{"id":"c30e9f1c-edcf-4ab0-9a0c-4259f161a773","arxiv_id":"2411.15362","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Unwanted couplings in NV-center and rubidium quantum memories can amplify retrieved signals past unity while apparent noise stays zero, signaling hidden quantum noise.","lead":"This paper shows that extra energy levels and unwanted couplings inside a quantum memory can amplify the stored signal, making the measured memory efficiency exceed 100 percent even when the apparent noise is zero. The result warns that simplified level models and noise checks based on no-input output can hide real quantum noise, so experimental fidelity must be characterized differently.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'negligible apparent noise' claim is an artifact of dropping quantum noise operators; a full quantum input–output calculation is needed to see if the effect survives.","rationale":"The reader's weakest_assumption correctly identified the semiclassical-to-quantum inference as the load-bearing step. I sharpen that concern: the problem is not merely that the added noise is uncomputed, but that the paper's central observable—'apparent noise'—is defined as the output with no input, and in the semiclassical equations this is identically zero because every Langevin noise operator is dropped. Any real gain mechanism (phase-insensitive or phase-sensitive) that amplifies a vacuum input must produce a nonzero mean photon number at the output (indeed, the amplifier-noise theorem the authors cite guarantees this). Therefore the 'negligible apparent noise' finding in Fig. 2 and the Abstract is not a surprising physical result; it is the expected behavior of a classical model from which noise has been removed. The only way the striking claim can be upheld is if a full quantum calculation shows the added noise to be negligible in the specific measured quadrature—a nontrivial requirement for a memory, which must preserve the input state. The proposed full quantum input–output test would settle this directly. No code or data is provided, and the paper itself states that a full quantum treatment is needed, so a conditional verdict is appropriate: the paper should either supply that calculation or rephrase the claim to make clear that 'negligible apparent noise' is a semiclassical artifact, not a physical prediction.","tokens_in":12357,"tokens_out":11713,"duration_ms":113314,"concrete_test":"Implement the quantum Langevin equations for the 9-level NV system of Eq. (1) (or the simplified 4-level system of Eq. (2)) at the parameters of Fig. 2, retaining all noise operators and the input vacuum noise, and compute: (i) the mean output photon number with vacuum input, N0 = <a_out† a_out>_vac; (ii) the mean output photon number for a single-photon-level coherent input, N1. If N0 is comparable to or larger than N1 in the regime where the semiclassical apparent efficiency exceeds unity, then 'negligible apparent noise' is an artifact; if N0 << N1, the claim survives. This settles whether the unconventional FWM path is genuinely noiseless in the measured quadrature or whether semiclassical gain overestimates quantum gain.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Dropping the Langevin noise operators in Eqs. (S3)–(S4) makes the no-input output identically zero, so the headline 'amplification despite negligible apparent noise' is true by construction. In any linear quantum amplifier with gain G>1 (including the phase-sensitive FWM path described by term 8 of Eq. (2) and Eq. (3)), vacuum input produces nonzero output photon number—for a phase-insensitive amplifier N_out(vac) ~ G-1, and for a phase-sensitive amplifier squeezed vacuum also has nonzero mean photon number. Thus, whenever the semiclassical model gives efficiency >1, a full quantum treatment of the same Hamiltonian with vacuum input should produce a nonzero noise output in the signal mode. The paper explicitly states it does not compute this noise, citing Refs. [22,25] to argue that amplification must imply noise. But that very noise should appear in the 'apparent noise' measurement, so the claim that the effect occurs with negligible apparent noise is not a physical prediction; it is a direct consequence of the semiclassical approximation. Unless a full quantum calculation shows the added noise to be extremely small (e.g., for a specific phase-sensitive configuration where the relevant quadrature is noiseless—which still leaves the conjugate quadrature noisy), the central 'striking' result in the Abstract and around Fig. 2 is unsupported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies an NV-center-based absorptive quantum memory using a semiclassical Maxwell-Bloch model that includes all nine relevant levels and both desired and unwanted couplings. It reports that the apparent memory efficiency can exceed unity due to unwanted couplings, even though the no-input output ('apparent noise') is negligible, and interprets this as evidence of quantum-level noise. The authors identify two critical unwanted couplings, G38 and Ω28, construct a reduced 4-level model, derive a semi-analytical equation whose last term produces exponential amplification, and extend the analysis to a hypothetical cavity-based rubidium memory. The central practical message is that characterizing memory fidelity via the no-input output is insufficient when unwanted couplings are present.","tokens_in":12575,"tokens_out":14877,"duration_ms":141557,"significance":"If the reported effect is physically real, the paper has clear practical importance: it challenges a common level-truncation practice in quantum-memory theory and warns that apparent-noise-based fidelity estimation can miss amplifier noise introduced by unwanted couplings. The paper's strengths include the use of published NV parameters with no parameter-fitting to produce the amplification, the reproduction of the effect in a reduced 4-level model, a semi-analytical equation with an explicit amplification parameter, and a second platform example. The amplification mechanism is concrete and falsifiable: the product G38Ω28/Δ8 controls the growth. However, the central quantum-noise conclusion is inferred rather than computed, and the semi-analytical model has a consistency issue that needs attention before the general claims can be accepted.","major_comments":[{"comment":"The claim that the effect occurs 'even when the apparent noise is negligible' is a semiclassical statement, not a demonstrated physical property of the device. The apparent noise is computed from Eqs. (S3)-(S4), which drop the quantum Langevin noise operators. In a full quantum treatment of the same Hamiltonian, a device with gain greater than unity necessarily produces nonzero output in the signal mode even for vacuum input, for both phase-insensitive and phase-sensitive amplification. The paper itself states that 'accurately quantifying the associated noise requires a full quantum treatment.' Without such a calculation, the title/abstract assertion that this is 'a clear indication of unwanted noise at the quantum level' is an inference from the amplifier-noise theorem rather than a computed prediction. The authors should either provide a quantum input-output calculation (at least for the 4-level model) that quantifies the added noise, or explicitly restrict the claim to the semiclassical model and reframe the conclusion as a methodological warning about semiclassical efficiency estimates.","section":"Abstract; Numerical estimations; Conclusion"},{"comment":"The 4-level equations of motion in Eq. (S5) are not a consistent reduction of the Hamiltonian in Eq. (1). The Hamiltonian includes the term -â G38 σ'83 e^{iδt} - H.c., so the Heisenberg equation for the cavity field should contain a back-action term iG38* σ'38 e^{-iδt}. Equation (S5) contains only iG29* σ'29 in the cavity-field equation. Since Eq. (2) is derived from Eq. (S5), the semi-analytical amplification term (term 8) and the growth parameter b in Eq. (4) may be artifacts of this truncation. The authors should either justify why the G38 back-action on the cavity mode can be neglected while the G38 terms in the atomic equations are retained, or correct Eq. (S5) and re-derive the 4-level results. This issue is load-bearing for the analytical explanation and for the rubidium example, which relies on the same simplified model.","section":"Supplement, Eq. (S5); Eq. (2)"},{"comment":"The statement that G38 and Ω28 are 'essential' for efficiencies above unity is supported only by a two-parameter scan in which all other unwanted couplings are held at their original values. This shows that these two are necessary in that hyperplane, but it does not rule out that other unwanted couplings, or other pairs of couplings, can also produce amplification in the full 9-level system. Given that the 4-level explanation currently rests on the inconsistent truncation described above, the 'essential' claim needs additional numerical evidence, such as a scan that sets G38=Ω28=0 while varying other couplings, or a demonstration that term 8 in Eq. (2) is the unique amplification mechanism in the full model.","section":"Fig. 3 and surrounding text"}],"minor_comments":[{"comment":"The phrase 'a clear indication of unwanted noise at the quantum level' should be qualified as an inference from the semiclassical gain, since the paper does not compute the quantum noise. Consider saying 'which, by the amplifier-noise theorem, indicates unwanted noise at the quantum level' to distinguish the theorem from a direct calculation.","section":"Introduction and Abstract"},{"comment":"The caption says 'we observe no output in the absence of an input,' while the text later says the apparent noise 'remains near zero.' Please state the actual numerical value or threshold used to define 'no apparent noise' or 'negligible apparent noise.'","section":"Fig. 2 caption"},{"comment":"The long equation in Eq. (2) is hard to parse with the term labels 'terms 1 and 3-5' and 'terms 2 and 6-8.' Please mark the term numbers directly in the equation, or add a table listing each term with its physical origin.","section":"Eq. (2)"},{"comment":"The temperature-dependent decay rate γd(T) has a missing citation: the text contains '[? ]' after the expression γd(T) = Γ(T)/2. Please add the reference.","section":"Supplement A"},{"comment":"The sentence 'previously reported impressive memory efficiencies in atomic memories may have been significantly influenced by amplification' is speculative. The Rb example is explicitly hypothetical, so please either soften this statement or provide a concrete experimental parameter set that would make the concern quantitative.","section":"Conclusion and Rb example"}],"recommendation":"major_revision","confidential_remarks":"The paper is best understood as a warning about the interpretation of semiclassical efficiency calculations in quantum-memory design. The numerical 9-level result is interesting, but the title and abstract currently overstate what is proven, and the semi-analytical model has a consistency error that should be fixed before publication. The authors should be encouraged to either add a quantum noise calculation or substantially temper the physical claims."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this paper identifies a new four-wave-mixing path that can push a quantum memory's apparent efficiency above unity while the no-input output stays negligible. The path requires unwanted couplings to both the control and the signal fields (here, G38 and Omega28), not just to the control field, which is what the earlier FWM-noise literature focused on. That is a real extension, and the authors demonstrate it concretely in a 9-level NV-center model, reproduce it in a simplified 4-level model with a semi-analytical equation, and give a rubidium example. The coupling values come from published dipole matrix elements; nothing is fitted to hit the amplification. That is solid workmanship.\n\nThe soft spots are proportional. The biggest one is stated plainly in the paper itself: the calculation is semiclassical, so the 'negligible apparent noise' is zero by construction because the Langevin noise operators are dropped. The authors do not compute the added quantum noise; they rely on the amplifier-noise theorem to argue that gain implies noise. The stress-test note says this makes the headline result an artifact. I think that is too harsh, because the paper's actual claim is not that a full quantum treatment shows zero noise, but that apparent-noise measurements are insufficient—and that point survives even if the quantitative gain changes when the noise operators are restored. Still, the absence of a full quantum input–output calculation is a genuine gap. The efficiency-above-unity number may shift, and the claim that previously reported high fidelities could be inflated is speculative, which the authors themselves acknowledge.\n\nMinor issues: no code or data are provided, and the semi-analytical reduction is derived from the same Hamiltonian, so it is internal consistency rather than an independent check. The storage-time oscillation is peripheral. None of this changes the core message.\n\nWho this is for: anyone who characterizes memory fidelity from no-input noise, and theorists designing multi-level memories. It deserves a serious referee—the mechanism is plausible, the calculations are standard, and the authors are unusually candid about the semiclassical limitation. I would not desk-reject it. For your own work, the paper is worth citing as a caution about apparent-noise-based fidelity estimation, though I would wait for a full quantum treatment before relying on the specific efficiency numbers.","headline":"A solid, transparent semiclassical study of a genuinely new FWM mechanism in quantum memories; the quantum-noise conclusion is inferred, not computed, but the paper itself says so.","tokens_in":13150,"tokens_out":1985,"would_cite":true,"duration_ms":20622,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["42.50.Gy","03.67.Hk"],"model":"deepseek-v4-flash","headline":"This paper shows that in an NV-center-based absorptive quantum memory, unwanted energy levels and couplings can amplify the stored signal so that measured efficiency exceeds unity, even when the output in the absence of an input—the usual…","keywords":["quantum memory","NV center","four-wave mixing","unwanted couplings","memory efficiency","apparent noise","semiclassical analysis","signal amplification"],"falsifier":"A full quantum input–output calculation of the 4-level memory that includes noise operators, computing the added noise in the retrieved temporal mode, would settle the matter: if the added noise does not grow with the predicted gain, or if the gain itself disappears when the field is quantized, the claim that amplification is a clear indication of quantum noise would need qualification. Experimentally, measuring the second-order correlation $g^2_{\\rm out}(0)$ of the retrieved light, as the paper itself suggests, can discriminate: $g^2_{\\rm out}(0) \\geq 1$ together with efficiency above unity supports the amplification picture, while a sub-Poissonian retrieved field would contradict it.","tokens_in":12136,"feed_emoji":"⚛️","tokens_out":6456,"duration_ms":55956,"temperature":0.7,"pith_summary":"The paper argues that when realistic energy-level structures are taken into account, quantum memories based on NV centers can do something unexpected: unwanted couplings to neighboring levels amplify the retrieved signal, so the apparent memory efficiency exceeds 100 percent. The amplification occurs even though the output with no input field—the 'apparent noise' used in typical fidelity checks—is essentially zero, meaning such checks can certify a memory as noiseless when it is actually adding noise at the single-photon level. The mechanism is a four-wave-mixing process that requires one unwanted coupling to the control field and one to the signal field, working through a nearby excited level. The same effect is shown in a cavity-based rubidium memory, so the caution extends beyond NV centers. A strategy of increasing level splittings and detunings, and minimizing the unwanted couplings, is proposed to reduce the amplification.","feed_headline":"Unwanted couplings amplify quantum memories with zero visible noise","feed_subtitle":"Efficiencies above unity reveal hidden four-wave mixing; apparent-noise fidelity checks can mislead.","key_machinery":"The central object is the 4-level reduction of the NV-center system, where the memory's spin coherence $\\sigma'_{32}$ evolves under an equation whose unwanted 'term 8' is proportional to $N G^*_{29} \\Omega_{39} G_{38} \\Omega^*_{28} \\sigma'_{23} / ((\\gamma_d + \\gamma_e - i\\Delta_8)\\alpha)$. Retaining only the desired term and this term yields the amplification equation with an exponentially growing solution $\\exp\\bigl(i\\delta + \\sqrt{b^2/(\\Gamma^2 + \\Delta_8^2)} - \\delta^2\\, t\\bigr)$, where $b = N G^*_{29} \\Omega_{39} G_{38} \\Omega^*_{28}/\\alpha$. This shows that the amplification arises from a four-wave-mixing process that needs both an unwanted coupling to the control field and an unwanted coupling to the signal field through a nearby excited level.","core_discovery":"The paper shows that unwanted energy levels and undesired couplings, typically neglected when modeling a memory as an ideal $\\Lambda$ system, can amplify the output of an NV-center-based absorptive memory. The apparent efficiency, defined as $E = \\int |a_{\\rm out}(t)|^2 dt$ for a normalized input, can exceed unity even when the apparent noise (the output with no input) is negligible and the apparent fidelity stays at unity. This amplification is a clear indication of quantum-level noise in the signal mode, even though the semiclassical model used here does not explicitly include the noise operators. Generalizing to a 4-level system, the authors identify a complex four-wave-mixing path involving the unwanted couplings $G_{38}$ (signal field) and $\\Omega_{28}$ (control field) as the origin, and they show that a cavity-based $^{87}$Rb memory exhibits the same behavior. They conclude that fidelity estimates based solely on apparent noise are insufficient and that mitigating the amplification requires increasing level splittings and detunings or minimizing the unwanted couplings.","pith_inferences":["A full quantum treatment that includes noise operators would likely confirm that the added noise in the retrieved temporal mode grows with the predicted gain, as required by the amplifier-noise theorem; the semiclassical method here captures the gain but not the noise itself.","The paper's proposed measurement of the second-order correlation $g^2_{\\rm out}(0)$ could be applied to existing warm-vapor or cavity memories: a retrieved field that becomes bunched or thermal as efficiency rises above unity would support the amplification picture, while a sub-Poissonian retrieved field would challenge it.","The same four-wave-mixing path may affect other multi-level memory platforms with weak selection rules, such as rare-earth-ion-doped crystals, where the $\\Lambda$-simplification is standard; the authors' findings suggest checking for similar amplification there.","If the four-wave-mixing gain turns out to be phase-sensitive in a full quantum treatment, the noise could be suppressed in one quadrature, meaning efficiency above unity would not automatically imply added noise for all input states—an experimentally testable distinction."],"forward_implications":["In NV-center memories, apparent efficiency greater than unity can occur with zero apparent noise, so apparent-noise measurements alone cannot certify quantum-memory fidelity.","The amplification requires unwanted couplings to both the control and signal fields; studies that consider only control-field unwanted couplings, as is common in four-wave-mixing noise analyses, will miss the effect.","Increasing the ground-state splitting $\\delta$ and the detuning $\\Delta_8$, or minimizing the ratio $G_{38}\\Omega_{28}/\\Delta_8$, reduces the amplification.","The effect is not limited to NV centers: a cavity-based $^{87}$Rb memory with realistic hyperfine levels shows apparent efficiency above unity with apparent fidelity of unity.","Previously reported memory efficiencies in atomic and solid-state systems may have been partly inflated by this amplification, depending on how noise or fidelity was characterized experimentally."],"supporting_citations":[{"why":"Supplies the theorem that an amplifier with gain must add noise, which is load-bearing for interpreting efficiency above unity as a sign of quantum-level noise.","marker":"[25]"},{"why":"Provides the prior framework for four-wave-mixing noise in memories, which the paper extends by showing that simultaneous unwanted control and signal couplings are needed.","marker":"[22]"},{"why":"Provides the semiclassical adiabatic-elimination framework and the efficiency definition used in the numerical and semi-analytical analyses.","marker":"[23]"},{"why":"Supplies the NV-center energy-level structure and optical selection rules used to build the 9-level numerical model.","marker":"[14]"},{"why":"Supplies the detailed NV-center Hamiltonian and eigenstates that determine the couplings and detunings in the model.","marker":"[28]"},{"why":"Supplies the rubidium D-line transition dipole data used to construct the cavity-based $^{87}$Rb memory example.","marker":"[32]"},{"why":"Supplies the $g^2_{\\rm out}(0)$ method for detecting four-wave-mixing noise, which the paper proposes as a way to isolate the amplification contribution.","marker":"[31]"},{"why":"Supplies the definition of the factor $f = \\Omega/\\Gamma$ distinguishing EIT and ATS regimes, used to choose operating parameters.","marker":"[29]"}],"fun_headline_variants":["Hidden couplings amplify quantum memory with no visible noise","Quantum memory efficiency exceeds unity despite zero apparent noise","Unwanted levels boost memory output past unity with silent noise","Apparent noise zero, but hidden four-wave mixing amplifies memory"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper's semiclassical analysis captures the gain but omits the quantum noise operators, so the conclusion that efficiency above unity implies added quantum noise rests on the amplifier-noise theorem applied to a classical gain that is never independently verified in a full quantum model.","fun_headline_variants_meta":{"raw":{"variants":["Hidden couplings amplify quantum memory with no visible noise","Quantum memory efficiency exceeds unity despite zero apparent noise","Unwanted levels boost memory output past unity with silent noise","Apparent noise zero, but hidden four-wave mixing amplifies memory"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000222,"raw_usage":{"total_tokens":1441,"prompt_tokens":920,"completion_tokens":521,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":536,"completion_tokens_details":{"reasoning_tokens":455}},"tokens_in":536,"tokens_out":521,"duration_ms":5260,"temperature":1.0,"reasoning_tokens":455,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:23:35.844154+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A full quantum input–output calculation of the 4-level memory that includes noise operators, computing the added noise in the retrieved temporal mode, would settle the matter: if the added noise does not grow with the predicted gain, or if the gain itself disappears when the field is quantized, the claim that amplification is a clear indication of quantum noise would need qualification. Experimentally, measuring the second-order correlation $g^2_{\\rm out}(0)$ of the retrieved light, as the paper itself suggests, can discriminate: $g^2_{\\rm out}(0) \\geq 1$ together with efficiency above unity supports the amplification picture, while a sub-Poissonian retrieved field would contradict it.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the prior framework for four-wave-mixing noise in memories, which the paper extends by showing that simultaneous unwanted control and signal couplings are needed."},{"cited_title":"Consequently, the following results apply to memory protocols based on the adiabatic elimination of absorption, such as EIT [29]","cited_arxiv_id":null,"evidence_quote":"Provides the semiclassical adiabatic-elimination framework and the efficiency definition used in the numerical and semi-analytical analyses."},{"cited_title":"Heshami, C","cited_arxiv_id":null,"evidence_quote":"Supplies the NV-center energy-level structure and optical selection rules used to build the 9-level numerical model."},{"cited_title":"For parameters see Fig","cited_arxiv_id":null,"evidence_quote":"Supplies the detailed NV-center Hamiltonian and eigenstates that determine the couplings and detunings in the model."},{"cited_title":"Doherty, F","cited_arxiv_id":null,"evidence_quote":"Supplies the $g^2_{\\rm out}(0)$ method for detecting four-wave-mixing noise, which the paper proposes as a way to isolate the amplification contribution."}],"review_version":1}