)-(15) that \"the process of instantaneous renormalization is equivalent to restricting to a fixed excitation number sector\" and that postselecting to this sector is equivalent to measuring n_a(t), n_b(t), and g^(1)(t) is not valid for initial states with more than one excitation. Under the Lindblad equation, a loss jump transfers population from the N-excitation sector to the (N-1)-excitation sector, and the unconditional expectation values such as <c^dagger c>(t) are sums over all sectors. Dividing by <N> weights each sector by its excitation number, not by the no-jump probability, so the renormalized quantities in Figs. 2-5 are sector mixtures rather than the fixed-N no-jump trajectory generated by H_PT. A concrete counterexample is an initial state |2,0>: after one c-loss the system is in |1,0>, so <c^dagger c>/<N> mixes the N=2 and N=1 sectors, while the non-Hermitian evolution stays in the N=2 sector. Consequently, the statement that \"Lindblad dynamics does not rise to a steady-state behavior ... solely due to the post-selection scheme\" does not follow from the calculations presented. The exact single-excitation identity in Eq. (12) is a special case and does not generalize. Please either replace the renormalization procedure with an explicit no-jump quantum-trajectory postselection, which would make the comparison exact, or revise the interpretation of Figs. 2-5 to state explicitly that the Lindblad results include sector-mixing corrections that are not part of the PT-dimer comparison. This change is also needed to resolve the internal inconsistency with the Conclusion, which states that multi-excitation Lindblad and non-Hermitian dynamics are \"not identical.\"","section":"II B, Eqs. (13)-(15)"},{"comment":"The PT-symmetric and PT-broken phases are introduced by construction: Eq. (8) is rewritten in a fixed-N sector as a common loss term plus H_PT, whose eigenvalues are ±sqrt(g^2 - Gamma^2). The threshold g=|Gamma| is therefore a property of the effective dimer by definition. Identifying the strong-to-weak-coupling transition in the optomechanical state-transfer protocol with this threshold, as done in the abstract and in the discussion of Fig. 6, is mathematically correct but is a relabeling of the known spectrum rather than an independent prediction. The finite-temperature Langevin dynamics of Fig. 6 are not themselves governed by a PT-symmetric Hamiltonian; they are full open-system dynamics. Please state explicitly that the mapping to a passive PT dimer is an exact algebraic correspondence in the no-loss/no-jump sector and that the finite-temperature results are an experimentally accessible signature of the same threshold, not a direct realization of H_PT. This clarification is important for calibrating the paper's novelty.","section":"II A, Eq. (9) and Fig. 6"}],"minor_comments":[{"comment":"The sentence \"Figure 3 shows the real (blue) and imaginary (blue) parts\" contains a typo; the imaginary part is plotted in red according to the figure caption, so it should read \"imaginary (red).\"","section":"II B, text before Fig. 3"},{"comment":"The notation n_a(t) and n_b(t) is used for raw expectation values in the single-excitation discussion and for renormalized quantities defined in Eqs. (13)-(14) in the multi-excitation sections. Please define in the figure captions whether the plotted quantities are raw, normalized by <N>, or conditional on no loss, to avoid ambiguity.","section":"II B, Figs. 1-5"},{"comment":"The claim that the coherence is \"maximum at the exceptional point\" should be quantified, since g^(1)(t) is complex and time-dependent; please specify in what norm or at what time the maximum is evaluated.","section":"II B, Fig. 5"},{"comment":"There is a typesetting issue in the phrase \"PT -symmetric orPT -symmetric broken phases\"; the space after \"or\" is missing and should be corrected.","section":"Introduction, first paragraph"},{"comment":"Please clarify that at zero temperature the D[c^dagger] and D[d^dagger] gain terms in Eq. (3) vanish because they are multiplied by ar n_a and ar n_b, respectively; the text currently leaves implicit how the zero-temperature Lindblad equation is obtained from Eq. (3).","section":"II A, Eqs. (5)-(6)"}],"recommendation":"major_revision","confidential_remarks":"The paper's most defensible core is the exact single-excitation equivalence and the mapping of the lossy beam-splitter Hamiltonian to a passive PT dimer in the no-jump sector. The multi-excitation interpretation via instantaneous renormalization is, as written, not valid, and the figures built on it do not support the stated conclusions. I would not recommend acceptance until the authors either provide genuine no-jump trajectory postselection comparisons or substantially weaken the claims about Figs. 2-5. The PT terminology applied to a passive lossy dimer is standard in the field, so the novelty rests on the optomechanical realization and the finite-temperature state-transfer threshold; the authors should frame their contribution accordingly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe core mapping is correct: the linearized red-sideband optomechanical beam splitter with unequal mode losses maps onto the passive PT dimer H_PT = g sigma_x - i Gamma sigma_z, Gamma=(gamma_a-gamma_b)/4. The single-excitation identity in Eq. (12) is exact, and the finite-temperature Langevin threshold at g=|Gamma| is a genuine, testable prediction: in one device, varying the drive strength should take you through mode hybridization to damped dynamics, with the transition at the exceptional point. That is the most useful part of the paper.\n\nWhat is new is the numerical comparison of Lindblad versus non-Hermitian dynamics for multi-excitation Fock and N00N states. The problem is the interpretative claim in Sec. II B, after Eqs. (13)-(15): the authors assert that instantaneously renormalizing Lindblad moments by the total excitation number is equivalent to postselecting on a fixed excitation sector. It is not. A loss jump removes one excitation but the remaining N-1 excitations continue evolving under the beam-splitter Hamiltonian, so the renormalized Lindblad quantities are mixtures over excitation sectors. The non-Hermitian evolution stays in the original sector. For N=1 the two coincide; for N>1 they generally do not. Figures 2-5 therefore do not give a clean signature of the PT transition; they show the PT-dimer dynamics plus sector-mixing corrections. The paper itself later says the multi-excitation dynamics 'are not identical', which contradicts the earlier equivalence claim. That is a load-bearing soft spot because the abstract and conclusion lean on the PT-phase interpretation of these numerics.\n\nOutside that, the paper is honest and well grounded: the derivation is transparent, the relevant passive-PT literature is cited, and the experimental parameters are realistic. The lack of shipped code is a minor reproducibility issue. The finite-temperature state-transfer result deserves emphasis and could be published after the postselection claim is corrected or explicitly qualified to the single-excitation case. I would send this to peer review rather than desk-reject: the core mapping is useful, the finite-T prediction is interesting, and the flaw is fixable with a revision.\n\nBest","headline":"Useful mapping of lossy optomechanical state transfer to a passive PT dimer, with an exact single-excitation correspondence, but the multi-excitation postselection equivalence does not hold and the PT-phase interpretation of the multi-excitation numerics needs rework.","tokens_in":10163,"tokens_out":4611,"would_cite":true,"duration_ms":48492,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that an optomechanical state-transfer protocol realizes a passive PT-symmetric dimer whose strong-to-weak coupling transition marks the PT-breaking threshold.","keywords":["PT symmetry","optomechanics","non-Hermitian Hamiltonian","Lindblad master equation","quantum Langevin equation","exceptional point","beam splitter","state transfer"],"falsifier":"Prepare a two-mode lossy beam splitter with unequal decay rates in the zero-temperature Fock state $|2,0\\rangle$ and compare the renormalized occupation and coherence dynamics with the noiseless $H_{\\mathrm{PT}}$ trajectory: exact agreement in this multi-excitation sector would disprove the claim that the proven equivalence is confined to single excitations, while a clear difference in the PT-broken regime would show that renormalized Lindblad observables are sector mixtures rather than the fixed-$N$ non-Hermitian dynamics.","tokens_in":9118,"feed_emoji":"⚛️","tokens_out":8093,"duration_ms":78279,"temperature":0.7,"pith_summary":"The paper argues that a standard optomechanical state-transfer protocol, once linearized around a strong classical drive, is a lossy quantum beam splitter whose fluctuation dynamics realize a passive parity-time (PT) symmetric dimer. In the fixed-excitation-sector reduction, the relevant Hamiltonian is $H_{\\mathrm{PT}}=g\\sigma_x - i\\Gamma\\sigma_z$ with $\\Gamma=(\\gamma_a-\\gamma_b)/4$, so the coupling-to-loss ratio $g/|\\Gamma|$ separates PT-symmetric ($g>|\\Gamma|$), exceptional-point ($g=|\\Gamma|$), and PT-broken ($g<|\\Gamma|$) regimes. The paper claims that the protocol's familiar transition from mode hybridization with state transfer to damped, effectively uncoupled dynamics is exactly this PT-breaking transition. It backs the claim by comparing full Lindblad/Langevin open-system dynamics with noiseless non-Hermitian evolution, finding exact agreement in the single-excitation subspace and the same three dynamical regimes for multi-excitation and N00N states, as well as at room temperature with thermal states. The result matters because it connects realistic open quantum systems to non-Hermitian physics and makes the PT transition controllable through the pump amplitude in a single device.","feed_headline":"Lossy optomechanical beam splitter realizes passive PT-symmetric dimer","feed_subtitle":"A single device can be swept across the exceptional point by changing the drive strength.","key_machinery":"The central object is the passive PT-symmetric quantum dimer Hamiltonian $H_{\\mathrm{PT}}=g\\sigma_x - i\\Gamma\\sigma_z$, obtained from the lossy fluctuation Hamiltonian by dropping the common decay term and projecting to a fixed total-excitation sector. This two-level non-Hermitian Hamiltonian carries the argument through its spectral phases: real eigenvalues with oscillating, hybridized dynamics when $|\\Gamma|<g$, a coalesced exceptional point when $|\\Gamma|=g$, and complex-conjugate eigenvalue pairs with a slowly decaying eigenmode when $|\\Gamma|>g$. The comparison between open-system and non-Hermitian dynamics is made through the Lindblad master equation and the equivalent quantum Langevin equations, with the observables being the renormalized occupation numbers and first-order coherence defined in Eqs. (13)-(15).","core_discovery":"The central claim is that the linearized optomechanical beam-splitter Hamiltonian, with unequal electromagnetic and mechanical decay rates, is equivalent to the passive PT-symmetric dimer $H_{\\mathrm{PT}}=g\\sigma_x - i\\Gamma\\sigma_z$ when restricted to a fixed total-excitation sector, where $\\Gamma=(\\gamma_a-\\gamma_b)/4$. Consequently the strong-coupling regime $g>|\\Gamma|$ is the PT-symmetric phase, with hybridized oscillatory dynamics and equal eigenmode decay rates; $g=|\\Gamma|$ is the exceptional point, where the eigenmodes coalesce and the approach to steady state becomes power-law; and $g<|\\Gamma|$ is the PT-broken phase, with a slowly decaying eigenmode and damped, mostly independent mode dynamics. The paper demonstrates that at zero temperature the Lindblad master equation and the non-Hermitian Hamiltonian evolution are exactly identical for initial states in the single-excitation subspace, because the nonlinear loss terms collapse to linear ones there. For higher-excitation Fock states and N00N states the two dynamics differ in detail, but after instantaneous renormalization the full Lindblad dynamics still displays the same three regimes; at finite temperature the full Langevin dynamics for thermal states shows that state transfer occurs only for $g>|\\Gamma|$, providing an experimentally accessible signature of the PT transition.","pith_inferences":["A quantum-trajectory simulation with number-resolved photodetection could test whether the PT signature survives when one conditions on no-jump trajectories rather than on instantaneous renormalization; the paper's equivalence is proven only for the single-excitation sector.","The sector-mixing induced by loss jumps means that the renormalized Lindblad observables of the paper are weighted mixtures over excitation sectors; extending the comparison to second-order coherence or full counting statistics would reveal whether the multi-excitation agreement is a feature of the observables chosen or of the underlying dynamics.","If the coherence maximum at the exceptional point persists with both channels lossy and at higher temperatures, it could be used as a coherence-preserving operating point for quantum state transfer, a potential resource not explored in the paper.","The same reduction should apply to other red-sideband optomechanical protocols and to coupled waveguide arrays with unequal losses, so the passive PT dimer may be a common effective description of lossy beam splitters."],"forward_implications":["In the single-excitation subspace, zero-temperature Lindblad dynamics and non-Hermitian evolution are exactly identical, so single-photon states realize the passive PT dimer without any postselection ambiguity.","For multi-excitation Fock and N00N states, the renormalized open-system dynamics show the same PT-symmetric, exceptional-point, and PT-broken regimes, so the PT-breaking signature is robust beyond the exactly solvable sector.","At room temperature with thermal initial states, the presence or absence of optomechanical state transfer marks the PT-symmetric versus PT-broken phase, with the transition at $g=|\\Gamma|$.","Because the effective coupling $g=g_0|\\alpha|$ is set by the pump, a single optomechanical device can be swept across the exceptional point by adjusting the driving strength.","The asymptotic coherence $g^{(1)}(t)$ is maximal at the exceptional point, suggesting that the PT transition point may be the best place to preserve intermodal coherence."],"supporting_citations":[{"why":"Provides the fixed-excitation-sector reduction and the passive PT dimer description in a lossy waveguide beam splitter.","marker":"[12]"},{"why":"Supplies the spectral-phase classification of passive PT-symmetry breaking without exceptional points.","marker":"[31]"},{"why":"Establishes the parametric normal-mode splitting that underlies the strong-coupling hybridization in optomechanics.","marker":"[19]"},{"why":"Gives the standard linearized optomechanical Hamiltonian and state-transfer framework.","marker":"[20]"},{"why":"Shows how the steady-state optical amplitude enhances the effective optomechanical coupling, $g=g_0|\\alpha|$.","marker":"[25]"},{"why":"Provides the equivalent Langevin-equation treatment and the numerical integration method used at finite temperature.","marker":"[28]"},{"why":"Supplies the experimental optomechanical parameters used for the numerical simulations.","marker":"[32]"},{"why":"Supports the identification of dissipative PT-breaking transitions in a physical system with losses.","marker":"[10]"}],"fun_headline_variants":["Optomechanical loss mimics passive PT symmetry","Sweep drive to cross exceptional point in optomechanics","Quantum optomechanics realizes PT-broken phases","Loss imbalance in beam splitter yields PT phases","Lindblad dynamics expose PT symmetry in cavities"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that instantaneously renormalizing the Lindblad density matrix is equivalent to postselecting on a fixed total-excitation sector; the paper proves this only for single-excitation states, because loss jumps move population between sectors for multi-excitation initial states.","fun_headline_variants_meta":{"raw":{"variants":["Optomechanical loss mimics passive PT symmetry","Sweep drive to cross exceptional point in optomechanics","Quantum optomechanics realizes PT-broken phases","Loss imbalance in beam splitter yields PT phases","Lindblad dynamics expose PT symmetry in cavities"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.0007,"raw_usage":{"total_tokens":3185,"prompt_tokens":994,"completion_tokens":2191,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":610,"completion_tokens_details":{"reasoning_tokens":2118}},"tokens_in":610,"tokens_out":2191,"duration_ms":19338,"temperature":1.0,"reasoning_tokens":2118,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:20:44.087040+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Prepare a two-mode lossy beam splitter with unequal decay rates in the zero-temperature Fock state $|2,0\\rangle$ and compare the renormalized occupation and coherence dynamics with the noiseless $H_{\\mathrm{PT}}$ trajectory: exact agreement in this multi-excitation sector would disprove the claim that the proven equivalence is confined to single excitations, while a clear difference in the PT-broken regime would show that renormalized Lindblad observables are sector mixtures rather than the fixed-$N$ non-Hermitian dynamics.","supporting_citations":[{"cited_title":"Exceptional points of any order in a single, lossy, waveguide beamsplitter by photon-number-resolved detection","cited_arxiv_id":"1905.06993","evidence_quote":"Provides the fixed-excitation-sector reduction and the passive PT dimer description in a lossy waveguide beam splitter."},{"cited_title":"Passive $\\mathcal{PT}$-symmetry breaking transitions without exceptional points in dissipative photonic systems","cited_arxiv_id":"1807.10330","evidence_quote":"Supplies the spectral-phase classification of passive PT-symmetry breaking without exceptional points."},{"cited_title":"Parametric Normal-Mode Splitting in Cavity Optomechanics","cited_arxiv_id":"0805.2528","evidence_quote":"Establishes the parametric normal-mode splitting that underlies the strong-coupling hybridization in optomechanics."},{"cited_title":"Robust optomechanical state transfer under composite phase driving","cited_arxiv_id":"1806.04266","evidence_quote":"Provides the equivalent Langevin-equation treatment and the numerical integration method used at finite temperature."},{"cited_title":"Phonon counting and intensity interferometry of a nanomechanical resonator","cited_arxiv_id":"1410.1047","evidence_quote":"Supplies the experimental optomechanical parameters used for the numerical simulations."},{"cited_title":"Observation of parity-time symmetry breaking transitions in a dissipative Floquet system of ultracold atoms","cited_arxiv_id":"1608.05061","evidence_quote":"Supports the identification of dissipative PT-breaking transitions in a physical system with losses."}],"review_version":1}