Entanglement Dimensionality of Continuous Variable States From Phase-Space Quasi-Probabilities

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: Hi, I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Entanglement Dimensionality of Continuous Variable States From Phase-Space Quasi-Probabilities".

Mira: Here's the gist of the paper. The gist The work introduces Schmidt number witnesses for continuous variable (CV) systems based directly on covariances of infinite dimensional Bloch operators that are readily accessible in experiments, showing that direct estimation leads to increased robustness and versatility compared to first discretising the system and using canonical discrete variable techniques<ref:2509.02743#pg2> Significant effort has been devoted to characterizing entanglement in continuous-variable (CV) quantum systems

1, 2: in experimentally accessible ways, despite the complication posed by the formally infinitedimensional Hilbert space<ref:2509.02743#pg2> Most prominently, conditions on the covariance matrix of the quadratures have been investigated

3, 4: , as well as approaches based on the moment of the mode operators

5–7: <ref:2509.02743#pg2> Other approaches are instead based on phase-space quasi-probability distributions, which can be seen as arising from correlations of continuous-variable Local Orthogonal Observables (CVLOOs), such as displacements or displaced parity operators

12, 13: <ref:2509.02743#pg2> However, these bounds are not directly applicable to the so-called entanglement dimensionality, which is an important quantifier of the effective dimension that is required to describe the observed correlations<ref:2509.02743#pg2> For a bipartite system in the pure state ψ⟩ = Pr k=1 √λk uk⟩ vk⟩, this is typically quantified through the Schmidt rank SN(ψ) that is defined as the number of nonzero Schmidt coefficients λk, i.e., the rank of the single-particle reduced density matrix<ref:2509.02743#pg2> This entanglement monotone can be extended to mixed states by considering a worst-case scenario over all decompositions

17: : SN(ϱ):= inf D(ϱ) max ψi⟩∈D(ϱ) SN(ψi), (1)<ref:2509.02743#pg2> This defines the so-called Schmidt number for bipartite density matrices, a quantity that plays a central role in various single-shot tasks and stands as one of the few entanglement measures linked to quantum computational advantage

18–20: <ref:2509.02743#pg2> Theoretical Framework The canonical fidelity-based witness for discrete systems are given in terms of an upper-bound to the fidelity with respect to a target state ΨT ⟩ = P k √λk kk⟩ (λk ordered non-increasingly), that typically corresponds to the ideal (high-dimensional entangled) preparation expected in an experiment<ref:2509.02743#pg2> More precisely, the following upper-bound F (ϱ, ΨT ⟩) ≤ X r k=1 λk, (2) holds for all density matrices ϱ with Schmidt number upperbounded by r, where the fidelity (to a pure state) is defined as F (ϱ, ΨT ⟩) = ⟨ΨT ϱ ΨT ⟩<ref:2509.02743#pg2> Here, building upon the finite-dimensional framework

39: , we consider a set of CVLOOs, i.e., an orthonormal set of operators Q(α) where α ∈ C, that are such that they span the observables’ space of a bosonic mode<ref:2509.02743#pg2> For example, the displacement operator D(α) = e αaˆ†−α∗aˆ, where α ∈ C, defines the characteristic function χϱ(α) = tr(D(α)ϱ), i.e., it provides a full description of the density matrix and can be used to construct examples of CVLOOs<ref:2509.02743#pg2> New Criteria for Entanglement Dimensionality The authors extend existing results by both generalizing these bounds to detect the entanglement dimensionality, and by considering another criterion containing nonlinear terms, that make it strictly stronger than the CCNR<ref:2509.02743#pg2> A first way to extend Eq.

Kai: (4) to detect the Schmidt number would be to use the inequalities R(ϱ) ≤ r and TA(ϱ) ≤ r, which hold for all states with Schmidt number bounded from above by a given r

49–51: , obtaining an expression of the form Z C ⟨OA(α) ⊗ O˜ B(α)⟩ d2α ⩽ r, (5)<ref:2509.02743#pg2> They derive another condition that is strictly stronger than a bound on the realigned density matrix and can potentially detect a wider set of states than the bounds in Eq.

Mira: (5)<ref:2509.02743#pg2> This criterion is based on crosscovariances between two observables, which are defined as Γϱ(QA, QB):= ⟨QA ⊗ QB⟩ϱ − ⟨QA⟩ϱ⟨QB⟩ϱ<ref:2509.02743#pg2> Based on these quantities, they prove that for any state with Schmidt number not larger than a given r, the following inequality holds Z C Xϱ(α) d2α − vt Y n=A,B 1 − Z C ⟨Qn(α)⟩ 2d2α!

Kai: + 1 ≤ r<ref:2509.02743#pg2> Applications and Experimental Implementation The method is illustrated by considering a symmetric two-mode squeezed thermal state where the Schmidt number detected by their method is r ⩾ e 2ξ, which is a finite number<ref:2509.02743#pg2> For practical scenarios, the authors show that their criterion outperforms the fidelity witness when truncating the Hilbert space at some finite dimension d<ref:2509.02743#pg2> They demonstrate experimental implementations using hybrid CV-DV systems where characteristic functions can be measured through a two-level system<ref:2509.02743#pg2> Specifically, they present two implementations for Eqs.

Paper summary: Kai: So, to wrap up this paper, "Entanglement Dimensionality of Continuous Variable States From Phase-Space Quasi-Probabilities" by Liu and his team, they've shown that using inherently CV methods can lead to a better visibility of the entanglement dimensionality. Mira They are making the characterization of this dimensionality much more tolerant to noise than those typical witnesses based on fidelities that we usually use.

Lev: And they’ve also derived new conditions, specifically a criterion involving cross-covariances between two observables, that can detect states that were invisible to the CCNR criterion before. Kai So it's about moving past the standard methods and using what's naturally available in a continuous setting to measure this entanglement dimension.

Mira: The implication is that for real-world CV quantum systems, especially when noise is involved, this approach makes characterizing entanglement dimensionality much more effective and readily implementable. Lev It’s about improving how we actually measure the resource we care about in these experiments.

Kai: They also showed that their threshold condition for a mixture of a two-mode squeezed state with single-mode thermal noise becomes strictly worse than what you get from other methods. Mira That means their method is more accurate and robust when dealing with those realistic noise conditions we have to contend with.

Lev: So the paper points toward using these direct phase-space correlations, like the crosscovariances, as a way to certify entanglement dimensions in CV systems without needing that initial discretization step.

Conclusion: Kai: So, we’re wrapping up this look at "Entanglement Dimensionality of Continuous Variable States From Phase-Space Quasi-Probabilities."

Mira: They’re using phase-space quasi-probabilities to figure out how complex the entanglement actually is in these continuous variable systems.

Kai: It seems like they've moved away from just looking at standard fidelity measures and are trying to measure something more fundamental about the state itself.

Mira: Yeah, they’re building witnesses directly from correlations and covariances of those phase-space observables that you can actually measure in an experiment.

Lev: From my side, if this works, it means we don't have to immediately discretize everything into a finite grid just to get *any* bound on the entanglement rank.

Kai: Exactly. The authors show that this direct approach is more robust and versatile than taking the system and chopping it up first with canonical discrete variable techniques.

Mira: They test this against some very specific states, like Gaussian states, where they find their method gives a finite Schmidt number even when the expected value is infinite.

Lev: That’s a good sign for hardware. If we can certify high dimensionality directly on the CV state, it cuts down on the overhead of running error correction protocols just to figure out what's inside.

Kai: And they even showed that their new nonlinear criterion can detect states that previous methods missed, which is a pretty big deal for testing the limits of what we can actually build.

Mira: It really suggests that leveraging the continuous nature of the system itself is key to getting a better picture of its entanglement structure than relying on older bounds.

Lev: So, what this means for us in terms of real-world implementation is that our experimental setups could be much more reliable when dealing with noisy states.