Liouville Fock state lattices and potential simulators

Caio B. Naves, Jonas Larson · 2025

Read Liouville Fock state lattices and potential simulators by Caio B. Naves, Jonas Larson (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-22047.txt.

Introduction

Open quantum systems leak energy and information to their surroundings, which makes their math harder than that of closed systems. Caio B Naves and Jonas Larson propose a fresh way to picture that math in this 2025 paper. They recast the standard equation for such systems as a grid, called a Liouville Fock state lattice, where the state moves much like particles on a classical lattice.

On this grid, states drift, gather, and drain in ways that closed-system states never do, because the grid's rules are not even-handed. The authors also try a second picture that looks more like a spread of chance values. They show that endless sets of steady states can arise from a kind of frustration in the grid. The framework may help design simulators that mimic open quantum behavior.

Catalog record

CREATORCaio B. Naves, Jonas Larson
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE90 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-22047

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Liouville Fock state lattices and potential simulators Caio B. Naves, Jonas Larson Abstract We introduce Liouville Fock state lattices (LFSLs) as a framework for visualizing open quantum systems through matrix representations of the Lindblad master equation (LME). By vectorizing the LME, the state evolves in a doubled Hilbert space, naturally forming a synthetic lattice. Unlike the unitary evolution of pure states, LFSL states exhibit nontrivial dynamics due to the non-Hermitian Liouvillian, featuring population drifts, sources, and sinks—paralleling stochastic classical lattices. We explore these ”classical simulators” in both the Fock representation and alternative positive semidefinite representations, which more closely resemble classical probability distributions. We further demonstrate how infinite steady state manifolds can derive from frustration in the LFSL. Introduction – Lattice models are essential tools for describing a wide range of physical phenomena across various branches of physics, including statistical mechanics, optics, condensed matter, and high-energy physics. A notable example is quantum simulators, which often rely on discretized lattice frameworks [ 1 , 2 ] . The precise control achievable in ultracold atomic and ionic systems has made these simulators a practical reality. Over the past two decades, quantum simulators have enabled groundbreaking experiments, from realizing canonical condensed matter models [ 3 , 4 , 5 ] to exploring quantum thermalization and many-body localization [ 6 , 7 , 8 ] . Recently, quantum simulation has expanded further to incorporate photonic systems and address applications in high-energy physics and quantum chemistry [ 9 ] . Building on the flexibility of these quantum systems, researchers have proposed methods for creating synthetic dimensions by exploiting internal degrees of freedom [ 10 , 11 , 12 , 13 , 14 ] . These synthetic dimensions allow for the study of phenomena unique to higher-dimensional systems, which are typically absent in traditional three-dimensional setups [ 15 , 16 , 17 , 18 , 19 ] . Synthetic dimensions have been proposed and experimentally realized in various platforms, including cold atoms [ 10 , 11 , 20 , 21 , 22 , 23 , 24 , 25 ] , trapped ions [ 26 , 27 ] , magnons [ 28 ] , plasmonics [ 29 ] , cavity and circuit QED systems [ 12 , 30 , 17 , 32 , 33 , 14 , 34 ] , electrical circuits [ 35 ] , and optical systems [ 36 , 37 , 38 , 13 , 39 , 40 , 41 , 42 ] . Typically, these synthetic dimensions correspond to internal quantum states, forming what are known as Fock state lattices (FSLs). In many systems, particularly in quantum optics, such lattices lack translational invariance due to state-dependent tunneling rates. Nevertheless, essential properties such as topology are governed by the overall lattice geometry rather than by translational symmetry [ 30 , 43 , 42 , 32 , 17 , 14 ] . This makes FSLs a powerful platform for exploring physical phenomena in state space rather than real space. While the FSL approach often reduces the problem to a single-particle framework, it has been experimentally demonstrated that computationally challenging problems can also be addressed using FSLs [ 44 ] , further showcasing their potential for quantum simulation. In this Letter, we extend the concept of FSLs to open quantum systems by introducing LFSLs derived from the LME. Extending phenomena from closed to open quantum systems presents unique challenges, particularly due to the role of quantum coherences in mixed states [ 45 ] . These challenges are well recognized in areas such as entanglement and topological properties of mixed states [ 46 , 47 , 48 , 49 ] , coherent control of open systems [ 50 ] , and defining thermodynamic quantities [ 51 ] . The difficulty arises from the fact that general density matrices \hat{\rho} span a much larger state space than pure states |\psi\rangle . Unlike traditional FSLs, where vectors are represented in terms of probability amplitudes, the vectors in LFSLs require a fundamentally different physical interpretation. While these differences introduce additional complexities, they also offer new opportunities beyond Hermitian systems to explore novel phenomena, like anomalous transport [ 52 , 53 ] and classical lattices/networks with biased complex weights [ 54 , 55 , 56 , 57 ] . To demonstrate the power of this approach, we provide examples, including the emergence of frustration within the extended space of LFSLs. Liouville-Fock-state lattices – Our approach is based on the LME [ 58 , 59 ] , given by \frac{d\hat{\rho}}{dt}=\mathcal{L}\left[\hat{\rho}\right]\equiv-i\left[\hat{H},\hat{\rho}\right]+\sum_{j}\gamma_{j}\left(\hat{L}_{j}\hat{\rho}\hat{L}_{j}^{\dagger}-\frac{1}{2}\left\{\hat{L}_{j}^{\dagger}\hat{L}_{j},\hat{\rho}\right\}\right), (1) where \mathcal{L} is the Liouvillian superoperator, \hat{H} is the system Hamiltonian, \hat{L}_{j} are the Lindblad jump operators, and \gamma_{j} are the corresponding decay rates (we use hats on normal operators and calligraphic font for superoperators). The Liouvillian, being a superoperator that maps operators to other operators, makes it challenging to interpret the LME in terms of a FSL. To address this, we adopt the vectorized form of the LME. This approach involves mapping the density matrix \hat{\rho} to a ”superstate” \mathinner{|{\rho}\rangle\!\rangle} by transforming bras into kets, such that \outerproduct{\psi_n}{\psi_m}\mapsto\ket{\psi_n}\otimes\ket{\psi_m}^{*} . This mapping is a specific instance of the more general Choi–Jamiołkowski isomorphism [ 60 , 61 ] . In this representation, the vectorized state \mathinner{|{\rho}\rangle\!\rangle} resides in the doubled Hilbert space \mathfrak{L}=\mathfrak{H}\otimes\mathfrak{H}^{*} , commonly referred to as the Liouville space [ 1 , 2 ] . To proceed, we assume the existence of a natural number basis \{|n\rangle;\,n\in\mathbb{N}\} , corresponding to Fock states. For a single degree of freedom, the vectorized density matrix is then expanded as \mathinner{|{\rho}\rangle\!\rangle}=\sum_{n,m}\rho_{n,m}\mathinner{|{n,m}\rangle\!\rangle} , where the Fock basis for \mathfrak{L} is \{\mathinner{|{n,m}\rangle\!\rangle}=|n\rangle|m\rangle^{*}\} . Within this vectorized framework, the LME is written as \partial_{t}\mathinner{|{\rho}\rangle\!\rangle}=\bar{\mathcal{L}}\mathinner{|{\rho}\rangle\!\rangle} , where the vectorized Liouvillian is given by \bar{\mathcal{L}}=-i\mathinner{\left[\mkern-2.0mu\left[{\hat{H},\hat{I}}\right]\kern-0.72226pt\right]}_{-}+\sum_{j}\gamma_{j}\left(\hat{L}_{j}\otimes\hat{L}_{j}^{*}-\frac{1}{2}\mathinner{\left[\mkern-2.0mu\left[{\hat{L}_{j}^{\dagger}\hat{L}_{j},\hat{I}}\right]\kern-0.89905pt\right]}_{+}\right), (2) and the super-(anti)commutator is defined as \mathinner{\left[\mkern-2.0mu\left[{\hat{A},\hat{B}}\right]\kern-0.72226pt\right]}_{\pm}=\hat{A}\otimes\hat{B}^{T}\pm\hat{B}\otimes\hat{A}^{T} . Here, \hat{I} represents the identity operator. Throughout this work, we use an overhead bar to denote the vectorized form of superoperators, \mathcal{S}\rightarrow\bar{\mathcal{S}} . The matrix elements of the Liouvillian read \bar{\mathcal{L}}_{j,k}^{n,m}=\mathinner{\langle\!\langle{n,m}|}\bar{\mathcal{L}}\mathinner{|{j,k}\rangle\!\rangle} . While the vector \mathinner{|{\rho}\rangle\!\rangle} belongs to the Hilbert space \mathfrak{L} , its physical interpretation differs fundamentally from that of pure states in the conventional Hilbert space \mathfrak{H} . For instance, the norm of \mathinner{|{\rho}\rangle\!\rangle} , defined via the Hilbert-Schmidt inner product \mathinner{\langle\!\langle{A}|{B}\rangle\!\rangle}=\mathrm{Tr}(A^{\dagger}B) , corresponds to the purity of the state, given by \mathinner{\langle\!\langle{\rho}|{\rho}\rangle\!\rangle}=\mathrm{Tr}(\rho^{2})\leq 1 . Probability conservation for \mathinner{|{\rho}\rangle\!\rangle} is instead ensured by the condition \mathinner{\langle\!\langle{\mathbb{I}}|{\rho}\rangle\!\rangle}=1 , where \mathinner{\langle\!\langle{\mathbb{I}}|} represents the vectorized identity operator. Furthermore, physical expectation values in the vectorized formalism are computed as \langle\hat{A}\rangle=\mathrm{Tr}(\hat{A}\hat{\rho})=\mathinner{\langle\!\langle{A}|{\rho}\rangle\!\rangle} . Note also, that in this framework, the components of \mathinner{|{\rho}\rangle\!\rangle} correspond to the “populations” of individual lattice sites in the LFSL. Since these components are generally complex, they cannot represent valid populations at each site, even though we will refer to them as such for convenience. It should be clear that the matrix \bar{\mathcal{L}} defines the LFSL in the same way the Hamiltonian defines the regular FSLs. For a finite-dimensional Hilbert space \mathfrak{H} with \mathrm{Dim}(\mathfrak{H})=D , the corresponding Liouville space \mathfrak{L} has dimension \mathrm{Dim}(\mathfrak{L})=D^{2} . Since \bar{\mathcal{L}} is not a Hermitian matrix, the evolution it generates is not unitary. Determining the lattice structure from \bar{\mathcal{L}} , such as its dimensions and bond identifications, is not unique. However, in many cases, there exists a natural choice, which typically

Read the complete text