Using a Recurrent Neural Network to Reconstruct Quantum Dynamics of a Superconducting Qubit from Physical Observations

Abstract

At its core, Quantum Mechanics is a theory developed to describe fundamental observations in the spectroscopy of solids and gases. Despite these practical roots, however, quantum theory is infamous for being highly counterintuitive, largely due to its intrinsically probabilistic nature. Neural networks have recently emerged as a powerful tool that can extract non-trivial correlations in vast datasets. They routinely outperform state-of-the-art techniques in language translation, medical diagnosis and image recognition. It remains to be seen if neural networks can be trained to predict stochastic quantum evolution without a priori specifying the rules of quantum theory. Here, we demonstrate that a recurrent neural network can be trained in real time to infer the individual quantum trajectories associated with the evolution of a superconducting qubit under unitary evolution, decoherence and continuous measurement from raw observations only. The network extracts the system Hamiltonian, measurement operators and physical parameters. It is also able to perform tomography of an unknown initial state without any prior calibration. This method has potential to greatly simplify and enhance tasks in quantum systems such as noise characterization, parameter estimation, feedback and optimization of quantum control.

References in corpus (29)

• LSTM: A Search Space Odyssey
• Charge insensitive qubit design derived from the Cooper pair box
• Solving the Quantum Many-Body Problem with Artificial Neural Networks
• Machine learning phases of matter
• Observation of high coherence in Josephson junction qubits measured in a three-dimensional circuit QED architecture
• Many-body quantum state tomography with neural networks
• Learning phase transitions by confusion
• Informational derivation of Quantum Theory
• Discovering Phase Transitions with Unsupervised Learning
• Observing single quantum trajectories of a superconducting qubit
• Quantum trajectory approach to circuit QED: Quantum jumps and the Zeno effect
• Dispersive magnetometry with a quantum limited SQUID parametric amplifier
• Reinforcement Learning with Neural Networks for Quantum Feedback
• Mapping the optimal route between two quantum states
• A Neural Decoder for Topological Codes
• Dynamics of simultaneously measured non-commuting observables
• Past quantum states
• Machine-learning-assisted correction of correlated qubit errors in a topological code
• Machine learning for discriminating quantum measurement trajectories and improving readout
• Time-Symmetric Quantum Theory of Smoothing
• Deep Neural Network Probabilistic Decoder for Stabilizer Codes
• Prediction and retrodiction for a continuously monitored superconducting qubit
• Observing interferences between past and future quantum states in resonance fluorescence
• Efficient Quantum Filtering for Quantum Feedback Control
• Dynamics of a qubit while simultaneously monitoring its relaxation and dephasing
• Using Recurrent Neural Networks to Optimize Dynamical Decoupling for Quantum Memory
• Quantum State Smoothing
• Quantum state tomography with non-instantaneous measurements, imperfections and decoherence
• Quantum smoothing for classical mixtures

Cited by in corpus (45)

• Prospects and applications of photonic neural networks
• Artificial Intelligence and Machine Learning for Quantum Technologies
• Model-Free Quantum Control with Reinforcement Learning
• Machine Learning on Neutron and X-Ray Scattering
• Integrated tool-set for Control, Calibration and Characterization of quantum devices applied to superconducting qubits
• Deep Reinforcement Learning for Quantum State Preparation with Weak Nonlinear Measurements
• Hilbert Space Fragmentation and Exact Scars of Generalized Fredkin Spin Chains
• Learning models of quantum systems from experiments
• Neural-network quantum state tomography
• A differentiable programming method for quantum control
• Reservoir Computing Approach to Quantum State Measurement
• Predicting non-Markovian superconducting qubit dynamics from tomographic reconstruction
• Learning Quantum Hamiltonians from Single-qubit Measurements
• Learning quantum dynamics with latent neural ODEs
• Machine Learning and Quantum Devices
• Quantum-tailored machine-learning characterization of a superconducting qubit
• Deep Learning of Quantum Many-Body Dynamics via Random Driving
• Quantum circuit fidelity estimation using machine learning
• Recurrent Quantum Neural Networks
• Control of Stochastic Quantum Dynamics by Differentiable Programming
• Neural networks for on-the-fly single-shot state classification
• Anti-crosstalk high-fidelity state discrimination for superconducting qubits
• Accurate phonon blockade detector composed of a quadratically coupled optomechanical system
• Machine classification for probe based quantum thermometry
• Time-dependent atomic magnetometry with a recurrent neural network
• Ground state search by local and sequential updates of neural network quantum states
• Bridging the reality gap in quantum devices with physics-aware machine learning
• Retrieving past quantum features with deep hybrid classical-quantum reservoir computing
• Deep learning of many-body observables and quantum information scrambling
• Quantum Next Generation Reservoir Computing: An Efficient Quantum Algorithm for Forecasting Quantum Dynamics
• Classification and reconstruction of optical quantum states with deep neural networks
• Learning Order Parameters from Videos of Dynamical Phases for Skyrmions with Neural Networks
• Neural networks in quantum many-body physics: a hands-on tutorial
• Quantum Model Learning Agent: characterisation of quantum systems through machine learning
• Quantum Reservoir Computing for Speckle-Disorder Potentials
• Monitoring-induced Entanglement Entropy and Sampling Complexity
• Multiparameter estimation of continuous-time Quantum Walk Hamiltonians through Machine Learning
• Dual-Capability Machine Learning Models for Quantum Hamiltonian Parameter Estimation and Dynamics Prediction
• Hamiltonian Learning using Machine Learning Models Trained with Continuous Measurements
• Differentiable master equation solver for quantum device characterisation
• Hamiltonian-learning quantum magnets with non-local impurity tomography
• Modeling and simulating the excited-state dynamics of a system with condensed phases: A machine learning approach
• Hamiltonian Learning via Inverse Physics-Informed Neural Networks
• Multiclass classification of dephasing channels
• Emulating ultrafast dissipative quantum dynamics with deep neural networks