arXiv subjects

Yikai Yang

Publications and source records attributed to Yikai Yang.

10 recordsLinked to original sources

arXiv2026

Nuclear quadrupole interaction and zero first-order Zeeman transitions of $^{167}$Er$^{3+}$ in CaWO$_4$

We report microwave spectroscopy of $^{167}$Er$^{3+}$ doped in CaWO$_4$ which reveals the hyperfine splitting of the erbium electronic ground state ($Z_1$, $J_\mathrm{eff.}$=15/2) induced by the $I$=7/2 nuclear spin. From spectra measured below$\sim$50 mK in magnetic fields up to 200 mT, we extract spin Hamiltonian parameters including the electron $\textbf{g}$, hyperfine $\textbf{A}$, and nuclear electric quadrupolar $\textbf{Q}$ tensors. Crucially, our analysis demonstrate unambiguously, that the previously unobserved nuclear electric quadrupolar moment is essential to reproduce the experimental data. With these refined parameters, we identify zero first-order Zeeman (ZEFOZ) transitions at zero magnetic field. Extending the analysis to finite fields, we uncover that ZEFOZ points lie either along the $c$ axis or within the $a$-$b$ plane. These results establish CaWO$_4$ as a promising host for long lifetime quantum memories.

cond-mat.mtrl-sci↗

arXiv2025

Virtual Gates Enabled by Digital Surrogate of Quantum Dot Devices

Advances in quantum technologies are often limited by slow device characterization, complex tuning requirements, and scalability challenges. Spin qubits in electrostatically defined quantum dots provide a promising platform but are not exempt from these limitations. Simulations enhance our understanding of such devices, and in many cases, rapid feedback between measurements and simulations can guide the development of optimal design and control strategies. Here, we introduce a modular, graph-based simulator that acts as a digital surrogate for a semiconductor quantum dot device, where computationally expensive processes are accelerated using deep learning. We demonstrate its potential by estimating crosstalk effects between gate electrodes and applying these estimates to construct virtual gates in a quantum dot device. We validate our approach through comparison with experiments on a double quantum dot defined in a Ge/SiGe heterostructure. We envision that this simulation framework will advance semiconductor-based quantum technologies by enabling more efficient design, characterization, and control of complex devices.

cond-mat.mes-hall↗

arXiv2025

Rethinking the Detail-Preserved Completion of Complex Tubular Structures based on Point Cloud: a Dataset and a Benchmark

Complex tubular structures are essential in medical imaging and computer-assisted diagnosis, where their integrity enhances anatomical visualization and lesion detection. However, existing segmentation algorithms struggle with structural discontinuities, particularly in severe clinical cases such as coronary artery stenosis and vessel occlusions, which leads to undesired discontinuity and compromising downstream diagnostic accuracy. Therefore, it is imperative to reconnect discontinuous structures to ensure their completeness. In this study, we explore the tubular structure completion based on point cloud for the first time and establish a Point Cloud-based Coronary Artery Completion (PC-CAC) dataset, which is derived from real clinical data. This dataset provides a novel benchmark for tubular structure completion. Additionally, we propose TSRNet, a Tubular Structure Reconnection Network that integrates a detail-preservated feature extractor, a multiple dense refinement strategy, and a global-to-local loss function to ensure accurate reconnection while maintaining structural integrity. Comprehensive experiments on our PC-CAC and two additional public datasets (PC-ImageCAS and PC-PTR) demonstrate that our method consistently outperforms state-of-the-art approaches across multiple evaluation metrics, setting a new benchmark for point cloud-based tubular structure reconstruction. Our benchmark is available at https://github.com/YaoleiQi/PCCAC.

cs.CV↗

arXiv2025

Evidence of rotational and tilting disorder of ReO6 octahedra in single crystals of a 5d1 double perovskite Ba2CaReO6

We present results of an experimental study on single crystals of a 5d1 double perovskite Ba2CaReO6. Magnetization measurements reveal a weak splitting between zero-field-cooled and field-cooled protocols below 12 K. At magnetic fields above 1 T the splitting is absent and the magnetic susceptibility is featureless. A detailed specific heat study in a wide temperature range and comprising different heat pulses did not reveal any indication of a thermodynamic phase transition. At low temperatures we do observe specific heat deviating from a phonon background, leading to a total electronic entropy release of ~Rln2. Resonant and non-resonant x-ray diffraction of characteristic Bragg peaks indicates a significant presence of disorder, potentially related to random tilts and rotations of rigid ReO6 octahedra.

cond-mat.str-el↗

arXiv2025

On-Policy Optimization with Group Equivalent Preference for Multi-Programming Language Understanding

Large language models (LLMs) achieve remarkable performance in code generation tasks. However, a significant performance disparity persists between popular programming languages (e.g., Python, C++) and others. To address this capability gap, we leverage the code translation task to train LLMs, thereby facilitating the transfer of coding proficiency across diverse programming languages. Moreover, we introduce OORL for training, a novel reinforcement learning (RL) framework that integrates on-policy and off-policy strategies. Within OORL, on-policy RL is applied during code translation, guided by a rule-based reward signal derived from unit tests. Complementing this coarse-grained rule-based reward, we propose Group Equivalent Preference Optimization (GEPO), a novel preference optimization method. Specifically, GEPO trains the LLM using intermediate representations (IRs) groups. LLMs can be guided to discern IRs equivalent to the source code from inequivalent ones, while also utilizing signals about the mutual equivalence between IRs within the group. This process allows LLMs to capture nuanced aspects of code functionality. By employing OORL for training with code translation tasks, LLMs improve their recognition of code functionality and their understanding of the relationships between code implemented in different languages. Extensive experiments demonstrate that our OORL for LLMs training with code translation tasks achieves significant performance improvements on code benchmarks across multiple programming languages.

cs.CL↗

arXiv2024

Rapid optimal work extraction from a quantum-dot information engine

The conversion of thermal energy into work is usually more efficient in the slow-driving regime, where the power output is vanishingly small. Efficient work extraction for fast driving protocols remains an outstanding challenge at the nanoscale, where fluctuations play a significant role. In this Letter, we use a quantum-dot Szilard engine to extract work from thermal fluctuations with maximum efficiency over two decades of driving speed. We design and implement a family of optimised protocols ranging from the slow- to the fast-driving regime, and measure the engine's efficiency as well as the mean and variance of its power output in each case. These optimised protocols exhibit significant improvements in power and efficiency compared to the naive approach. Our results also show that, when optimising for efficiency, boosting the power output of a Szilard engine inevitably comes at the cost of increased power fluctuations.

quant-ph↗

arXiv2024

Cavity-Magnon-Polariton spectroscopy of strongly hybridized electro-nuclear spin excitations in LiHoF4

We first present a formalism that incorporates the input-output formalism and the linear response theory to employ cavity-magnon-polariton coupling as a spectroscopic tool for investigating strongly hybridized electro-nuclear spin excitations. A microscopic relation between the generalized susceptibility and the scattering parameter |S11| in strongly hybridized cavity-magnon-polariton systems has been derived without resorting to semi-classical approximations. The formalism is then applied to both analyze and simulate a specific systems comprising a model quantum Ising magnet (LiHoF4) and a high-finesse 3D re-entrant cavity resonator. Quantitative information on the electro-nuclear spin states in LiHoF4 is extracted, and the experimental observations across a broad parameter range were numerically reproduced, including an external magnetic field titraversing a quantum critical point. The method potentially opens a new avenue not only for further studies on the quantum phase transition in LiHoF4 but also for a wide range of complex magnetic systems.

cond-mat.str-el↗

arXiv2021

Superconductivity with High Upper Critical Field in the Cubic Centrosymmetric $η$-Carbide Nb$_4$Rh$_2$C$_{1-δ}$

The upper critical field is a fundamental measure of the strength of superconductivity in a material. It is also a cornerstone for the realization of superconducting magnet applications. The critical field arises because of the Copper pair breaking at a limiting field, which is due to the Pauli paramagnetism of the electrons. The maximal possible magnetic field strength for this effect is commonly known as the Pauli paramagnetic limit given as $μ_0 H_{\rm Pauli} \approx 1.86{\rm [T/K]} \cdot T_{\rm c}$ for a weak-coupling BCS superconductor. The violation of this limit is only rarely observed. Exceptions include some low-temperature heavy fermion and some strongly anisotropic superconductors. Here, we report on the superconductivity at 9.75 K in the centrosymmetric, cubic $η$-carbide-type compound Nb$_4$Rh$_2$C$_{1-δ}$, with a normalized specific heat jump of $ΔC/γT_{\rm c} =$ 1.64. We find that this material has a remarkably high upper critical field of $μ_0 H_{\rm c2}{\rm (0)}$ =~28.5~T, which is exceeding by far its weak-coupling BCS Pauli paramagnetic limit of $μ_0 H_{\rm Pauli}$~=~18.1 T. Determination of the origin and consequences of this effect will represent a significant new direction in the study of critical fields in superconductors.

cond-mat.supr-con↗

arXiv2021

Continental generalization of an AI system for clinical seizure recognition

Electroencephalogram (EEG) monitoring and objective seizure identification is an essential clinical investigation for some patients with epilepsy. Accurate annotation is done through a time-consuming process by EEG specialists. Computer-assisted systems for seizure detection currently lack extensive clinical utility due to retrospective, patient-specific, and/or irreproducible studies that result in low sensitivity or high false positives in clinical tests. We aim to significantly reduce the time and resources on data annotation by demonstrating a continental generalization of seizure detection that balances sensitivity and specificity. This is a prospective inference test of artificial intelligence on nearly 14,590 hours of adult EEG data from patients with epilepsy between 2011 and 2019 in a hospital in Sydney, Australia. The inference set includes patients with different types and frequencies of seizures across a wide range of ages and EEG recording hours. We validated our inference model in an AI-assisted mode with a human expert arbiter and a result review panel of expert neurologists and EEG specialists on 66 sessions to demonstrate achievement of the same performance with over an order-of-magnitude reduction in time. Our inference on 1,006 EEG recording sessions on the Australian dataset achieved 76.68% with nearly 56 [0, 115] false alarms per 24 hours on average, against legacy ground-truth annotations by human experts, conducted independently over nine years. Our pilot test of 66 sessions with a human arbiter, and reviewed ground truth by a panel of experts, confirmed an identical human performance of 92.19% with an AI-assisted system, while the time requirements reduce significantly from 90 to 7.62 minutes on average.

physics.med-ph↗

arXiv2020

Epileptic Seizure Forecasting: Probabilistic seizure-risk assessment and data-fusion

Epileptic seizure forecasting, combined with the delivery of preventative therapies, holds the potential to greatly improve the quality of life for epilepsy patients and their caregivers. Forecasting seizures could prevent some potentially catastrophic consequences such as injury and death in addition to a long list of potential clinical benefits it may provide for patient care in hospitals. The challenge of seizure forecasting lies within the seemingly unpredictable transitions of brain dynamics into the ictal state. The main body of computational research on determining seizure risk has been focused solely on prediction algorithms, which involves a remarkable issue of balancing accuracy and false-alarms. In this paper, we developed a seizure-risk warning system that employs Bayesian convolutional neural network (BCNN) to provide meaningful information to the patient and provide a greater opportunity for him/her to be potentially more in charge of his/her health. We use scalp electroencephalogram (EEG) signals and release information on the certainty of our automatic seizure-risk assessment. In the process, we pave the ground-work towards incorporating auxiliary signals to improve our EEG-based seizure-risk assessment system. Our previous CNN results show an average AUC of 74.65% while we could achieve on an EEG-only BCNN an average AUC of 68.70%. This drop in performance is the cost of providing richer information to the patient at this stage of this research.

eess.SP↗