Abstract:The decoding of error syndromes of surface codes with classical algorithms may slow down quantum computation. To overcome this problem it is possible to implement decoding algorithms based on artificial neural networks. This work reports a study of decoders based on convolutional neural networks, tested on different code distances and noise models. The results show that decoders based on convolutional neural networks have good performance and can adapt to different noise models. Moreover, explainable machine learning techniques have been applied to the neural network of the decoder to better understand the behaviour and errors of the algorithm, in order to produce a more robust and performing algorithm.

What problem does this paper attempt to address?

The paper primarily explores how to efficiently decode error syndromes in surface codes using a Convolutional Neural Network (CNN)-based decoder to overcome the speed bottlenecks that traditional classical algorithms may encounter in quantum computing. Specifically, the paper investigates the performance of CNN decoders under different code distances and noise models, and applies interpretable machine learning techniques to better understand the behavior and error sources of these decoders, thereby improving their stability and performance. The main contributions of the paper include: 1. **Proposing a CNN Decoder**: To address the issue of significantly increased execution time of classical algorithms for surface code decoding as the number of physical qubits increases, the authors propose a CNN-based decoder and test its performance under different code distances and noise models. 2. **Adaptability to Different Noise Models**: The paper not only studies the depolarizing noise model but also considers a more complex noise model that includes measurement errors. Experimental results show that the CNN decoder can adapt well to these different noise models. 3. **Application of Interpretable Machine Learning**: To improve the performance and stability of the decoder, the paper applies interpretable machine learning methods to analyze the behavior of the decoder. Through this approach, researchers can better understand the working principles of the decoder and identify potential error sources. 4. **Impact of Training Set Selection**: The paper also explores the impact of different training set error probabilities on the performance of the decoder. The results indicate that using training sets with higher error probabilities generally leads to better decoding performance, but excessively high error probabilities may cause training difficulties. In summary, this paper aims to accelerate the error decoding process in surface codes by designing and optimizing a CNN-based decoder, especially in the face of complex noise environments, and further enhances the reliability and efficiency of the decoder by introducing interpretability techniques.

Convolutional neural network based decoders for surface codes

Decoding surface code with a distributed neural network based decoder

A scalable and fast artificial neural network syndrome decoder for surface codes

Scalable Neural Decoder for Topological Surface Codes

Learning to Decode the Surface Code with a Recurrent, Transformer-Based Neural Network

Comparing neural network based decoders for the surface code

Neural network decoder for near-term surface-code experiments

Neural-Network Decoders for Quantum Error Correction using Surface Codes:A Space Exploration of the Hardware Cost-Performance Trade-Offs

Decoding algorithms for surface codes

A Scalable, Fast and Programmable Neural Decoder for Fault-Tolerant Quantum Computation Using Surface Codes

CIM-Based Parallel Fully FFNN Surface Code High-Level Decoder for Quantum Error Correction

Optimal adaptation of surface-code decoders to local noise

ecurrent neural network decoding of rotated surface codes based on distributed strategy

Improved decoding of circuit noise and fragile boundaries of tailored surface codes

Efficient soft-output decoders for the surface code

Testing the Accuracy of Surface Code Decoders

The END: An Equivariant Neural Decoder for Quantum Error Correction

Data-driven decoding of quantum error correcting codes using graph neural networks

Minimising surface-code failures using a color-code decoder

Deep Quantum Error Correction

Decoding surface codes with deep reinforcement learning and probabilistic policy reuse