A5019021341- Ferroelectric and Negative Capacitance Devices
- MXene and MAX Phase Materials
- Privacy-Preserving Technologies in Data
- Ferroelectric and Piezoelectric Materials
- Semiconductor materials and devices
- Physical Unclonable Functions (PUFs) and Hardware Security
- Adversarial Robustness in Machine Learning
- Analog and Mixed-Signal Circuit Design
- Integrated Circuits and Semiconductor Failure Analysis
- Advanced Memory and Neural Computing
- Security and Verification in Computing
- Wildlife Ecology and Conservation
- Species Distribution and Climate Change
- Advancements in PLL and VCO Technologies
- Advanced Neural Network Applications
- Catalysis and Oxidation Reactions
- Infrared Target Detection Methodologies
- Ecology and Vegetation Dynamics Studies
- Stochastic Gradient Optimization Techniques
- Model Reduction and Neural Networks
- Heavy metals in environment
- 2D Materials and Applications
- CCD and CMOS Imaging Sensors
- Tensor decomposition and applications
- Acoustic Wave Resonator Technologies
Soochow University
2025
Shandong First Medical University
2025
Shandong Provincial QianFoShan Hospital
2025
Nanjing University
2025
Harbin Medical University
2024
University of Southern California
2020-2024
East China Normal University
2023-2024
Hainan University
2024
Tsinghua University
2019-2023
State Key Joint Laboratory of Environment Simulation and Pollution Control
2019-2023
HfO_{2}-based ferroelectric thin films are promising for their application in devices. Predicting the ultimate magnitude of polarization and understanding its switching mechanism critical to realize optimal performance these Here, a generalized solid-state variable cell nudged elastic band method is employed predict pathway associated with domain-wall motion (Hf,Zr)O_{2} ferroelectrics. It found that reversal pathway, where threefold coordinated O atoms pass across nominal unit-cell...
Threatened and endangered species are more vulnerable to climate change due small population specific geographical distribution. Therefore, identifying incorporating the biological processes underlying a species' adaptation its environment important for determining whether they can persist in situ. Correlative models widely used predict distribution changes, but generally fail capture buffering capacity of organisms. Giant pandas (Ailuropoda melanoleuca) live topographically complex...
The discovery of ferroelectric polarization in HfO2-based ultrathin films has spawned a lot interest due to their potential applications data storage. Recently, new R3m rhombohedral phase was proposed be responsible for the emergence ferroelectricity [111]-oriented Hf0.5Zr0.5O2 thin films, but fundamental mechanism such remains poorly understood. In this paper, we employ density-functional-theory calculations investigate structural and properties HfO2 phase. We find that film thickness...
Background: There is still no conclusive understanding of whether the glucokinase regulator (GCKR) gene rs780094 and rs1260326 polymorphisms predispose to gestational diabetes mellitus (GDM). Objective: This systematic review meta-analysis aimed determine effect GCKR on GDM susceptibility. Methods: Seven literature databases were searched (from inception February 17, 2024) locate relevant studies included in further meta-analysis. Odds ratio (OR) 95% confidence intervals (CI) pooled...
Reasoning LLMs such as OpenAI o1, o3 and DeepSeek R1 have made significant progress in mathematics coding, yet find challenging advanced tasks International Mathematical Olympiad (IMO) combinatorics problems, Abstraction Corpus (ARC) puzzles, Humanity's Last Exam (HLE) questions. We use a diverse inference approach that combines multiple models methods at test time. verifying code rejection sampling on other problems is simple effective. automatically verify correctness of solutions to IMO...
Twisted moir\'e supercells, which can be approximated as a combination of sliding bilayers and constitute various topologically nontrivial polarization patterns, attract extensive attention recently. However, because the excessive size supercell, most studies are based on effective models lack results first-principles calculation. In this work, we use machine learning to determine topological structure pattern in twisted strained bilayer hexagonal boron nitride (h-BN). We further confirm...
Abstract This study presents a novel method using spin quantum sensors to explore temporal variations of fundamental constants, significantly expanding the frequency range and providing constraints on scalar dark matter.
Abstract Rapid industrial development increases soil contamination risks in suburban agricultural areas; however, the extent and sources of heavy metal(loid)s pollution during this process are not well understood. This study investigated concentrations, spatial distribution, potential (As, Hg, Cd, Cr, Pb) wheat samples from an area. The found that Cd was main pollutant with highest enrichment factor 15.9 21.6% having concentrations exceeding China's limit 0.1 mg/kg. Soil As, Pb were...
As an emerging magnetic phase, altermagnets with compensated order and nonrelativistic spin-splitting have attracted widespread attention. Currently, strain engineering is considered to be effective method for inducing valley polarization in altermagnets; however, achieving controllable switching of extremely challenging. Herein, combined the tight-binding model first-principles calculations, we propose that interlayer sliding can used successfully induce effectively manipulate large...
Ferroelectric tunnel junctions (FTJs) have recently aroused significant interest due to the interesting physics controlling their properties and potential application in nonvolatile memory devices. In this work, we propose a new concept design high-performance FTJs based on ferroelectric/polar-oxide composite barriers. Using density functional theory calculations, model electronic transport of LaNiO3/PbTiO3/LaAlO3/LaNiO3 demonstrate that an ultrathin polar LaAlO3(001) layer strongly enhances...
The SAT attack has shown to be efficient against most combinational logic encryption methods. It can extended sequential techniques by leveraging circuit unrolling and model checking However, with no guidance on the number of times that a needs unrolled find correct key, tends solve many time-consuming Boolean satisfiability (SAT) problems, which significantly hamper its efficiency. In this paper, we introduce Fun-SAT, functional corruptibility-guided SAT-based decrease solving time...
Large number of ReLU and MAC operations Deep neural networks make them ill-suited for latency compute-efficient private inference. In this paper, we present a model optimization method that allows to learn be shallow. particular, leverage the sensitivity convolutional block remove layer merge its succeeding preceding convolution layers shallow block. Unlike existing reduction methods, our joint can yield models with improved both ReLUs linear by up 1.73× 1.47×, respectively, evaluated...
The drug supervision methods based on near-infrared spectroscopy analysis are heavily dependent the chemometrics model which characterizes relationship between spectral data and categories. preliminary application of convolution neural network in demonstrates excellent end-to-end prediction ability, but it is sensitive to hyper-parameters network. transformer a deep-learning self-attention mechanism that compares convolutional networks (CNNs) predictive performance has an easy-to-design...
Recently, private inference (PI) has addressed the rising concern over data and model privacy in machine learning as a service. However, existing PI frameworks suffer from high computational communication overheads due to expensive multi-party computation (MPC) protocols, particularly for large models such vision transformers (ViT). The majority of this overhead is encrypted softmax operation each self-attention layer. In work, we present SAL-ViT with two novel techniques boost efficiency on...
The Boolean satisfiability (SAT) attack is an oracle-guided that can break most combinational logic locking schemes by efficiently pruning out all the wrong keys from search space. Extending such to sequential requires multiple time-consuming rounds of SAT solving, performed using "unrolled" version circuit, and model checking, used determine successful termination attack. This article addresses these challenges formally characterizing relation between minimum unrolling depth required prune...