To maximize the potential of quantum error-correcting codes, it is essential to develop high-performance decoders. The subsystem many-hypercube (MHC) codes have been developed to achieve both high encoding rates and low-weight syndrome-measurements, but the introduction of gauge degrees of freedom makes decoding more challenging. In this work, we develop neural-network-based decoders for the subsystem MHC codes in a circuit-level noise model. We demonstrate that even the gauge-measurement information can be utilized for decoding by carefully arranging the syndrome-measurement sequence, improving the decoding performance. We further show that recurrent neural decoders outperform simple fully connected neural decoders, and can decode syndrome-measurement sequences longer than those used during training.
To maximize the potential of quantum error-correcting codes, it is essential to develop high-performance decoders. The subsystem many-hypercube (MHC) codes have been developed to achieve both high encoding rates and low-weight syndrome-measurements, but the introduction of gauge degrees of freedom makes decoding more challenging. In this work, we develop neural-network-based decoders for the subsystem MHC codes in a circuit-level noise model. We demonstrate that even the gauge-measurement information can be utilized for decoding by carefully arranging the syndrome-measurement sequence, improving the decoding performance. We further show that recurrent neural decoders outperform simple fully connected neural decoders, and can decode syndrome-measurement sequences longer than those used during training.
Quantum error-correcting codes (QECCs) require high encoding rates in addition to high thresholds unless a sufficiently large number of physical qubits are available. Many-hypercube (MHC) codes, defined as concatenations of the [ [ 6 , 4 , 2 ] ] quantum error-detecting code, have been proposed as high-performance and high-encoding-rate QECCs. However, concatenated codes have the disadvantage that the syndrome weight grows exponentially with respect to the concatenation level. To address this issue, here we propose subsystem quantum codes based on MHC codes. In particular, we study the smallest subsystem MHC codes, namely, subsystem codes derived from concatenated [ [ 4 , 2 , 2 ] ] error-detecting codes. The resulting codes have a constant syndrome-measurement weight of 4, while keeping high encoding rates. We build block-MAP and neural-network decoders and show that they demonstrate superior performance to the bounded-distance decoder.
Abstract. Suppressing qubit errors is one of the main challenges in building fault-tolerant quantum computers, requiring quantum error correction to protect fragile quantum information during computation cycles. The surface code is a leading approach because of its high error tolerance and compatibility with realistic hardware, but it requires a decoder that can accurately and quickly use the syndrome measurements to discern whether a logical error has occurred. This study investigates whether a feedforward neural-network can serve as a practical decoder for the surface code under circuit-level noise. Using the Stim software package, this study generates surface code circuits at multiple code distances and physical error rates and compiles them into detector samplers, then trains a PyTorch neural-network to predict whether a logical observable flip occurred from the syndrome data. The neural-network decoder is compared against Minimum-Weight Perfect Matching (MWPM), focusing on metrics including logical error rate and inference speed. The results show that the neural-network decoder performs competitively with MWPM for code distances 3 and 5 and provides inference time advantages at higher physical error rates as its inference time remains relatively constant over increasing physical error rates. However, MWPM outperforms the neural-network at distance 7, suggesting that scalability remains a major limitation for neural-network decoders. These results show that neural-networks are a promising approach for tackling quantum error correction, although future work is still needed to improve these models for real systems at larger code distances.
Fault-tolerant quantum computation critically depends on the ability to detect and correct physical qubit errors faster than they accumulate. The surface code (SC) has emerged as the most experimentally promising quantum error correction (QEC) framework due to its high threshold and local stabilizer structure, yet its practical performance is bottlenecked by decoder efficiency. Traditional decoding approaches such as minimum-weight perfect matching (MWPM) achieve near-optimal accuracy under independent noise but struggle with correlated errors, circuit-level noise, and real-time latency constraints. This paper investigates how neural network (NN)-based decoders reduce logical error rates in SC quantum systems by learning complex syndrome-to-correction mappings from data. We propose a hybrid convolutional-recurrent neural architecture that processes stabilizer measurement syndromes across both spatial and temporal dimensions, enabling the decoder to capture correlated error patterns inherent in realistic hardware. Systematic simulations under depolarizing, biased, and circuit-level noise models demonstrate that the proposed neural decoder achieves a logical error rate reduction of up to 41.3% compared to MWPM at a physical error rate of p = 0.008, while maintaining sub-millisecond decoding latency compatible with superconducting qubit cycle times. These results establish that data-driven neural decoders represent a scalable and hardware-compatible pathway toward achieving the logical error rates required for practical fault-tolerant quantum computing.
Quantum computers has the potential to outperform classical computers for certain tasks, but the practical use of quantum computers is limited by noise and errors in quantum systems. Quantum Error Correction (QEC) provides a way for detecting and correcting these errors, but in the presence of complex and correlated noise, efficient decoding still remains a significant challenge. In this thesis, neural network-based decoders are explored as an approach to decode QEC data. In particular, recurrent neural network (RNN) architectures, including Gated Recurrent Unit (GRU) and Long Short-Term Memory (LSTM), are investigated as their are able to capture temporal dependencies and model sequential data. GRU and LSTM models are trained using the data generated with the help of the Stim simulator for quantum circuits that are based on experimental setups. The models are trained to classify error occurrences from detection event data and evaluated under different configurations, including varying number of rounds, code distance and input representation. The results show that both GRU and LSTM models can learn error patterns for shorter sequences and provide competitive decoding performance. However, for longer sequences the models struggle. Overall, this work demonstrates that relatively simple neural network models can be used as alternatives to traditional decoding methods and highlights the potential of machine learning approaches for improving quantum error correction.
Qudits provide a larger local Hilbert space than qubits and have recently attracted interest as building blocks for efficient fault-tolerant quantum computation (FTQC). In this work, we extend many-hypercube (MHC) codes, high-rate concatenated quantum codes, from qubits to qudits of prime dimension q. We construct the level-2 and level-3 MHC codes with parameters [[62,42,22]]q and [[63,43,23]]q, respectively, and generalize the level-by-level minimum-distance (LLMD) decoder originally proposed for the qubit MHC codes. In a qudit X-error model, we evaluate the code-capacity performance of the qudit MHC codes for dimensions up to q=13. We find that the error threshold increases monotonically with the local dimension q, reaching 10.0% for q=13, nearly double the value of 5.1% for qubits. Furthermore, for sufficiently large local dimensions, we observe a pronounced waterfall regime in which the logical error rate decreases substantially faster than expected from conventional distance-based estimates. To understand the origin of the observed qudit advantage, we analyze the decoding performance in terms of the entropy of the underlying q-ary noise channel. This analysis reveals a competition between the increased syndrome information available at larger q and the simultaneous growth of the typical physical error weight at fixed entropy. These results demonstrate that higher-dimensional quantum systems can substantially improve the performance of high-rate concatenated quantum codes and motivate further development of qudit-based FTQC architectures.
Neural network decoders can achieve a lower logical error rate compared to conventional decoders, like minimum-weight perfect matching, when decoding the surface code. Furthermore, these decoders require no prior information about the physical error rates, making them highly adaptable. In this study, we investigate the performance of such a decoder using both simulated and experimental data obtained from a transmon-qubit processor, focusing on small-distance surface codes. We first show that the neural network typically outperforms the matching decoder due to better handling of errors leading to multiple correlated syndrome defects, such as Y errors. When applied to the experimental data of Google Quantum AI [R. Acharya , ], the neural network decoder achieves logical error rates approximately 25 % lower than minimum-weight perfect matching, approaching the performance of a maximum-likelihood decoder. To demonstrate the flexibility of this decoder, we incorporate the soft information available in the analog readout of transmon qubits and evaluate the performance of this decoder in simulation using a symmetric Gaussian-noise model. Considering the soft information leads to an approximately 10 % lower logical error rate, depending on the probability of a measurement error. The good logical performance, flexibility, and computational efficiency make neural network decoders well-suited for near-term demonstrations of quantum memories.