arXiv:2610.02734
SMP: A General Hyperedge-Based Framework for Circuit-Level Quantum Error Correction
Siying Wang, Yue Yan, Zhixin Xia, Canwei Shi, Hannuo Yuan, Xiang-Bin Wang
Abstract
Hyperedge faults remain a major challenge in circuit-level quantum error correction. Accurately exploiting hyperedge correlations often requires substantial computational resources, which complicates real-time decoding. Matching-based decoders are fast and scalable, but their pairwise graph representation limits the use of higher-order fault information. Here, we introduce Syndrome Motif Projection (SMP), a general framework for incorporating hyperedge information into matching-based decoding. The key idea is that a hyperedge fault produces a characteristic local syndrome motif. Observing such a motif provides evidence for the corresponding fault. SMP extracts this information directly from the measured syndrome with linear-complexity preprocessing while preserving the matching backend. It can therefore serve as a lightweight front end to existing scalable and real-time matching decoders. We demonstrate SMP across surface code memories, color code memories, and logical gate circuits. On Google's Willow experimental surface code data, SMP improves both standard and correlated matching and outperforms the tested belief matching baseline. For color code memories, SMP combined with Chromobius reduces logical failure probabilities by up to 77.8\%. For logical gate circuits, SMP reduces the failure probability of a six-CNOT circuit by 35.5\% relative to logical observable matching while achieving performance comparable to iterative belief matching. These results demonstrate that SMP provides a lightweight and transferable approach to hyperedge decoding, with a clear path toward integration into scalable real-time decoding architectures for fault-tolerant quantum computation.
AI Summary
Matching decoders replace circuit-level hyperedge faults with pairwise edges, potentially losing correlations. Syndrome Motif Projection (SMP) addresses how to recover some of this information without replacing the matching backend. It identifies detector patterns associated with candidate faults, estimates local evidence from complete or partially observed motifs, and uses that evidence to interpolate edge weights toward the candidate’s joint fault cost. Overlapping proposals are combined edgewise rather than through a joint posterior. For the bounded-locality surface-code circuits considered, preprocessing and storage scale as $O(d^2r)$. Tests cover public Willow data and simulated surface-code memories, color-code memories, and transversal-CNOT circuits. On Willow at $d=7$, SMP reduces fitted per-round logical error rates by about 22% over standard MWPM and 6.0–7.9% over correlated MWPM. At 13 rounds, it has lower observed failure probabilities than the tested ten-iteration belief-matching baseline, but not Harmony-51. Surface-code simulations extend to $d=19$, with 13.6–17.4% reductions over correlated MWPM at one reported noise strength. With Chromobius, reductions reach 77.8%; one color-code case falls from 0.036% to 0.008%. A six-CNOT, $d=5$ simulation falls from 0.535% to 0.345% circuit failure on 20,000 shared shots. The evidence models and weight updates are approximations with application-specific conventions. Some color-code estimates involve few failures; the gate benchmark lacks independent validation. BP-assisted gate decoding has lower observed failure rates. No execution-time measurements establish the accuracy–latency trade-off, and hardware integration remains untested.
Error found: The abstract's linear-complexity preprocessing claim needs bounded-locality assumptions that the paper establishes for its surface-code setting, not for arbitrary applications of the general framework.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.