KIAS Quantum Information Workshop 2023 (July 3-5, Korea Institute for Advanced Study, Room 8101) |
| Program | Home > Program |
Time table
.png)
3 Monday
Invited talk 1. Seongwook Shin (Seoul National University):
Analyzing quantum machine learning with tensor networks
Quantum machine learning (QML), which employs a quantum circuit as a computational model for machine learning, is considered one of the most promising applications for near-term quantum devices. However, fairly comparing the QML model with the classical ML model is challenging due to their inherently different structures. In this study, we present a consolidated approach to comparing classical and quantum machine learning models within the unified framework of tensor networks. By representing QML models as tensor network (TN) models, we identify the model as a featured linear model (FLM) with a constrained coefficient where the feature map is given by the efficiently manageable tensor products. This allows us to create the same feature map classically and efficiently using only the same amount of pre-processing as QML. Using this feature map we create the classical TN ML model residing within the identical function space, spanned by the same basis functions as those of QML models. By representing the coefficient components of the models using matrix product states (MPS), we compare the function classes of the two models and determine the conditions for efficient approximation of QML models by classical models. Finally, we compare the two models in the context of function regression based on kernel methods and variational optimization, highlighting the distinct characteristics between them.
Lecture 1. Issac Kim (UC Davis):
Entanglement Bootstrap Program
Entanglement bootstrap program is an approach to study many-body topological invariants of gapped ground states from a basic principle rooted in quantum information theory. In this lecture series, I will explain the philosophy that underlies the entanglement bootstrap program and discuss some notable recent developments.
4 Tuesday
Invited talk 2. Hyukgun Kwon (Seoul National University):
Efficacy of virtual purification-based error mitigation on quantum metrology
Recently there has been research applying Virtual-Purification (VPEM) to quantum metrology to reduce bias caused by unknown noise. However, while they presented a framework of VPEM for quantum metrology and provided examples in which VPEM effectively reduces a bias, the general applicability of VPEM to quantum metrology has not been fully understood.
In this work, we study when VPEM can effectively reduce the bias caused by an unknown noise. In particular, we identify two crucial factors that determine whether VPEM can reduce bias. The first one is how close the dominant eigenvector of a noisy state and the (noiseless) ideal state are. More specifically, the closeness decides the achievable amount of reduced bias. The second one is the reference point of a parameter that we want to estimate. Assuming that the unknown parameter to be estimated is small (often called local parameter estimation), the reference point is the one around which the unknown parameter varies. In local parameter estimation, we show that one has to carefully choose the reference point that gives the smallest bias before applying VPEM to quantum parameter estimation. Otherwise, even if the expectation values of an observable over the ideal quantum probe and the dominant eigenvector are the same, the bias of the mitigated case could be larger than the non-mitigated case. We emphasize that a strategy of choosing a reference point is a unique feature of quantum metrology that has not been considered in the previous studies.
Detailed discussions can be found in arXiv: 2303.15838, (2023)
Invited talk 3. Donghoon Kim (KAIST):
Entanglement and Screening Cloud in Exotic Quantum Impurity Systems
The quantum coherent screening of a local spin in metals is a long-standing issue in quantum many-body problems. In this work, we analyze the screening cloud, a spatial structure of electrons screening a local impurity spin-1/2, in multichannel Kondo effects, pseudogap Kondo effects and two-impurity Kondo effects. We characterize the screening cloud by its quantum entanglement with the impurity spin and compute it by boundary conformal field theory and tensor network methods. We show that distinct (non-)Fermi liquids coexist in the cloud, forming concentric shells centered at the impurity. Outer shells are suppressed one by one as temperature increases, and the remaining outermost shell determines the thermal non-Fermi liquid phase of the metal at that temperature. We propose how to observe the entanglement and the screening cloud structure, based on a recent experimental setup.
Lecture 2. ‪Myung-Joong Hwang (Duke Kunshan University):
Quantum Light and Matter
Coherent interaction between quantum light and matter is a fundamental physical process that underlies most of quantum technological platforms. In the first part of this lecture, I will introduce quantum theory of light-matter interaction and show how they are utilized as operational principles for quantum information processing. In the second part, I will introduce exciting opportunities to synthesize exotic phases of interacting quantum light and matter using quantum devices and theoretical methods to understand such novel quantum phases realized in intrinsically driven open quantum systems.
5 Wednesday
Invited talk 4. Mijung So (Korea University): Generic Decoherence Free subspace of Non-Interacting Open Quantum System
We consider a non-interacting open quantum system that has permutation symmetry, and found the decoherence free subspace that the relative capacity asymptotically approaches one.
Lecture 3. Changhun Oh (University of Chicago):
Quantum computational advantage using noisy intermediate-scale devices
Over the last few years, we have seen the first plausible quantum computational advantage experiments using random circuit sampling with qubits and Gaussian boson sampling. In this lecture series, I will talk about the theoretical background of such sampling problems and the experimental progress so far. I will also discuss important open problems in this topic.