Seminars
Information
Valid until the end of the 2022/23 academic year
Hour: 14h00 (14pm WET)
Room: 0.05 (Department of Informatics)
Date: 23 June 2023
Speaker: André Sequeira
Title: Barren Plateaus in Quantum Policy Gradients [slides]
Abstract: Over the past few years, the use of Variational Quantum Algorithms (VQAs) in the context of Reinforcement Learning (RL) has been proposed by multiple studies. The effectiveness of VQAs has been empirically substantiated across numerous benchmark environments in both value-based [1-3] and policy-based [4,5] RL frameworks. Notably, Jerbi et.al [4] outlined a quadratic separation in the optimization process of a policy-based RL agent when employing Parametrized Quantum Circuits (PQC)-based policies instead of an entirely classical agent. In a separate study [6], the authors introduced quantum neural network designs with orthogonal and compound layers for policy and value functions for the problem of hedging that are devoid of Barren Plateaus (BPs). Nevertheless, a more comprehensive understanding of the impact of the BP phenomenon within RL is necessary, particularly in arbitrary PQC-based policies, as it might hinder any acceleration in the optimization process. Consequently, this study focuses on scrutinizing cost-function dependent BPs within the REINFORCE policy-based RL algorithm proposed by Williams [7], considering two families of PQC-based policies: Born and softmax policies.
References: [1] - Chen et.al. Variational Quantum Circuits for Deep Reinforcement Learning - https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9144562
[2] - Skolik et.al. Robustness of quantum reinforcement learning under hardware errors - https://epjquantumtechnology.springeropen.com/articles/10.1140/epjqt/s40507-023-00166-1
[3] - Skolik et.al. Quantum agents in the Gym: a variational quantum algorithm for deep Q-learning - https://quantum-journal.org/papers/q-2022-05-24-720/
[4] - Jerbi et.al. Parametrized Quantum Policies for Reinforcement Learning - https://proceedings.neurips.cc/paper/2021/file/eec96a7f788e88184c0e713456026f3f-Paper.pdf
[5] -Sequeira et.al. Policy gradients using variational quantum circuits - https://link.springer.com/article/10.1007/s42484-023-00101-8
[6] - Cherrat et.al. Quantum Deep Hedging - https://arxiv.org/abs/2303.16585
[7] - Williams Simple statistical gradient-following algorithms for connectionist reinforcement learning - https://link.springer.com/article/10.1007/BF00992696
Date: 26 May 2023
Speaker: Tomás Carneiro
Title: iQbricks: Integration of a fully-featured quantum language in the framework Qbricks
Abstract: Quantum Computing has noticeably grown over the last two decades, making it a revolutionary field of investigation in the current era of technological research. Such a growth has been leading to an increasing demand in research by several big enterprises such as IBM, Google and Microsoft, paving the way for a richer ecosystem and untold benefits among the Quantum Computing community. Verification is a crucial aspect of software development, as it ensures that a program performs as intended and reduces the risk of introducing errors. This is especially important in the field of Quantum Computing, where the complexity of programs is high and the behavior of quantum systems is often counterintuitive. Verification of quantum programs can help detect errors that may lead to incorrect results, which is of utmost importance when dealing with quantum algorithms and quantum simulations. As a result, having a formal verification framework for quantum programs can greatly benefit the development of reliable and accurate quantum software. Qbricks is a verification framework for building quantum programs, and corresponds to the framework on which this project has been integrated. During the course of this thesis, iQbricks – an intuitive and user-friendly language to build and formally verify quantum programs – was developed, along with a framework to translate and generate verifiable Qbricks programs from iQbricks. This project’s main achievements were: (1) the design and implementation of a high-level programming language for describing quantum circuits in an intuitive and user-friendly way and (2) the implementation of a translator, embedded in Qbricks’ framework, that converts iQbricks programs to Qbricks ones. The developed framework was evaluated against two different quantum algorithms: the Quantum Fourier Transform and Grover’s algorithm. This project was accompanied by an internship at the Commissariat à l’énergie atomique et aux énergies alternatives (CEA) - LSL, where this implementation was developed in direct involvement with Qbricks’ team of investigators.
Date: 19 May 2023
Speaker: Mario Silva (Université de Lorraine) [slides]
Title: A Programming Language Characterizing Quantum Polynomial Time
Abstract: We introduce a first-order quantum programming language, named FOQ, whose terminating programs are reversible. We restrict FOQ to a strict and tractable subset, named PFOQ, of terminating programs with bounded width, that provides a first programming language-based characterization of the quantum complexity class FBQP. We finally present a tractable semantics-preserving algorithm compiling a PFOQ program to a quantum circuit of size polynomial in the number of input qubits.
Date: 28 April 2023
Speaker: Alexandra Alves [slides]
Title: Noise Resilient Quantum Amplitude Estimation
Abstract: Quantum amplitude estimation (QAE) is a key routine offering a quadratic quantum speed-up. Notable applications include numerical integration, machine learning, estimating expectation values of observables, and database searching. However, owing to experimental issues, present-day quantum devices struggle to realize this potential. Algorithm design plays a pivotal role in overcoming this issue. In this talk, I will present and discuss several QAE algorithms, with an emphasis on their susceptibility to noise.
Date: 14 April 2023
Speaker: Rodrigo Coelho [slides]
Title: The effect of Data Re-Uploading in Variational Q-Learning
Abstract: A Reinforcement Learning (RL) agent learns by interacting with the surrounding environment in a feedback loop, with only a reward signal guiding its actions [1]. However, state-of-the-art RL algorithms are extremely data-hungry and often not efficiently learnable. Variational Quantum Circuits (VQCs) allow for implementations that require minimal quantum resources, making them suitable for near-term applications [2]. Recently, research has shown that VQCs can be used as function approximators in RL algorithms, reducing the total number of trainable parameters of parameterized models for some environments [3] [4]. Furthermore, by repeating simple data encoding gates multiple times, a technique named data re-uploading, these quantum models can access increasingly rich frequency spectra [5]. This project’s primary goal is to study the effect of data re-uploading in VQC’s applied to Q-Learning [6], a RL algorithm.
References:
[1] R. S. Sutton and A. G. Barto, Reinforcement learning: An introduction. MIT press, 2018.
[2] M. Cerezo, A. Arrasmith, R. Babbush, S. C. Benjamin, S. Endo, K. Fujii, J. R. McClean, K. Mitarai, X. Yuan, L. Cincio et al., “Variational quantum algorithms,” Nature Reviews Physics, vol. 3, no. 9, pp. 625–644, 2021.
[3] A. Skolik, S. Jerbi, and V. Dunjko, “Quantum agents in the gym: a variational quantum algorithm for deep q-learning,” Quantum, vol. 6, p. 720, 2022.
[4] S. Y.-C. Chen, C.-H. H. Yang, J. Qi, P.-Y. Chen, X. Ma, and H.-S. Goan, “Variational quantum circuits for deep reinforcement learning,” IEEE Access, vol. 8, pp. 141 007–141 024, 2020.
[5] M. Schuld, R. Sweke, and J. J. Meyer, “Effect of data encoding on the expressive power of variational quantum-machine-learning models,” Physical Review A, vol. 103, no. 3, p. 032430, 2021.
[6] C. J. Watkins and P. Dayan, “Q-learning,” Machine learning, vol. 8, pp. 279–292, 1992.
Date: 14 April 2023
Speaker: Ioannis Kolotouros (University of Edinburgh) [slides]
Title: Adiabatic quantum computing with parameterized quantum circuits
Abstract: Adiabatic quantum computing is a universal model for quantum computing. Standard error correction methods require overhead that makes their application prohibitive for near-term devices. To mitigate the limitations of near-term devices, a number of hybrid approaches have been pursued in which a parameterized quantum circuit prepares and measures quantum states and a classical optimization algorithm minimizes an objective function that encompasses the solution to the problem of interest. In this work, we propose a different approach starting by analyzing how a small perturbation of a Hamiltonian affects the parameters that minimize the energy within a family of parameterized quantum states. We derive a set of equations that allow us to compute the new minimum by solving a constrained linear system of equations obtained by measuring a series of observables on the unperturbed system. We then propose a discrete version of adiabatic quantum computing which can be implemented with NISQ devices while at the same time is insensitive to the initialization of the parameters and to other limitations hindered in the optimization part of variational quantum algorithms. We also derive a lower bound on the number of discrete steps needed to guarantee success. We compare our proposed algorithm with the Variational Quantum Eigensolver on two classical optimization problems, namely MaxCut and Number Partitioning, and on a quantum spin-configuration problem, the Transverse-Field Ising Chain model, and confirm that our approach demonstrates superior performance.
Date: 31 March 2023
Speaker: Leander Reascos [slides]
Title: Quantum simulation of spin systems on quantum computers
Abstract: Quantum computing has shown great promise in simulating quantum systems [1, 2]. Therefore, this project focuses on simulating and studying the properties of spin systems, specifically their chirality. These systems have multiple applications in various fields, including skyrmions, spintronics, and high-temperature superconductors [3]. The aim is to measure scalar spin chirality non-destructively using various algorithms based on different techniques, such as the Hadamard test, Linear combination of unitaries (LCU) [4], and quantum phase estimation. These algorithms use indirect measurement to avoid disrupting the state and measure expectation values or eigenvalues [5].
References:
[1] S. Lloyd, “Universal quantum simulators,” vol. 273, no. 5278, pp. 1073–1078, 1996. [Online]. Available: https://www.science.org/doi/10.1126/science.273.5278.1073
[2] A. J. Daley, I. Bloch, C. Kokail, S. Flannigan, N. Pearson, M. Troyer, and P. Zoller, “Practical quantum advantage in quantum simulation,” vol. 607, no. 7920, pp. 667–676, 2022, number: 7920 Publisher: Nature Publishing Group. [Online]. Available: https://www.nature.com/articles/s41586-022- 04940-6
[3] X. G. Wen, F. Wilczek, and A. Zee, “Chiral spin states and superconductivity,” vol. 39, no. 16, pp. 11 413–11 423, 1989. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevB.39.11413
[4] A. M. Childs and N. Wiebe, “Hamiltonian simulation using linear combina- tions of unitary operations,” arXiv preprint arXiv:1202.5822, 2012.
[5] K. Mitarai and K. Fujii, “Methodology for replacing indirect measurements with direct measurements,” Physical Review Research, vol. 1, no. 1, p. 013006, 2019.
Date: 24 March 2023
Speaker: André Sequeira
Title: Quantum Reinforcement Learning
Abstract: In the context of Quantum Reinforcement Learning (QRL), provably efficient exploration [1] and a quadratic separation in gradient estimation were recently established for QRL models with oracle access to environment dynamics [2]. However, once oracle access is removed, even though Parameterized Quantum Circuits (PQCs) are being used as versatile QRL models, only empirical advantage relative to a contrived set of classical models was established thus far [3,4]. This work introduces quantum policy gradients using PQCs and presents the quantum-enhanced natural policy gradient algorithm. Numerical results are reported, showcasing an improved performance in a series of classical control benchmarking environments. In addition, regret improvements for natural policy gradients [5] are discussed using the quantum fisher information [6].
References:
[1] - Provably Efficient Exploration in Quantum Reinforcement Learning with Logarithmic Worst-Case Regret - H.Zhong et.al 2023
[2] - Quantum policy gradient algorithms - S.Jerbi et.al 2022
[3] - Policy gradients using variational quantum circuits - A.Sequeira et.al 2022
[4] - Parameterized quantum policies for reinforcement learning - S.Jerbi et.al
[5] - On the theory of policy gradients - optimality, approximation and distribution shift - A.Agarwal et.al 2020
[6] - Fisher information in noisy intermediate-scale quantum applications - J.Meyer 2021
Date: 10 March 2023
Speaker: José Diogo Guimarães [slides]
Title: Noise-assisted digital quantum simulation of open systems
Abstract: Quantum systems are inherently open and subject to environmental noise, which can have both detrimental and beneficial effects on their dynamics. In particular, noise has been observed to enable novel functionalities in bio-molecular systems, making the simulation of their dynamics an important target for digital and analog quantum simulation. However, current quantum devices are typically noisy, limiting their computational capabilities. In this work, we propose a novel approach that leverages the intrinsic noise of a quantum device to reduce the quantum computational resources required for simulating open quantum systems. We achieve this by combining quantum noise characterization methods with quantum error mitigation techniques, which allow us to transform and control the intrinsic noise in a quantum circuit. Specifically, we selectively enhance or reduce decoherence rates in the quantum circuit to achieve the desired simulation of open system dynamics. We describe our methods in detail and report on the results of noise characterization and quantum error mitigation on real and emulated IBM Quantum computers. We also provide estimates of the experimental resource requirements for our techniques. We believe that this approach can pave the way for new simulation techniques in Noisy Intermediate-Scale Quantum (NISQ) devices, where their intrinsic noise can be harnessed to assist quantum computations.
Date: 3 March 2023
Speakers: Inês Dias and Vitor Fernandes [slides]
Title: Towards quantum concurrency (Part 1)
Abstract: In the classical setting, concurrency speeds up the execution of programs, which decreases their execution time [1]. Currently, quantum computers must handle fault-tolerant hardware [2]. This faulty-hardware is prone to noise, which affects the superposition time of qubits (the superposition time of qubits is in the order of a few milliseconds [3]).
We then aim to bring the benefits of concurrency to the quantum setting. In order to study to what extent concurrency can help to mitigate noise in quantum programming, our goal is to define and implement the semantics of a concurrent quantum language, which will in turn allow to simulate quantum concurrent programs. To do that, we based our work on a concurrent imperative language developed by Brookes [4]. In fact, our concurrent quantum language is an extension of said language, by adding basic quantum features to it.
In this talk, we focus on the language developed by Brookes. We present its syntax and operational semantics. Additionally, we discuss the use of Parsec [5], a package of Haskell, in order to build a parser for this language, and also the implementation of its operational semantics in Haskell. Furthermore, we briefly discuss the syntax and the operational semantics of the quantum concurrent language.
References:
[1] Cantrill, Bryan, and Jeff Bonwick. “Real-world concurrency.” Communications of the ACM 51.11 (2008): 34-39.
[2] Nielsen, Michael A., and Isaac Chuang. “Quantum computation and quantum information.” (2002): 558-559.
[3] Zhang, Yu, et al. “Optimizing quantum programs against decoherence: delaying qubits into quantum superposition.” 2019 International Symposium on Theoretical Aspects of Software Engineering (TASE). IEEE, 2019.
[4] Stephen Brookes. Full abstraction for a shared-variable parallel language. Information and Computation, 127(2):145–163, 1996. URL: https://www.sciencedirect.com/science/ article/pii/S0890540196900565, doi:https://doi.org/10.1006/inco.1996.0056.
[5] Daan Leijen, Paolo Martini, and Antoine Latter. parsec: Monadic parser combinators, 2023. [link]
Date: 24 February 2023
Speaker: Henrique Faria [slides]
Title: Multi-Party Computations: Providing a Post-Quantum Solution (Part 1)
Abstract: In recent years, the development of quantum computers led the world in a search for post-quantum resistant algorithms. During the last round of the National Institute of Standards and Technology’s Post-Quantum Cryptography contest (NIST-PQC)[1] it was concluded that the proposed quantum-resistant digital signatures lacked variability, thus requiring further candidates. During the NIST-PQC’s third round, NIST withdrew PICNIC, a Multi-Party Computation (MPC) candidate, from the schemes to be standardized due to concerns with its security. However, NIST stated that upon further research and development, MPC schemes could provide promising candidates for standardization. The fact is that during the NIST-PQC’s contest duration, several MPC schemes that improved upon PICNIC’s performance and security appeared [2,3]. A big problem with their implementations is that these often detach themselves from the original scheme leading to doubts about their security. To solve this issue, the toolchain Jasmin-EasyCrypt[4,5] emerged allowing one to link optimized implementations to the original scheme’s security claims. In this talk, we will detail what an MPC scheme is and how the Jasmin-EasyCrypt toolchain can provide high-speed high-assurance software implementations with proofs over these concrete implementations instead of proofs over a theoretical scheme.
References:
[1] Status Report on the Third Round of the NIST Post-Quantum Cryptography Standardization Process. Gorjan Alagic, Daniel Apon, David Cooper, Quynh Dang, Thinh Dang, John Kelsey, Jacob Lichtinger, Yi-Kai Liu, Carl Miller, Dustin Moody, Rene Peralta, Ray Perlner, Angela Robinson, Daniel Smith-Tone.
[2] Banquet: Short and Fast Signatures from AES. Carsten Baum, Cyprien Delpech de Saint Guilhem, Daniel Kales, Emmanuela Orsini, Peter Scholl and Greg Zaverucha. In: Cryptology ePrint Archive, Paper 2021/068.
[3] LegRoast: Efficient post-quantum signatures from the Legendre PRF. Ward Beullens, Cyprien Delpech de Saint Guilhem. In: Cryptology ePrint Archive, Paper 2020/128.
[4] Jasmin: High-Assurance and High-Speed Cryptography. José Bacelar Almeida, Manuel Barbosa, Gilles Barthe, Arthur Blot, Benjamin Grégoire, Vincent Laporte, Tiago Oliveira, Hugo Pacheco, Benedikt Schmidt, Pierre-Yves Strub. In: CCS’17.
[5] [https://github.com/EasyCrypt/easycrypt].
Date: 17 February 2023
Speaker: Andrea D’Urbano [slides]
Title: Resource theory of local operations and shared randomness - a primer
Abstract: In [1], Schmidt et al. presented a method to analyse and formalise scenarios involving multiple parties sharing a common cause but no cause-effect relationship. This scenario could prove to be useful to quantify the advantage, if any, in using quantum resources in distributed quantum information processing, and in particular secure multi-party computation. The resource-theory discussion in [1] and the general framework presented in [2] and [3], allow to describe the nonclassicality in such scenarios, taking into account the causal structure of the experiment. We will describe this approach and try to apply the methodologies presented in [3] to the easiest and the most fundamental example of quantum multi-party computation: quantum secret sharing [4].
References:
[1] Schmid, David, Denis Rosset, and Francesco Buscemi. “The type-independent resource theory of local operations and shared randomness.” Quantum 4 (2020): 262.
[2] Wolfe, Elie, et al. “Quantifying Bell: The resource theory of nonclassicality of common-cause boxes.” Quantum 4 (2020): 280.
[3] Zjawin, Beata, et al. “Quantifying EPR: the resource theory of nonclassicality of common-cause assemblages.” arXiv preprint arXiv:2111.10244 (2021).
[4] Hillery, Mark, Vladimír Bužek, and André Berthiaume. “Quantum secret sharing.” Physical Review A 59.3 (1999): 1829.