{"id":"67a25e1b-1538-4e67-b8ec-fe1cf1a08a19","arxiv_id":"2412.13730","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"Dispersive qubit readout with squeezed light gives exponential temperature-precision improvement only in zero-temperature, zero-time, or zero-photon limits; a claimed Heisenberg 1/N scaling for N bath-coupled qubits rests on a questionable noise approximation.","lead":"This paper analyzes whether squeezed light can improve the precision of temperature measurements made by reading out a qubit dispersively coupled to a cavity. It finds exponential improvements only in special limits when the qubit is isolated, and claims a Heisenberg-style 1/N scaling when many qubits stay in contact with the bath.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Algebraic inconsistencies in the steady-state signal S_T (Eq. 40 vs C10, and vs the derivative of Eq. C9) leave the derivation of Eq. 43 unverified; the claimed 1/N scaling needs a corrected derivation.","rationale":"The paper's central claim, Eq. (43), is a Heisenberg-limited thermometry formula for N bath-coupled qubits. A careful re-derivation from the given equations reveals two concrete defects: the signal S_T is stated inconsistently in Eq. (40) versus Eq. (C10), and neither version matches the derivative of Eq. (C9) when the T-dependence of the denominator is retained. This is load-bearing because Eq. (43) is obtained by combining the noise variance with one of these signal expressions; if the correct signal has a different prefactor or scaling, the central quantitative claim changes. The reader's specific objection about a 1/√N quantum-Fisher-information bound is not the strongest route: the output light is an external probe whose quantum Fisher information can scale as N² even for independent qubits, so a collective phase measurement can in principle yield 1/N scaling. The paper's own algebra, however, must be corrected before the claim is credible. A second independent issue is that the appendix replaces the sum over N qubit noise operators with N times a single common noise, overestimating the projection-noise variance by a factor of N; this affects the regime of validity of the large-κ approximation. The isolated-qubit sections are a genuine strength: the quantum Fisher information saturation in Sections III-IV is coherent and supports the SQL result for independent qubits. The recommendation is therefore to keep the verdict conditional: the claimed Heisenberg scaling may survive a corrected derivation, but the current manuscript does not provide a consistent derivation of Eq. (43).","tokens_in":12889,"tokens_out":25611,"duration_ms":235712,"concrete_test":"Recompute S_T by differentiating Eq. (C9) with Φ=π/2 without assuming the denominator is T-independent, and compare the large-κ limit with Eq. (40), Eq. (C10), and the signal implied by Eq. (43). Independently recompute the qubit-noise contribution to ⟨δa†δa⟩ and ⟨δa²⟩ in Appendix C while keeping independent noise operators for each qubit (δ_jk correlations), and check whether the projection-noise variance scales as N or N²; then test whether the condition κ ≫ 2Nχe^r/(2n+1) is sufficient for Eq. (43) to hold.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central formula Eq. (43) is not actually derived by the equations preceding it. The steady-state signal is given inconsistently: Eq. (40) in the main text has 2√κα Nχ|∂_T n|(2n+1)/D, while Eq. (C10) in the appendix has √κα N²χ|∂_T n|(2n+1)/D, where D = N²χ² + (2n+1)²κ²/4. Neither equals the derivative of Eq. (C9). From Eq. (C9) with Φ=π/2, ⟨Q⟩ = −2√κα Nχ u/(N²χ² + u²κ²/4) with u=2n+1, so ∂_T⟨Q⟩ = −2√κα Nχ u′ (N²χ² − u²κ²/4)/(N²χ² + u²κ²/4)². The T-dependence of the denominator is therefore not negligible and changes the prefactor of the signal in the large-κ limit, even though the signal remains linear in N. Because Eq. (43) is built from the noise variance divided by this signal, the claimed Heisenberg scaling is not reliably supported by the written derivation. In addition, Eqs. (C3) and (C6) replace the sum over N independent qubit noise operators by N times a single common noise operator; this overestimates the qubit-projection-noise contribution to the output variance by a factor of N and alters the regime condition κ ≫ 2Nχe^r/(2n+1). The reader's quantum-Fisher-information objection is not the most decisive point, since an external probe can carry an N²-scaling Fisher information even for independent qubits; the resolvable defect is the internal algebraic inconsistency.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies dispersive qubit readout used as a thermometer. For a single thermalized qubit isolated from the bath, Sections II and III derive the temperature uncertainty obtained from homodyne detection with injected external squeezing (IES) and with IES plus intracavity squeezing (ICS), concluding that exponential improvement is possible only when the temperature, the measurement time, or the input photon number tends to zero. Section IV treats N independent isolated qubits and recovers the standard quantum limit, 1/sqrt(N). The central new claim is in Section V: when N qubits remain in contact with the thermal bath and the cavity loss rate is large while the qubit-cavity coupling is weak, the temperature precision scales as 1/N (Eq. 43), with an additional exponential improvement from squeezing. The paper also reports that for sufficiently large N the precision degrades (Eq. 44).","tokens_in":13355,"tokens_out":15901,"duration_ms":135758,"significance":"The question addressed is relevant: squeezing-enhanced dispersive readout is an active experimental topic, and extending it to thermometry is a natural and potentially useful step. The manuscript has genuine strengths: the isolated-qubit calculation is self-contained, and Eqs. (23) and (27) correctly show that the dispersive readout saturates the single-qubit quantum Fisher information bound in the appropriate limit; Section IV also correctly identifies the 1/sqrt(N) standard quantum limit for independent qubits. I do not regard the N-scaling of the single-qubit QFI as an automatic obstruction to the claimed 1/N result, because the cavity output is a collective probe that can in principle carry an N^2-scaling signal even for independent qubits. However, the written derivation of the Section V claim contains internal algebraic inconsistencies that are load-bearing, so the 1/N result is not established as it stands.","major_comments":[{"comment":"The signal S_T^m is not the derivative of the steady-state quadrature. From Eq. (C9) with Phi=pi/2 one obtains d_T<Q> = sqrt(kappa alpha_in) N chi (d_T n) [(2n+1)^2 kappa^2 - 4N^2 chi^2] / [N^2 chi^2 + (2n+1)^2 kappa^2/4]^2, which is neither Eq. (40) nor Eq. (C10); moreover Eq. (40) and Eq. (C10) differ from each other by a factor N/2. Since Eq. (43) divides the output noise by this signal, the prefactor and the regime condition of the claimed 1/N scaling are not derived as written. In the large-kappa limit the exact derivative differs from the signal used in Eq. (43) by a factor such that the correct delta_T is ((2n+1)/2) times the value in Eq. (43), so even the final formula is not correct as stated.","section":"Section V, Eqs. (40), (C9), (C10)"},{"comment":"The derivation replaces the sum sum_j delta_sigma_{jz} appearing in Eq. (C1) by N times a single operator delta_sigma_{jz}, and Eq. (C6) then contains N chi times one noise operator sigma_z_in. The qubit noise operators in Eq. (C4) are independent for different j, so the cavity-fluctuation variance should contain a sum of N independent single-qubit contributions, i.e., a factor N rather than N^2. Using N^2 chi^2 in Eqs. (C7)-(C8) overestimates the qubit-projection-noise contribution to the output variance by a factor N and therefore changes the regime condition kappa >> 2N chi e^r/(2n+1) and the large-N degradation formula in Eq. (44). This point is load-bearing for the 1/N claim and must be redone with the independent qubit noises.","section":"Appendix C, Eqs. (C3) and (C6)"},{"comment":"The stated condition 'arctan(2 chi/kappa) = n pi' is not satisfiable for nonzero chi, because tan(n pi) = 0 forces 2 chi/kappa = 0. This condition is used to obtain the exponential factor e^{-2r} via cos(4 psi) = 1, so Eqs. (14), (15), (A6), (A7), (A10), and (A11) do not apply to the physical dispersive regime chi != 0. The authors should either give the correct phase-matching condition for real chi and kappa, or show that the qualitative conclusion of Sections II and III (no exponential improvement unless T, alpha_in, or tau tends to zero) survives a correct optimization.","section":"Section II and Appendix A, Eqs. (14), (15), (A6)-(A11)"}],"minor_comments":[{"comment":"There are several typographical errors: 'eqiuation' in Eq. (44), 'limt' in Appendix A, and 'Qauntum' in Ref. [28]; these should be corrected.","section":"General"},{"comment":"Reference [6] appears to be a duplicate of Ref. [5] and lacks full bibliographic information; the authors should either supply the complete reference or remove the duplicate.","section":"References"},{"comment":"The caption of Fig. 2 does not state whether the curves are computed from Eq. (43), Eq. (44), or the general numerical solution, which makes the plot difficult to interpret.","section":"Figure 2"},{"comment":"The symbol tau is used both as the upper limit of the integrated quadrature M and as the measurement time; the authors should define this consistently in one place.","section":"Notation"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of the journal and the central idea is worth pursuing, but the Section V derivation contains internal algebraic inconsistencies (signal formula mismatch and an unjustified common-noise replacement) that must be corrected before the Heisenberg-scaling claim can be assessed. The issues appear fixable by re-derivation, and the physical claim may survive correction, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nHere's my read of arXiv:2412.13730. The first half, on a single thermalized qubit read out by a squeezed cavity, is coherent and worth a look. The result that squeezed light only gives exponential temperature-precision gains in the limits T→0, τ→0, or α_in→0, and that the steady-state precision saturates the quantum Fisher information bound, is cleanly derived. That part is a legitimate new application of the Barzanjeh/Qin readout machinery to thermometry.\n\nThe second half, the claimed 1/N Heisenberg scaling for N qubits in contact with the bath, does not hold up as written. There are two concrete problems. First, the steady-state signal S_T is given inconsistently: Eq. (40) in the main text has N, Eq. (C10) in the appendix has N², and neither equals the derivative of ⟨Q⟩ from Eq. (C9). Taking that derivative gives a term proportional to Nχ|∂_T n| (N²χ² − u²κ²/4)/D², so the T-dependence of the denominator is not negligible. The prefactor in Eq. (43) is therefore not actually derived. Second, in the appendix the sum over N independent qubit noise operators is replaced by N times a single common noise operator, which overestimates the projection-noise contribution to the output variance by a factor of N. That changes the regime condition under which the output noise is N-independent.\n\nThe reader's worry about the quantum Fisher information is not the decisive point: an external probe can carry N² Fisher information even when the qubits are independent, if the signal grows linearly with N while the noise stays flat. The decisive problem is that the written derivation doesn't get you to Eq. (43). With the correct derivative, the signal still scales linearly with N in the large-κ limit, so the Heisenberg scaling might survive a corrected derivation, but the prefactor and the regime condition will change.\n\nThis paper is for people working on squeezed-light readout and quantum thermometry. It deserves a serious referee, but the referee should demand a corrected Appendix C and a reconciled signal formula before the central claim is accepted. I would not cite it in its current form.\n\nSend it to review with the expectation of major revision.","headline":"Solid isolated-qubit thermometry, but the Heisenberg-scaling claim for bath-coupled qubits is not supported by the written derivation due to inconsistent signal formulas and an overestimated noise term.","tokens_in":13783,"tokens_out":12388,"would_cite":false,"duration_ms":94484,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Dispersive qubit readout reaches Heisenberg scaling for temperature when N bath-coupled qubits are read out through a lossy, weakly coupled cavity, with squeezing adding an exponential boost.","keywords":["dispersive qubit readout","quantum thermometry","squeezed light","Heisenberg scaling","homodyne detection","quantum Fisher information","cavity quantum electrodynamics","parameter estimation"],"falsifier":"Compute the quantum Fisher information of the full steady state of the cavity coupled to $N$ bath-contacted qubits; if that information grows only as $N$ rather than as $N^2$, then $\\delta T\\propto 1/N$ cannot be a fundamental precision limit and Eq. (43) would overestimate what is achievable.","tokens_in":12666,"feed_emoji":"🌡️","tokens_out":15108,"duration_ms":115496,"temperature":0.7,"pith_summary":"This paper asks whether the exponential gains that squeezed light brings to dispersive qubit readout can also improve the precision of thermometry. For a single qubit that is thermalized and then isolated from the bath, the answer is mostly no: thermal fluctuations wash out the squeezing advantage, and exponential improvement appears only in the limits of zero temperature, zero measurement time, or zero input photon number. The main positive claim is that when $N$ qubits remain in contact with the thermal bath and are read out through a high-loss, weakly coupled cavity, the temperature uncertainty reaches Heisenberg scaling, $\\delta T\\propto 1/N$, and injected squeezing multiplies the precision by $e^r$. If correct, this gives a concrete route to beating the standard quantum limit in quantum thermometry with existing dispersive-cavity hardware.","feed_headline":"Qubit thermometry hits Heisenberg scaling in a lossy cavity","feed_subtitle":"Squeezed light boosts precision exponentially, while N thermalized qubits cut uncertainty as 1/N.","key_machinery":"The object that carries the argument is the steady-state homodyne readout of the cavity field under the Hamiltonian $H_c=\\sum_{j=1}^N\\chi\\,\\sigma_j^z a^\\dagger a$. Temperature enters through the mean qubit polarization $\\langle\\sigma_j^z\\rangle=-1/(2n+1)$, which shifts the cavity amplitude $\\langle a\\rangle$; detecting the output quadrature at phase $\\Phi=\\pi/2$ gives a temperature signal $S_T^m=\\frac{\\sqrt{\\kappa\\alpha_{\\rm in}}\\,N^2\\chi(2n+1)|\\partial_T n|}{N^2\\chi^2+(2n+1)^2\\kappa^2/4}$. The error-propagation formula $\\delta T=\\sqrt{\\langle\\Delta^2 Q\\rangle}/S_T^m$ converts the output noise into temperature uncertainty. In the limit $\\kappa\\gg 2N\\chi e^r/(2n+1)$, the squeezed-vacuum term dominates the variance $\\langle\\Delta^2 Q\\rangle$ and the qubit-projection-noise term is dropped, leaving the noise $N$-independent while the signal grows with $N$; injected squeezing supplies the $e^{-r}$ reduction in noise.","core_discovery":"The paper's central claim is that dispersive qubit readout—a scheme in which the state of a qubit shifts the resonance frequency of an optical cavity and is inferred from a phase-sensitive measurement of the output light—can be turned into a thermometer with Heisenberg scaling. If $N$ qubits remain in contact with a thermal bath while a strongly damped, weakly coupled cavity is read out in steady state, the temperature uncertainty is claimed to be\n$$\n\\delta T\\simeq \\frac{(2n+1)\\$kappa^{2}$ $e^{{-r}}$}{8\\sqrt{\\kappa\\alpha_{\\rm in}}\\,N\\chi\\,|\\partial_T n|},\n$$\nwith $n=(e^{\\omega_q/T}-1)^{-1}$, so precision improves as $1/N$ and injected external squeezing improves it exponentially via $e^{-r}$. For a single qubit thermalized and then isolated from the bath, the paper finds that squeezing does not give an exponential gain unless temperature, measurement time, or input photon number tends to zero; for $N$ independent such qubits the precision recovers the standard quantum limit $\\delta T\\propto 1/\\sqrt{N}$.","pith_inferences":["The claimed $1/N$ precision would imply that the full cavity-plus-qubits steady state carries $N^2$ worth of Fisher information; computing that quantity directly would connect the mechanism to standard quantum-metrology bounds.","The condition $\\kappa\\gg 2N\\chi e^r/(2n+1)$ makes the useful qubit number depend on the ratio $\\kappa/\\chi$ and shrink as squeezing grows, so real devices should show a crossover $N^*$ beyond which both more qubits and stronger squeezing degrade precision.","The mechanism suggests a design principle for quantum thermometers: keep the probe coupled to the sample during the measurement and read it out through a fast-decaying, weakly coupled cavity, a configuration testable with existing circuit-QED hardware."],"forward_implications":["In the bath-coupled regime, the Heisenberg scaling $\\delta T\\propto 1/N$ holds only up to a crossover qubit number; beyond it the uncertainty grows linearly with $N$ (Eq. (44), Fig. 2).","Increasing the input photon number $\\alpha_{\\rm in}$ improves precision in every regime considered, so a stronger drive is a universal resource for this thermometer.","In the isolated-qubit case, even ideal readout saturates the quantum Cramér-Rao bound only in the limits $\\alpha_{\\rm in}\\tau\\to\\infty$ or $r\\to\\infty$; otherwise thermal fluctuations prevent squeezed light from helping.","In the steady-state bath-coupled case, using intracavity squeezing gives the same precision as injected external squeezing alone, so the simpler IES-only setup is sufficient."],"supporting_citations":[{"why":"Supplies the IES-plus-ICS readout scheme whose exponential SNR gain this paper tests for thermometry.","marker":"[12]"},{"why":"Establishes the Heisenberg-limited dispersive readout baseline with two-mode squeezed light.","marker":"[10]"},{"why":"Provides the dispersive cavity-qubit Hamiltonian that is the starting model for the readout.","marker":"[23]"},{"why":"Gives the SNR and error-propagation definitions used to turn output quadrature statistics into temperature uncertainty.","marker":"[24]"},{"why":"Underlies the quantum Cramér-Rao bound used to compare the readout precision with the optimal bound.","marker":"[25]"},{"why":"Supplies the Markovian input-noise correlation formalism used in the quantum Langevin derivation.","marker":"[28]"}],"fun_headline_variants":["Squeezed light enables Heisenberg-limited qubit thermometry","Qubit temperature readout reaches quantum limit with squeezing","Lossy cavity turns qubits into Heisenberg-limited thermometer","Squeezed light exponentially boosts qubit thermometry precision","Thermalized qubits in lossy cavity achieve Heisenberg precision"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that when the cavity decay is fast and the qubit-cavity coupling is weak, the output noise is dominated by the squeezed vacuum and the fluctuations of the $N$ thermalized qubits can be neglected, even though $N$ independent thermalized qubits normally limit precision to $1/\\sqrt{N}$.","fun_headline_variants_meta":{"raw":{"variants":["Squeezed light enables Heisenberg-limited qubit thermometry","Qubit temperature readout reaches quantum limit with squeezing","Lossy cavity turns qubits into Heisenberg-limited thermometer","Squeezed light exponentially boosts qubit thermometry precision","Thermalized qubits in lossy cavity achieve Heisenberg precision"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001023,"raw_usage":{"total_tokens":4302,"prompt_tokens":919,"completion_tokens":3383,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":535,"completion_tokens_details":{"reasoning_tokens":3299}},"tokens_in":535,"tokens_out":3383,"duration_ms":19299,"temperature":1.0,"reasoning_tokens":3299,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T12:53:29.074295+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the quantum Fisher information of the full steady state of the cavity coupled to $N$ bath-contacted qubits; if that information grows only as $N$ rather than as $N^2$, then $\\delta T\\propto 1/N$ cannot be a fundamental precision limit and Eq. (43) would overestimate what is achievable.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the IES-plus-ICS readout scheme whose exponential SNR gain this paper tests for thermometry."},{"cited_title":"Didier, A","cited_arxiv_id":null,"evidence_quote":"Establishes the Heisenberg-limited dispersive readout baseline with two-mode squeezed light."},{"cited_title":"Ebadi, T","cited_arxiv_id":null,"evidence_quote":"Gives the SNR and error-propagation definitions used to turn output quadrature statistics into temperature uncertainty."},{"cited_title":"Rao, Linear Statistical Inference and Its Applica - tions (Wiley, NewYork, 1973)","cited_arxiv_id":null,"evidence_quote":"Supplies the Markovian input-noise correlation formalism used in the quantum Langevin derivation."}],"review_version":1}