arXiv subjects

Lei Wang

Publications and source records attributed to Lei Wang.

At least 649 records · Page 36Linked to original sources

arXiv2024

User Behavior Simulation with Large Language Model based Agents

Simulating high quality user behavior data has always been a fundamental problem in human-centered applications, where the major difficulty originates from the intricate mechanism of human decision process. Recently, substantial evidences have suggested that by learning huge amounts of web knowledge, large language models (LLMs) can achieve human-like intelligence. We believe these models can provide significant opportunities to more believable user behavior simulation. To inspire such direction, we propose an LLM-based agent framework and design a sandbox environment to simulate real user behaviors. Based on extensive experiments, we find that the simulated behaviors of our method are very close to the ones of real humans. Concerning potential applications, we simulate and study two social phenomenons including (1) information cocoons and (2) user conformity behaviors. This research provides novel simulation paradigms for human-centered applications.

cs.IR↗

arXiv2024

Super stealth dicing of transparent solids with nanometric precision

Laser cutting of semiconductor wafers and transparent dielectrics has become a dominant process in manufacturing industries, encompassing a wide range of applications from flat display panels to microelectronic chips. Limited by the diffraction barrier imposed on the beam width and its longitudinal extend of laser focus, a trade-off must be made between cutting accuracy and aspect ratio in conventional laser processing, with accuracy typically approaching a micron and the aspect ratio on the order of $10^2$. Herein, we propose a method to circumvent this limitation. It is based on the laser modification induced by a back-scattering interference crawling mechanism, which creates a positive feedback for homogenizing longitudinal energy deposition and lateral sub-wavelength light confinement during laser-matter interaction. Consequently, cutting width on the scale of tens of nanometers and aspect ratio $10^3 \sim 10^4$ were simultaneously achieved. We refer to this technique as ``super stealth dicing'', which is validated through numerical simulations, ensuring its broad applicability. It can be applied to various transparent functional solids, such as glass, laser crystal, ferroelectric, and semiconductor, and is elevating the precision of future advanced laser dicing, patterning, and drilling into the nanometric era.

physics.optics↗

arXiv2024

The $U(1)_{L_μ-L_τ}$ breaking phase transition, muon $g-2$, dark matter, collider and gravitational wave

Combining the dark matter and muon $g-2$ anomaly, we study the $U(1)_{L_μ-L_τ}$ breaking phase transition, gravitational wave spectra, and the direct detection at the LHC in an extra $U(1)_{L_μ-L_τ}$ gauge symmetry extension of the standard model. The new fields includes vector-like leptons ($E_1,~ E_2,~ N$), $U(1)_{L_μ-L_τ}$ breaking scalar $S$ and gauge boson $Z'$, as well as the dark matter candidate $X_I$ and its heavy partner $X_R$. A joint explanation of the dark matter relic density and muon $g-2$ anomaly excludes the region where both $min(m_{E_1},m_{E_2},m_N,m_{X_R})$ and $min(m_{Z'},m_S)$ are much larger than $m_{X_I}$. In the parameter space accommodating the DM relic density and muon $g-2$ anomaly, the model can achieve a first order $U(1)_{L_μ-L_τ}$ breaking phase transition, whose strength is sensitive to the parameters of Higgs potential. The corresponding gravitational wave spectra can reach the sensitivity of U-DECIGO. In addition, the direct searches at the LHC impose stringent bound on the mass spectra of the vector-like leptons and dark matter.

hep-ph↗

arXiv2024

Electroweak baryogenesis, dark matter, and dark CP-symmetry

With the motivation of explaining the dark matter and achieving the electroweak baryogenesis via the spontaneous CP-violation at high temperature, we propose a complex singlet scalar ($S=\frac{χ+iη_s}{\sqrt{2}}$) extension of the two-Higgs-doublet model respecting a discrete dark CP-symmetry: $S\to -S^*$. The dark CP-symmetry guarantees $χ$ to be a dark matter candidate on one hand and on the other hand allows $η_s$ to have mixings with the pseudoscalars of the Higgs doublet fields, which play key roles in generating the CP-violation sources needed by the electroweak baryogenesis at high temperature. The universe undergoes multi-step phase transitions, including a strongly first-order electroweak phase transition during which the baryon number is produced. At the present temperature, the observed vacuum is produced and the CP-symmetry is restored so that the stringent electric dipole moment experimental bounds are satisfied. Considering relevant constraints, we study the simple scenario of $m_χ$ around the Higgs resonance region, and find that the dark matter relic abundance and the baryon asymmetry can be simultaneously explained. Finally, we briefly discuss the gravitational wave signatures at future space-based detectors and the LHC signatures.

hep-ph↗

arXiv2024

CAFE: Carbon-Aware Federated Learning in Geographically Distributed Data Centers

Training large-scale artificial intelligence (AI) models demands significant computational power and energy, leading to increased carbon footprint with potential environmental repercussions. This paper delves into the challenges of training AI models across geographically distributed (geo-distributed) data centers, emphasizing the balance between learning performance and carbon footprint. We consider Federated Learning (FL) as a solution, which prioritizes model parameter exchange over raw data, ensuring data privacy and compliance with local regulations. Given the variability in carbon intensity across regions, we propose a new framework called CAFE (short for Carbon-Aware Federated Learning) to optimize training within a fixed carbon footprint budget. Our approach incorporates coreset selection to assess learning performance, employs the Lyapunov drift-plus-penalty framework to address the unpredictability of future carbon intensity, and devises an efficient algorithm to address the combinatorial complexity of the data center selection. Through extensive simulations using real-world carbon intensity data, we demonstrate the efficacy of our algorithm, highlighting its superiority over existing methods in optimizing learning performance while minimizing environmental impact.

cs.LG↗

arXiv2024

A Systematic Evaluation of GPT-4V's Multimodal Capability for Medical Image Analysis

This work conducts an evaluation of GPT-4V's multimodal capability for medical image analysis, with a focus on three representative tasks of radiology report generation, medical visual question answering, and medical visual grounding. For the evaluation, a set of prompts is designed for each task to induce the corresponding capability of GPT-4V to produce sufficiently good outputs. Three evaluation ways including quantitative analysis, human evaluation, and case study are employed to achieve an in-depth and extensive evaluation. Our evaluation shows that GPT-4V excels in understanding medical images and is able to generate high-quality radiology reports and effectively answer questions about medical images. Meanwhile, it is found that its performance for medical visual grounding needs to be substantially improved. In addition, we observe the discrepancy between the evaluation outcome from quantitative analysis and that from human evaluation. This discrepancy suggests the limitations of conventional metrics in assessing the performance of large language models like GPT-4V and the necessity of developing new metrics for automatic quantitative analysis.

cs.CV↗

arXiv2024

Localization of Dummy Data Injection Attacks in Power Systems Considering Incomplete Topological Information: A Spatio-Temporal Graph Wavelet Convolutional Neural Network Approach

The emergence of novel the dummy data injection attack (DDIA) poses a severe threat to the secure and stable operation of power systems. These attacks are particularly perilous due to the minimal Euclidean spatial separation between the injected malicious data and legitimate data, rendering their precise detection challenging using conventional distance-based methods. Furthermore, existing research predominantly focuses on various machine learning techniques, often analyzing the temporal data sequences post-attack or relying solely on Euclidean spatial characteristics. Unfortunately, this approach tends to overlook the inherent topological correlations within the non-Euclidean spatial attributes of power grid data, consequently leading to diminished accuracy in attack localization. To address this issue, this study takes a comprehensive approach. Initially, it examines the underlying principles of these new DDIAs on power systems. Here, an intricate mathematical model of the DDIA is designed, accounting for incomplete topological knowledge and alternating current (AC) state estimation from an attacker's perspective. Subsequently, by integrating a priori knowledge of grid topology and considering the temporal correlations within measurement data and the topology-dependent attributes of the power grid, this study introduces temporal and spatial attention matrices. These matrices adaptively capture the spatio-temporal correlations within the attacks. Leveraging gated stacked causal convolution and graph wavelet sparse convolution, the study jointly extracts spatio-temporal DDIA features. Finally, the research proposes a DDIA localization method based on spatio-temporal graph neural networks. The accuracy and effectiveness of the DDIA model are rigorously demonstrated through comprehensive analytical cases.

eess.SY↗

arXiv2024

C2Ideas: Supporting Creative Interior Color Design Ideation with Large Language Model

Interior color design is a creative process that endeavors to allocate colors to furniture and other elements within an interior space. While much research focuses on generating realistic interior designs, these automated approaches often misalign with user intention and disregard design rationales. Informed by a need-finding preliminary study, we develop C2Ideas, an innovative system for designers to creatively ideate color schemes enabled by an intent-aligned and domain-oriented large language model. C2Ideas integrates a three-stage process: Idea Prompting stage distills user intentions into color linguistic prompts; Word-Color Association stage transforms the prompts into semantically and stylistically coherent color schemes; and Interior Coloring stage assigns colors to interior elements complying with design principles. We also develop an interactive interface that enables flexible user refinement and interpretable reasoning. C2Ideas has undergone a series of indoor cases and user studies, demonstrating its effectiveness and high recognition of interactive functionality by designers.

cs.HC↗

arXiv2024

Inference Attacks Against Face Recognition Model without Classification Layers

Face recognition (FR) has been applied to nearly every aspect of daily life, but it is always accompanied by the underlying risk of leaking private information. At present, almost all attack models against FR rely heavily on the presence of a classification layer. However, in practice, the FR model can obtain complex features of the input via the model backbone, and then compare it with the target for inference, which does not explicitly involve the outputs of the classification layer adopting logit or other losses. In this work, we advocate a novel inference attack composed of two stages for practical FR models without a classification layer. The first stage is the membership inference attack. Specifically, We analyze the distances between the intermediate features and batch normalization (BN) parameters. The results indicate that this distance is a critical metric for membership inference. We thus design a simple but effective attack model that can determine whether a face image is from the training dataset or not. The second stage is the model inversion attack, where sensitive private data is reconstructed using a pre-trained generative adversarial network (GAN) guided by the attack model in the first stage. To the best of our knowledge, the proposed attack model is the very first in the literature developed for FR models without a classification layer. We illustrate the application of the proposed attack model in the establishment of privacy-preserving FR techniques.

cs.CV↗

arXiv2024

A Decentralized Proximal Gradient Tracking Algorithm for Composite Optimization on Riemannian Manifolds

This paper focuses on minimizing a smooth function combined with a nonsmooth regularization term on a compact Riemannian submanifold embedded in the Euclidean space under a decentralized setting. Typically, there are two types of approaches at present for tackling such composite optimization problems. The first, subgradient-based approaches, rely on subgradient information of the objective function to update variables, achieving an iteration complexity of $\mathcal{O}(ε^{-4}\log^2(ε^{-2}))$. The second, smoothing approaches, involve constructing a smooth approximation of the nonsmooth regularization term, resulting in an iteration complexity of $\mathcal{O}(ε^{-4})$. This paper proposes a proximal gradient type algorithm that fully exploits the composite structure. The global convergence to a stationary point is established with a significantly improved iteration complexity of $\mathcal{O}(ε^{-2})$. To validate the effectiveness and efficiency of our proposed method, we present numerical results in real-world applications, showcasing its superior performance.

math.OC↗

arXiv2024

ConceptThread: Visualizing Threaded Concepts in MOOC Videos

Massive Open Online Courses (MOOCs) platforms are becoming increasingly popular in recent years. Online learners need to watch the whole course video on MOOC platforms to learn the underlying new knowledge, which is often tedious and time-consuming due to the lack of a quick overview of the covered knowledge and their structures. In this paper, we propose ConceptThread, a visual analytics approach to effectively show the concepts and the relations among them to facilitate effective online learning. Specifically, given that the majority of MOOC videos contain slides, we first leverage video processing and speech analysis techniques, including shot recognition, speech recognition and topic modeling, to extract core knowledge concepts and construct the hierarchical and temporal relations among them. Then, by using a metaphor of thread, we present a novel visualization to intuitively display the concepts based on video sequential flow, and enable learners to perform interactive visual exploration of concepts. We conducted a quantitative study, two case studies, and a user study to extensively evaluate ConceptThread. The results demonstrate the effectiveness and usability of ConceptThread in providing online learners with a quick understanding of the knowledge content of MOOC videos.

cs.HC↗

arXiv2024

A Fast, Performant, Secure Distributed Training Framework For Large Language Model

The distributed (federated) LLM is an important method for co-training the domain-specific LLM using siloed data. However, maliciously stealing model parameters and data from the server or client side has become an urgent problem to be solved. In this paper, we propose a secure distributed LLM based on model slicing. In this case, we deploy the Trusted Execution Environment (TEE) on both the client and server side, and put the fine-tuned structure (LoRA or embedding of P-tuning v2) into the TEE. Then, secure communication is executed in the TEE and general environments through lightweight encryption. In order to further reduce the equipment cost as well as increase the model performance and accuracy, we propose a split fine-tuning scheme. In particular, we split the LLM by layers and place the latter layers in a server-side TEE (the client does not need a TEE). We then combine the proposed Sparsification Parameter Fine-tuning (SPF) with the LoRA part to improve the accuracy of the downstream task. Numerous experiments have shown that our method guarantees accuracy while maintaining security.

cs.LG↗

arXiv2024

Asymptotic behavior of solution of the non-resistive 2D MHD equations on the half space

In this paper, we obtain the global well-posedness and the asymptotic behavior of solution of non-resistive 2D MHD problem on the half space. We overcome the difficulty of zero spectrum gap by building the relationship between half space and the whole space, and get the resolvent estimate for the weak diffusion system. We use the two-tier energy method that couples the boundedness of high-order $(H^3)$ energy to the decay of low-order energy, the latter of which is necessary to control the growth of the highest energy.

math.AP↗

arXiv2024

AdvSV: An Over-the-Air Adversarial Attack Dataset for Speaker Verification

It is known that deep neural networks are vulnerable to adversarial attacks. Although Automatic Speaker Verification (ASV) built on top of deep neural networks exhibits robust performance in controlled scenarios, many studies confirm that ASV is vulnerable to adversarial attacks. The lack of a standard dataset is a bottleneck for further research, especially reproducible research. In this study, we developed an open-source adversarial attack dataset for speaker verification research. As an initial step, we focused on the over-the-air attack. An over-the-air adversarial attack involves a perturbation generation algorithm, a loudspeaker, a microphone, and an acoustic environment. The variations in the recording configurations make it very challenging to reproduce previous research. The AdvSV dataset is constructed using the Voxceleb1 Verification test set as its foundation. This dataset employs representative ASV models subjected to adversarial attacks and records adversarial samples to simulate over-the-air attack settings. The scope of the dataset can be easily extended to include more types of adversarial attacks. The dataset will be released to the public under the CC BY-SA 4.0. In addition, we also provide a detection baseline for reproducible research.

cs.SD↗

arXiv2024

A Hypernetwork Based Framework for Non-Stationary Channel Prediction

In order to break through the development bottleneck of modern wireless communication networks, a critical issue is the out-of-date channel state information (CSI) in high mobility scenarios. In general, non-stationary CSI has statistical properties which vary with time, implying that the data distribution changes continuously over time. This temporal distribution shift behavior undermines the accurate channel prediction and it is still an open problem in the related literature. In this paper, a hypernetwork based framework is proposed for non-stationary channel prediction. The framework aims to dynamically update the neural network (NN) parameters as the wireless channel changes to automatically adapt to various input CSI distributions. Based on this framework, we focus on low-complexity hypernetwork design and present a deep learning (DL) based channel prediction method, termed as LPCNet, which improves the CSI prediction accuracy with acceptable complexity. Moreover, to maximize the achievable downlink spectral efficiency (SE), a joint channel prediction and beamforming (BF) method is developed, termed as JLPCNet, which seeks to predict the BF vector. Our numerical results showcase the effectiveness and flexibility of the proposed framework, and demonstrate the superior performance of LPCNet and JLPCNet in various scenarios for fixed and varying user speeds.

eess.SP↗

arXiv2024

Federated Learning with Instance-Dependent Noisy Label

Federated learning (FL) with noisy labels poses a significant challenge. Existing methods designed for handling noisy labels in centralized learning tend to lose their effectiveness in the FL setting, mainly due to the small dataset size and the heterogeneity of client data. While some attempts have been made to tackle FL with noisy labels, they primarily focused on scenarios involving class-conditional noise. In this paper, we study the more challenging and practical issue of instance-dependent noise (IDN) in FL. We introduce a novel algorithm called FedBeat (Federated Learning with Bayesian Ensemble-Assisted Transition Matrix Estimation). FedBeat aims to build a global statistically consistent classifier using the IDN transition matrix (IDNTM), which encompasses three synergistic steps: (1) A federated data extraction step that constructs a weak global model and extracts high-confidence data using a Bayesian model ensemble method. (2) A federated transition matrix estimation step in which clients collaboratively train an IDNTM estimation network based on the extracted data. (3) A federated classifier correction step that enhances the global model's performance by training it using a loss function tailored for noisy labels, leveraging the IDNTM. Experiments conducted on CIFAR-10 and SVHN verify that the proposed method significantly outperforms state-of-the-art methods.

cs.LG↗

arXiv2024

An Inexact Preconditioned Zeroth-order Proximal Method for Composite Optimization

In this paper, we consider the composite optimization problem, where the objective function integrates a continuously differentiable loss function with a nonsmooth regularization term. Moreover, only the function values for the differentiable part of the objective function are available. To efficiently solve this composite optimization problem, we propose a preconditioned zeroth-order proximal gradient method in which the gradients and preconditioners are estimated by finite-difference schemes based on the function values at the same trial points. We establish the global convergence and worst-case complexity for our proposed method. Numerical experiments exhibit the superiority of our developed method.

math.OC↗

arXiv2024

An Initial Investigation of Neural Replay Simulator for Over-the-Air Adversarial Perturbations to Automatic Speaker Verification

Deep Learning has advanced Automatic Speaker Verification (ASV) in the past few years. Although it is known that deep learning-based ASV systems are vulnerable to adversarial examples in digital access, there are few studies on adversarial attacks in the context of physical access, where a replay process (i.e., over the air) is involved. An over-the-air attack involves a loudspeaker, a microphone, and a replaying environment that impacts the movement of the sound wave. Our initial experiment confirms that the replay process impacts the effectiveness of the over-the-air attack performance. This study performs an initial investigation towards utilizing a neural replay simulator to improve over-the-air adversarial attack robustness. This is achieved by using a neural waveform synthesizer to simulate the replay process when estimating the adversarial perturbations. Experiments conducted on the ASVspoof2019 dataset confirm that the neural replay simulator can considerably increase the success rates of over-the-air adversarial attacks. This raises the concern for adversarial attacks on speaker verification in physical access applications.

cs.SD↗