Gravitational Cat State: Theory & Applications

• Gravitational cat state is a quantum superposition of a massive particle localized in two distinct positions via a symmetric double-well potential.
• Effective Hamiltonians and thermal Gibbs states model gravcat dynamics, revealing gravity-mediated interactions and the hierarchy of quantum correlations.
• Applications include force detection, quantum metrology, and quantum communication, bridging the gap between quantum mechanics and gravitational physics.

A gravitational cat state, or “gravcat,” is a quantum superposition of two spatially distinct configurations of a massive particle, typically realized by a mass confined in a symmetric double-well potential and coherently delocalized between the two minima. In the simplest Newtonian description, a single motionless massive particle can in principle be in a superposition state of two spatially-separated locations, so that the mass distribution sourcing the Newtonian field is itself “quantum,” leading to fluctuations in the Newtonian force exerted on a nearby test particle (Anastopoulos et al., 2015). In the many-body and two-qubit literature, the term also denotes pairs of interacting massive cat states whose effective dynamics are governed by gravity-induced couplings, thermal Gibbs states, and open-system channels (Anastopoulos et al., 2020). More recent effective-field-theory treatments recast the same setup as a matter-wave interferometer in which matter and the graviton are treated on an equal footing at a perturbative level, with the graviton vacuum displaced into branch-dependent coherent states (Mazumdar et al., 6 May 2026).

1. Definition and state-space constructions

The canonical gravcat construction begins with a massive particle in a symmetric double-well potential with minima at $x=\pm L/2$. In the low-energy double-well approximation one keeps only the two lowest states, denoted either by $|g\rangle,|e\rangle$ or by $|0\rangle,|1\rangle$, and defines localized cat states through

$$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$

or equivalently

$$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$

These states are associated with the two wells and define an effective qubit degree of freedom (Rojas et al., 2021).

A single gravcat is therefore a “Schrödinger-cat” superposition of two distinct spatial configurations of a massive particle in a double-well potential. In one formulation, the general two-state gravitational system is

$$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$

while in another the cat state is written as the equal superposition

$|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$

For two gravcats, the composite Hilbert space is $\mathbb C^2\otimes\mathbb C^2$, with computational basis $\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$ (Anastopoulos et al., 2015).

The same idea extends beyond the single-particle two-level model. Anastopoulos and Hu considered a pair of single-particle double-well qubits and also a Bose–Einstein condensate description, where a gravcat is a mean-field solution of a gravitational Gross–Pitaevskii equation. In that setting, the gravcat concept interpolates between the usual GP equation and the Newton–Schrödinger limit, while preserving the central picture of a massive object in a coherent superposition of distinct spatial configurations (Anastopoulos et al., 2020).

A distinct but related construction appears in perturbative quantum gravity. Mazumdar and Zhou model a single non-relativistic mass $M$ in a matter-wave interferometer,

$|g\rangle,|e\rangle$0

with left and right Gaussian wavepackets centered at $|g\rangle,|e\rangle$1 and $|g\rangle,|e\rangle$2. In that formulation, the gravcat is simultaneously a cat state of matter and a branch-dependent coherent displacement of the graviton vacuum (Mazumdar et al., 6 May 2026).

2. Effective Hamiltonians and solvable regimes

The standard effective Hamiltonian for two interacting gravcats in the weak-field, nonrelativistic limit is

$|g\rangle,|e\rangle$3

where $|g\rangle,|e\rangle$4 is the single-particle level splitting and

$|g\rangle,|e\rangle$5

is the gravity-mediated coupling strength (Jaloum et al., 2024). Equivalent notations $|g\rangle,|e\rangle$6 or $|g\rangle,|e\rangle$7 are also used for the gravitational coupling in closely related models (Rojas et al., 2021, Hadipour et al., 2024).

In the computational basis, one convenient matrix representation is

$|g\rangle,|e\rangle$8

with eigenvalues

$|g\rangle,|e\rangle$9

This diagonal structure underlies most closed-form results for thermal correlations, metrology, and thermodynamics (Bachain et al., 19 Mar 2026).

A more general two-gravcat Hamiltonian includes an external magnetic field with homogeneous part $|0\rangle,|1\rangle$0 and inhomogeneity $|0\rangle,|1\rangle$1,

$|0\rangle,|1\rangle$2

In this model, diagonalization yields

$|0\rangle,|1\rangle$3

with eigenvectors in the $|0\rangle,|1\rangle$4-$|0\rangle,|1\rangle$5 and $|0\rangle,|1\rangle$6-$|0\rangle,|1\rangle$7 sectors parameterized by mixing angles $|0\rangle,|1\rangle$8 and $|0\rangle,|1\rangle$9 (Houça et al., 2021).

At the single-gravcat level, the observable of interest is often the Newtonian force on a nearby probe. Projecting onto the two-state subspace gives

$$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$0

or, in the far-field approximation,

$$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$1

These formulas connect the abstract two-state model to force sensing and continuous measurement protocols (Anastopoulos et al., 2015, Derakhshani et al., 2016).

The dominant approximation throughout this literature is the two-level truncation. Additional assumptions are quasi-instantaneous or Newtonian gravitational coupling, uniform or $$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$2-directed external fields when present, and neglect of higher motional levels. This suggests that gravcat theory is best regarded as an effective low-energy framework, not a complete microscopic theory of quantum gravity.

3. Thermal states and the hierarchy of quantum correlations

For two gravcats in contact with a heat bath at temperature $$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$3, the equilibrium state is typically taken to be Gibbsian,

$$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$4

with partition function

$$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$5

for the basic $$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$6 model, or

$$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$7

for the inhomogeneous-field model (Bachain et al., 19 Mar 2026, Houça et al., 2021). In the computational basis, $$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$8 has the standard $$|+\rangle=\frac{|g\rangle+|e\rangle}{\sqrt2},\qquad |-\rangle=\frac{|g\rangle-|e\rangle}{\sqrt2},$$9-state form, which allows analytic treatments of concurrence, coherence, steering, discord, and local quantum uncertainty.

Several correlation quantifiers have been studied systematically. For entanglement, the standard measure is Wootters concurrence; for coherence, the $$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$0-norm; for discord-type correlations, geometric quantum discord or local quantum uncertainty (LQU); and for nonlocal correlations, Bell non-locality and steering (Rojas et al., 2021, Jaloum et al., 2024). In the LQU formulation of Houça et al.,

$$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$1

or equivalently for the $$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$2-state,

$$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$3

with $$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$4 and $$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$5 expressed analytically in terms of the thermal density-matrix entries (Houça et al., 2021).

The thermal behavior is strongly structured. As $$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$6 increases, off-diagonal terms shrink, concurrence can vanish at a finite entanglement-death temperature, and the state approaches the completely mixed state in the infinite-temperature limit (Jaloum et al., 2024). Rojas and Lobo found that the thermal concurrence and $$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$7-norm can be significantly optimized by increasing the masses or decreasing the distance between them, and that thermal fluctuations raise non-entangled quantum correlations when entanglement suddenly drops (Rojas et al., 2021).

A robust hierarchy also emerges. In the two-gravcat thermal model studied in 2024, steering disappears first at $$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$8, entanglement vanishes next at $$|+\rangle=\frac{|0\rangle+|1\rangle}{\sqrt2},\qquad |-\rangle=\frac{|0\rangle-|1\rangle}{\sqrt2}.$$9, and geometric discord remains strictly $$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$0 even when $$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$1, decaying to zero only in the completely mixed limit (Jaloum et al., 2024). Houça et al. similarly report that the thermal LQU captures a stronger quantum correlation than the entanglement, especially for low external magnetic field levels combined with low field inhomogeneity or high-temperature domains, and that the states become non-entangled and separable when the gap between the fundamental level and the first excited level becomes large (Houça et al., 2021).

The influence of control parameters is equally central. Increasing uniform field $$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$2 or inhomogeneity $$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$3 suppresses both LQU and concurrence at low $$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$4, enlarging a zero-correlation window, while low $$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$5, low $$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$6, and low $$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$7 produce maximal correlations (Houça et al., 2021). In the basic $$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$8 model, a large energy gap $$|\psi\rangle=c_+|+\rangle+c_-|-\rangle,\qquad |c_+|^2+|c_-|^2=1,$$9 accelerates the loss of quantum correlations under thermal effects and can render the state separable (Jaloum et al., 2024).

4. Probing schemes, decoherence channels, and experimental platforms

The earliest gravcat proposals focus on force detection. For a classical probe, the recorded signal is a fluctuating Newtonian force with

$|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$0

where $|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$1 is set by the tunneling rate $|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$2 and the probe’s temporal resolution $|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$3 (Anastopoulos et al., 2015). Anastopoulos and Hu emphasize that a classical probe generically records a non-Markovian fluctuating force, while a quantum harmonic-oscillator probe may undergo Rabi oscillations in a strong coupling regime (Anastopoulos et al., 2015).

Feasibility analyses have centered on superconducting microspheres, optomechanical probes, and micromechanical resonators. Derakhshani’s progress report describes a levitated superconducting microsphere of mass $|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$4 and a Si$|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$5N$|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$6 trampoline resonator with projected force sensitivity near $|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$7, but concludes that the direct gravcat force signal remains many orders of magnitude below projected sensitivity in currently analyzed configurations (Derakhshani et al., 2016). By contrast, larger-separation protocols discussed in the collapse-theory comparison paper target $|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$8 and move the characteristic force scale into the $|Cat\rangle=\frac{|+\rangle+|-\rangle}{\sqrt2}.$9–$\mathbb C^2\otimes\mathbb C^2$0 range (Derakhshani, 2016).

Open-system studies treat the thermal Gibbs state as an initial resource exposed to stochastic, decaying, power-law, or correlated dephasing channels. Rahman et al. analyze thermal, classical stochastic, general decaying, and power-law noisy fields, and also introduce a weak-measurement-reversal protocol with

$\mathbb C^2\otimes\mathbb C^2$1

showing that a suitable choice of $\mathbb C^2\otimes\mathbb C^2$2 can substantially boost one-way steerability, Bell non-locality, concurrence, and purity (Rahman et al., 2023). Haddadi et al. consider a classically correlated dephasing channel with memory parameter $\mathbb C^2\otimes\mathbb C^2$3, for which the thermal $\mathbb C^2\otimes\mathbb C^2$4-state retains the same form except that the off-diagonals are rescaled by

$\mathbb C^2\otimes\mathbb C^2$5

In that model, $\mathbb C^2\otimes\mathbb C^2$6 gives $\mathbb C^2\otimes\mathbb C^2$7, so $\mathbb C^2\otimes\mathbb C^2$8, i.e. no decoherence (Haddadi et al., 2024).

Gravcat ideas have also migrated to more aggressive state-preparation schemes. One proposal exponentially expands an initially small superposition via Gaussian dynamics, obtaining a position separation

$\mathbb C^2\otimes\mathbb C^2$9

and predicts that gravitationally induced entanglement grows exponentially with the expansion time $\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$0 (Braccini et al., 2024). Another combines Earth’s gravitational acceleration and diamagnetic repulsion to generate mass-independent large spatial superpositions, reporting an increase from $\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$1 to $\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$2 in less than $\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$3 for diamond nanocrystals hosting NV centers (Zhou et al., 2022).

Gravcat experiments are also relevant to gravity-related decoherence. Measurements on a $\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$4 mechanical Schrödinger cat state rule out the Diósi–Penrose prediction associated with the standard cutoff $\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$5 for that platform, while setting a lower bound on $\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$6 from the absence of excess decoherence (Fadel, 2023). Although that experiment is not itself a double-well gravcat test, it constrains one class of models frequently discussed in gravcat phenomenology.

5. Metrology, thermodynamics, gravimetry, and quantum communication

The two-gravcat Gibbs state has become a platform for multiparameter quantum estimation. In the metrological formulation, one estimates

$\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$7

through the quantum Fisher information matrix,

$\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$8

with the symmetric logarithmic derivatives $\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\}$9 defined by

$M$0

The covariance matrix satisfies the quantum Cramér–Rao bound

$M$1

and the gravcat model exhibits optimal estimation regions in which the precision is significantly enhanced (Bachain et al., 19 Mar 2026).

The same paper develops a quantum Stirling cycle with $M$2 as the control parameter. The four strokes are two isothermals and two isochores, with heat exchanges

$M$3

$M$4

and overall work and efficiency

$M$5

The reported behavior is that both $M$6 and $M$7 can be made positive over a finite $M$8-range, larger thermal gradients increase $M$9 and $|g\rangle,|e\rangle$00, and increasing $|g\rangle,|e\rangle$01 typically broadens the $|g\rangle,|e\rangle$02-window over which $|g\rangle,|e\rangle$03 is close to Carnot efficiency (Bachain et al., 19 Mar 2026).

Work extraction has also been analyzed via ergotropy. After decoupling the system from the bath, the maximum cyclic-unitary work is

$|g\rangle,|e\rangle$04

where $|g\rangle,|e\rangle$05 is the passive state unitarily related to $|g\rangle,|e\rangle$06 (Hadipour et al., 2024). The principal result is that the increase in temperature and the interaction between states decrease the amount of work that can be extracted from gravitational cat states; $|g\rangle,|e\rangle$07 vanishes at exactly zero temperature because the ground state is passive, and it also vanishes in the fully mixed limit, producing a moderate-temperature sweet spot (Hadipour et al., 2024).

Quantum-information protocols have been formulated directly on the thermal gravcat resource. In dense coding, Alice and Bob share the gravcat Gibbs state, Alice applies one of $|g\rangle,|e\rangle$08, and the dense-coding capacity is

$|g\rangle,|e\rangle$09

For the gravcat channel, $|g\rangle,|e\rangle$10 is the quantum-advantage condition, $|g\rangle,|e\rangle$11 is maximal, and weak-measurement protection can enlarge the parameter region in which the advantage survives thermal mixing (Haddadi et al., 2023).

A different line of work replaces ancillary qubits with mechanical qubits. In “Quantum gravimetry with mechanical qubits,” a mechanical cat qubit built from even and odd coherent-state cats,

$|g\rangle,|e\rangle$12

is used as the gravity sensor. The reported sensitivity scales as

$|g\rangle,|e\rangle$13

where $|g\rangle,|e\rangle$14 is the particle mass and $|g\rangle,|e\rangle$15 is the mean phonon number, leading to the “double standard quantum limits” with $|g\rangle,|e\rangle$16 and $|g\rangle,|e\rangle$17 simultaneously (Huo et al., 16 Apr 2026). This is not the same physical model as the double-well gravcat, but it shows how cat-state mechanics and gravity sensing are being unified under closely related terminology.

6. Conceptual interpretation, competing theories, and limitations

A central interpretive issue is what gravcat phenomena do and do not establish about gravity. One line of argument holds that detecting gravity-induced entanglement between two massive superposed objects would provide strong—though not yet fully model-independent—evidence that the gravitational interaction must carry quantum degrees of freedom (Rojas et al., 2021). Anastopoulos and Hu, however, emphasize a narrower conclusion: in the nonrelativistic weak-field regime, all predicted Rabi oscillations, entanglement, and energy transfer originate from the Newtonian potential term, and Coulomb-mediated entanglement in QED already shows that mediator-induced entanglement need not directly witness field quantization. On that view, gravcat entanglement via the Newtonian $|g\rangle,|e\rangle$18 term is agnostic about the quantum nature of gravity itself (Anastopoulos et al., 2020).

This debate is sharpened by comparisons with collapse models. Derakhshani analyzes canonical quantum theory, semiclassical objective collapse theories, and quantized-gravity extensions of objective collapse theories in the gravcat setting. In the canonical and quantized-gravity-collapse descriptions, the probe records random force jumps $|g\rangle,|e\rangle$19 with exponentially decaying two-point correlations; in semiclassical collapse models, the probe instead sees a static classical force or no force, depending on the preparation regime (Derakhshani, 2016). Gravcat experiments are therefore proposed not only as gravity tests, but also as discriminators among GRW, CSL, Diósi–Penrose, and Karolyhazy-type modifications.

A separate conceptual development concerns semiclassical gravity beyond coherent states. Ahmed, Lima, and Martínez identify families of field cat states for which the semiclassical Einstein equation remains reliable because the Kuo–Ford estimator is small. In the large-cat regime, with $|g\rangle,|e\rangle$20 and negligible overlap, the stress-tensor moments are suppressed by $|g\rangle,|e\rangle$21; in the $|g\rangle,|e\rangle$22-phase regime,

$|g\rangle,|e\rangle$23

all central moments vanish exactly and $|g\rangle,|e\rangle$24 (Ahmed et al., 2023). This suggests that not every gravitational cat state lies outside semiclassical gravity; some cat families remain within its domain of validity.

The perturbative-EFT perspective deepens the geometric interpretation. In the matter-wave interferometer model, the joint state takes the form

$|g\rangle,|e\rangle$25

where $|g\rangle,|e\rangle$26 are branch-dependent coherent graviton states. Their overlap

$|g\rangle,|e\rangle$27

defines the gravitational contrast, and the reduced matter entropy

$|g\rangle,|e\rangle$28

quantifies matter-graviton entanglement (Mazumdar et al., 6 May 2026). In that language, a gravcat is not only a superposed mass distribution but also a superposition of quantum geometries led by coherent graviton states.

Across these formulations, the main limitations recur: two-level truncation, instantaneous thermalization or quasi-instantaneous Newtonian coupling, omission of realistic dissipative decoherence in some thermodynamic and work-extraction models, neglect of higher multipoles and non-Gaussian environmental couplings in expansion protocols, and the restriction to weak-field or perturbative regimes (Hadipour et al., 2024, Braccini et al., 2024, Mazumdar et al., 6 May 2026). A plausible implication is that “gravcat” functions less as a single model than as a family of effective frameworks for studying massive superposition, gravity-mediated interaction, and the boundary between quantum mechanics, thermodynamics, and gravitational physics.