arXiv subjects

Yuan Zhang

Publications and source records attributed to Yuan Zhang.

At least 181 records · Page 10Linked to original sources

arXiv2025

DH-TRNG: A Dynamic Hybrid TRNG with Ultra-High Throughput and Area-Energy Efficiency

As a vital security primitive, the true random number generator (TRNG) is a mandatory component to build roots of trust for any encryption system. However, existing TRNGs suffer from bottlenecks of low throughput and high area-energy consumption. In this work, we propose DH-TRNG, a dynamic hybrid TRNG circuitry architecture with ultra-high throughput and area-energy efficiency. Our DH-TRNG exhibits portability to distinct process FPGAs and passes both NIST and AIS-31 tests without any post-processing. The experiments show it incurs only 8 slices with the highest throughput of 670Mbps and 620Mbps on Xilinx Virtex-6 and Artix-7, respectively. Compared to the state-of-the-art TRNGs, our proposed design has the highest Throughput/SlicesPower with a 2.63 times increase.

cs.CR↗

arXiv2025

On $\overline\partial$ homotopy formulae for product domains: Nijenhuis-Woolf's formulae and optimal Sobolev estimates

We construct homotopy formulae $f=\overline\partial\mathcal H_qf+\mathcal H_{q+1}\overline\partial f$ for $(0,q)$ forms on the product domain $Ω_1\times\dots\timesΩ_m$, where each $Ω_j$ is either a bounded Lipschitz domain in $\mathbb C^1$, a bounded strongly pseudoconvex domain with $C^2$ boundary, or a smooth convex domain of finite type. Such homotopy operators $\mathcal H_q$ yield solutions to the $\overline\partial$ equation with optimal Sobolev regularity $W^{k,p}\to W^{k,p}$ simultaneously for all $k\in\mathbb Z$ and $1<p<\infty$.

math.CV↗

arXiv2025

You Can't Eat Your Cake and Have It Too: The Performance Degradation of LLMs with Jailbreak Defense

With the rise of generative large language models (LLMs) like LLaMA and ChatGPT, these models have significantly transformed daily life and work by providing advanced insights. However, as jailbreak attacks continue to circumvent built-in safety mechanisms, exploiting carefully crafted scenarios or tokens, the safety risks of LLMs have come into focus. While numerous defense strategies--such as prompt detection, modification, and model fine-tuning--have been proposed to counter these attacks, a critical question arises: do these defenses compromise the utility and usability of LLMs for legitimate users? Existing research predominantly focuses on the effectiveness of defense strategies without thoroughly examining their impact on performance, leaving a gap in understanding the trade-offs between LLM safety and performance. Our research addresses this gap by conducting a comprehensive study on the utility degradation, safety elevation, and exaggerated-safety escalation of LLMs with jailbreak defense strategies. We propose USEBench, a novel benchmark designed to evaluate these aspects, along with USEIndex, a comprehensive metric for assessing overall model performance. Through experiments on seven state-of-the-art LLMs, we found that mainstream jailbreak defenses fail to ensure both safety and performance simultaneously. Although model-finetuning performs the best overall, their effectiveness varies across LLMs. Furthermore, vertical comparisons reveal that developers commonly prioritize performance over safety when iterating or fine-tuning their LLMs.

cs.CR↗

arXiv2025

CS-Eval: A Comprehensive Large Language Model Benchmark for CyberSecurity

Over the past year, there has been a notable rise in the use of large language models (LLMs) for academic research and industrial practices within the cybersecurity field. However, it remains a lack of comprehensive and publicly accessible benchmarks to evaluate the performance of LLMs on cybersecurity tasks. To address this gap, we introduce CS-Eval, a publicly accessible, comprehensive and bilingual LLM benchmark specifically designed for cybersecurity. CS-Eval synthesizes the research hotspots from academia and practical applications from industry, curating a diverse set of high-quality questions across 42 categories within cybersecurity, systematically organized into three cognitive levels: knowledge, ability, and application. Through an extensive evaluation of a wide range of LLMs using CS-Eval, we have uncovered valuable insights. For instance, while GPT-4 generally excels overall, other models may outperform it in certain specific subcategories. Additionally, by conducting evaluations over several months, we observed significant improvements in many LLMs' abilities to solve cybersecurity tasks. The benchmarks are now publicly available at https://github.com/CS-EVAL/CS-Eval.

cs.CR↗

arXiv2025

EDNet: Edge-Optimized Small Target Detection in UAV Imagery -- Faster Context Attention, Better Feature Fusion, and Hardware Acceleration

Detecting small targets in drone imagery is challenging due to low resolution, complex backgrounds, and dynamic scenes. We propose EDNet, a novel edge-target detection framework built on an enhanced YOLOv10 architecture, optimized for real-time applications without post-processing. EDNet incorporates an XSmall detection head and a Cross Concat strategy to improve feature fusion and multi-scale context awareness for detecting tiny targets in diverse environments. Our unique C2f-FCA block employs Faster Context Attention to enhance feature extraction while reducing computational complexity. The WIoU loss function is employed for improved bounding box regression. With seven model sizes ranging from Tiny to XL, EDNet accommodates various deployment environments, enabling local real-time inference and ensuring data privacy. Notably, EDNet achieves up to a 5.6% gain in mAP@50 with significantly fewer parameters. On an iPhone 12, EDNet variants operate at speeds ranging from 16 to 55 FPS, providing a scalable and efficient solution for edge-based object detection in challenging drone imagery. The source code and pre-trained models are available at: https://github.com/zsniko/EDNet.

cs.CV↗

arXiv2025

Point-wise doubling indices of measures and its application to bi-Lipschitz classification of Bedford-McMullen carpets

Doubling measure was introduced by Beurling and Ahlfors in 1956 and now it becomes a basic concept in analysis on metric space. In this paper, for a measure which is not doubling, we introduce a notion of point-wise doubling index, and calculate the point-wise doubling indices of uniform Bernoulli measures on Bedford-McMullen carpets. As an application, we show that, except a small class of Bedford-McMullen carpets, if two Bedford-McMullen carpets are bi-Lipschitz equivalent, then they have the same fiber sequence up to a permutation.

math.DS↗

arXiv2024

Robust distribution-free tests for the linear model

Recently, there has been growing concern about heavy-tailed and skewed noise in biological data. We introduce RobustPALMRT, a flexible permutation framework for testing the association of a covariate of interest adjusted for control covariates. RobustPALMRT controls type I error rate for finite-samples, even in the presence of heavy-tailed or skewed noise. The new framework expands the scope of state-of-the-art tests in three directions. First, our method applies to robust and quantile regressions, even with the necessary hyper-parameter tuning. Second, by separating model-fitting and model-evaluation, we discover that performance improves when using a robust loss function in the model-evaluation step, regardless of how the model is fit. Third, we allow fitting multiple models to detect specialized features of interest in a distribution. To demonstrate this, we introduce DispersionPALRMT, which tests for differences in dispersion between treatment and control groups. We establish theoretical guarantees, identify settings where our method has greater power than existing methods, and analyze existing immunological data on Long-COVID patients. Using RobustPALMRT, we unveil novel differences between Long-COVID patients and others even in the presence of highly skewed noise.

stat.ME↗

arXiv2024

Casevo: A Cognitive Agents and Social Evolution Simulator

In this paper, we introduce a multi-agent simulation framework Casevo (Cognitive Agents and Social Evolution Simulator), that integrates large language models (LLMs) to simulate complex social phenomena and decision-making processes. Casevo is designed as a discrete-event simulator driven by agents with features such as Chain of Thoughts (CoT), Retrieval-Augmented Generation (RAG), and Customizable Memory Mechanism. Casevo enables dynamic social modeling, which can support various scenarios such as social network analysis, public opinion dynamics, and behavior prediction in complex social systems. To demonstrate the effectiveness of Casevo, we utilize one of the U.S. 2020 midterm election TV debates as a simulation example. Our results show that Casevo facilitates more realistic and flexible agent interactions, improving the quality of dynamic social phenomena simulation. This work contributes to the field by providing a robust system for studying large-scale, high-fidelity social behaviors with advanced LLM-driven agents, expanding the capabilities of traditional agent-based modeling (ABM). The open-source code repository address of casevo is https://github.com/rgCASS/casevo.

cs.SI↗

arXiv2024

DRDM: A Disentangled Representations Diffusion Model for Synthesizing Realistic Person Images

Person image synthesis with controllable body poses and appearances is an essential task owing to the practical needs in the context of virtual try-on, image editing and video production. However, existing methods face significant challenges with details missing, limbs distortion and the garment style deviation. To address these issues, we propose a Disentangled Representations Diffusion Model (DRDM) to generate photo-realistic images from source portraits in specific desired poses and appearances. First, a pose encoder is responsible for encoding pose features into a high-dimensional space to guide the generation of person images. Second, a body-part subspace decoupling block (BSDB) disentangles features from the different body parts of a source figure and feeds them to the various layers of the noise prediction block, thereby supplying the network with rich disentangled features for generating a realistic target image. Moreover, during inference, we develop a parsing map-based disentangled classifier-free guided sampling method, which amplifies the conditional signals of texture and pose. Extensive experimental results on the Deepfashion dataset demonstrate the effectiveness of our approach in achieving pose transfer and appearance control.

cs.CV↗

arXiv2024

Public emotions on Internet: In case of AIGC

The proliferation of interactive AI like ChatGPT has fueled intense public discourse surrounding AI- generated content (AIGC). While some fear job displacement, others anticipate productivity gains. Social media provides a rich source of data reflecting public opinion, attitudes, and behaviors. By examining the factors influencing collective sentiment toward AIGC on various platforms, we can refine products, marketing, and AI models themselves. Our research utilized a novel system for real-time tracking and detailed visualization of public mood related to AIGC. This system enabled analysis of the dynamics shaping opinions on nine AIGC products across China's three leading social media sites. Our findings reveal a negative correlation between user demographics (age and education) and positive sentiment towards AIGC on Douyin, contrasting with Weibo's susceptibility to the rapid spread of extreme viewpoints. This work uniquely connects group dynamics theory with social media sentiment, offering valuable guidance for managing online opinion and tailoring targeted campaigns.

cs.SI↗

arXiv2024

Deep learning assisted SERS detection of prolines and hydroxylated prolines using nitrilotriacetic acid functionalized gold nanopillars

Proline (Pro) is one kind of proteinogenic amino acid and an important signaling molecule in the process of metabolism. Hydroxyproline (Hyp) is a product on Pro oxygen sensing post-translational modification (PTM), which is efficiently modulated tumor cells for angiogenesis. Distinguishing between Pro and Hyp is crucial for diagnosing connective tissue disorders, as elevated levels of Hyp can indicate abnormal collagen metabolism, often associated with diseases like osteogenesis imperfecta or fibrosis. However, there is a very small difference between molecular structures of Pro and Hyp, which is a big challenge for current detection technologies to distinguish them. For surface-enhanced Raman scattering (SERS) sensors, the similar molecule structure leads to similar Raman spectra that are difficult to distinguish. Furthermore, another problem is the weak affinity between amino acids sample and SERS-active substrates by physical adsorption. The selecting capturing of Pro and Hyp in the mixture of amino acids is not easy to achieve. In this work, we designed a new method for Pro and Hyp specifical detection and recognition by using gold nanopillars as the SERS substrate and combing nitrilotriacetic acid (NTA) with nickel (Ni) to form NTA-Ni structure as a specifical affinity agent. One side of NTA-Ni was attached to gold nanopillars through thiol binding. Another side captured the amino acids using reversible binding by receptor-ligand interaction between Ni and amino acids. Because of the different binding time with NTA-Ni and amino acids, the sensor can recognize Pro and Hyp from amino acids mixture. Then we used automatic peak assignment program for data analysis and machine learning model to distinguish between Pro and Hyp. The label-free SERS detection of amino acids PTM using gold nanopillars provides a potential method to further biomolecule detection and specifical capture.

q-bio.BM↗

arXiv2024

Contribution of soundscape appropriateness to soundscape quality assessment in space: a mediating variable affecting acoustic comfort

Soundscape appropriateness (SA) provides supplemental information on the matching degree between auditory information and the surrounding scene in soundscape perception. This indicator has been integrated into the standard ISO process for collecting soundscape data, forming a component of the sound quality assessment questionnaire. However, its role in soundscape quality assessment has not been fully understood. Herein, we present the findings from soundscape data collected from Beiling Park in Shenyang, China. A method was developed that integrates mediation effect models with multiscale geographically weighted regression models to explore the mediating role of SA in the impact of sound source types on soundscape quality, as well as the spatial heterogeneity of this mediation effect. The results confirm that SA does mediates the influence of sound source types on acoustics comfort (AC). Specifically, natural sounds (indirect effect/total effect = .19/.19), traffic sounds (indirect effect/total effect = -.46/-.65), and commercial sounds (indirect effect/total effect = -.25/-.12) impact the perception of AC by either enhancing or reducing SA. Moreover, the relationships among variables depicted in this model demonstrate spatial heterogeneity, demonstrating that in urban open spaces with complex constructures, local spatial models may be needed for soundscape assessment. The research reaffirms the significance of SA in urban open spaces. In terms of practical implications for urban and landscape planners, when sound sources cannot be controlled or altered, coordinating between the sound and the surrounding environment through landscape optimisation could also improve the quality of the soundscape through enhancing SA and help achieve the goal of creating healthy urban open spaces.

cs.SD↗

arXiv2024

Rank Resilience of Pattern Matrices against Structured Perturbations with Applications

In structured system theory, a pattern matrix is a matrix with entries either fixed to zero or free to take arbitrary numbers. The (generic) rank of a pattern matrix is the rank of almost all its realizations. The resilience of various system properties is closely tied to the rank resilience of the corresponding pattern matrices. Yet, existing literature predominantly explores the latter aspect by focusing on perturbations that change the zero-nonzero structure of pattern matrices, corresponding to link additions/deletions. In this paper, we consider the rank resilience of pattern matrices against structured perturbations that can arbitrarily alter the values of a prescribed set of entries, corresponding to link weight variations. We say a pattern matrix is structurally rank $r$ resilient against a perturbation pattern if almost all realizations of this pattern matrix have a rank not less than $r$ under arbitrary complex-valued realizations of the perturbation pattern. We establish a generic property in this concept. We then present combinatorial necessary and sufficient conditions for a rectangular pattern matrix to lose full rank under given perturbation patterns. We also generalize them to obtain a sufficient condition and a necessary one for losing an arbitrarily prescribed rank. We finally show our results can be applied to the resilience analysis of various properties of structured (descriptor) systems, including controllability and input-state observability, as well as characterizing zero structurally fixed modes.

cs.IT↗

arXiv2024

SASE: A Searching Architecture for Squeeze and Excitation Operations

In the past few years, channel-wise and spatial-wise attention blocks have been widely adopted as supplementary modules in deep neural networks, enhancing network representational abilities while introducing low complexity. Most attention modules follow a squeeze-and-excitation paradigm. However, to design such attention modules, requires a substantial amount of experiments and computational resources. Neural Architecture Search (NAS), meanwhile, is able to automate the design of neural networks and spares the numerous experiments required for an optimal architecture. This motivates us to design a search architecture that can automatically find near-optimal attention modules through NAS. We propose SASE, a Searching Architecture for Squeeze and Excitation operations, to form a plug-and-play attention block by searching within certain search space. The search space is separated into 4 different sets, each corresponds to the squeeze or excitation operation along the channel or spatial dimension. Additionally, the search sets include not only existing attention blocks but also other operations that have not been utilized in attention mechanisms before. To the best of our knowledge, SASE is the first attempt to subdivide the attention search space and search for architectures beyond currently known attention modules. The searched attention module is tested with extensive experiments across a range of visual tasks. Experimental results indicate that visual backbone networks (ResNet-50/101) using the SASE attention module achieved the best performance compared to those using the current state-of-the-art attention modules. Codes are included in the supplementary material, and they will be made public later.

cs.CV↗

arXiv2024

RadioDiff: An Effective Generative Diffusion Model for Sampling-Free Dynamic Radio Map Construction

Radio map (RM) is a promising technology that can obtain pathloss based on only location, which is significant for 6G network applications to reduce the communication costs for pathloss estimation. However, the construction of RM in traditional is either computationally intensive or depends on costly sampling-based pathloss measurements. Although the neural network (NN)-based method can efficiently construct the RM without sampling, its performance is still suboptimal. This is primarily due to the misalignment between the generative characteristics of the RM construction problem and the discrimination modeling exploited by existing NN-based methods. Thus, to enhance RM construction performance, in this paper, the sampling-free RM construction is modeled as a conditional generative problem, where a denoised diffusion-based method, named RadioDiff, is proposed to achieve high-quality RM construction. In addition, to enhance the diffusion model's capability of extracting features from dynamic environments, an attention U-Net with an adaptive fast Fourier transform module is employed as the backbone network to improve the dynamic environmental features extracting capability. Meanwhile, the decoupled diffusion model is utilized to further enhance the construction performance of RMs. Moreover, a comprehensive theoretical analysis of why the RM construction is a generative problem is provided for the first time, from both perspectives of data features and NN training methods. Experimental results show that the proposed RadioDiff achieves state-of-the-art performance in all three metrics of accuracy, structural similarity, and peak signal-to-noise ratio. The code is available at https://github.com/UNIC-Lab/RadioDiff.

cs.LG↗

arXiv2024

DSDE: Using Proportion Estimation to Improve Model Selection for Out-of-Distribution Detection

Model library is an effective tool for improving the performance of single-model Out-of-Distribution (OoD) detector, mainly through model selection and detector fusion. However, existing methods in the literature do not provide uncertainty quantification for model selection results. Additionally, the model ensemble process primarily focuses on controlling the True Positive Rate (TPR) while neglecting the False Positive Rate (FPR). In this paper, we emphasize the significance of the proportion of models in the library that identify the test sample as an OoD sample. This proportion holds crucial information and directly influences the error rate of OoD detection.To address this, we propose inverting the commonly-used sequential p-value strategies. We define the rejection region initially and then estimate the error rate. Furthermore, we introduce a novel perspective from change-point detection and propose an approach for proportion estimation with automatic hyperparameter selection. We name the proposed approach as DOS-Storey-based Detector Ensemble (DSDE). Experimental results on CIFAR10 and CIFAR100 demonstrate the effectiveness of our approach in tackling OoD detection challenges. Specifically, the CIFAR10 experiments show that DSDE reduces the FPR from 11.07% to 3.31% compared to the top-performing single-model detector.

stat.ML↗

arXiv2024

Human-AI Collaborative Multi-modal Multi-rater Learning for Endometriosis Diagnosis

Endometriosis, affecting about 10% of individuals assigned female at birth, is challenging to diagnose and manage. Diagnosis typically involves the identification of various signs of the disease using either laparoscopic surgery or the analysis of T1/T2 MRI images, with the latter being quicker and cheaper but less accurate. A key diagnostic sign of endometriosis is the obliteration of the Pouch of Douglas (POD). However, even experienced clinicians struggle with accurately classifying POD obliteration from MRI images, which complicates the training of reliable AI models. In this paper, we introduce the Human-AI Collaborative Multi-modal Multi-rater Learning (HAICOMM) methodology to address the challenge above. HAICOMM is the first method that explores three important aspects of this problem: 1) multi-rater learning to extract a cleaner label from the multiple "noisy" labels available per training sample; 2) multi-modal learning to leverage the presence of T1/T2 MRI images for training and testing; and 3) human-AI collaboration to build a system that leverages the predictions from clinicians and the AI model to provide more accurate classification than standalone clinicians and AI models. Presenting results on the multi-rater T1/T2 MRI endometriosis dataset that we collected to validate our methodology, the proposed HAICOMM model outperforms an ensemble of clinicians, noisy-label learning models, and multi-rater learning methods.

cs.CV↗

arXiv2024

Enhancing Zeroth-order Fine-tuning for Language Models with Low-rank Structures

Parameter-efficient fine-tuning (PEFT) significantly reduces memory costs when adapting large language models (LLMs) for downstream applications. However, traditional first-order (FO) fine-tuning algorithms incur substantial memory overhead due to the need to store activation values for back-propagation during gradient computation, particularly in long-context fine-tuning tasks. Zeroth-order (ZO) algorithms offer a promising alternative by approximating gradients using finite differences of function values, thus eliminating the need for activation storage. Nevertheless, existing ZO methods struggle to capture the low-rank gradient structure common in LLM fine-tuning, leading to suboptimal performance. This paper proposes a low-rank ZO gradient estimator and introduces a novel low-rank ZO algorithm (LOZO) that effectively captures this structure in LLMs. We provide convergence guarantees for LOZO by framing it as a subspace optimization method. Additionally, its low-rank nature enables LOZO to integrate with momentum techniques while incurring negligible extra memory costs. Extensive experiments across various model sizes and downstream tasks demonstrate that LOZO and its momentum-based variant outperform existing ZO methods and closely approach the performance of FO algorithms.

cs.LG↗