Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2410.02818 · Physical Review Applied

Recovery of quantum correlations using machine learning

Edward W. Steele, Donald R. Reising, Tian Li

Published Mar 27, 2025• 5 citations
arXivHTMLPDF

Abstract

Quantum sources with strong correlations are essential but delicate resources in quantum information science and engineering. Decoherence and loss are the primary factors that degrade nonclassical quantum correlations, with scattering playing a role in both processes. In this work, we present a method that leverages long short-term memory, a machine-learning technique known for its effectiveness in time-series prediction, to mitigate the detrimental impact of scattering in quantum systems. Our setup involves generating two-mode squeezed light via four-wave mixing in warm rubidium vapor, with one mode subjected to a scatterer to disrupt quantum correlations. Mutual information and intensity-difference squeezing between the two modes are used as metrics for quantum correlations. We demonstrate a 74.7% recovery of mutual information and an 87.7% recovery of two-mode squeezing, despite significant photon loss that would otherwise eliminate quantum correlations. This approach marks a significant step toward recovering quantum correlations from random disruptions without the need for hardware modifications, paving the way for practical applications of quantum protocols.

DOI
10.1103/physrevapplied.23.034083

Related papers

Mitigating Scattering in a Quantum System Using Only an Integrating Sphere

Zhenfei Jiang, Tian Li, Matthew L. Boone +4 more

Strong quantum correlated sources are essential but delicate resources for quantum information science and engineering protocols. Decoherence and loss are the two main disruptive processes that lead to the loss of nonclassical behavior in quantum correlations. In quantum systems, scattering can contribute to both decoherence and loss. In this work, we present an experimental scheme capable of significantly mitigating the adverse impact of scattering in quantum systems. Our quantum system is composed of a two-mode squeezed light generated with the four-wave-mixing process in hot rubidium vapor and a scatterer is introduced to one of the two modes. An integrating sphere is then placed after the scatterer to recollect the scattered photons. We use mutual information between the two modes as the measure of quantum correlations and demonstrate a 47.5% mutual information recovery from scattering, despite an enormous photon loss of greater than 85%. Our scheme is the very first step toward recovering quantum correlations from disruptive random processes and thus has the potential to bridge the gap between proof-of-principle demonstrations and practical real-world implementations of quantum protocols. Published by the American Physical Society 2024

arXiv:2405.15947•PRX Quantum•Sep 13, 2024•3 citations

Machine Learning using Quantum Light

Jorge García‐Beni

[eng] Quantum photonic technologies have achieved considerable success across a range of disciplines, including quantum computing, quantum communications, and quantum metrology. Conversely, in the current era of information, the ability to discern patterns within voluminous data sets is highly sought after, with Machine Learning (ML) having emerged as a powerful solution to address this challenge. The processing speed of photonic hardware, in conjunction with the potential for computation enhancement resulting from quantum properties, positions quantum photonics as a promising platform for the implementation of ML algorithms. In this thesis, we focus on reservoir computing (RC) as an ML framework that is suitable for time series processing. It provides fast and energy-efficient training in comparison to more conventional methods and is also appropriate for physical implementations. In this thesis, we explore the possibilities brought by quantum photonic setups in the field of quantum RC. Specifically, we propose and study different platform designs that are able to perform different temporal benchmark tasks by providing tunable memory and nonlinearity. Furthermore, we leverage the capabilities of state-of-the-art quantum photonic hardware and identify novel strategies to address key challenges in RC in quantum substrates, as well as expand the potential of quantum machine learning. Firstly, an approach is provided to overcome some of the limitations of quantum projective measurements, thereby enabling information processing in real time. This is achieved by generating a physical ensemble of quantum reservoirs that recirculate through an optical fiber inside a feedback loop. We investigate the detrimental effects of sampling noise due to finite ensembles and provide a strategy to mitigate them. Secondly, the consequences of increasing quantum squeezing in the dynamics of such setups were studied, and it was found that it monotonically improves the noise robustness of the platforms in realistic scenarios. Finally, we propose a novel design in which photonic cluster states, a specific kind of highly entangled quantum network, are used as resources to implement quantum RC. In this configuration, external input information is encoded into the quantum network via quantum teleportation by engineering the measurement basis of the protocol. Furthermore, the platform has the capacity to implement global quantum gates through local operations (single-mode measurements), thus establishing the foundation for distributed quantum ML.

•Jan 1, 2026•0 citations

Quantum Optical Experiments Modeled by Long Short-Term Memory

Thomas Adler, Manuel Erhard, Mario Krenn +3 more

We demonstrate how machine learning is able to model experiments in quantum physics. Quantum entanglement is a cornerstone for upcoming quantum technologies, such as quantum computation and quantum cryptography. Of particular interest are complex quantum states with more than two particles and a large number of entangled quantum levels. Given such a multiparticle high-dimensional quantum state, it is usually impossible to reconstruct an experimental setup that produces it. To search for interesting experiments, one thus has to randomly create millions of setups on a computer and calculate the respective output states. In this work, we show that machine learning models can provide significant improvement over random search. We demonstrate that a long short-term memory (LSTM) neural network can successfully learn to model quantum experiments by correctly predicting output state characteristics for given setups without the necessity of computing the states themselves. This approach not only allows for faster search, but is also an essential step towards the automated design of multiparticle high-dimensional quantum experiments using generative machine learning models.

arXiv:1910.13804•Photonics•Nov 26, 2021•10 citations

Experimental Generation of Multiple Quantum Correlated Beams from Hot Rubidium Vapor

Zhongzhong Qin, Leiming Cao, Hailong Wang +3 more

Quantum correlations and entanglement shared among multiple quantum modes are important for both fundamental science and the future development of quantum technologies. This development will also require an efficient quantum interface between multimode quantum light sources and atomic ensembles, which makes it necessary to implement multimode quantum light sources that match the atomic transitions. Here, we report on such a source that provides a method for generating quantum correlated beams that can be extended to a large number of modes by using multiple four-wave mixing (FWM) processes in hot rubidium vapor. Experimentally, we show that two cascaded FWM processes produce strong quantum correlations between three bright beams but not between any two of them. In addition, the intensity-difference squeezing is enhanced with the cascaded system to -7.0±0.1 dB from the -5.5±0.1/-4.5±0.1 dB squeezing obtained with only one FWM process. One of the main advantages of our system is that as the number of quantum modes increases, so does the total degree of quantum correlations. The proposed method is also immune to phase instabilities due to its phase insensitive nature, can easily be extended to multiple modes, and has potential applications in the production of multiple quantum correlated images.

arXiv:1407.3863•Physical Review Letters•Jul 10, 2014•192 citations

Atomic entanglement and decoherence.

Claudiu Genes

The generation of entanglement in atomic systems plays a central topic in the fields of quantum information storage and processing. Moreover, a special category of entangled states of multi-atom ensembles, spin squeezed states, have been proven to lead to considerable improvement in the sensitivity of precision measurements compared to systems involving uncorrelated atoms. A treatment of entanglement in open systems is, however, incomplete without a precise description of the process of decoherence which necessarily accompanies it. The theory of entanglement and decoherence are the two main topics of this thesis. Methods are described for the generation of strong correlations in large atomic ensembles using either cavity quantum electrodynamics or measurement outcome conditioned quantum dynamics. Moreover, the description of loss of entanglement resulting from the coupling to a noise reservoir (electromagnetic vacuum) is explored. A spin squeezing parameter is used throughout this thesis as both a measure of entanglement strength and as an indication of the sensitivity improvement above the so-called standard quantum limit (sensitivity obtained with uncorrelated particles) in metrology. The first scheme considered consists of a single mode cavity field interacting with a collection of atoms for which spin squeezing is produced in both resonant and off-resonant regimes. In the resonant case, transfer of squeezing from a field state to the atoms is analyzed, while in the off-resonant regime squeezing is produced via an effective nonlinear interaction (one-axis twisting Hamiltonian). A second, more experimentally realistic case, is one involving the interaction of free space atoms with laser pulses; a projective measurement of a source field originating from atomic fluctuations provides a means of preparing atomic collective states such as spin squeezed and Schrodinger cat states. A new unravelling is proposed, that employs the detection of photon number in a single output channel and is capable of producing mesoscopic Schrodinger cat states in a single step. As a first step to understanding the role of cooperative decoherence, of importance in the case of dense pencil-shaped atomic ensembles, the collective spin decoherence of a two multilevel atom system is derived. This calculation is also relevant to entanglement loss for two qubits manipulated using with reading/writing pulses. Finally, a scheme is proposed in which arbitrarily strong entanglement can be produced in a four-wave mixing setup, where the preparation of the atoms in a dark state limits the decoherence to negligible values.

•Jan 1, 2006•0 citations

All-optical correlated noisy channel and its application in recovering quantum coherence

Dan Lei, Disheng Guo, Jun Xin +1 more

Attenuation and amplification are the most common processes for optical communications. Amplification can be used to compensate the attenuation of the complex amplitude of an optical field, but is unable to recover the coherence lost, provided that the attenuation channel and the amplification channel are independent. In this paper, we show that the quantum coherence of an optical field can be regained if the attenuation channel and the amplification channel share correlated noise. We propose an all-optical correlated noisy channel relying on the four-wave mixing process and demonstrate its capability of recovering quantum coherence within continuous-variable systems. We quantitatively investigate the coherence recovery phenomena for coherent states and two-mode squeezed states. Moreover, we analyze the effect of other photon losses that are independent from the recovery channel on the performance of recovering coherence. Different from correlated noisy channels previously proposed based on electro-optic conversions, the correlated noisy channel in our protocol is all-optical and thus owns larger operational bandwidths.

arXiv:2310.16342•Physical Review A•Jun 10, 2024•4 citations
Quantum sources with strong correlations are essential but delicate resources in quantum information science and engineering. Decoherence and loss are the primary factors that degrade nonclassical quantum correlations, with scattering playing a role in both processes. In this work, we present a method that leverages long short-term memory, a machine-learning technique known for its effectiveness in time-series prediction, to mitigate the detrimental impact of scattering in quantum systems. Our setup involves generating two-mode squeezed light via four-wave mixing in warm rubidium vapor, with one mode subjected to a scatterer to disrupt quantum correlations. Mutual information and intensity-difference squeezing between the two modes are used as metrics for quantum correlations. We demonstrate a 74.7% recovery of mutual information and an 87.7% recovery of two-mode squeezing, despite significant photon loss that would otherwise eliminate quantum correlations. This approach marks a significant step toward recovering quantum correlations from random disruptions without the need for hardware modifications, paving the way for practical applications of quantum protocols.
Strong quantum correlated sources are essential but delicate resources for quantum information science and engineering protocols. Decoherence and loss are the two main disruptive processes that lead to the loss of nonclassical behavior in quantum correlations. In quantum systems, scattering can contribute to both decoherence and loss. In this work, we present an experimental scheme capable of significantly mitigating the adverse impact of scattering in quantum systems. Our quantum system is composed of a two-mode squeezed light generated with the four-wave-mixing process in hot rubidium vapor and a scatterer is introduced to one of the two modes. An integrating sphere is then placed after the scatterer to recollect the scattered photons. We use mutual information between the two modes as the measure of quantum correlations and demonstrate a 47.5% mutual information recovery from scattering, despite an enormous photon loss of greater than 85%. Our scheme is the very first step toward recovering quantum correlations from disruptive random processes and thus has the potential to bridge the gap between proof-of-principle demonstrations and practical real-world implementations of quantum protocols. Published by the American Physical Society 2024
[eng] Quantum photonic technologies have achieved considerable success across a range of disciplines, including quantum computing, quantum communications, and quantum metrology. Conversely, in the current era of information, the ability to discern patterns within voluminous data sets is highly sought after, with Machine Learning (ML) having emerged as a powerful solution to address this challenge. The processing speed of photonic hardware, in conjunction with the potential for computation enhancement resulting from quantum properties, positions quantum photonics as a promising platform for the implementation of ML algorithms. In this thesis, we focus on reservoir computing (RC) as an ML framework that is suitable for time series processing. It provides fast and energy-efficient training in comparison to more conventional methods and is also appropriate for physical implementations. In this thesis, we explore the possibilities brought by quantum photonic setups in the field of quantum RC. Specifically, we propose and study different platform designs that are able to perform different temporal benchmark tasks by providing tunable memory and nonlinearity. Furthermore, we leverage the capabilities of state-of-the-art quantum photonic hardware and identify novel strategies to address key challenges in RC in quantum substrates, as well as expand the potential of quantum machine learning. Firstly, an approach is provided to overcome some of the limitations of quantum projective measurements, thereby enabling information processing in real time. This is achieved by generating a physical ensemble of quantum reservoirs that recirculate through an optical fiber inside a feedback loop. We investigate the detrimental effects of sampling noise due to finite ensembles and provide a strategy to mitigate them. Secondly, the consequences of increasing quantum squeezing in the dynamics of such setups were studied, and it was found that it monotonically improves the noise robustness of the platforms in realistic scenarios. Finally, we propose a novel design in which photonic cluster states, a specific kind of highly entangled quantum network, are used as resources to implement quantum RC. In this configuration, external input information is encoded into the quantum network via quantum teleportation by engineering the measurement basis of the protocol. Furthermore, the platform has the capacity to implement global quantum gates through local operations (single-mode measurements), thus establishing the foundation for distributed quantum ML.
We demonstrate how machine learning is able to model experiments in quantum physics. Quantum entanglement is a cornerstone for upcoming quantum technologies, such as quantum computation and quantum cryptography. Of particular interest are complex quantum states with more than two particles and a large number of entangled quantum levels. Given such a multiparticle high-dimensional quantum state, it is usually impossible to reconstruct an experimental setup that produces it. To search for interesting experiments, one thus has to randomly create millions of setups on a computer and calculate the respective output states. In this work, we show that machine learning models can provide significant improvement over random search. We demonstrate that a long short-term memory (LSTM) neural network can successfully learn to model quantum experiments by correctly predicting output state characteristics for given setups without the necessity of computing the states themselves. This approach not only allows for faster search, but is also an essential step towards the automated design of multiparticle high-dimensional quantum experiments using generative machine learning models.
Quantum correlations and entanglement shared among multiple quantum modes are important for both fundamental science and the future development of quantum technologies. This development will also require an efficient quantum interface between multimode quantum light sources and atomic ensembles, which makes it necessary to implement multimode quantum light sources that match the atomic transitions. Here, we report on such a source that provides a method for generating quantum correlated beams that can be extended to a large number of modes by using multiple four-wave mixing (FWM) processes in hot rubidium vapor. Experimentally, we show that two cascaded FWM processes produce strong quantum correlations between three bright beams but not between any two of them. In addition, the intensity-difference squeezing is enhanced with the cascaded system to -7.0±0.1 dB from the -5.5±0.1/-4.5±0.1 dB squeezing obtained with only one FWM process. One of the main advantages of our system is that as the number of quantum modes increases, so does the total degree of quantum correlations. The proposed method is also immune to phase instabilities due to its phase insensitive nature, can easily be extended to multiple modes, and has potential applications in the production of multiple quantum correlated images.
The generation of entanglement in atomic systems plays a central topic in the fields of quantum information storage and processing. Moreover, a special category of entangled states of multi-atom ensembles, spin squeezed states, have been proven to lead to considerable improvement in the sensitivity of precision measurements compared to systems involving uncorrelated atoms. A treatment of entanglement in open systems is, however, incomplete without a precise description of the process of decoherence which necessarily accompanies it. The theory of entanglement and decoherence are the two main topics of this thesis. Methods are described for the generation of strong correlations in large atomic ensembles using either cavity quantum electrodynamics or measurement outcome conditioned quantum dynamics. Moreover, the description of loss of entanglement resulting from the coupling to a noise reservoir (electromagnetic vacuum) is explored. A spin squeezing parameter is used throughout this thesis as both a measure of entanglement strength and as an indication of the sensitivity improvement above the so-called standard quantum limit (sensitivity obtained with uncorrelated particles) in metrology. The first scheme considered consists of a single mode cavity field interacting with a collection of atoms for which spin squeezing is produced in both resonant and off-resonant regimes. In the resonant case, transfer of squeezing from a field state to the atoms is analyzed, while in the off-resonant regime squeezing is produced via an effective nonlinear interaction (one-axis twisting Hamiltonian). A second, more experimentally realistic case, is one involving the interaction of free space atoms with laser pulses; a projective measurement of a source field originating from atomic fluctuations provides a means of preparing atomic collective states such as spin squeezed and Schrodinger cat states. A new unravelling is proposed, that employs the detection of photon number in a single output channel and is capable of producing mesoscopic Schrodinger cat states in a single step. As a first step to understanding the role of cooperative decoherence, of importance in the case of dense pencil-shaped atomic ensembles, the collective spin decoherence of a two multilevel atom system is derived. This calculation is also relevant to entanglement loss for two qubits manipulated using with reading/writing pulses. Finally, a scheme is proposed in which arbitrarily strong entanglement can be produced in a four-wave mixing setup, where the preparation of the atoms in a dark state limits the decoherence to negligible values.
Attenuation and amplification are the most common processes for optical communications. Amplification can be used to compensate the attenuation of the complex amplitude of an optical field, but is unable to recover the coherence lost, provided that the attenuation channel and the amplification channel are independent. In this paper, we show that the quantum coherence of an optical field can be regained if the attenuation channel and the amplification channel share correlated noise. We propose an all-optical correlated noisy channel relying on the four-wave mixing process and demonstrate its capability of recovering quantum coherence within continuous-variable systems. We quantitatively investigate the coherence recovery phenomena for coherent states and two-mode squeezed states. Moreover, we analyze the effect of other photon losses that are independent from the recovery channel on the performance of recovering coherence. Different from correlated noisy channels previously proposed based on electro-optic conversions, the correlated noisy channel in our protocol is all-optical and thus owns larger operational bandwidths.