Enrico Bozzetto of Politecnico di Torino and Newcastle University and Jonte R. Hance of Newcastle University have established a new connection between mathematical quasiprobability representations and quantum error correction. Their work demonstrates that a phase-space representation of a quantum code is possible only when the code space satisfies specific conditions. The researchers found that only the Gottesman-Kitaev-Preskill family meets these criteria, revealing limitations in the design of certain quantum codes. The study also links a property in qudit stabiliser codes, for those with odd prime dimensions, directly to a quantum resource.

Phase-Space Tools for Quantum Error Correction

The structure of quantum error correction is directly linked to quasiprobability representations, enabling a new method for determining how error channels impact code-adapted phase space. A semi-functorial representation of error correction cycles is possible when a code space satisfies a specific set of mathematical conditions governing quasiprobability.

This connection allows for a way to determine the action of an error channel on a code-adapted phase space, but obtaining such a representation for any quantum error-correcting code is not straightforward. Among compact-phase-space bosonic codes, the Gottesman-Kitaev-Preskill family remains invariant under a lattice of displacements, a condition revealed by this new mapping method.

This limitation arises because the Brif and Mann construction, used to create a phase-space representation, requires the code space to be invariant under a subgroup of an extended Weyl-Heisenberg group. Codes outside this family are limited by their code-independent sign and phase freedoms. The work builds on prior structure theorems that fix the form of quasiprobability representations of generalized probabilistic theories, combining them with the Brif and Mann construction to map error channels onto the code’s phase space.

In finite dimensions, every qudit stabiliser code exhibits a negative volume that directly corresponds to an average across syndromes. For rotation-symmetric, one-photon, non-Pauli qudit and spin codes, only the ordinary kernel remains, with its negativity reflecting the physical carrier.

A sector-graded representation, built from recovery isometries, restores a resource reading, while the relationship between the “torus” and “correctable cell × logical lattice” further defines the code’s structure. “ $χ_{K}$ = $e^{i f}$ with f real and odd,” describes a key element of the kernel used in this representation, highlighting the mathematical precision of the approach.

Brif-Mann Construction and Code-Adapted Kernels

The Brif-Mann construction, when paired with a kernel satisfying the Stratonovich-Weyl axioms, yields a semi-functorial representation that meets the conditions outlined in a key structure theorem for quantum error correction. In finite dimension, this is every qudit stabiliser code, whose negative volume for odd prime dimension is the syndrome-averaged logical magic. Researchers have identified a clear division within quantum error-correcting codes based on this new mapping method. Codes allowing a code-adapted kernel differing from the standard Wigner kernel are limited to specific families, notably the ideal Gottesman-Kitaev-Preskill codes.

Every qudit stabiliser code also falls into this category, with a resource identification detailed in Theorem IV. 1 of the study. This classification is significant because it narrows the possibilities for designing codes with specific properties and resource requirements.

The work also details several specific achievements, including the proof that Brif-Mann representations with Stratonovich-Weyl kernels fulfill structure-theorem conditions, and the identification of the class of transformations on which the representation composes correctly. The research provides a closed-form calculation of magic dissipation and lifetime, as described in Proposition IV. 5, and introduces a code-adapted Kirkwood-Dirac layer.

Stratonovich-Weyl Axioms Restrict Kernel Forms

This connection is notable because it bridges the abstract realm of quasiprobability representations with the practical demands of maintaining quantum information. Every representation generated by a kernel adhering to the Stratonovich-Weyl axioms functions as an empirically adequate and linearity-preserving semi-functor. These findings collectively refine the understanding of how mathematical constraints shape viable quantum error-correcting codes and their associated representations.

GKP Codes and Unique Phase-Space Representation

A code’s ability to maintain a consistent phase-space representation hinges on its invariance under displacement lattices, a newly established link between abstract mathematics and practical quantum error correction. A code-adapted kernel differing from the ordinary Wigner kernel exists if and only if the code space is invariant under a lattice of displacements in the relevant Weyl-Heisenberg group. The study further reveals a distinction in how codes behave in continuous versus finite dimensions.

While the ideal Dirac-comb GKP family exclusively meets the displacement lattice requirement in continuous variable systems, every qudit stabiliser code satisfies it in finite dimensions. This difference is highlighted by the cat code, which lacks its own phase-space representation and relies on the standard Wigner function, normalised by a Jacobian factor.

Tracing out a stabiliser subsystem in the cat code achieves ideal decoding, a process now mirrored by a phase-space representation layer. This holds true when the reconstruction map is the identity, a condition met when codewords reside in $L^{2}$ space. The team’s work connects the structure theorem for quasiprobability representations to quantum error-correcting codes, effectively dividing the set of codes into two distinct groups based on their phase-space characteristics.

Among bosonic codes with compact phase space, the ideal Gottesman-Kitaev-Preskill family uniquely satisfies a condition related to its stabiliser lattice, acting as the fixed-point space for that lattice. The negative volume for odd prime dimension is the syndrome-averaged logical magic.

Limitations for Bosonic and Spin Code Kernels

The work demonstrates that a code-adapted phase-space negativity can only measure logical magic within this specific family, highlighting a narrow pathway for certain types of bosonic error correction. The analysis extends to finite-dimensional codes, establishing that every qudit stabiliser code possesses a kernel constructed at prime-power dimensions. However, this construction relies on two-step nilpotency making isotropy subgroups normal, a condition not universally met. For codes whose stabiliser is a lattice of displacements within a Weyl-Heisenberg group, a code-adapted kernel exists, but the study clarifies that this is contingent upon adherence to the Stratonovich-Weyl axioms.

Recovery Isometries Yield Sector-Graded Kernels

Researchers have demonstrated that constructing a representation from a code’s recovery isometries, rather than its inherent symmetry, yields a sector-graded logical kernel. This kernel retains the semi-functorial and compositional structure of the broader framework, offering a new approach to identifying encoded resources. This sector-graded kernel provides a syndrome-resolved witness of logical magic, functioning two-sidedly on pure logical states for codes encoding a qudit of odd prime dimension.

A key outcome of this work is the development of a channel atlas, alongside a closed-form calculation of magic lifetime under displacement noise. Further refinements include the introduction of a Kirkwood-Dirac layer and finite-energy corrections, enhancing the precision of the representation.

This identification fails for codes lacking this specific structure, highlighting a critical link between code design and resource preservation. The work provides a method for quantifying how quickly this “magic” dissipates under Gaussian displacement noise, a common source of error in quantum systems.

We add a channel atlas, a Kirkwood-Dirac layer, and finite-energy corrections. The team added a Kirkwood-Dirac layer and finite-energy corrections.

👉 More information
🗞 Phase-space representations of quantum error-correcting codes
✍️ Enrico Bozzetto and Jonte R. Hance
🧠 DOI: http://link.aps.org/doi/10.1103/7my9-zgfz

More like this