Pseudomode Mapping in Quantum Dynamics

• Pseudomode mapping is a technique that discretizes structured non-Markovian reservoirs into finite, damped auxiliary modes to simplify simulation and analysis.
• It leverages spectral decomposition and residue calculus to represent bath correlations as sums of exponentials, converting nonlocal memory effects into a Lindblad framework.
• Applications span quantum optics, condensed matter, and quantum impurity models, with extensions addressing nonlinear couplings and finite temperature effects.

Pseudomode Mapping

Pseudomode mapping is a mathematical and physical technique that replaces a structured, non-Markovian reservoir—typically a continuum of bosonic or fermionic modes—with a finite set of discrete, damped auxiliary modes called pseudomodes. This transformation renders an originally nonlocal, memory-kernel-driven open quantum dynamics into a Lindblad-type (or quasi-Lindblad) Markovian equation on an enlarged Hilbert space, enabling exact or highly accurate simulation, analytic insight, and efficient numerical implementation. The method is applicable to both quantum optical and condensed matter settings, with extensions to nonlinear system-bath couplings, finite temperature, and quantum impurity models. The foundational logic and rigorous justification are provided through contour integration over the bath spectral density and analytic properties of the bath correlation function, as formalized in the works of Garraway and many subsequent developments (Behzadi et al., 2017, Huang et al., 12 Jun 2025, Cirio et al., 2023, Pleasance et al., 2020, Pleasance et al., 2021, Thoenniss et al., 2024, Alford et al., 19 Sep 2025, Park et al., 2024, Menczel et al., 2024, Liang et al., 2024, Zhang et al., 25 Feb 2025).

1. Mathematical Framework and Spectral Decomposition

Pseudomode mapping is grounded on the analytic structure of the environmental spectral density $J(\omega)$, typically assumed meromorphic in the relevant half-plane. The total system-plus-environment Hamiltonian is

$$H = H_S + \sum_k \omega_k\,b_k^\dagger b_k + \sum_k \left( g_k\, L\, b_k^\dagger + g_k^*\, L^\dagger\, b_k \right)$$

where $H_S$ is the system Hamiltonian and $L$ its coupling operator. The reduced system dynamics is determined by the bath correlation function

$$C(t) = \langle X(t) X(0) \rangle = \int_0^\infty d\omega\, J(\omega)\, e^{-i\omega t}$$

with $X = \sum_k (g_k\, b_k^\dagger + g_k^*\, b_k)$. This $C(t)$ typically exhibits a sum-of-exponentials representation

$$C(t) \approx \sum_{j=1}^N w_j\, e^{-i\Omega_j t - \Gamma_j |t|}, \qquad w_j\in\mathbb{C}$$

for appropriate $N$, with $\Omega_j$ and $\Gamma_j$ traceable to the poles $z_j$ of $J(\omega)$ (or $J(\omega)\coth(\beta\omega/2)$ at finite temperature) in the complex plane (Pleasance et al., 2021, Behzadi et al., 2017, Park et al., 2024). This decomposition can be achieved analytically (residue calculus for rational $J(\omega)$), via Prony or ESPRIT fitting, or robust Loewner/SVD plus semidefinite programming (Huang et al., 12 Jun 2025, Thoenniss et al., 2024).

2. Construction and Properties of the Auxiliary Pseudomode Model

Each exponential component in $C(t)$ is mapped to a discrete auxiliary (pseudomode) oscillator $a_j$ with frequency $\Omega_j$ and damping rate $\Gamma_j$, linearly coupled to the system with strength $g_j = \sqrt{w_j}$ (for real-positive $w_j$). The enlarged Hamiltonian and Lindblad equations become

$$H_{\rm pm} = H_S + \sum_{j=1}^N \Omega_j a_j^\dagger a_j + \sum_{j=1}^N (g_j L a_j^\dagger + g_j^* L^\dagger a_j)$$

$$\dot\rho = -i [H_{\rm pm}, \rho] + \sum_{j=1}^N \Gamma_j \Big( a_j \rho a_j^\dagger - \frac{1}{2} \{ a_j^\dagger a_j, \rho \} \Big)$$

Tracing out the pseudomodes recovers the original system dynamics, including all non-Markovian memory effects as encoded in $C(t)$ (Pleasance et al., 2021, Cirio et al., 2023, Behzadi et al., 2017). For more complex structured environments, the pseudomodes may be coupled via a non-Hermitian matrix $W$; the effective spectral density is then

$$J_{\rm eff}(\omega) = 2\,\mathrm{Im}[\zeta^\dagger (i W - \omega )^{-1} \zeta]$$

with system–pseudomode coupling vector $\zeta$ and pseudomode generator $W$ (Alford et al., 19 Sep 2025, Pleasance et al., 2020).

3. Algorithmic and Numerical Aspects

The practical workflow for pseudomode construction is:

1. Compute or fit the bath correlation function $C(t)$ over the required time interval.
2. Approximate $C(t)$ by a (minimal) sum of exponentials to target error $\varepsilon$ (Prony/AAA/ESPRIT/Loewner-SVD/SDP).
3. Assign each term to a pseudomode: extract frequencies, dampings, and couplings.
4. Formulate the joint Markovian master equation for system+pseudomodes.
5. Simulate this Lindblad equation via master-equation solvers, quantum trajectory methods, tensor-network time evolution, or (in the case of fermionic baths) operator-state mapping (Thoenniss et al., 2024, Huang et al., 12 Jun 2025).
6. Trace out the pseudomodes to recover the observable system dynamics.

For analytic $J(\omega)$, the number of required pseudomodes scales as $N = O(\mathrm{polylog}(T/\varepsilon))$ for simulation time $T$ and tolerance $\varepsilon$, with rigorous error bounds on both correlation and state fidelity (Huang et al., 12 Jun 2025, Thoenniss et al., 2024). Further compression via interpolative decomposition or optimal gauge choices minimizes computational costs (Thoenniss et al., 2024, Alford et al., 19 Sep 2025).

4. Generalizations: Fermions, Nonlinear Couplings, and Stochastic Decomposition

Pseudomode mapping generalizes to structured fermionic baths (quantum impurity models), where the Feynman–Vernon hybridization kernel is decomposed into exponentials—each mapped onto a discrete fermionic pseudomode, yielding analogous Markovian Liouvillians for the impurity plus modes (Thoenniss et al., 2024). For nonlinear system-bath interactions,

$$H_{\rm SB} = s\, Q(X), \quad Q(X) = \sum_{n=1}^\infty \alpha_n X^n$$

the correlation structure entering high-order moments must be matched, requiring additional zero-frequency or "purified" modes to ensure correct reproduction of both $C(t)$ and $\langle X^2 \rangle$ (Zhang et al., 25 Feb 2025). In scenarios where representation of certain symmetric (classical) components or Matsubara terms of the bath correlation is needed, a quantum–classical decomposition is employed: the quantum part is handled by pseudomodes, and the symmetric remainder by a classical stochastic field driving the system within a stochastic master equation framework (Luo et al., 2023, Cirio et al., 2023).

5. Physical Interpretation, Applications, and Limitations

Pseudomodes serve as minimal, memory-storage intermediaries between system and "flat" residual baths, mediating all non-Markovian memory effects via their amplitude evolution (Ohyama et al., 2017, Roy et al., 3 Jul 2025). In spin-chain transport, inclusion of auxiliary pseudomodes produces decoherence-free subspaces and greatly enhances energy transfer efficiency, especially as the network size increases (Behzadi et al., 2017). In quantum impurity models, the mapping provides nearly-optimal complexity scaling and a platform for tensor-network or quantum-algorithm implementation (Thoenniss et al., 2024). In quantum thermodynamics, the approach enables the evaluation of heat, work, and entropy production in the strong-coupling regime via one-time expectation values of system and pseudomodes (Albarelli et al., 2024). The input–output pseudomode formalism and its purification support the analysis of non-Gaussian baths and multi-photon processes (Liang et al., 2024).

Limitations arise for spectral densities with branch cuts (non-meromorphic) or non-Lorentzian singularities, where completeness issues for the Lorentzian basis prevent uniform convergence; coupled or non-diagonalizable pseudomode arrangements can partially remedy this but do not guarantee full generality (Alford et al., 19 Sep 2025). For high-temperature baths or strong nonlinearity, additional care is needed to represent the requisite higher-order cumulants or to compensate for "unphysical" mode parameters via analytic continuation and measurement-based extrapolation (Cirio et al., 2023, Zhang et al., 25 Feb 2025).

6. Connections to Markovian Embedding, Quasi-Lindblad Extensions, and Scattering Theory

The pseudomode formalism systematically converts non-Markovian quantum processes into finite Markovian (GKSL-type) evolutions on an enlarged space, providing exact simulation for meromorphic $J(\omega)$ and a controlled approximation for analytic $J(\omega)$ (Pleasance et al., 2020, Huang et al., 12 Jun 2025). Recent formulations include quasi-Lindblad mappings, where the requirement of complete positivity on the global master equation is relaxed, so that the Lindblad dissipators may acquire non-CP components as long as the reduced system evolution remains physical and stable (Park et al., 2024, Menczel et al., 2024). The mapping possesses substantial gauge freedom in partitioning the bath correlation weights among Hamiltonian and dissipative terms, with implications for numerical stability and efficiency.

In non-interacting systems, the pseudomode mapping translates into a block-structured NEGF, with effective spectral densities and transmission coefficients expressed via pseudomode Green's functions (Alford et al., 19 Sep 2025). This embedding maintains the Landauer–Büttiker form for scattering observables, provided that the pseudomode parameters are optimally tuned to reproduce the true $J(\omega)$.

---

References:

• Garraway, B.M., Phys. Rev. A 55, 2290 (1997); (Behzadi et al., 2017)
• Tamascelli et al., Phys. Rev. Lett. 120, 030402 (2018); (Huang et al., 12 Jun 2025)
• Pleasance & Petruccione; (Pleasance et al., 2021, Pleasance et al., 2020)
• Thoenniss et al.; (Thoenniss et al., 2024)
• Alford et al.; (Alford et al., 19 Sep 2025)
• Menczel et al.; (Menczel et al., 2024)
• Liang et al.; (Zhang et al., 25 Feb 2025, Liang et al., 2024)
• Brask et al.; (Park et al., 2024)
• Cirio et al.; (Cirio et al., 2023)
• Teretenkov; (Teretenkov, 2019)