Product Weyl--Heisenberg covariant mutually unbiased bases and extremal non-stabilizerness

In this work, researchers investigate discrete structures in product Hilbert spaces by introducing a notion of magick analogous to magic for single-partite systems, proving that fiducial states for symmetric informationally complete measurements (SIC) and mutually unbiased bases (MUBs) are necessarily maximizers of this quantity when covariant with respect to the product Weyl-Heisenberg group. This investigation extends known constructions to multipartite, prime-power dimensional quantum systems, yielding explicit fiducial vectors for MUBs in dimensions where previous constructions were not available.

The Core Concept of Magick

The paper introduces a measure called magick, defined for a quantum state ρ of a composite system as the sum over the product Weyl-Heisenberg group operators:

M(ρ) = X(d−1)k,l=0 Tr[Wkl ρ]. This quantity is presented as an extension of the standard magic measure, and it exhibits several useful properties:

1. M(ρ) is invariant under product of Clifford operations V = N i Vi, i.e., M(VρV†) = M(ρ).

2. M(ρ) is convex.

3. For any pure state ψ⟩, M(ψ⟩) ≥ d with equality achieved only for the product of stabilizer states.

Connection to Known Structures (SIC and MUBs)

The paper establishes a crucial link between the magick measure and the divergence of a set of vectors from known structures:

- SIC Connection:

Lemma 2 shows that the divergence from a SIC is given by PSIC(Oψ⟩) = d(3 − 2 + d(2/√d + 1)) (1 − M(ψ⟩)). This implies that an orbit Oψ⟩ generates a SIC if and only if M(ψ⟩) = 1 + (d − 1)√d + 1.

- MUB Connection:

Lemma 3 shows that the divergence from a set of d mutually unbiased bases is PMUB(Oψ⟩) = d(3 − 2 + d√d(1 − M(ψ⟩))). This implies that an orbit Oψ⟩ generates MUBs if and only if M(ψ⟩) = 1 + (d − 1)√d.

Construction of Fiducial Vectors

The paper constructs explicit fiducial vectors corresponding to local-WH-covariant MUBs for prime-power dimensions d = p n, with p ≥ 3:

1. For p ≥ 5, the construction is given by fMUB,pn⟩ = p(Xn−1)j=0 ω trF[a j 3]p j⟩ in H⊗n p (Equation 21), where a ∈ Fpn and trF is the field trace.

2. For p = 3, the construction utilizes Galois rings GR(9, n) and its residue field F3n to circumvent known non-existence results for Alltop sequences over fields of characteristic 3n. The fiducial vector is fMUB,3n⟩ = p(Xn−1)j=0 ω trGR[a j 3]9 j⟩ in H⊗n cubed (Equation 24), where a ∈ GR(9, n).

Extremal Properties and Limitations

The study concludes by analyzing the maximal values of magick:

- SIC Maximization:

Lemma 4 proves that the maximal value of magick is achieved for the fiducial vector of a SIC measurement: M(ρ) ≤ M(fSIC⟩) = 1 + (d − 1)√d + 1.

- MUB Maximization:

Lemma 6 shows that the maximal value over equimodular vectors is achieved for the fiducial state fMUB⟩, yielding M(θ⟩) ≤ M(fMUB⟩) = 1 + (d − 1)√d.

Sporadic and Composite Case Results

The work also presents sporadic full sets of MUBs in dimensions d = 4 and d = 9 using different methods, including constructions derived from three fiducial states under a subgroup of the product-WH group. Furthermore, the paper notes that for p=2 (qubits), explicit examples exist for d=2 and d=2 squared, while a conjecture is stated regarding the non-existence of such constructions in dimension d = 2n for n > 2. The construction for p=3 also provides a novel way to circumvent known no-go theorems by using Galois rings instead of standard finite fields.

As a fastidious researcher, I have analyzed this paper, Product Weyl-Heisenberg covariant MUBs and Maximizers of Magick, which deals with constructing highly symmetric quantum states (Mutually Unbiased Bases (MUBs) and Symmetric Informationally Complete Measurements (SIC)) in multipartite systems using the framework of group theory.

The core contribution is the introduction of a new measure, magick, derived from the product Weyl-Heisenberg group, which serves as a powerful tool to characterize and maximize quantum designs. The paper provides explicit constructions for these states in prime-power dimensions, extending known results (like those for Klappenecker and Rötteler) to multipartite settings.

Here are the specific improvements that can be made to AI systems by leveraging the principles derived from this research:

The scientific framework presented in this paper suggests advancements in AI systems, particularly those focused on quantum simulation, high-dimensional data analysis, and complex optimization problems. The key takeaways are:

1. The existence of highly symmetric structures (MUBs/SIC) is guaranteed by specific fiducial states derived from group actions (Weyl-Heisenberg group).

2. These states maximize a measure of magick, providing an extremal property for quantum designs.

3. The construction methods are scalable, especially when moving from single-party to multipartite systems via product group covariance and Galois rings (for specific dimensions like p=3).

Here are the specific improvements:

1. Improve the efficiency and accuracy of high-dimensional quantum state tomography and characterization for complex, highly symmetric systems.

2. Develop new optimization algorithms capable of finding extremal configurations in massive, structured Hilbert spaces relevant to quantum machine learning (QML) or quantum chemistry simulations.

3. Enhance the robustness of AI models dealing with complex data structures by incorporating group-theoretic symmetries directly into their architecture or loss functions.

Specifically, here is what the improved AI systems can do:

1. Improved Quantum State Tomography and Characterization:

This system can precisely identify and characterize quantum states in large Hilbert spaces (e.g., those relevant to complex molecular simulations or large-scale quantum error correction).

• Instead of relying on generic measurement sequences, the AI will use the derived fiducial vectors (like those in Theorem 1 and Theorem 2) as optimized starting points.

• It can determine if a given experimental state is an isoentangled MUB/SIC fiducial state by calculating its magick value; if it matches the known maximum, the system has high confidence that the state is a highly symmetric design.

2. Advanced Optimization for Quantum Designs:

The magick measure provides a quantifiable metric for evaluating how close a set of quantum measurements (or states) is to being perfectly unbiased (MUBs) or informationally complete (SIC).

• The AI can be used to optimize experimental parameters in quantum experiments (e.g., pulse sequences, gate timings) to maximize the magick quantity, directly aiming for the most symmetric and robust quantum measurement possible.

• This is particularly useful in designing optimal measurement schemes for complex quantum algorithms where maximizing symmetry leads to better performance or higher fidelity.

3. Symmetry-Aware Quantum Machine Learning (QML):

The paper's focus on group covariance suggests a path toward building QML models that intrinsically respect underlying symmetries rather than learning them from scratch.

• The AI can be trained to recognize and utilize the product Weyl-Heisenberg symmetry in its feature space. This allows it to generalize patterns across different dimensions or multipartite subsystems more effectively than standard neural networks.

• For instance, in a quantum neural network, this could lead to architectures where the connectivity or weight structure is constrained by group orbits, leading to models that are inherently robust against certain types of noise (since stabilizer states maximize magick).

4. Novel State Generation for Quantum Simulation:

The explicit constructions (like Theorem 1 and Theorem 2) provide a blueprint for generating new, highly structured quantum states in prime-power dimensions that were previously unknown or difficult to find.

• The AI can be used as a generative model to propose new, maximally entangled states or MUBs in specific high-dimensional contexts (like those relevant to topological quantum computing), accelerating the discovery of novel quantum resources.

Discrete structures in product Hilbert spaces are investigated. For monopartite systems of size d one relies on the Weyl--Heisenberg group WH(d), while in the case of composite Hilbert spaces with the local dimensions d i we identify designs covariant with respect to the product group, i WH(d i). In analogy with magic -- a quantity attaining its maximum for states fiducial with respect to WH(d) -- we introduce a similar quantifier of product magic, defined with respect to the product group. The maximum of this quantity over all equimodular vectors yields fiducial states that generate d = product i d i a priori isoentangled mutually unbiased bases (MUBs), which, when supplemented by the identity, form their complete set. Such fiducial states are explicitly constructed in all prime-power dimensions d = p n with p 3. The result for p 5 extends the construction of Klappenecker and R ö tteler, whereas for p=3 it is mathematically distinct and is based on Galois rings. The global maximum of the quantifier of product magic for d=2 cubed yields fiducial states corresponding to the symmetric informationally complete (SIC) generalized measurement of Hoggar. Our approach feeds into a unifying perspective in which highly symmetric quantum designs emerge from fiducial states with extremal properties via structured group-orbit constructions.