{"id":"aa20c115-08d8-4f62-b18d-7fe8684ad898","arxiv_id":"quant-ph/0602202","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":5,"one_line_summary":"A two-mode reduction of continuous-wave Gaussian fields enables direct calculation of the conditioned non-Gaussian output state for three types of photon-counting measurements.","lead":"This paper develops a general method for calculating the quantum state of an output light mode after a continuous-wave squeezed light source is measured with a photon counter. The method helps design experiments that create non-Gaussian states, such as single photons and Schrödinger cat-like states, from practical continuous-wave sources.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The pre-detection state is taken from the unprobed steady-state source, but continuous no-click monitoring before t_c changes the prior; the Sec. V averaging argument is not valid in general.","rationale":"The reader's weakest assumption identifies exactly the same point, and the paper itself flags it in Sec. V. The two-mode Gaussian reduction and the projection step are standard and internally consistent; the issue is not the algebra but the prior state. I agree with the reader that at the paper's stated parameters (1% trigger transmission, weak driving) the no-click histories have negligible weight, so the numerical figures are likely fine. However, the paper's central claim is explicitly general, and the Sec. V averaging argument is a non-sequitur in the high-efficiency/high-flux regime: the ensemble of states immediately before clicks is the Palm distribution, not the stationary one. This is not a matter of outside-consensus disagreement; it is an identifiable gap between the claimed scope and the demonstrated scope. The concrete check (a Gaussian no-click trajectory simulation, which the paper says is easy) would settle it. If the discrepancy is small, the ACCEPT verdict stands; if it is significant, the formalism should be presented as conditional on weak/low-efficiency operation or completed with the no-click extension. Hence CONDITIONAL rather than REJECT: the paper is valuable and probably correct in the demonstrated regime, but the general prescription needs a boundary condition.","tokens_in":9539,"tokens_out":13249,"duration_ms":128487,"concrete_test":"Run the OPO example of Fig. 4 (ε=0.2γ1, γ2=0) with unit trigger detection efficiency in a quantum-trajectory simulation that keeps the full conditioned Gaussian state under continuous photon counting on the trigger channel: evolve with the no-click (vacuum-projection) Gaussian dynamics up to time t_c, then apply the click superoperator a ρ a†, and reconstruct the Wigner function of the output mode exp(-γ1|t-tc|/2). Compare with the paper's Fig. 4 obtained from the stationary covariance matrix. If the negativity at the origin or the one-photon overlap changes by more than about 5% relative, the steady-state averaging assumption fails and the central claim must be restricted to weak driving / low trigger efficiency or extended with the no-click formalism advertised in Sec. V. Varying trigger efficiency from 1% to 100% will locate the regime where the discrepancy appears.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central reduction (Sec. II) constructs the joint trigger+output state from the stationary covariance matrix of the unprobed cw source, and Eqs. (3)-(6) then apply a click projector. The actual state before a click at t_c, however, is conditioned by the continuous record of no clicks before t_c. Because the output mode function (e.g., exp(-γ1|t-tc|/2)) overlaps times before t_c, those null events carry information about mode 2. The paper's Sec. V defense--that averaging over all click events and no-click histories equals the steady-state average--is not generally true: in a stationary point process, the state at a jump time is the Palm distribution, size-biased by the click rate, not the stationary distribution. The bias vanishes only at low trigger efficiency (the paper's examples use 1% transmission) or in the weak-flux limit. For a high-flux OPO with high-efficiency triggering, the computed Wigner functions and fidelities are therefore not the actual heralded states. Since the paper advertises a general prescription for any cw Gaussian source, this unproven equivalence is load-bearing.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a general method for computing the quantum state of a designated temporal mode of the output of a continuous-wave Gaussian light source, conditioned on a non-Gaussian measurement (photon-number projection, on/off detection, or photoelectric click) on a trigger temporal mode. The method first reduces the infinite-dimensional multimode Gaussian state, characterized by the source correlation functions, to the covariance matrix and Wigner function of the two-mode (trigger+output) system. It then applies the measurement as a Wigner-function operation and integrates over the trigger variables. The formalism is applied to an OPO below threshold, with exponential mode functions, yielding Wigner functions, single-photon overlaps, and Wigner negativities for low and higher gain, with and without output loss.","tokens_in":9791,"tokens_out":13537,"duration_ms":147824,"significance":"If correct, the paper supplies a practical and fairly general toolbox for predicting the non-Gaussian states produced by postselecting cw Gaussian sources, a topic of direct relevance to single-photon and Schrödinger-cat-state preparation. The reduction from field correlations to a two-mode covariance matrix is elegant, and the analytic treatment of the three detector models is useful. The numerical examples give concrete, falsifiable predictions (for example, the 98.82% overlap with a one-photon state at ε=0.01γ1), and the paper makes a credible case that the two-mode Wigner-function method is simpler than the prolate-spheroid treatment of Sasaki and Suzuki. The main reservation concerns the treatment of continuous no-click conditioning, which is not rigorously justified in the general case.","major_comments":[{"comment":"The central new ingredient is the reduction of the cw Gaussian state to the two-mode trigger+output state, and the paper explicitly wants to describe a realistic continuous photodetection record: Sec. IV says that \"after an interval of no clicks, the counting device reports a click at time tc.\" The use of the stationary two-time correlation functions of the unprobed source is defended in Sec. V by the claim that averaging over all click events and no-click histories is equivalent to the steady-state average. This equivalence is not proven and is not generally true: in a stationary point process, the distribution at a randomly selected click time is the Palm distribution, which is size-biased by the click rate and differs from the stationary distribution unless the clicks form a Poisson process. Because the chosen output mode (for example, exp(-α|t-tc|)) overlaps times before tc, the null events before tc carry information about the output mode and modify the conditioned state. The numerical examples use only 1% trigger transmission, where the bias is small, but the paper advertises a general prescription for any cw Gaussian source and any detection efficiency (Secs. I and III). Please either supply a rigorous proof of the averaging equivalence under stated conditions, or explicitly restrict the general claim and include the no-click (vacuum-projection) Gaussian update that the paper itself notes in Sec. V, or clarify that the calculation conditions only on the instantaneous trigger event at tc rather than on the full counting record.","section":"Sec. V (and Secs. II–IV)"}],"minor_comments":[{"comment":"The correlation function is missing a closing angle bracket: \"⟨a(t)a(t′) = ...\" should read \"⟨a(t)a(t′)⟩ = ...\".","section":"Eq. (7)"},{"comment":"The text \"The Wignerfunction is shown in Fig.3\" has a missing space, and the figure panels are not explicitly labeled (a)/(b) in the printed figure.","section":"Sec. IV"},{"comment":"The output mode function is written as \"u(t) ∝ exp(−α |t − tc|\" with an unmatched parenthesis; the closing brace is missing.","section":"Sec. IV"},{"comment":"It would improve reproducibility to display explicitly the covariance matrix elements V_ij in terms of the mode functions f_i(t), f_j(t), and the source correlation functions, including the vacuum-noise contributions from Eq. (1).","section":"Sec. II"},{"comment":"The reported \"overlap\" with the one-photon state should state whether these numbers are Hilbert-space fidelities or Wigner-function overlaps, since the two quantities differ by normalization conventions.","section":"Sec. IV"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is an older preprint being refereed now; the main issue is conceptual rather than computational. I would encourage the editor to request a revision that either proves the averaging claim or narrows the scope of the claimed generality, rather than rejecting, because the formalism and the numerical examples are valuable and can be made correct by a modest addition."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Klaus,\n\nThe useful thing about this paper is the recipe: take the two-time correlation functions of a cw Gaussian source, integrate them against your chosen trigger and output mode functions, and you get a two-mode Gaussian Wigner function. Then any non-Gaussian measurement on the trigger mode (photon counting, on/off, click) becomes an integral over the two-mode Wigner function. That is a clean, practical generalization of the earlier two-mode results from Dakna, Kim, and Sasaki/Suzuki, and it is new. The OPO examples with loss and mode-shape optimization show the formalism in action, and the numbers (98.82% overlap with a single photon at low flux) look right.\n\nThe soft spot, which the stress-test correctly identifies, is the steady-state averaging argument in Sec. V. The pre-click state of the output mode is actually conditioned on the continuous no-click record before tc, not just on the click at tc. The paper's defense — that averaging over all click events and no-click histories equals the unprobed steady state — is the Palm distribution pitfall: for a stationary point process the state at a jump is size-biased by the click rate. The bias disappears at low trigger transmission (the examples use 1%) and low OPO flux, so the plotted Wigner functions are fine. But the paper advertises a general prescription for any cw Gaussian source, and for high-efficiency triggering at higher flux the predicted fidelities would not match the actual heralded states.\n\nTo the paper's credit, it does not hide this. Sec. V explicitly says that if you require no clicks for a time before tc, the steady-state correlation functions are not enough, and it notes that the no-click projection is Gaussian and can be folded into the formalism. So the limitation is flagged and a route around it is sketched, though not worked out. A referee would want that route made explicit for the high-efficiency regime, and perhaps a demonstration that the Palm bias actually vanishes in the low-flux limit.\n\nThe central derivation is standard Gaussian Wigner manipulation, and the citation pattern is fair. Self-citations to the earlier Gaussian-state work are appropriate.\n\nWho is this for? Anyone designing or analyzing heralded non-Gaussian state generation from OPOs or similar cw sources. It deserves a serious referee and a publication in a good quantum optics journal. I would use it and cite it.\n\nRecommendation: engage with the paper, and request a clarifying treatment of the no-click conditioning — but accept.","headline":"A genuinely useful formalism for computing heralded non-Gaussian states from cw Gaussian sources, with the steady-state averaging caveat honestly acknowledged in the paper itself.","tokens_in":10261,"tokens_out":2182,"would_cite":true,"duration_ms":20273,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["03.65.Wj","03.67.-a","42.50.Dv"],"model":"deepseek-v4-flash","headline":"A two-mode Wigner-function recipe predicts the non-Gaussian output state of any continuous-wave Gaussian source after a photon-counting trigger, including a 98.82% one-photon state from an OPO.","keywords":["non-Gaussian states","continuous-wave squeezed light","Wigner function","photon subtraction","temporal modes","covariance matrix","optical parametric oscillator","single-photon state"],"falsifier":"Run the same OPO experiment with a fast, high-efficiency trigger detector and reconstruct the output mode by homodyne tomography, but sort the data by the duration of the no-click interval immediately before the click; if the reconstructed Wigner function depends systematically on that duration or deviates from the steady-state prediction for long intervals, the steady-state averaging assumption fails.","tokens_in":9363,"feed_emoji":"⚛️","tokens_out":7436,"duration_ms":65403,"temperature":0.7,"pith_summary":"This paper gives a general recipe for computing the quantum state of a single temporal output mode of a continuous-wave Gaussian light source after a non-Gaussian measurement, such as a photon count, has been performed on a separate trigger mode of the same source. The recipe starts from the source's two-time correlation functions, extracts the covariance matrix for the two relevant modes, and applies the measurement as an operation on the two-mode Wigner function, followed by integration over the trigger. This makes analytically tractable the calculation of Wigner functions, single-photon overlaps, and non-classical character for realistic setups. Applied to an optical parametric oscillator, the recipe yields an output mode with 98.82% overlap with a single-photon state in the low-flux limit and negative Wigner functions at higher flux, suggesting a practical route to non-Gaussian states from standard squeezed-light sources.","feed_headline":"98.8% one-photon states from a click on squeezed light","feed_subtitle":"A two-mode recipe predicts the conditioned output state of any cw Gaussian source after a trigger click; testable by tomography.","key_machinery":"The load-bearing object is the two-mode Gaussian Wigner function $W_V(\\mathbf{y})$ with a $4\\times4$ covariance matrix $V$ derived from the source correlations and the temporal mode functions $f_1(t), f_2(t)$; the non-Gaussian measurement is implemented as a projection (number states, on/off subtraction of the vacuum, or the annihilation-operator click) expressed as an operation on $W_V$, after which the trigger variables are integrated out. What makes the argument work is that every one of these operations maps a Gaussian in the output variables to a Gaussian times a low-order polynomial.","core_discovery":"The central claim is that the infinite-mode continuous-wave Gaussian field, which cannot be handled in a number-state picture, can be reduced without approximation to a Gaussian two-mode state describing only the trigger and output modes, with all other modes traced out. The two-mode covariance matrix is built from the mode functions defining the two temporal modes and from the stationary correlation functions $\\langle \\hat a(t)\\hat a(t')\\rangle$ and $\\langle \\hat a^\\dagger(t)\\hat a(t')\\rangle$. Any realistic non-Gaussian measurement on the trigger mode, whether perfect photon-number resolution, an on/off detector, or a real click event whose back action is the annihilation operator, is then written as a finite-order differential or polynomial operation on the Gaussian Wigner function. Since Gaussian integrals of polynomials are elementary, the conditioned output state is always a Gaussian multiplied by a second-order polynomial in the output quadratures, and losses and detector inefficiencies are absorbed into a simple transformation of the covariance matrix.","pith_inferences":["Inference: Because the formalism only uses the stationary two-time correlations, it should apply equally to thermal or amplified-spontaneous-emission Gaussian sources, predicting conditioned non-Gaussian output states wherever a click can be registered, not only for squeezed light.","Inference: The no-click-history equivalence could be probed numerically by simulating the full conditional dynamics for a high-efficiency trigger and comparing with the steady-state two-time prediction; a mismatch would identify a regime where the recipe needs a history-dependent correction.","Inference: The same two-mode reduction might serve as a module in continuous-variable quantum-information processing, where the trigger measurement acts as a non-Gaussian gate on the output mode and the mode functions are chosen to implement a desired operation."],"forward_implications":["The temporal mode shapes of trigger and output can be treated as variational functions, so one can numerically optimize them for single-photon overlap, Wigner negativity, or Schrödinger-cat character.","Loss, detector inefficiency, dark counts, and frequency filtering enter only through the covariance matrix, allowing direct quantitative comparison with homodyne tomography of the conditioned output.","The same reduction applies when several Gaussian cw or pulsed sources are involved, so coincidence-based proposals for single-photon generation with suppressed higher-number components can be analyzed under true cw conditions.","At higher OPO gain, the click-conditioned output is predicted to carry Schrödinger-cat-like structure with a negative Wigner function at the origin, which can be searched for experimentally by varying the output mode function."],"supporting_citations":[{"why":"Supplies the special case of a pure single-mode squeezed state after beam splitting with number-state conditioning, which the present formalism generalizes to cw sources.","marker":"[8]"},{"why":"Provides the experimental demonstration of photon subtraction from squeezed light that motivates the setup.","marker":"[9]"},{"why":"Companion experimental report of an odd-photon-number superposition state, an application of the formalism.","marker":"[10]"},{"why":"Gives an alternative multimode treatment of the on/off-detector case against which the present two-mode reduction is compared.","marker":"[12]"},{"why":"Provides the analytical form of the one-photon-conditioned Wigner function and its negativity at the origin, used for mode optimization.","marker":"[13]"},{"why":"Proposes a single-photon source using linear optics and squeezed states, the target of the multi-source generalization.","marker":"[14]"},{"why":"Prior Gaussian description of continuous measurements that the present work extends beyond Gaussian operations.","marker":"[17]"},{"why":"Defines the click back action as the field annihilation operator in photodetection theory.","marker":"[19]"},{"why":"Gives the OPO correlation functions used in the example calculations.","marker":"[20]"}],"fun_headline_variants":["One click on squeezed light yields non-Gaussian states","Infinite modes to two: new theory for non-Gaussian states","Click on squeezed light: non-Gaussian states on demand","From cw Gaussian to non-Gaussian with a single click","Any cw Gaussian source, one click, non-Gaussian states"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculation assumes that averaging over all possible no-click histories before the trigger click reproduces the steady-state statistics of the unprobed source, so the two-time correlation functions of the bare source can be used to describe the conditioned state at the click time.","fun_headline_variants_meta":{"raw":{"variants":["One click on squeezed light yields non-Gaussian states","Infinite modes to two: new theory for non-Gaussian states","Click on squeezed light: non-Gaussian states on demand","From cw Gaussian to non-Gaussian with a single click","Any cw Gaussian source, one click, non-Gaussian states"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001202,"raw_usage":{"total_tokens":4885,"prompt_tokens":808,"completion_tokens":4077,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":424,"completion_tokens_details":{"reasoning_tokens":3992}},"tokens_in":424,"tokens_out":4077,"duration_ms":26895,"temperature":1.0,"reasoning_tokens":3992,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-28T19:57:25.892451+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same OPO experiment with a fast, high-efficiency trigger detector and reconstruct the output mode by homodyne tomography, but sort the data by the duration of the no-click interval immediately before the click; if the reconstructed Wigner function depends systematically on that duration or deviates from the steady-state prediction for long intervals, the steady-state averaging assumption fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the special case of a pure single-mode squeezed state after beam splitting with number-state conditioning, which the present formalism generalizes to cw sources."},{"cited_title":"Keller, B","cited_arxiv_id":null,"evidence_quote":"Provides the experimental demonstration of photon subtraction from squeezed light that motivates the setup."},{"cited_title":"Oxborrow and A","cited_arxiv_id":null,"evidence_quote":"Companion experimental report of an odd-photon-number superposition state, an application of the formalism."},{"cited_title":"Wenger, R","cited_arxiv_id":null,"evidence_quote":"Gives an alternative multimode treatment of the on/off-detector case against which the present two-mode reduction is compared."},{"cited_title":"Grosshans and P","cited_arxiv_id":null,"evidence_quote":"Proposes a single-photon source using linear optics and squeezed states, the target of the multi-source generalization."},{"cited_title":"A Single Photon Source With Linear Optics and Squeezed States","cited_arxiv_id":"quant-ph/0408194","evidence_quote":"Prior Gaussian description of continuous measurements that the present work extends beyond Gaussian operations."},{"cited_title":"Eisert and M","cited_arxiv_id":null,"evidence_quote":"Defines the click back action as the field annihilation operator in photodetection theory."},{"cited_title":"Continuous measurements on continuous variable quantum systems: The Gaussian description","cited_arxiv_id":"quant-ph/0511154","evidence_quote":"Gives the OPO correlation functions used in the example calculations."}],"review_version":1}