Qumode Formalism Overview

• Qumode formalism is a continuous-variable quantum framework that uses infinite-dimensional Fock spaces of harmonic oscillators to encode and process information.
• It employs coherent and squeezed state representations alongside universal Gaussian and non-Gaussian gate sets for simulating complex quantum systems such as quantum chemistry and lattice gauge theories.
• Hybrid qubit–qumode implementations and variational ansatz circuits like SNAP–displacement provide enhanced expressivity and resource efficiency while mitigating noise effects.

The qumode formalism is a continuous-variable (CV) quantum information framework grounded in the physics of quantum harmonic oscillators and bosonic modes, referred to as "qumodes." Unlike qubit-based models with two-level systems, the qumode approach exploits the infinite-dimensional Fock space structure to encode, process, and simulate quantum information and quantum fields. Qumode formalism underpins a range of photonic quantum computing, quantum simulation, and hybrid discrete-continuous architectures, with core applications in quantum chemistry, lattice gauge theory, and quantum memory. Its defining ingredients are the bosonic creation and annihilation operators, associated phase-space representations (coherent and squeezed states), universal gate sets tailored to continuous variables, and encodings that leverage the Fock basis to efficiently represent exponentially large discrete Hilbert spaces.

1. Foundations of the Qumode Hilbert Space and Operators

A single qumode is mathematically described as a quantum harmonic oscillator with Hilbert space

$$\mathcal H = \mathrm{span}\bigl\{\,|n\rangle\,\big|\,n = 0,1,2,\dots\bigr\}$$

where $\{|n\rangle\}$ is the Fock basis (photon number states). The canonical bosonic operators $a$, $a^\dagger$ act as

$$a\,|n\rangle = \sqrt{n}\,|n-1\rangle,\quad a^\dagger\,|n\rangle = \sqrt{n+1}\,|n+1\rangle,$$

and obey $[a,\,a^\dagger]=1$ (Krishna, 4 Jul 2025, Dutta et al., 2024, Dutta et al., 5 Sep 2025, Jost et al., 15 Apr 2026).

Associated quadratures are

$$x = \frac{a + a^\dagger}{\sqrt{2}},\quad p = \frac{a - a^\dagger}{\sqrt{2}i},$$

with $[x,p]=i$. States in this Hilbert space can be represented in Fock, position, or momentum bases, with the phase-space (Wigner function) description providing a central tool for analysis and visualization (Krishna, 4 Jul 2025, Liu et al., 2015).

2. State Representations: Coherent and Squeezed States, Displacement

Key continuous-variable states used in qumode formalism are:

• Coherent states: Created by the displacement operator acting on vacuum,

$$D(\alpha) = \exp\bigl(\alpha\,a^\dagger-\alpha^*\,a\bigr),\quad |\alpha\rangle = D(\alpha)|0\rangle.$$

The phase-space representation of $|\alpha\rangle$ is a minimum-uncertainty Gaussian centered at $\{|n\rangle\}$0 with $\{|n\rangle\}$1 (Krishna, 4 Jul 2025, Liu et al., 2015).

• Squeezed states: Formed by the squeezing operator,

$\{|n\rangle\}$2

This state has reduced (squeezed) variance in one quadrature with a corresponding increase in the conjugate quadrature, quantified by squeezing factor $\{|n\rangle\}$3 (Liu et al., 2015).

Both types of states are instrumental for CV quantum phase estimation, quantum simulation tasks, and as carriers of quantum information (Liu et al., 2015, Dutta et al., 2024, Krishna, 4 Jul 2025).

3. Quantum Information Processing with Qumodes: Universal Gate Sets and Encodings

Qumode-based quantum processors are equipped with a universal gate set comprising:

• Gaussian gates: Displacement $\{|n\rangle\}$4, phase-space rotation $\{|n\rangle\}$5, single-mode squeezers $\{|n\rangle\}$6, two-mode beam-splitters $\{|n\rangle\}$7.
• Non-Gaussian gates: Selective-number-dependent arbitrary phase (SNAP) $\{|n\rangle\}$8, cubic-phase, and hybrid qubit–qumode controlled gates.
• Ancilla-assisted gates: Echoed conditional displacement (ECD) and controlled-beam-splitter (CBS) operations in hybrid circuits (Dutta et al., 5 Sep 2025, Dutta et al., 2024, Ale et al., 18 Nov 2025).

Typical encodings include mapping qubit or fermion spaces into the Fock subspace of one or more qumodes, for instance using binary-to-integer maps: $\{|n\rangle\}$9 which embeds a $a$0-qubit space in a single qumode of dimension $a$1 (Dutta et al., 5 Sep 2025, Jost et al., 15 Apr 2026). Particle number conservation can be enforced via Hamming-weight filters on the Fock basis (Jost et al., 15 Apr 2026).

4. Applications in Quantum Image Storage, Simulation, and Quantum Chemistry

The formalism finds pivotal use in diverse applications:

• Quantum memory and image storage: Grayscale image intensities $a$2 are mapped to displacement amplitudes $a$3. Memory-efficient storage exploits "delta evolution," storing only changes $a$4 between frames,

$a$5

Frame indexing is achieved using von Neumann entropy $a$6 after noise admixture, serving as a non-destructive, frame-unique fingerprint. Retrieval accuracy is assessed via phase-space fidelity between Wigner functions, with explicit formulae for displaced coherent states (Krishna, 4 Jul 2025).

• Lattice gauge theory and field simulation: Qumodes describe bosonic U(1) gauge fields via canonical quadratures. Hybrid qubit-qumode lattice Hamiltonians enforce gauge constraints either via squeezing-type projections or penalty Hamiltonians. Exact gauge invariance is preserved, and the system is decomposable into experimentally universal qubit-qumode gates (Ale et al., 18 Nov 2025).
• Quantum chemistry simulation: Electronic or vibrational Hamiltonians are mapped either directly or via Jordan–Wigner to Fock subspaces, enabling VQE and excited-state algorithms (QSS-VQE, QumVQD) utilizing qumode-native gates. Particle number conservation reduces complexity via Hamming-weight filtering, with advantages in ansatz expressivity, circuit depth, and resource scaling (Dutta et al., 2024, Dutta et al., 5 Sep 2025, Jost et al., 15 Apr 2026).

5. Variational Ansatz Construction and Expressivity

A central practical advantage is the compactness and universality of variational ansatz circuits:

• SNAP–displacement ansatz: Layers of

$a$7

are sufficient for universal control in the finite Fock subspace. Multimode extensions employ interleaving beam-splitters (Dutta et al., 5 Sep 2025, Jost et al., 15 Apr 2026).

• ER ansatz (ECD–rotation):

$a$8

is also universal and compatible with hybrid cQED implementations (Dutta et al., 2024).

• Deflation for excited states: QumVQD applies deflation through a penalty term in the cost, maintaining orthogonality to previously found eigenstates (Jost et al., 15 Apr 2026).

Empirical benchmarks show drastic reductions in parameter count and circuit depth versus qubit-based TwoLocal circuits for electronic and vibrational problems, particularly in excited-state spectroscopy (Dutta et al., 5 Sep 2025, Jost et al., 15 Apr 2026).

6. Error, Noise, and Resource Analysis

Qumode processors admit noise analysis tailored to bosonic hardware:

• Gate-fidelity model: With $a$9 two-mode gates of fidelity $a^\dagger$0, the leading energy error scales as $a^\dagger$1. Qumode circuits typically use fewer entangling gates than their qubit counterparts to reach chemical accuracy (Jost et al., 15 Apr 2026).
• Photon loss (amplitude-damping/Kraus channel): Dynamics during time $a^\dagger$2 at decay rate $a^\dagger$3 is captured by

$a^\dagger$4

with Kraus operators accounting for excitation decay. Chemical accuracy in VQE requires per-gate loss parameters $a^\dagger$5 (Jost et al., 15 Apr 2026).

• Resource metrics (in phase estimation): Precision in phase estimation is governed by the product $a^\dagger$6 (squeezing-time-precision tradeoff). For Shor-like factoring, squeezing must scale exponentially; DQC1-like trace estimation requires only squeezer in a coherent state ($a^\dagger$7) (Liu et al., 2015).

Expressivity, parameter efficiency, and error resilience make qumode formalisms particularly suitable for problems requiring compact large-Hilbert-space representations and low-depth gate protocols, especially where decoherence dominates (Dutta et al., 5 Sep 2025, Jost et al., 15 Apr 2026).

7. Hybrid Models and Extensions

Qumode formalism naturally supports hybrid models, combining discrete-variable (qubit) registers with continuous-variable (qumode) registers. In lattice gauge theory, matter fields are encoded as qubits and gauge fields as pairs of qumodes, with hybrid Hamiltonians built from local generators enforcing gauge invariance (Ale et al., 18 Nov 2025). In quantum chemistry and quantum simulation, hybrid circuits exploit cQED platforms coupling microwave resonator qumodes to superconducting transmon qubits, implementing universal sets including ECD and SNAP gates (Dutta et al., 2024, Dutta et al., 5 Sep 2025).

Squeezing-based constraint implementation, ancilla-assisted non-Gaussian routines, and deflation/fragmentation techniques (e.g., Bogoliubov transforms for vibrational modes) define the ongoing development of qumode-based computational paradigms (Ale et al., 18 Nov 2025, Jost et al., 15 Apr 2026).

---

Taken as a whole, the qumode formalism constitutes a rigorous and general framework for continuous-variable quantum information processing, simulation, and storage, characterized by infinite-dimensional state spaces, phase-space control, and deep connections to both quantum optics and high-dimensional quantum computation (Krishna, 4 Jul 2025, Liu et al., 2015, Dutta et al., 2024, Dutta et al., 5 Sep 2025, Ale et al., 18 Nov 2025, Jost et al., 15 Apr 2026).