arXiv subjects

Li-Yi Hsu

Publications and source records attributed to Li-Yi Hsu.

At least 19 recordsLinked to original sources

arXiv2026

Affine-Profile Stabilizer Thresholds for Magic in Codeword-Stabilized Quantum Codes

Codeword-stabilized quantum codes give a unified graph-state description of stabilizer and nonadditive quantum error-correcting codes. Although each individual CWS word state is stabilizer, coherent superpositions of different word states can be nonstabilizer. We develop a CWS-adapted magic-witness framework that isolates this codeword coherence and converts it into certified lower bounds on robustness of magic. The main result is an exact reduction of the stabilizer threshold of a natural CWS coherence witness to a finite-geometric problem over the classical CWS word set. For general weighted superpositions, the threshold is computed by enumerating affine intersections and affine-quadratic phases. For equal-weight superpositions, the phase optimization collapses, and the threshold is determined entirely by how many CWS words can lie in affine flats of each dimension. Thus a quantum optimization over stabilizer states becomes a classical affine-incidence problem. This reduction yields a fixed-parameter algorithm, an analytic lower bound for an infinite union-stabilizer family, and exact rational certificates for several standard nonadditive CWS examples. The framework also clarifies why exact enumeration fails for large structured families and identifies the remaining task as an affine-intersection problem. The result provides a geometric mechanism by which nonlinear CWS word sets generate certifiable magic.

quant-ph↗

arXiv2026

Quantum Transfer Learning Shows Improved Robustness in Low-Data Regimes

Transfer learning under limited data is a challenging setting, where models must adapt to new tasks with minimal supervision. Prior work has primarily focused on improving absolute accuracy in transfer learning. However, empirical evidence comparing quantum and classical models in realistic transfer learning settings remains limited, especially in low-data regimes. In this work, we systematically study the robustness of quantum models under reduced training data. We evaluate multiple quantum and classical architectures across diverse transfer tasks and retraining configurations, and quantify robustness using accuracy degradation and relative performance retention (RPR). Our results show that, although classical models often achieve higher peak performance, they exhibit significantly larger degradation when training data is limited. In contrast, quantum models maintain more stable performance across data regimes, indicating improved robustness and data efficiency. These findings provide empirical evidence that quantum models can offer improved robustness in low-resource transfer learning scenarios.

quant-ph↗

arXiv2025

Unsupervised Feature Extraction and Reconstruction Using Parameterized Quantum Circuits

Autoencoders are fundamental tools in classical computing for unsupervised feature extraction, dimensionality reduction, and generative learning. The Quantum Autoencoder (QAE), introduced by Romero J.[2017 Quantum Sci. Technol. 2 045001], extends this concept to quantum systems and has been primarily applied to tasks like anomaly detection. Despite its potential, QAE has not been extensively evaluated in basic classification tasks such as handwritten digit classification, which could provide deeper insights into its capabilities. In this work, we systematically investigate the performance of QAE in unsupervised feature extraction and reconstruction tasks. Using various encoder and decoder architectures, we explore QAE's ability to classify MNIST and analyze its reconstruction effectiveness. Notably, with a QCNN-based encoder, QAE achieves an average accuracy of 97.59% in binary classification of MNIST (0 and 1), demonstrating the feasibility of QAE for such tasks. This result highlights the synergy between QAE and QCNN in achieving optimal feature extraction performance. However, we also identify limitations in QAE's reconstruction capability, underscoring the need for further advancements in encoder and decoder designs. Our findings provide a foundation for future research on leveraging QAE in practical quantum machine learning applications.

quant-ph↗

arXiv2025

Correlation strengths in hybrid networks

In a generic hybrid network, classical, quantum, and no-signaling sources emit local hidden variables, stabilizer states, and no-signaling systems, respectively. We investigate the maximal correlation strength as the non-classical feature in this network. Given the associated fully-quantum network of a hybrid network, we exploit the stabilizing operators of the distributed quantum state to construct segmented Bell operators and the Bell inequalities tailored to the state. We derive the upper bounds of the maximal correlation strengths in the associated full-classical, full-quantum, and fully-no-signaling networks as the benchmarks. Our study shows that the achievable correlation strength depends on the number of type-A measurements and that of nonlocal sources. We also introduce the t-nonlocality criteria, indicating that the achievable maximal correlation strength cannot modeled by the network with at least t observers with local hidden variables performing type-A measurements.

quant-ph↗

arXiv2024

Self-Testing Quantum Error Correcting Codes: Analyzing Computational Hardness

We present a generalization of the tilted Bell inequality for quantum [[n,k,d]] error-correcting codes and explicitly utilize the simplest perfect code, the [[5,1,3]] code, the Steane [[7,1,3]] code, and Shor's [[9,1,3]] code, to demonstrate the self-testing property of their respective codespaces. Additionally, we establish a framework for the proof of self-testing, as detailed in \cite{baccari2020device}, which can be generalized to the codespace of CSS stabilizers. Our method provides a self-testing scheme for $\cos\theta \lvert \bar{0} \rangle + \sin\theta \lvert \bar{1} \rangle$, where $\theta \in [0, \frac{\pi}{2}]$, and also discusses its experimental application. We also investigate whether such property can be generalized to qudit and show one no-go theorem. We then define a computational problem called ISSELFTEST and describe how this problem formulation can be interpreted as a statement that maximal violation of a specific Bell-type inequality can self-test a particular entanglement subspace. We also discuss the computational complexity of ISSELFTEST in comparison to other classical complexity challenges and some related open problems.

quant-ph↗

arXiv2023

Statistical link between Bell nonlocality and uncertainty relations

Bell nonlocality and uncertainty relations are distinct features of quantum theory from classical physics. Bell nonlocality concerns the correlation strength among local observables on different quantum particles, whereas the uncertainty relations set the lower bound of the sum or product of the variance square of observables. Here we establish the statistical link between these two quantum characters using the Aharonov-Vaidman identity. Therein, the upper bounds of Bell-type inequalities are expressed in terms of the product of the local sum of the variance square. On the other hand, instead of evaluating local uncertainty relations, the uncertainty relations on two or more quantum systems are upper-bounded by the amount of Bell nonlocality therein.

quant-ph↗

arXiv2022

Genuine Bell locality and nonlocality in the networks

In the literature on $K$-locality ($K\geq2$) networks, the local hidden variables are strictly distributed in the specific observers rather than the whole ones. Regarding genuine Bell locality, all local hidden variables, as classical objects that allow for perfect cloning in classical physics, should be cloned and then spread throughout the networks. More correlators are involved in the proposed linear and non-linear Bell-type inequalities, where their upper bounds are specified by the pre-determined output probability distribution. As for the quantum version, the no-clone theorem limits the broadcast of quantum correlations. To explore genuine Bell nonlocality in variant particle distributions in the networks, the Pauli operators stabilizing the two-qubit Bell states or multi-qubit Greenberger--Horne--Zeilinger states (GHZ states) play an essential role in designing the proposed linear and non-linear Bell tests and assigning the local incompatible measurements for the spatially separated observers. We prove the maximal violations of the proposed Bell-type inequalities quantum networks. In the end, how entanglement swapping replaces the joint measurements in the Bell tests is demonstrated.

quant-ph↗

arXiv2022

Any four orthogonal ququad-ququad maximally entangled states are locally markable

In quantum state discrimination, the observers are given a quantum system and aim to verify its state from the two or more possible target states. In the local quantum state marking as an extension of quantum state discrimination, there are N composite quantum systems and N possible orthogonal target quantum states. Distant Alice and Bob are asked to correctly mark the states of the given quantum systems via local operations and classical communication. Here we investigate the local state marking with N 4 ${\otimes}$ 4 systems, N=4, 5, 6, and 7. Therein, Alice and Bob allow for three local operations: measuring the local observable either ${\sigma}_{z}$ or ${\sigma}_{x}$ simultaneously, and entanglement swapping. It shows that, given arbitrary four 4 ${\otimes}$ 4 systems, Alice and Bob can perform the perfect local quantum state marking. In the N=5, 6 cases, they can perform perfect local state marking with specific target states. We conjecture the impossibility of the local quantum state marking given any seven target states since Alice and Bob cannot fulfill the task in the simplest case.

quant-ph↗

arXiv2021

Exploring Bell nonlocality of quantum networks with stabilizing and logical operators

In practical quantum networks, a variety of multi-qubit stabilized states emitted from independent sources are distributed among the agents, and the correlations across the entire network can be derived from each agent's local measurements on the shared composite quantum systems. To reveal the Bell non-locality in such cases as a quantum feature, minimal knowledge of the emitted stabilizer state is required. Here, we demonstrate that knowing the stabilizing and logical operators indeed provides a new way of exploring Bell non-locality in quantum networks. For the qubit distribution in quantum networks, the associated nonlinear Bell inequalities are derived. On the other hand, to violate these inequalities, one can design local incompatible observables using minimal knowledge of the emitted states. The tilted nonlinear Bell inequalities tailored for specific non-maximal entangled stabilizer states and a way of achieving the maximal violation are also explored.

quant-ph↗

arXiv2020

Carrying an arbitrarily large amount of information using a single quantum particle

Theoretically speaking, a photon can travel arbitrarily long before it enters into a detector, resulting a click. How much information can a photon carry? We study a bipartite asymmetric "two-way signaling" protocol as an extension of that proposed by Del Santo and Daki\'{c}. Suppose that Alice and Bob are distant from each other and each of them has an $n$-bit string. They are tasked to exchange the information of their local n-bit strings with each other, using only a single photon during the communication. It has been shown that the superposition of different spatial locations in a Mach-Zehnder (MZ) interferometer enables bipartite local encodings. We show that, after the travel of a photon through a cascade of $n$-level MZ interferometers in our protocol, the one of Alice or Bob whose detector clicks can access the other's full information of $n$-bit string, while the other can gain one-bit of information. That is, the wave-particle duality makes two-way signaling possible, and a single photon can carry arbitrarily large (but finite) information.

quant-ph↗

arXiv2016

Entropic uncertainty relation based on generalized uncertainty principle

We explore the modification of the entropic formulation of uncertainty principle in quantum mechanics which measures the incompatibility of measurements in terms of Shannon entropy. The deformation in question is the type so called generalized uncertainty principle that is motivated by thought experiments in quantum gravity and string theory and is characterized by a parameter of Planck scale. The corrections are evaluated for small deformation parameters by use of the Gaussian wave function and numerical calculation. As the generalized uncertainty principle has proven to be useful in the study of the quantum nature of black holes, this study would be a step toward introducing an information theory viewpoint to black hole physics.

hep-th↗

arXiv2014

Informatic error-disturbance relation in the qubit case

In 1927, Heisenberg heuristically disclosed the tradeoff between the error in the measurement and the caused disturbance on another complementary observable. In the quantum theory, most of uncertainty relations are proposed to reveal the amount of unavoidable uncertainty in the measuring process. In this paper, we study the error-disturbance relation from the information viewpoint. We ask how much information, rather than how much uncertainty, can be obtained during the two sequential measurements. To achieve optimal information gain, we argue that the strategy for the "intelligent" prior apparatus is to clone the unknown state, and for the posterior one is to perform the swapping operation. We propose the coarse-grained random access code, and therein information causality as a physical principle can be exploited for deriving the upper-bound of information gain. Finally, we conjecture the information gain of measuring the position and momentum of a quantum object in the coarse-grained way.

quant-ph↗

arXiv2011

Monogamy of Information Causality

We consider the information causality in the multi-receiver random access codes. Therein, no receiver can gain any information only from classical communication. We claim the following statement. Information causality still holds even with the help of the multi-partite physical non-local resource. That is, the summation of all revivers' information gain cannot be greater than the amount of classical communication. The distributive multi-party physical nonlocal resource can be exploited only for information splitting. It is proved that such trade-off leads to the monogamy of entanglement. Finally the connection between information causality and spin-glass Bethe lattice is discussed.

quant-ph↗

arXiv2010

Information Causality and Noisy Computations

We reformulate the information causality in a more general framework by adopting the results of signal propagation and computation in a noisy circuit. In our framework, the information causality leads to a broad class of Tsirelson inequalities. This fact allows us to subject information causality to experimental scrutiny. A no-go theorem for reliable nonlocal computation is also derived. Information causality prevents any physical circuit from performing reliable computations.

quant-ph↗

arXiv2010

Information causality and non-locality swapping are equivalent from emergence of quantum correlations

Is information causality a new physical principle? To answer this question, we first analytically derive the criteria of emergence of quantum correlations from information causality. Then it is shown that, as emergence criteria of quantum correlations, information causality and uselessness of coupler-based non-locality swapping can be regarded equivalent. Therefore, incapability of non-locality swapping using a coupler is as powerful as information causality in the single-out of quantum physics from generalized non-signalling models.

quant-ph↗

arXiv2009

High performance entanglement-assisted quantum LDPC codes need little entanglement

Though the entanglement-assisted formalism provides a universal connection between a classical linear code and an entanglement-assisted quantum error-correcting code (EAQECC), the issue of maintaining large amount of pure maximally entangled states in constructing EAQECCs is a practical obstacle to its use. It is also conjectured that the power of entanglement-assisted formalism to convert those good classical codes comes from massive consumption of maximally entangled states. We show that the above conjecture is wrong by providing families of EAQECCs with an entanglement consumption rate that diminishes linearly as a function of the code length. Notably, two families of EAQECCs constructed in the paper require only one copy of maximally entangled state no matter how large the code length is. These families of EAQECCs that are constructed from classical finite geometric LDPC codes perform very well according to our numerical simulations. Our work indicates that EAQECCs are not only theoretically interesting, but also physically implementable. Finally, these high performance entanglement-assisted LDPC codes with low entanglement consumption rates allow one to construct high-performance standard QECCs with very similar parameters.

quant-ph↗

arXiv2009

Optimal Nonadditive Quantum Error-Detecting Code

In this paper, we investigate the optimal nonadditive quantum error-detecting codes with distance two. The the numerical simulation shows that, with n being can be 5, 6, 7, 8, 10 and 12, such the n-qubit quantum error-detecting codes with maximal number of codewords can be found. Therein, except the n=7 case, the n-vertex loop graphs help find the optimal quantum codes.

quant-ph↗

arXiv2008

Undetermined states: how to find them and their applications

We investigate the undetermined sets consisting of two-level, multi-partite pure quantum states, whose reduced density matrices give absolutely no information of their original states. Two approached of finding these quantum states are proposed. One is to establish the relation between codewords of the stabilizer quantum error correction codes (SQECCs) and the undetermined states. The other is to study the local complementation rules of the graph states. As an application, the undetermined states can be exploited in the quantum secret sharing scheme. The security is guaranteed by their undetermineness.

quant-ph↗