Unified Fluctuation–Response Relation

• The paper introduces a unified fluctuation–response relation that bounds observable responses using dynamical activity via the Cramér–Rao bound.
• It employs both kinetic and entropic perturbations in Markov jump processes and quantum Lindblad systems to derive rigorous fluctuation–response inequalities.
• The framework bridges the fluctuation–dissipation theorem and thermodynamic uncertainty relations, offering tight constraints for finite-time, non-steady-state observables.

A unified fluctuation–response relation generalizes the equilibrium fluctuation–dissipation theorem (FDT) to nonequilibrium classical and quantum stochastic systems, linking the response of observables under perturbations to ensemble fluctuations in the presence of both kinetic (symmetric) and entropic (antisymmetric) parameter variations. In Markov jump processes, this relation yields rigorous fluctuation–response inequalities (FRIs) bounded by dynamical activity (traffic), and extends naturally to open quantum systems governed by Lindblad master equations, where the quantum FRI similarly features quantum dynamical activity as the central control parameter. This unifies static and dynamic responses, bridging the FDT and thermodynamic uncertainty relations, and is applicable to finite-time, non-steady-state, and a broad class of observables.

1. Framework and Mathematical Formulation

Consider a continuous-time Markov jump process with transition rates $W_{ij}$ ($i,j \in \{1,...,N\}$), steady-state distribution $\pi_i$, and a general time-extensive observable: $$\Theta(\tau) = \int_0^\tau dt \left[\sum_i g_i\,\eta_i(t) + \sum_{i\neq j} \Lambda_{ij}\, \dot N_{ij}(t)\right],$$ where $\eta_i(t) = \delta_{s(t), i}$ and $\dot N_{ij}(t)$ represents the instantaneous rate for jumps $j \to i$. Two classes of rate-parameter perturbations are distinguished:

• Kinetic (symmetric) perturbations: $B_{ij} = B_{ji}$ enter as $$W_{ij} = \exp\bigl(B_{ij}(\epsilon) + F_{ij}(\eta)/2\bigr)$$.
• Entropic (antisymmetric) perturbations: $F_{ij} = -F_{ji}$.

The finite-time response to these perturbations is defined as

$$R_{B_{ij}}(\tau) = \left. \frac{\partial}{\partial B_{ij}}\langle \Theta(\tau) \rangle \right|_0, \quad R_{F_{ij}}(\tau) = \left. \frac{\partial}{\partial F_{ij}}\langle \Theta(\tau) \rangle \right|_0.$$

Define the steady-state traffic (dynamical activity) on edge $(i<j)$,

$$a_{ij} = W_{ij} \pi_j + W_{ji} \pi_i, \quad \dot A = \sum_{i<j} a_{ij}.$$

2. Unified Fluctuation–Response Inequality

The central result is the pair of fluctuation–response inequalities: $$\sum_{i<j} \frac{R_{B_{ij}}(\tau)^2}{\tau\, a_{ij}} \leq \mathrm{Var}[\Theta(\tau)], \qquad \sum_{i<j} \frac{R_{F_{ij}}(\tau)^2}{\tau\, a_{ij}/4} \leq \mathrm{Var}[\Theta(\tau)].$$ These bounds are obtained via the Cramér–Rao bound for path-space Fisher information using the path probability: $$\mathcal{P}[\Gamma_\tau] = \pi_{s(0)} \exp\left(-\int_0^\tau dt \sum_{i\neq j} \left[\eta_j(t) W_{ij} - \dot N_{ij}(t) \ln W_{ij}\right]\right).$$ The Fisher information is diagonal, with the relevant entries for $B_{ij}$ and $F_{ij}$ being $\tau a_{ij}$ and $\tau a_{ij}/4$, respectively.

3. Static, Dynamic, and Observable Generality

The FRIs:

• Apply for arbitrary finite $\tau$ (dynamic response) as well as for $\tau\to\infty$ (static/steady-state susceptibility).
• Require only time-homogeneous rates, initial steady-state sampling, and linear response in the perturbation parameter.
• Hold for all $\Theta$, including current-like ($g_i=0$, $\Lambda_{ij}=-\Lambda_{ji}$) and state-dependent ($\Lambda_{ij}=0$) observables.

Saturation occurs for current-like observables in both short $\tau \to 0$ and long $\tau \to \infty$ limits; state-dependent observables generally only saturate for large $\tau$.

4. Bounds for Collective Perturbations

Under collective kinetic perturbations $B_{ij} \to B_{ij} + \epsilon f_{ij}$,

$$\mathrm{Var}[\Theta(\tau)] \geq \frac{R_\epsilon(\tau)^2}{\tau \sum_{i<j} f_{ij}^2 a_{ij}} \geq \frac{R_\epsilon(\tau)^2}{\tau f_\mathrm{max}^2 \dot A},$$

where $$R_\epsilon(\tau) = \sum_{i<j} f_{ij} R_{B_{ij}}(\tau)$$.

For entropic perturbations $F_{ij} \to F_{ij} + \eta h_{ij}$,

$$\mathrm{Var}[\Theta(\tau)] \geq \frac{4 R_\eta(\tau)^2}{\tau \sum_{i<j} h_{ij}^2 a_{ij}} \geq \frac{4 R_\eta(\tau)^2}{\tau h_\mathrm{max}^2 \dot A},$$

where $R_\eta(\tau) = \sum_{i<j} h_{ij} R_{F_{ij}}(\tau)$.

These forms yield practical kinetic and entropic uncertainty bounds on the achievable response of the system given fixed dynamical activity and connect to the thermodynamic uncertainty relation (TUR) structure in the far-from-equilibrium regime.

5. Quantum Generalization: Lindblad Evolution

For open quantum systems governed by Lindblad quantum master equations,

$$\dot\rho = -i[H, \rho] + \sum_{k=1}^{K} \mathcal{D}[L_k^{\theta_k}]\rho, \qquad L_k^{\theta_k} = e^{\theta_k/2} L_k,$$

the quantum dynamical activity for each channel is

$$a_k = \mathrm{tr}[L_k^{\theta_k} \rho_\mathrm{ss} (L_k^{\theta_k})^\dagger], \quad \dot A_Q= \sum_{k=1}^K a_k.$$

The quantum Fisher matrix is diagonal: $[\mathcal{I}_Q]_{kk} = \tau a_k$. The quantum FRI is

$$\sum_{k=1}^K \frac{R_{\theta_k}(\tau)^2}{\tau a_k} \leq \mathrm{Var}[\Theta(\tau)],$$

demonstrating the equivalence of the role of dynamical activity in bounding response in both classical and quantum cases.

6. Equilibrium, Far-From-Equilibrium, and TUR Limits

• In the near-equilibrium limit ($\dot\Sigma / \dot A \ll 1$), the combined kinetic and entropic bounds reduce to the equilibrium FDT.
• Far from equilibrium ($\dot\Sigma/\dot A \gg 1$), the bounds reduce to a TUR-type constraint.
• The result interpolates smoothly between FDT and TUR, unifying the two under a single information-theoretic framework rooted in the Cramér–Rao bound.

7. Physical Interpretation and Impact

Dynamical activity (traffic) emerges as the fundamental resource governing the tightness of the fluctuation–response bound for both classical and quantum systems. This places kinetic “frenesy” on equal footing with entropy production for constraining system sensitivity, regardless of the observable or perturbation. The inequalities are directly applicable to finite-time, non-steady-state, and quantum Markovian systems, and recover standard fluctuation–dissipation relations, nonequilibrium current-response equalities, and thermodynamic uncertainty relations as special cases. No additional model-specific structure is required, emphasizing the universality and model-independence of these unified fluctuation–response relations (Kwon et al., 2024).