arXiv:2610.05901
Fusion-assisted decoding of non-Abelian topological order
Rohith Sajith, Yabo Li, Zijian Song, Tsung-Cheng Lu, Carolyn Zhang
Abstract
Maximum likelihood (ML) decoders perform quantum error correction by measuring a set of syndromes, classically computing an optimal recovery based on these syndromes and the noise model, and then performing that recovery. ML decoders are well understood for the toric code and generalize straightforwardly to other surface codes with Abelian anyons, where they can saturate information-theoretic decoding thresholds. On the other hand, codes with non-Abelian anyons have become increasingly relevant because they can support transversal or constant-depth non-Clifford gates and admit efficient preparation protocols in modern quantum devices. However, they present new challenges for decoding: non-Abelian anyons cannot in general be deterministically annihilated, and the logical operators cannot be classified by Abelian homology. In this work, we show that refined syndrome sets, including measurements that resolve non-Abelian anyon fusion outcomes, systematically improve the performance of ML decoders. We demonstrate this concretely by constructing explicit decoders for the $D_4$ quantum double model under charge noise, and obtaining numerical results for thresholds via mappings to classical statistical-mechanics models. We show that there is a nonzero noise threshold below which the entire logical qudit can be recovered with a single round of measurements.
AI Summary
Non-Abelian topological memories require decoding information beyond local anyon positions: collective fusion states affect recovery, while fusion measurements can themselves reveal logical information. This paper develops syndrome-dependent maximum-likelihood decoding for quantum-double models under charge noise, specializing to independent non-Abelian $E$-charge noise in the $D_4$ model on a cylinder. The five-dimensional ground space contains a protected “central qubit” associated with the group center. Three nested syndrome records resolve $E$ positions, all onsite charges, and additionally pair-fusion outcomes along matching-selected paths. Exact syndrome probabilities and winding-sector weights map to classical statistical-mechanics partition functions. Sampling physical syndrome records and contracting these partition functions exactly gives estimated central-qubit thresholds $p_c^{(1)}\simeq0.111$, $p_c^{(2)}\simeq0.165$, and $p_c^{(3)}\simeq0.214$. Abelian grading extends the central-sector construction to other groups with nontrivial centers. Path-dependent fusion measurements preserve the central qubit but can measure noncentral logical sectors when recovery paths close noncontractibly with physical errors. A construction also argues for full-qudit recovery below a matching threshold, quoting $p_{\mathrm{full}}\geq p_{\mathrm{HMWPM}}\geq p_{\mathrm{MWPM}}\simeq0.103$, with Abelian-cleanup confinement assumed. Here “single-shot” means one syndrome-measurement stage with potentially nonlocal fusion measurements and perfect measurements and recovery—not tolerance to measurement noise. The explicit threshold improvements concern a restricted charge-noise model and logical subspace; flux noise, efficient likelihood evaluation, and broader optimal recovery remain open. Rényi-2 coherent-information calculations indicate a transition near $p\simeq0.45$–$0.46$ for both logical inputs, but do not establish the intrinsic von Neumann or optimal-recovery thresholds.
Central claim may not hold: The advertised full-qudit single-round recovery result, particularly the bound p_full ≥ p_HMWPM ≥ 0.103, relies on an explicitly unestablished confinement assumption for Abelian cleanup, while the measurement-reordering identity alone does not establish that assumption or the threshold bound.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 29 + 32 pages, 16 + 1 figures