arXiv subjects

Wenhui Zhang

Publications and source records attributed to Wenhui Zhang.

34 records · Page 2Linked to original sources

arXiv2024

Introducing v0.5 of the AI Safety Benchmark from MLCommons

This paper introduces v0.5 of the AI Safety Benchmark, which has been created by the MLCommons AI Safety Working Group. The AI Safety Benchmark has been designed to assess the safety risks of AI systems that use chat-tuned language models. We introduce a principled approach to specifying and constructing the benchmark, which for v0.5 covers only a single use case (an adult chatting to a general-purpose assistant in English), and a limited set of personas (i.e., typical users, malicious users, and vulnerable users). We created a new taxonomy of 13 hazard categories, of which 7 have tests in the v0.5 benchmark. We plan to release version 1.0 of the AI Safety Benchmark by the end of 2024. The v1.0 benchmark will provide meaningful insights into the safety of AI systems. However, the v0.5 benchmark should not be used to assess the safety of AI systems. We have sought to fully document the limitations, flaws, and challenges of v0.5. This release of v0.5 of the AI Safety Benchmark includes (1) a principled approach to specifying and constructing the benchmark, which comprises use cases, types of systems under test (SUTs), language and context, personas, tests, and test items; (2) a taxonomy of 13 hazard categories with definitions and subcategories; (3) tests for seven of the hazard categories, each comprising a unique set of test items, i.e., prompts. There are 43,090 test items in total, which we created with templates; (4) a grading system for AI systems against the benchmark; (5) an openly available platform, and downloadable tool, called ModelBench that can be used to evaluate the safety of AI systems on the benchmark; (6) an example evaluation report which benchmarks the performance of over a dozen openly available chat-tuned language models; (7) a test specification for the benchmark.

cs.CL↗

arXiv2024

Evaluation of General Large Language Models in Contextually Assessing Semantic Concepts Extracted from Adult Critical Care Electronic Health Record Notes

The field of healthcare has increasingly turned its focus towards Large Language Models (LLMs) due to their remarkable performance. However, their performance in actual clinical applications has been underexplored. Traditional evaluations based on question-answering tasks don't fully capture the nuanced contexts. This gap highlights the need for more in-depth and practical assessments of LLMs in real-world healthcare settings. Objective: We sought to evaluate the performance of LLMs in the complex clinical context of adult critical care medicine using systematic and comprehensible analytic methods, including clinician annotation and adjudication. Methods: We investigated the performance of three general LLMs in understanding and processing real-world clinical notes. Concepts from 150 clinical notes were identified by MetaMap and then labeled by 9 clinicians. Each LLM's proficiency was evaluated by identifying the temporality and negation of these concepts using different prompts for an in-depth analysis. Results: GPT-4 showed overall superior performance compared to other LLMs. In contrast, both GPT-3.5 and text-davinci-003 exhibit enhanced performance when the appropriate prompting strategies are employed. The GPT family models have demonstrated considerable efficiency, evidenced by their cost-effectiveness and time-saving capabilities. Conclusion: A comprehensive qualitative performance evaluation framework for LLMs is developed and operationalized. This framework goes beyond singular performance aspects. With expert annotations, this methodology not only validates LLMs' capabilities in processing complex medical data but also establishes a benchmark for future LLM evaluations across specialized domains.

cs.CL↗

arXiv2023

Self-guided Few-shot Semantic Segmentation for Remote Sensing Imagery Based on Large Vision Models

The Segment Anything Model (SAM) exhibits remarkable versatility and zero-shot learning abilities, owing largely to its extensive training data (SA-1B). Recognizing SAM's dependency on manual guidance given its category-agnostic nature, we identified unexplored potential within few-shot semantic segmentation tasks for remote sensing imagery. This research introduces a structured framework designed for the automation of few-shot semantic segmentation. It utilizes the SAM model and facilitates a more efficient generation of semantically discernible segmentation outcomes. Central to our methodology is a novel automatic prompt learning approach, leveraging prior guided masks to produce coarse pixel-wise prompts for SAM. Extensive experiments on the DLRSD datasets underline the superiority of our approach, outperforming other available few-shot methodologies.

cs.CV↗

arXiv2023

Novel $π$-type vortex in a nanoscale extreme type-II superconductor: Induced by quantum-size effect

By numerically solving the Bogoliubov-de Gennes equations, we report a novel $π$-type vortex state whose order parameter near the core undergoes an extraordinary $π$-phase change for a quantum-confined extreme type-II $s$-wave superconductor. Its supercurrent behaves as the cube of the radial coordinate near the core, and its local density of states spectrum exhibits a significant negative-shifted zero-bias peak. Such $π$-type vortex state is induced by quantum-size effect, and can survive thermal smearing at temperatures up to a critical value $T_τ$. The Anderson's approximation indicates that the $π$-type vortex may remain stable under sufficiently week magnetic field in the case less deep in the type-II limit. Moreover, we find that its appearance is governed by the sample size and $k_Fξ_0$ with $k_F$ the Fermi wave number and $ξ_0$ the zero-temperature coherence length. Similar effects may be expected in quantum-confined ultracold superfluid Fermi gasses, or even high-$T_c$ superconductors with proper $k_Fξ_0$ value.

cond-mat.supr-con↗

arXiv2022

Worst-case Design for RIS-aided Over-the-air Computation with Imperfect CSI

Over-the-air computation (AirComp) enables fast wireless data aggregation at the receiver through concurrent transmission by sensors in the application of Internet-of-Things (IoT). To further improve the performance of AirComp under unfavorable propagation channel conditions, we consider the problem of computation distortion minimization in a reconfigurable intelligent surface (RIS)-aided AirComp system. In particular, we take into account an additive bounded uncertainty of the channel state information (CSI) and the total power constraint, and jointly optimize the transceiver (Tx-Rx) and the RIS phase design from the perspective of worst-case robustness by minimizing the mean squared error (MSE) of the computation. To solve this intractable nonconvex problem, we develop an efficient alternating algorithm where both solutions to the robust sub-problem and to the joint design of Tx-Rx and RIS are obtained in closed forms. Simulation results demonstrate the effectiveness of the proposed method.

cs.IT↗

arXiv2021

Enhanced countering adversarial attacks via input denoising and feature restoring

Despite the fact that deep neural networks (DNNs) have achieved prominent performance in various applications, it is well known that DNNs are vulnerable to adversarial examples/samples (AEs) with imperceptible perturbations in clean/original samples. To overcome the weakness of the existing defense methods against adversarial attacks, which damages the information on the original samples, leading to the decrease of the target classifier accuracy, this paper presents an enhanced countering adversarial attack method IDFR (via Input Denoising and Feature Restoring). The proposed IDFR is made up of an enhanced input denoiser (ID) and a hidden lossy feature restorer (FR) based on the convex hull optimization. Extensive experiments conducted on benchmark datasets show that the proposed IDFR outperforms the various state-of-the-art defense methods, and is highly effective for protecting target models against various adversarial black-box or white-box attacks. \footnote{Souce code is released at: \href{https://github.com/ID-FR/IDFR}{https://github.com/ID-FR/IDFR}}

cs.CV↗

arXiv2021

Analyzing the Overhead of Filesystem Protection Using Linux Security Modules

Over the years, the complexity of the Linux Security Module (LSM) is keeping increasing, and the count of the authorization hooks is nearly doubled. It is important to provide up-to-date measurement results of LSM for system practitioners so that they can make prudent trade-offs between security and performance. This work evaluates the overhead of LSM for file accesses on Linux v5.3.0. We build a performance evaluation framework for LSM. It has two parts, an extension of LMBench2.5 to evaluate the overhead of file operations for different security modules, and a security module with tunable latency for policy enforcement to study the impact of the latency of policy enforcement on the end-to-end latency of file operations. In our evaluation, we find opening a file would see about 87% (Linux v5.3) performance drop when the kernel is integrated with SELinux hooks (policy enforcement disabled) than without, while the figure was 27% (Linux v2.4.2). We found that performance of the above downgrade is affected by two parts, policy enforcement and hook placement. To further investigate the impact of policy enforcement and hook placement respectively, we build a Policy Testing Module, which reuses hook placements of LSM, while alternating latency of policy enforcement. With this module, we are able to quantitatively estimate the impact of the latency of policy enforcement on the end-to-end latency of file operations by using a multiple linear regression model and count policy authorization frequencies for each syscall. We then discuss and justify the evaluation results with static analysis on our enhanced syscalls' call graphs.

cs.CR↗

arXiv2020

Cascaded Channel Estimation for IRS-assisted Mmwave Multi-antenna with Quantized Beamforming

In this letter, we optimize the channel estimator of the cascaded channel in an intelligent reflecting surface (IRS)-assisted millimeter wave (mmWave) multi-antenna system. In this system, the receiver is equipped with a hybrid architecture adopting quantized beamforming. Different from traditional multiple-input multiple-output (MIMO) systems, the design of channel estimation is challenging since the IRS is usually a passive array with limited signal processing capability. We derive the optimized channel estimator in a closed form by reformulating the problem of cascaded channel estimation in this system, leveraging the typical mean-squared error (MSE) criterion. Considering the presence of possible channel sparsity in mmWave channels, we generalize the proposed method by exploiting the channel sparsity for further performance enhancement and computational complexity reduction. Simulation results verify that the proposed estimator significantly outperforms the existing ones.

cs.IT↗

arXiv2019

SvTPM: A Secure and Efficient vTPM in the Cloud

Virtual Trusted Platform Modules (vTPMs) have been widely used in commercial cloud platforms (e.g. Google Cloud, VMware Cloud, and Microsoft Azure) to provide virtual root-of-trust for virtual machines. Unfortunately, current state-of-the-art vTPM implementations are suffering from confidential data leakage and high performance overhead. In this paper, we present SvTPM, a secure and efficient software-based vTPM implementation based on hardware-rooted Trusted Execution Environment (TEE), providing a whole life cycle protection of vTPMs in the cloud. SvTPM offers strong isolation protection, so that cloud tenants or even cloud administrators cannot get vTPM's private keys or any other sensitive data. In SvTPM, we identify and solve a couple of critical security challenges for vTPM protection with SGX, such as NVRAM replacement attack, rollback attacks, trust establishment, and a fine-grained trusted clock. We implement a prototype of SvTPM on both QEMU and KVM. Performance evaluation results show that SvTPM achieves orders of magnitude of performance gains comparing to the vTPMs protected with physical TPM. The launch time of SvTPM is 2600$\times$ faster than vTPMs built upon hardware TPM. In the micro-benchmarks evaluation, we find that the command execution latency of SvTPM is smaller than or equal to the existing schemes.

cs.CR↗

arXiv2019

Parallel implementation of w-projection wide-field imaging

w-Projection is a wide-field imaging technique that is widely used in radio synthesis arrays. Processing the wide-field big data generated by the future Square Kilometre Array (SKA) will require significant updates to current methods to significantly reduce the time consumed on data processing. Data loading and gridding are found to be two major time-consuming tasks in w-projection. In this paper, we investigate two parallel methods of accelerating w-projection processing on multiple nodes: the hybrid Message Passing Interface (MPI) and Open Multi-Processing (OpenMP) method based on multicore Central Processing Units (CPUs) and the hybrid MPI and Compute Unified Device Architecture (CUDA) method based on Graphics Processing Units (GPUs). Both methods are successfully employed and operated in various computational environments, confirming their robustness. The experimental results show that the total runtime of both MPI + OpenMP and MPI + CUDA methods is significantly shorter than that of single-thread processing. MPI + CUDA generally shows faster performance when running on multiple nodes than MPI + OpenMP, especially on large numbers of nodes. The single-precision GPU-based processing yields faster computation than the double-precision processing; while the single- and doubleprecision CPU-based processing shows consistent computational performance. The gridding time remarkably increases when the support size of the convolution kernel is larger than 8 and the image size is larger than 2,048 pixels. The present research offers useful guidance for developing SKA imaging pipelines.

astro-ph.IM↗

arXiv2018

End to End Video Segmentation for Driving : Lane Detection For Autonomous Car

Safety and decline of road traffic accidents remain important issues of autonomous driving. Statistics show that unintended lane departure is a leading cause of worldwide motor vehicle collisions, making lane detection the most promising and challenge task for self-driving. Today, numerous groups are combining deep learning techniques with computer vision problems to solve self-driving problems. In this paper, a Global Convolution Networks (GCN) model is used to address both classification and localization issues for semantic segmentation of lane. We are using color-based segmentation is presented and the usability of the model is evaluated. A residual-based boundary refinement and Adam optimization is also used to achieve state-of-art performance. As normal cars could not afford GPUs on the car, and training session for a particular road could be shared by several cars. We propose a framework to get it work in real world. We build a real time video transfer system to get video from the car, get the model trained in edge server (which is equipped with GPUs), and send the trained model back to the car.

cs.CV↗

arXiv2018

Medical Volume Reconstruction Techniques

Medical visualization is the use of computers to create 3D images from medical imaging data sets, almost all surgery and cancer treatment in the developed world relies on it.Volume visualization techniques includes iso-surface visualization, mesh visualization and point cloud visualization techniques, these techniques have revolutionized medicine. Much of modern medicine relies on the 3D imaging that is possible with magnetic resonance imaging (MRI) scanners, functional magnetic resonance imaging (fMRI)scanners, positron emission tomography (PET) scanners, ultrasound imaging (US) scanners, X-Ray scanners, bio-marker microscopy imaging scanners and computed tomography (CT) scanners, which make 3D images out of 2D slices. The primary goal of this report is the application-oriented optimization of existing volume rendering methods providing interactive frame rates. Techniques are presented for traditional alpha-blending rendering, surface-shaded display, maximum intensity projection (MIP), and fast previewing with fully interactive parameter control. Different preprocessing strategies are proposed for interactive iso-surface rendering and fast previewing, such as the well-known marching cube algorithm.

cs.GR↗

arXiv2017

On Completeness Results of Hoare Logic Relative to the Standard Model

The general completeness problem of Hoare logic relative to the standard model $N$ of Peano arithmetic has been studied by Cook, and it allows for the use of arbitrary arithmetical formulas as assertions. In practice, the assertions would be simple arithmetical formulas, e.g. of a low level in the arithmetical hierarchy. In addition, we find that, by restricting inputs to $N$, the complexity of the minimal assertion theory for the completeness of Hoare logic to hold can be reduced. This paper further studies the completeness of Hoare Logic relative to $N$ by restricting assertions to subclasses of arithmetical formulas (and by restricting inputs to $N$). Our completeness results refine Cook's result by reducing the complexity of the assertion theory.

cs.LO↗

arXiv2017

Completeness of Hoare Logic over Nonstandard Models

The nonstandard approach to program semantics has successfully resolved the completeness problem of Floyd-Hoare logic. The known versions of nonstandard semantics, the Hungary semantics and axiomatic semantics, are so general that they are absent either from mathematical elegance or from practical usefulness. The aim of this paper is to exhibit a not only mathematically elegant but also practically useful nonstandard semantics. A basic property of computable functions in the standard model $N$ of Peano arithmetic $PA$ is $Σ_1$-definability. However, the functions induced by the standard interpretation of while-programs $S$ in nonstandard models $M$ of $PA$ are not always arithmetical. The problem consists in that the standard termination of $S$ in $M$ uses the finiteness in $N$, which is not the finiteness in $M$. To this end, we shall give a new interpretation of $S$ in $M$ such that the termination of $S$ uses $M$-finiteness, and the functions produced by $S$ in all models of $PA$ have the uniform $Σ_1$-definability. Then we define, based on the new semantics of while-programs, a new semantics of Hoare logic in nonstandard models of $PA$, and show that the standard axiom system of Hoare logic is sound and complete w.r.t. the new semantics. It will be established, in $PA$, that the Hungary semantics and axiomatic semantics coincide with the new semantics of while-programs. Moreover, various comparisons with the previous results, usefulness of the nonstandard semantics, and remarks on the completeness issues are presented.

cs.LO↗

arXiv2015

Correlations of neutron multiplicity and gamma multiplicity with fragment mass and total kinetic energy in spontaneous fission of 252Cf

The dependence of correlations of neutron multiplicity ν and gamma-ray multiplicity Mg in spontaneous fission of 252Cf on fragment mass A* and total kinetic energy TKE has been investigated employing the ratio of Mg/ν and the form of Mg(ν). We show for the first time that Mg and ν have a complex correlation for heavy fragment masses, while there is a positive dependence of Mg(ν) for light fragment masses and for near-symmetric mass splits. The ratio Mg/ν exhibits strong shell effects for the neutron magic number N = 50 and near the doubly magic number shell closure at Z = 50 and N = 82. The gamma-ray multiplicity Mg has a maximum for TKE = 165-170 MeV. Above 170 MeV Mg(TKE) is approximately linear, while it deviates significantly from a linear dependence at lower TKE. The correlation between the average neutron and gamma-ray multiplicities can be partly reproduced by model calculations.

nucl-ex↗

arXiv2015

Enhancing the Security of Protocols against Actor Key Compromise Problems

Security of complex systems is an important issue in software engineering. For complex computer systems involving many actors, security protocols are often used for the communication of sensitive data. Actor key compromise (AKC) denotes a situation where the long-term secret key of an actor may be known to an adversary for some reasons. Many protocols are not secure enough for ensuring security in such a situation. In this paper, we further study this problem by looking at potential types of attacks, defining their formal properties and providing solutions to enhance the level of security. As case studies, we analyze the vulnerabilities (with respect to potential AKC attacks) of practical protocols, including PKMv2RSA and Kerberos, and provide solutions to enhance the level of security of such protocols.

cs.CR↗