Approximate Quantum Error Correction at Chiral Topological Edges

AI summary

Overview

Research area: Quantum information theory at the interface of condensed matter physics — approximate quantum error correction, topological order, and chiral conformal field theory (CFT).

Technical level: Advanced. The paper assumes familiarity with quantum error-correcting codes, coherent information, anyon models / unitary modular tensor categories, and 1+1D conformal field theory.

Scope: The paper proposes and analyzes "chiral edge codes," a family of approximate quantum error-correcting (AQEC) codes built from the gapless chiral edge modes of two-dimensional topologically ordered phases on a cylinder, and derives power-law scaling of coherent-information loss, a robustness hierarchy versus dimensionally reduced CFT codes, and a power-law-range recovery map for Abelian code subspaces.

What This Paper Is About

Gapped topological phases give exact quantum error correction through nonlocal encoding, but exact protection really only holds in the zero-correlation-length limit; gapless CFTs give approximate codes with power-law local distinguishability, but a CFT only exists if a lattice Hamiltonian is fine-tuned to criticality. The authors ask whether conformal degrees of freedom can provide approximate quantum error correction when their existence is enforced by a stable phase of matter rather than by tuning. Their answer is to use the ungapped chiral edge of a chiral topologically ordered bulk, combining the robustness of the gapped bulk with the flexibility of the edge CFT.

Key Contributions

1. A new code family: chiral edge codes (Definition 4). The authors define AQEC codes on a cylinder whose code subspace is the span of a finite set of "cylinder primary states" |𝔞, φ^I_𝔞(L), φ^J_{𝔞̄}(R)⟩, labeled by a bulk anyon sector 𝔞 and by a choice of primary state on each of the two physical edges. The code interpolates between a TQFT code and a CFT code, inheriting topological protection from the bulk and CFT-like local distinguishability from the edge.

2. An exact relative-entropy formula for coherent-information loss (Theorem 7). Using topological-sector orthogonality, the authors show that when the complement of the erased region contains a noncontractible bulk annulus, that annulus resolves the anyon sector up to the finite-correlation-length correction, and the coherent-information loss can be written as a single average of relative entropies on the erased region. This makes the quantity UV-finite and reduces recoverability to universal properties of the edge theory.

3. A robustness hierarchy against local erasure (Proposition 8). For geometrically local erasures near a single edge, the two-dimensional chiral edge code is at least as robust as its dimensionally reduced one-dimensional CFT counterpart, and strictly more robust in several representative examples.

4. A power-law-range recovery theorem for Abelian code subspaces (Theorem 9). An O(1)-sized erasure adjacent to one edge can be approximately recovered by a channel supported on the erased region together with a buffer whose length grows subextensively with the cylinder's circumference. The recovery map is a universal twirled Petz map built from the maximally mixed code state, so it depends only on the code subspace and the erased region, not on the unknown encoded state or the specific noise channel.

Main Findings

• Hybrid protection. A noncontractible bulk annulus can resolve the encoded anyon sector, and a contractible region detached from the edge is locally indistinguishable up to a finite-size correction; but a local region on the physical edge can distinguish different cylinder primary states, with information leakage vanishing only as a power law. Chiral edge codes therefore show algebraic suppression of local distinguishability along the edges together with the stronger exponential suppression in the gapped bulk.

• Diagnostic: coherent-information loss (Definition 2). The paper defines Δ(A;𝕍) as the mutual information I(A:R) between the erased region A and a reference purifying the maximally mixed code state. It equals the coherent-information change under complete erasure, is channel-independent, and upper bounds the coherent-information deficit ε(A;𝕍) generated by any channel supported on A (Proposition 1), with the bound saturated for complete corruption. An equivalent form is Δ(A;𝕍) = (S_A + S_Q − S_B)_{ρ̄_Q} (Lemma 3).

• Three power-law exponents. For an erasure Y near a single edge, Δ_2D ∼ x^γ, where x ≪ 1 is the angular size. The dimensionally reduced 1D CFT code has Δ_1D ∼ x^{min{α,β}} for an erased interval of the same angular size. The exponents α and β are defined by relative entropies of small full-column regions X and their complements on the cylinder, and admit a purely 1D CFT interpretation under the dimensional-reduction assumption (Assumption 5); γ is intrinsic to a genuinely two-dimensional region localized near a single edge.

• Robustness hierarchy γ ≥ α ≥ min{α,β}. The physical reason is that dimensional reduction treats a full column of the cylinder as a local interval, whereas a local two-dimensional error near one edge cannot simultaneously access the opposite edge through the gapped bulk.

• Numerical support in two examples. The chiral semion code (realized by bosonic Laughlin-type states, dimensionally reduced to the compactified free-boson CFT) and the chiral Ising code (dimensionally reduced to the Ising CFT). For the corresponding 1D CFT codes the authors extract α and β numerically and compare them with analytic CFT predictions for the relative entropies; for the chiral semion code they additionally estimate γ directly from bosonic Laughlin-state wave functions on finite cylinders. The results support the hierarchy with strict separation for several code subspaces. The specific numerical values of the exponents are not reported in the portion of the paper provided here.

• Buffer range and asymptotic error exponents. For Abelian code subspaces, the authors also estimate the CFT exponents controlling the buffer range and the asymptotic error in the power-law-range recovery theorem; the numerical values are not reported in the provided text.

• Effective code distance (Eq. 16). Because AQEC codes have no sharp distance threshold, the authors use d_eff(δ) := min{|A| : Δ(A;𝕍) ≥ δ}, the smallest erasure size at which coherent-information loss stays above tolerance δ, echoing prior approximate-distance proposals.

• Contrast with exact codes. For an [[n,k,d]] stabilizer code with |A| ≤ d−1, Δ(A;𝕍) = 0 exactly. For a TQFT code on a torus (e.g., the toric code with its four ground states labeled 1, e, m, ε), a local disk erasure is perfectly correctable via the entanglement bootstrap axiom A0, and a local recovery channel R_{AB₋} exists; such errors are neglected as exponentially small.

Methodology in Plain English

The authors set up the problem on a cylinder: a chiral topological order in the bulk, with counter-propagating chiral edge modes on the two physical edges and a gapped bulk in between. They pick a code subspace by selecting a finite set of anyon sectors, choosing one cylinder primary state in each, and taking their span. To measure how much logical information leaks to a local region, they use coherent-information loss under erasure, which is a basis-independent, channel-independent quantity. They then exploit a topological fact — a noncontractible annulus in the bulk can identify the anyon sector — to convert this loss into a relative-entropy average over the erased region only, which is finite and computable. An additional assumption (dimensional reduction) identifies this relative-entropy average with the corresponding quantity in a 1D CFT when the erased region is a full column, so standard CFT techniques give the scaling exponents. For the recovery theorem, the authors combine full-boundary entanglement bootstrap conditions with the decay of mutual information between separated intervals in the dimensionally reduced CFT, and construct an explicit twirled Petz map. Finally, they check the predicted power laws numerically: exact diagonalization-type CFT computations for α and β, and finite-cylinder bosonic Laughlin wave functions for γ.

Why This Matters

Impact on research. The work supplies a construction in which conformal (power-law) error-correcting behavior is protected by a gapped topological phase rather than by fine tuning, filling a gap between topological codes with exponential local indistinguishability and critical CFT codes that require tuning. It also gives a general information-theoretic toolkit — coherent-information loss, an exact relative-entropy reduction, and an effective code distance — that can be applied to other gapless many-body encodings, and it separates the question of whether recovery exists from the question of whether recovery can be made geometrically local, which general Petz-type theorems do not address.

Potential applications (as directions implied by the framework, not demonstrated devices):

• Quantum memories built on platforms hosting chiral topological order, such as fractional quantum Hall systems and chiral spin liquids, where the edge mode is guaranteed by the bulk rather than by tuning.
• Benchmarking and designing codes in gapless or edge-dominated architectures where exact code distance is not available and power-law diagnostics are more informative.
• Judging whether a physically realistic recovery operation can be implemented locally, using the power-law-range recovery map as a template.
• Providing quantitative criteria (effective code distance at a chosen tolerance) for comparing candidate topological platforms as quantum memories.

Industry relevance. The results are relevant to quantum-computing efforts that rely on topological protection and to hardware platforms where chiral edge physics appears, since they clarify what kind of protection such hardware can offer beyond exact topological codes. The locality constraints on recovery maps also bear on the practical decoding question of how much of a device must be read out and acted upon to correct a local error.

Future Directions

• Extending the power-law-range recovery theorem beyond Abelian code subspaces, since Theorem 9 assumes only Abelian anyon sectors in the code subspace; non-Abelian chiral edges (for example, Ising-type sectors with d_𝔞 > 1) are not covered by the locality result as stated.
• Obtaining a systematic derivation, rather than an assumption, of the dimensional-reduction correspondence (Assumption 5) that links full-column relative-entropy averages on the cylinder to the 1D CFT.
• Determining how the exponents α, β, γ and the buffer-length scaling behave across a wider range of chiral topological orders and code-subspace choices, and whether strict separation from the 1D counterpart is generic or restricted to specific examples.
• Assessing robustness of the proposed codes against perturbations and imperfections in realistic lattice realizations, and clarifying how the neglected finite-correlation-length corrections to anyon-sector orthogonality affect the scaling predictions.
• Addressing the open problems listed in the paper's concluding section (Section VIII), which are not reproduced in the portion of the text available here.

Target Audience

Quantum information theorists and condensed matter physicists working on quantum error correction, topological order, and conformal field theory; graduate students and researchers interested in approximate quantum error correction in many-body systems; and practitioners evaluating topological or gapless platforms for quantum memory who need quantitative, geometry-dependent measures of protection and recovery locality.

Authors’ abstract

Topologically ordered phases naturally realize quantum error correction through nonlocal encoding of quantum information. More recently, conformal field theories have been shown to realize approximate quantum error-correcting codes, but such constructions generally require fine tuning to criticality. Here we introduce a family of approximate quantum error-correcting codes realized by the chiral edges of two-dimensional topologically ordered phases. The proposed encoding combines the robustness of a gapped topological bulk with the flexibility of gapless edge conformal field theories. To characterize its robustness, we study coherent-information loss under local erasure. We derive an exact expression relating coherent-information loss to relative entropy, reducing the recoverability problem to universal properties of the edge theory. This leads to power-law scaling of coherent-information loss with the size of the erased region. We further show that, for geometrically local erasures near one edge, the two-dimensional chiral edge code is at least as robust as the dimensionally reduced CFT code, and is strictly more robust in several representative examples. For Abelian code subspaces, we further construct a power-law-range recovery map supported on the erased region together with a power-law-range buffer; this recovery map depends only on the code subspace, not on the unknown encoded state. We provide numerical calculations for lattice realizations of compact free boson and Ising CFT examples that support the theoretical predictions of the power-law exponents.