{"id":"af2c5d71-4ecf-4a07-b416-f8ec32551a56","arxiv_id":"2507.08661","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Steady-state susceptibility to a control parameter yields a universal lower bound on relaxation and second-order correlation times of driven-dissipative quantum systems.","lead":"A quantum metrology bound is repurposed to show that the rate at which an open quantum system's steady state responds to a parameter change lower-bounds both its relaxation time and the correlation time of its emitted light. This gives experimentalists a way to estimate ultrafast correlation times from static measurements, and theorists a route to bound dynamics of many-body models without solving their evolution.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (7) is conditional on assumption (*), which the paper asserts but does not characterize; outside the regime where (*) holds the optimized bound is trivial, so the claimed 'general' steady-state lower bound is not universal.","rationale":"The paper's derivation is a corollary of the metrological bounds in refs [31-33], and the cavity example provides an independent check that the bound is not violated in a solvable linear model. The soft spot is not the QFI inequality itself but the step where Eq. (7) is extracted from the optimized inequality: the interior optimum exists only when (*) holds. The reader identified this as the weakest assumption, and I agree. The abstract and main text present Eq. (7) as a general steady-state formula, while the condition is only mentioned in the derivation and asserted to be 'typical' at criticality without proof. A Kerr-oscillator test would settle whether the bound survives in a nonlinear model where condition (*) can be checked and the exact τc is computable. Since the paper explicitly flags (*) and supplies a numerical validation in the DPA case, the appropriate verdict remains CONDITIONAL rather than REJECT or ACCEPT.","tokens_in":12473,"tokens_out":40520,"duration_ms":467835,"concrete_test":"Simulate a driven-dissipative Kerr oscillator, H=ωa†a+χ a†²a²+Ω(a+a†), with loss γD[a] (a model satisfying all assumptions in Eq. (1)), and use the quantum regression theorem to compute the exact g(2)(τ) and τc for a detuning where condition (*) holds. Compare this exact τc with the right-hand side of Eq. (7) computed from the steady-state photon number and its derivative. If an instance violates the inequality, the bound is not universal; if no violation is found across the parameter scan, the remaining concern is purely the domain restriction (*).","verdict_should_be":"UNCHANGED","load_bearing_attack":"Equation (7) follows from Eq. (5) only after optimizing over the split γ0+γ1=γ and adopting the interior solution γ1=2⟨G0⟩ss/|∂ω⟨G0⟩ss|. This requires condition (*), 2⟨G0⟩ss/|∂ω⟨G0⟩ss| ≤ γ. If (*) fails, the optimum is at γ1=γ, γ0=0, and the QFI constraint (4) is vacuous, so the method produces no nontrivial bound. The paper states that (*) is 'typically satisfied' at critical points, but gives no general argument or counterexample characterization. Consequently, a user who applies Eq. (7) to a generic driven-dissipative system must first verify a condition that depends on the unknown steady-state susceptibility, and no bound of the advertised closed form exists when the condition is violated. This does not invalidate the derivation, but it makes the headline formula a conditional statement rather than a universal steady-state bound.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper derives lower bounds on two dynamical timescales of Markovian driven-dissipative systems: the relaxation time tau_ss and the second-order correlation time tau_c = integral(g^(2)(tau)-1)d tau. The starting point is a quantum-Fisher-information bound for Markovian noise (Eq. (2)), which the authors combine with the signal-to-noise ratio of counting monitored output photons. By splitting the total dissipation rate gamma into an unmonitored part gamma_0 and a monitored part gamma_1, they obtain Eq. (3) for tau_ss and Eq. (7) for tau_c, expressed solely through steady-state expectation values and their derivatives with respect to a parameter omega. A generalization to arbitrary observables is stated as Eq. (10). The method is benchmarked on a lossy parametric resonator, where the exact g^(2) is computed in End Matter III and the bound matches the exact scaling near criticality, and on the infinite-range dissipative transverse-field Ising model, whose steady state is known exactly but whose dynamics is not analytically tractable.","tokens_in":12627,"tokens_out":16418,"duration_ms":182276,"significance":"If the bounds are valid as stated, the paper provides a practical and conceptually interesting bridge: correlation times can be estimated from steady-state susceptibility data, without time-resolved detection or solving the dynamics. The parametric-resonator validation is a genuine strength, since the bound is compared with a fully analytical exact solution and is shown to be tight up to leading order in the relevant regime. The Ising example demonstrates that the method can produce nontrivial statements about a model whose dynamics are otherwise inaccessible. The derivation does not fit any free parameters: the external QFI inequalities are used to constrain the time-integrated correlations, and the examples provide concrete, checkable predictions. The main weakness is that the headline formula Eq. (7) is conditional on an uncharacterized assumption, so the advertised generality is not fully established.","major_comments":[{"comment":"Eq. (7) is not a universal steady-state lower bound as the abstract implies. The optimization over gamma_1 has an interior solution gamma_1 = 2<G0>_ss/|d_omega<G0>_ss| only when condition (*) holds. As the authors note, when (*) fails the optimum moves to the boundary gamma_1=gamma, where gamma_0=0 and Eq. (4) becomes vacuous, so the method produces no nontrivial bound. The statement that (*) is 'typically satisfied' at critical points is not backed by a general criterion or by an explicit characterization of the failure region. Since a user must verify a condition involving the same steady-state susceptibility that the method is intended to exploit, the paper should present Eq. (7) as a conditional theorem, discuss the boundary case explicitly, and provide at least a criterion or an example showing when (*) fails and the bound becomes trivial.","section":"Main results, Eq. (7)"},{"comment":"The derivation of the relaxation-time bound identifies the estimation time t in Eq. (2) with the relaxation time tau_ss without giving an operational definition of tau_ss. If tau_ss is a hitting time at which the state exactly reaches the steady state, then Eq. (3) follows from Eq. (2) applied at t=tau_ss, but the use of <O>_ss and Delta^2_ss O requires the state to be exactly steady at that time. If tau_ss is instead intended as a dynamical timescale such as the inverse Liouvillian gap, the inequality does not follow directly. Please state precisely what Eq. (3) bounds and how tau_ss is defined; the example comparison is only at the level of scaling and does not resolve this ambiguity.","section":"Main results, Eq. (3)"},{"comment":"Eq. (10) is listed as one of the three main results, but its hypotheses are not stated. It is unclear what class of observables A is allowed (a system operator, an output-field operator, or both), how the constant C[d_omega H,{L_i}] is computed from the Lindblad operators, and under what conditions the steady-state autocorrelation integral is finite or the inequality holds. A reader cannot apply or assess Eq. (10) without returning to the metrology literature. The authors should either derive Eq. (10) in the End Matter or state it as a theorem with explicit conditions and a clear reference to the corresponding result in Refs. [31-33].","section":"General observable of the output field, Eq. (10)"}],"minor_comments":[{"comment":"The derivation of Eq. (5) uses gamma in place of gamma_1 in the variance formula, while Eq. (5) and the preceding text use gamma_1 for the monitored output mode. As printed, the appendix does not reproduce Eq. (5); the notation for the measured output mode should be made consistent.","section":"End Matter I, Eqs. (23)-(24)"},{"comment":"The phrase 'general lower bounds' in the abstract overstates the scope of Eq. (7), which requires condition (*). The abstract should either mention the condition or the main text should qualify the claim in the introduction.","section":"Abstract and introduction"},{"comment":"The caption states that the density plot was generated using 'Eqs. (14) and (S39) of [40]'; this appears to be a typo, since Eq. (14) of the present paper is the relaxation-time bound, not the correlation-time formula plotted. It should likely refer to Eq. (19) of this paper and Eq. (S39) of Ref. [40].","section":"Fig. 3 caption"},{"comment":"The inset caption contains a duplicated word: 'calculated for for fixed values'. In addition, the claim that the variance of m converges to a finite constant in the thermodynamic limit with Delta^2_ss m approximately 0.16 is stated without derivation or citation; a brief justification would be helpful.","section":"Fig. 3 inset caption"},{"comment":"The role of the Lindbladian M[rho] is described ambiguously: the text first says it models the coupling to the measurement apparatus, then says it accounts for effects related to interaction with a controlled field. These two descriptions should be reconciled, since the distinction between L and M is later used to make Eq. (9) independent of M.","section":"General setting, Eq. (8)"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of a quantum-physics journal and the self-citation to Ref. [33] is appropriate, as it is prior published work that supplies the underlying metrology bound. The central derivation is a legitimate rearrangement of known QFI inequalities, and the examples are concrete and reproducible. The main obstacle to acceptance is the gap between the abstract's claim of generality and the conditional nature of Eq. (7); I do not see grounds for rejection if the authors make the hypotheses precise and clarify the status of Eq. (3) and Eq. (10)."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a careful read. The main result—lower bounds on the relaxation time τss and the second-order correlation time τc from steady-state expectation values and their parameter derivatives—is genuinely new. The derivation is a legitimate application of QFI bounds from refs [31–33], not a circular repackaging: the target times appear only after rearranging the inequality. The resonator example is the right kind of calibration: they compute the exact τc, show the bound scales correctly, and even check the difference is nonnegative in a short inequality. The Ising example demonstrates the tool on a model where only the steady state is solvable. That is real value.\n\nThe soft spot the stress-test flags is real. Eq. (7) is obtained by optimizing over the split γ0+γ1=γ and choosing the interior solution γ1=2⟨G0⟩ss/|∂ω⟨G0⟩ss|. That requires condition (*): 2⟨G0⟩ss/|∂ω⟨G0⟩ss| ≤ γ. The paper asserts (*) is 'typically satisfied' near critical points but gives no characterization of where it fails. If (*) fails, the optimum is at γ1=γ, the QFI constraint is vacuous, and the advertised closed-form bound does not exist. A user applying Eq. (7) to a generic driven-dissipative system must first verify a condition that depends on the unknown steady-state susceptibility—exactly the kind of quantity the method is supposed to avoid. This is a genuine limitation, not a fatal one; the derivation is sound and the general bound could be stated in a less clean form.\n\nTwo minor issues: the identification of the measurement time with τss in the derivation of Eq. (3) is heuristic, and the SNR formula relies on an exponential-decay approximation for g(2)−1. These are standard but deserve a clearer caveat. The citation pattern is fine: the only self-citation is ref [33] for the underlying QFI bound, which is prior published work, and there are no fitted parameters.\n\nThis paper deserves a serious referee. I would send it to review with a request that the authors sharpen the conditions for Eq. (7) and state the general bound. I would not cite it in my own work until that is clarified.","headline":"Genuinely useful conditional bound on correlation times from steady-state data, but Eq. (7)'s condition (*) is under-characterized and the paper oversells its universality.","tokens_in":13188,"tokens_out":2937,"would_cite":false,"duration_ms":29341,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Steady-state measurements alone bound a system's correlation times","keywords":["steady-state correlation bounds","second-order correlation time","relaxation time","quantum metrology","driven-dissipative systems","critical quantum systems","open quantum systems"],"falsifier":"Compute the left-hand side of Eq. (7) from the exact dynamics of a Markovian driven-dissipative system (e.g., a parametrically driven Kerr resonator) over a range of parameters where the steady-state derivative $|\\partial_\\omega \\langle G_0\\rangle_{ss}|$ is large but the condition $(*)$ is violated; if the inequality is ever found to be violated when $(*)$ holds, the claim would be falsified. Alternatively, check the bound against a numerically exact solution for a many-body model where the steady state is known but the dynamics is solvable by time-dependent variational Monte Carlo.","tokens_in":12222,"feed_emoji":"⏱️","tokens_out":1839,"duration_ms":20410,"temperature":0.7,"pith_summary":"This paper establishes that for a broad class of Markovian open quantum systems, the relaxation time and the second-order correlation time of emitted light are lower-bounded by quantities computable entirely from steady-state expectation values and their derivatives with respect to a control parameter. The bounds follow from quantum metrology's fundamental precision limits, reinterpreted as a statement about temporal correlations: if a steady state is more sensitive to a parameter change than the uncorrelated-measurement bound allows, that excess sensitivity must be carried by temporal correlations. The authors demonstrate the bounds on a driven-dissipative resonator, where they nearly saturate the exact correlation time, and on the infinite-range dissipative Ising model, where the dynamics are otherwise intractable but the steady state is known.","feed_headline":"Steady-state data alone reveal a system's correlation times","feed_subtitle":"New bounds on relaxation and bunching times use only the steady-state photon flux and its parameter derivative.","key_machinery":"The central mechanism is the quantum metrology bound on signal-to-noise ratio for continuously monitored Markovian systems: the quantum Fisher information accumulated over time $t$ is bounded by $4/\\gamma_0 \\int_0^t \\langle G_0\\rangle_{t'} dt'$ for the unobserved fraction of photons. The paper treats the same physical system as coupled to an unmonitored environment (rate $\\gamma_0$) and an ancilla representing measurable photons (rate $\\gamma_1$), with $\\gamma_0+\\gamma_1=\\gamma$. If the measurement outcomes were temporally uncorrelated, the SNR would grow only linearly with the number of samples; any super-linear growth forced by the bound must be supplied by the second-order correlation function. Inverting this logic yields the lower bounds on $\\tau_{ss}$ and $\\tau_c$. The optimization over $\\gamma_1$ gives the closed-form bound of Eq. (7) when the condition $2\\langle G_0\\rangle_{ss}/|\\partial_\\omega \\langle G_0\\rangle_{ss}| \\le \\gamma$ holds, which the authors argue is typical at critical points.","core_discovery":"The paper derives a universal relation between steady-state parameter sensitivity and two key timescales. For a Hamiltonian $H = \\omega G_0 + G_1$ with $G_0 \\ge 0$ and photon emission rate $\\gamma\\langle G_0\\rangle$, the relaxation time is bounded by $\\tau_{ss} \\ge \\frac{\\gamma}{4\\langle G_0\\rangle_{\\max}} \\frac{|\\partial_\\omega \\langle O\\rangle_{ss}|^2}{\\Delta^2_{ss} O}$ for any observable $O$. For the emitted field, the second-order correlation time $\\tau_c := \\int_{-\\infty}^{+\\infty} [g^{(2)}(\\tau)-1]d\\tau$ satisfies, under a stated condition, $\\tau_c \\ge \\frac{1}{\\langle G_0\\rangle_{ss}}\\left[\\frac{\\gamma |\\partial_\\omega \\langle G_0\\rangle_{ss}|^2}{4\\langle G_0\\rangle_{ss}^2} - \\frac{|\\partial_\\omega \\langle G_0\\rangle_{ss}|}{\\langle G_0\\rangle_{ss}}\\right]$. Both formulas involve only steady-state values and parameter derivatives, so they can be evaluated without solving the dynamics. The paper shows that near critical points the bound typically diverges, signaling strong bunching, and it generalizes the argument to arbitrary observables via a constant $C[\\partial_\\omega H, \\{L_i\\}]$ that depends only on the Hamiltonian perturbation and Lindblad operators.","pith_inferences":["A testable extension is to apply the bound to a driven-dissipative system where the steady state is known only numerically and compare the bound against direct Monte Carlo simulations of the dynamics; the inequality should hold for all parameters, providing a consistency check.","The method might be inverted: if an experiment can resolve the correlation time but not the steady-state derivative, the bound gives a nontrivial constraint on the susceptibility, effectively a thermodynamic uncertainty relation.","The condition $2\\langle G_0\\rangle_{ss}/|\\partial_\\omega \\langle G_0\\rangle_{ss}| \\le \\gamma$ is the regime where the optimized $\\gamma_1$ is in the interior; outside it, the bound takes a different, more complicated form, but the underlying SNR inequality still holds, so a weaker bound remains available.","The general observable bound of Eq. (10) may be useful for non-photon observables such as order parameters in dissipative phase transitions, connecting critical metrology sensitivity to the autocorrelation time of the order parameter itself."],"forward_implications":["Experimentalists can estimate correlation times of ultrafast systems by measuring only steady-state photon flux and its derivative with respect to a tunable parameter, bypassing sub-picosecond detector resolution.","For many-body models whose steady state is solvable but whose dynamics are not, the bounds provide analytic constraints on relaxation and bunching timescales without solving the master equation.","Near quantum critical points, the divergent steady-state susceptibility automatically implies divergent correlation and relaxation times, making the connection between criticality and long-lived temporal correlations quantitative.","The method generalizes to autocorrelations of arbitrary observables, so the same technique can bound time-correlation properties beyond photon counting, such as spin or charge autocorrelations.","The bounds are universal in the sense that they do not depend on the specific form of the driving Hamiltonian $G_1$, only on its effect on the steady state."],"supporting_citations":[{"why":"Provides the fundamental bound on adaptive quantum metrology under Markovian noise, which is the basis for Eq. (2).","marker":"[31]"},{"why":"Supplies a related bound on adaptive quantum metrology under Markovian noise, complementing Ref. [31].","marker":"[32]"},{"why":"The authors' own recent work on bosonic noisy quantum metrology, giving the specific bound used for the photon-counting observable.","marker":"[33]"},{"why":"Provides the exact steady-state solution of the infinite-range dissipative transverse-field Ising model, which the paper uses to evaluate its bounds.","marker":"[40]"},{"why":"Provides the analytical solution for the lossy parametric resonator used to validate the relaxation-time bound.","marker":"[39]"},{"why":"Supplies the standard quantum-optics input-output formalism and the expression for the thermal correlation function used as a comparison.","marker":"[44]"}],"fun_headline_variants":["Correlation times from steady-state data alone","Bounds on correlation times without dynamics","Steady-state measurements unlock correlation timescales","Relaxation times from steady-state parameter sensitivity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The headline bound on $\\tau_c$ assumes the condition $2\\langle G_0\\rangle_{ss}/|\\partial_\\omega \\langle G_0\\rangle_{ss}| \\le \\gamma$, which the paper says is typically satisfied at critical points but does not systematically characterize when it fails.","fun_headline_variants_meta":{"raw":{"variants":["Correlation times from steady-state data alone","Bounds on correlation times without dynamics","Steady-state measurements unlock correlation timescales","Relaxation times from steady-state parameter sensitivity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000585,"raw_usage":{"total_tokens":2789,"prompt_tokens":1020,"completion_tokens":1769,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":636,"completion_tokens_details":{"reasoning_tokens":1714}},"tokens_in":636,"tokens_out":1769,"duration_ms":16225,"temperature":1.0,"reasoning_tokens":1714,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:15:06.449893+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the left-hand side of Eq. (7) from the exact dynamics of a Markovian driven-dissipative system (e.g., a parametrically driven Kerr resonator) over a range of parameters where the steady-state derivative $|\\partial_\\omega \\langle G_0\\rangle_{ss}|$ is large but the condition $(*)$ is violated; if the inequality is ever found to be violated when $(*)$ holds, the claim would be falsified. Alternatively, check the bound against a numerically exact solution for a many-body model where the steady state is known but the dynamics is solvable by time-dependent variational Monte Carlo.","supporting_citations":[{"cited_title":"Demkowicz-Dobrza ´nski, J","cited_arxiv_id":null,"evidence_quote":"Provides the fundamental bound on adaptive quantum metrology under Markovian noise, which is the basis for Eq. (2)."},{"cited_title":"Wan and R","cited_arxiv_id":null,"evidence_quote":"Supplies a related bound on adaptive quantum metrology under Markovian noise, complementing Ref. [31]."},{"cited_title":"G ´orecki, F","cited_arxiv_id":null,"evidence_quote":"The authors' own recent work on bosonic noisy quantum metrology, giving the specific bound used for the photon-counting observable."},{"cited_title":"Roberts and A","cited_arxiv_id":null,"evidence_quote":"Provides the exact steady-state solution of the infinite-range dissipative transverse-field Ising model, which the paper uses to evaluate its bounds."},{"cited_title":"Alushi, W","cited_arxiv_id":null,"evidence_quote":"Provides the analytical solution for the lossy parametric resonator used to validate the relaxation-time bound."}],"review_version":1}