arXiv subjects

Jie Zhang

Publications and source records attributed to Jie Zhang.

At least 379 records · Page 21Linked to original sources

arXiv2025

RFWNet: A Lightweight Remote Sensing Object Detector Integrating Multiscale Receptive Fields and Foreground Focus Mechanism

Challenges in remote sensing object detection(RSOD), such as high interclass similarity, imbalanced foreground-background distribution, and the small size of objects in remote sensing images, significantly hinder detection accuracy. Moreover, the tradeoff between model accuracy and computational complexity poses additional constraints on the application of RSOD algorithms. To address these issues, this study proposes an efficient and lightweight RSOD algorithm integrating multiscale receptive fields and foreground focus mechanism, named robust foreground weighted network(RFWNet). Specifically, we proposed a lightweight backbone network receptive field adaptive selection network (RFASNet), leveraging the rich context information of remote sensing images to enhance class separability. Additionally, we developed a foreground-background separation module(FBSM)consisting of a background redundant information filtering module (BRIFM) and a foreground information enhancement module (FIEM) to emphasize critical regions within images while filtering redundant background information. Finally, we designed a loss function, the weighted CIoU-Wasserstein loss (LWCW),which weights the IoU-based loss by using the normalized Wasserstein distance to mitigate model sensitivity to small object position deviations. The comprehensive experimental results demonstrate that RFWNet achieved 95.3% and 73.2% mean average precision (mAP) with 6.0 M parameters on the DOTA V1.0 and NWPU VHR-10 datasets, respectively, with an inference speed of 52 FPS.

cs.CV↗

arXiv2025

Generation of high power spatially-structured laser pulses via forward Raman amplification in plasma

Spatially-structured light with tunable intensity, wavelength, and spatiotemporal profiles has demonstrated significant potentials for fundamental and applied science, including the ultrafast and high-field physics. Nevertheless, the generation or amplification of such light towards extremely high power remains challenging due to the limitations of conventional gain media. Building upon our recently proposed forward Raman amplification (FRA) mechanism [Lei et al., Phys. Rev. Lett. 134, 255001 (2025)], here we develop a universal plasma-based amplification scheme that is capable of generating high-power structured laser beams, including vortex, Bessel, and Airy beams. Through theoretical modeling and multi-dimensional particle-in-cell simulations, we demonstrate that a near-infrared structured seed laser with an initial intensity of 1e12 W/cm2 can achieve 1e4~1e5-fold intensity amplification via FRA, and subsequently be self-compressed to sub-cycle duration with petawatt-level peak power. Benefiting from its exceptionally high amplification growth rate, the FRA process requires only femtosecond-scale interaction time and submillimeter propagation distance in plasma, effectively suppressing concomitant plasma instabilities. The high output intensity 1e17 W/cm2, compactness (<500 um), high temporal contrast, universal applicability to diverse structured beams, and relatively easy implementation with the co-propagating configuration combine to make the FRA a disruptive approach to the generation of petawatt-class spatially-structured light, enabling unprecedented applications in high-field physics and ultrafast science.

physics.plasm-ph↗

arXiv2025

Staggered nonlinear spin generations in centrosymmetric altermagnets under electric current

Current-induced spin generations are of significant importance for electrically controllable magnetization. Due to symmetry constraints, linear spin generation is absent in centrosymmetric magnets and nonlinear contributions become crucial. However, nonlinear spin generations have few examples in centrosymmetric compensated magnets with opposite-spin sublattices, which hinders electric control of associated magnetization. Here, we study nonlinear spin generations in altermagnets with opposite-spin sublattices. In a square altermagnetic model, both staggered and uniform nonlinear spin generations appear at opposite-spin sublattices. They vary as the magnetization direction rotates, with emerging out-of-plane components that can be utilized in perpendicular magnetization switching of high-density storage devices. By first-principles calculations, out-of-plane, staggered nonlinear spin generations are found to be considerable in a typical altermagnet, Fe$_2$Se$_2$O monolayer. Our findings provide opportunities for electrically manipulating magnetization and designing energy-efficient magnetic devices based on compensated magnets.

cond-mat.mes-hall↗

arXiv2025

Programming Geotechnical Reliability Algorithms using Generative AI

Programming reliability algorithms is crucial for risk assessment in geotechnical engineering. This study explores the possibility of automating and accelerating this task using Generative AI based on Large Language Models (LLMs). Specifically, the most popular LLM, i.e., ChatGPT, is used to test the ability to generate MATLAB codes for four classical reliability algorithms. The four specific examples considered in this study are: (1) First Order Reliability Method (FORM); (2) Subset simulation; (3) Random field simulation; and (4) Bayesian update using Gibbs sampling. The results obtained using the generated codes are compared with benchmark methods. It is found that the use of LLMs can be promising for generating reliability codes. Failure, limitations, and challenges of adopting LLMs are also discussed. Overall, this study demonstrates that existing LLMs can be leveraged powerfully and can contribute toward accelerating the adoption of reliability techniques in routine geotechnical engineering.

stat.AP↗

arXiv2025

FullLoRA: Efficiently Boosting the Robustness of Pretrained Vision Transformers

In recent years, the Vision Transformer (ViT) model has gradually become mainstream in various computer vision tasks, and the robustness of the model has received increasing attention. However, existing large models tend to prioritize performance during training, potentially neglecting the robustness, which may lead to serious security concerns. In this paper, we establish a new challenge: exploring how to use a small number of additional parameters for adversarial finetuning to quickly and effectively enhance the adversarial robustness of a standardly trained model. To address this challenge, we develop novel LNLoRA module, incorporating a learnable layer normalization before the conventional LoRA module, which helps mitigate magnitude differences in parameters between the adversarial and standard training paradigms. Furthermore, we propose the FullLoRA framework by integrating the learnable LNLoRA modules into all key components of ViT-based models while keeping the pretrained model frozen, which can significantly improve the model robustness via adversarial finetuning in a parameter-efficient manner. Extensive experiments on several datasets demonstrate the superiority of our proposed FullLoRA framework. It achieves comparable robustness with full finetuning while only requiring about 5\% of the learnable parameters. This also effectively addresses concerns regarding extra model storage space and enormous training time caused by adversarial finetuning.

cs.CV↗

arXiv2025

Wireless-Friendly Window Position Optimization for RIS-Aided Outdoor-to-Indoor Networks based on Multi-Modal Large Language Model

This paper aims to simultaneously optimize indoor wireless and daylight performance by adjusting the positions of windows and the beam directions of window-deployed reconfigurable intelligent surfaces (RISs) for RIS-aided outdoor-to-indoor (O2I) networks utilizing large language models (LLM) as optimizers. Firstly, we illustrate the wireless and daylight system models of RIS-aided O2I networks and formulate a joint optimization problem to enhance both wireless traffic sum rate and daylight illumination performance. Then, we present a multi-modal LLM-based window optimization (LMWO) framework, accompanied by a prompt construction template to optimize the overall performance in a zero-shot fashion, functioning as both an architect and a wireless network planner. Finally, we analyze the optimization performance of the LMWO framework and the impact of the number of windows, room size, number of RIS units, and daylight factor. Numerical results demonstrate that our proposed LMWO framework can achieve outstanding optimization performance in terms of initial performance, convergence speed, final outcomes, and time complexity, compared with classic optimization methods. The building's wireless performance can be significantly enhanced while ensuring indoor daylight performance.

cs.NI↗

arXiv2025

Jamming as a topological satisfiability transition with contact number hyperuniformity and criticality

The jamming transition between flow and amorphous-solid states exhibits paradoxical properties characterized by hyperuniformity (suppressed spatial fluctuations) and criticality (hyperfluctuations), whose origin remains unclear. Here we model the jamming transition by a topological satisfiability transition in a minimum network model with simultaneously hyperuniform distributions of contacts, diverging length scales and scale-free clusters. We show that these phenomena stem from isostaticity and mechanical stability: the former imposes a global equality, and the latter local inequalities on arbitrary sub-systems. This dual constraint bounds contact number fluctuations from both above and below, limiting them to scale with the surface area. The hyperuniform and critical exponents of the network model align with those of frictionless jamming, suggesting a new universality class of non-equilibrium phase transitions. Our results provide a minimal, dynamics-independent framework for jamming criticality and hyperuniformity in disordered systems.

cond-mat.soft↗

arXiv2025

Stress-stress correlations in two-dimensional amorphous and crystalline solids

Stress-stress correlations in crystalline solids with long-range order can be straightforwardly derived using elasticity theory. In contrast, the `emergent elasticity' of amorphous solids, rigid materials characterized by an underlying disordered structure, defies direct explanation within traditional theoretical frameworks. To address this challenge, tensor gauge theories have been recently proposed as a promising approach to describe the emergent elasticity of disordered solids and predict their stress-stress correlations. In this work, we revisit this problem in two-dimensional amorphous and crystalline solids by employing a canonical elasticity theory approach, supported by experimental and simulation data. We demonstrate that, with respect to static stress-stress correlations, the response of a 2D disordered solid is indistinguishable from that of a 2D isotropic crystalline solid and it is well predicted by vanilla elasticity theory. Moreover, we show that the presence of pinch-point singularities in the stress response is not an exclusive feature of amorphous solids. Our results confirm previous observations about the universal character of static stress-stress correlations in crystalline and amorphous packings.

cond-mat.soft↗

arXiv2025

Textual Unlearning Gives a False Sense of Unlearning

Language Models (LMs) are prone to ''memorizing'' training data, including substantial sensitive user information. To mitigate privacy risks and safeguard the right to be forgotten, machine unlearning has emerged as a promising approach for enabling LMs to efficiently ''forget'' specific texts. However, despite the good intentions, is textual unlearning really as effective and reliable as expected? To address the concern, we first propose Unlearning Likelihood Ratio Attack+ (U-LiRA+), a rigorous textual unlearning auditing method, and find that unlearned texts can still be detected with very high confidence after unlearning. Further, we conduct an in-depth investigation on the privacy risks of textual unlearning mechanisms in deployment and present the Textual Unlearning Leakage Attack (TULA), along with its variants in both black- and white-box scenarios. We show that textual unlearning mechanisms could instead reveal more about the unlearned texts, exposing them to significant membership inference and data reconstruction risks. Our findings highlight that existing textual unlearning actually gives a false sense of unlearning, underscoring the need for more robust and secure unlearning mechanisms.

cs.CR↗

arXiv2025

A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deployment

The remarkable success of Large Language Models (LLMs) has illuminated a promising pathway toward achieving Artificial General Intelligence for both academic and industrial communities, owing to their unprecedented performance across various applications. As LLMs continue to gain prominence in both research and commercial domains, their security and safety implications have become a growing concern, not only for researchers and corporations but also for every nation. Currently, existing surveys on LLM safety primarily focus on specific stages of the LLM lifecycle, e.g., deployment phase or fine-tuning phase, lacking a comprehensive understanding of the entire "lifechain" of LLMs. To address this gap, this paper introduces, for the first time, the concept of "full-stack" safety to systematically consider safety issues throughout the entire process of LLM training, deployment, and eventual commercialization. Compared to the off-the-shelf LLM safety surveys, our work demonstrates several distinctive advantages: (I) Comprehensive Perspective. We define the complete LLM lifecycle as encompassing data preparation, pre-training, post-training, deployment and final commercialization. To our knowledge, this represents the first safety survey to encompass the entire lifecycle of LLMs. (II) Extensive Literature Support. Our research is grounded in an exhaustive review of over 800+ papers, ensuring comprehensive coverage and systematic organization of security issues within a more holistic understanding. (III) Unique Insights. Through systematic literature analysis, we have developed reliable roadmaps and perspectives for each chapter. Our work identifies promising research directions, including safety in data generation, alignment techniques, model editing, and LLM-based agent systems. These insights provide valuable guidance for researchers pursuing future work in this field.

cs.CR↗

arXiv2025

DTVM: Revolutionizing Smart Contract Execution with Determinism and Compatibility

We introduce the DeTerministic Virtual Machine (DTVM) Stack, a next-generation smart contract execution framework designed to address critical performance, determinism, and ecosystem compatibility challenges in blockchain networks. Building upon WebAssembly (Wasm) while maintaining full Ethereum Virtual Machine (EVM) ABI compatibility, DTVM introduces a Deterministic Middle Intermediate Representation (dMIR) and a hybrid lazy-JIT compilation engine to balance compilation speed and execution efficiency. DTVM further accommodates diverse instruction set architectures (e.g., EVM, RISC-V) through modular adaptation layers. This enables seamless integration with DTVM's hybrid lazy-JIT compilation engine, which dynamically optimizes performance while preserving deterministic execution guarantees across heterogeneous environments. The key contributions including: 1). The framework achieves up to 2$\times$ acceleration over evmone in dominant Ethereum contract (e.g. ERC20/721/1155) execution and reduces fibonacci computation latency by 11.8$\sim$40.5% compared to Wasm based VMs. 2). A novel trampoline hot-switch mechanism enables sub-millisecond (0.95ms) post-deployment invocation times, outperforming up to about 23$\times$ in compilation and invocation efficiency. 3). It supports multi-language development (Solidity, C++, Rust, Java, Go, and AssemblyScript) through unified bytecode conversion while maintaining EVM ABI compatibility for seamless invocation. It reduces machine code object sizes by 30.0$\sim$72.6%, coupled with a minimized Trusted Computing Base. 4). It offers SmartCogent, an AI-driven full-stack development experience, leveraging fine-tuned LLMs and retrieval-augmented generation to automate tasks across the smart contract lifecycle: development, debugging, security auditing, and deployment. DTVM Stack has been open-sourced (https://github.com/DTVMStack).

cs.DC↗

arXiv2025

Membership Inference Attacks on Sequence Models

Sequence models, such as Large Language Models (LLMs) and autoregressive image generators, have a tendency to memorize and inadvertently leak sensitive information. While this tendency has critical legal implications, existing tools are insufficient to audit the resulting risks. We hypothesize that those tools' shortcomings are due to mismatched assumptions. Thus, we argue that effectively measuring privacy leakage in sequence models requires leveraging the correlations inherent in sequential generation. To illustrate this, we adapt a state-of-the-art membership inference attack to explicitly model within-sequence correlations, thereby demonstrating how a strong existing attack can be naturally extended to suit the structure of sequence models. Through a case study, we show that our adaptations consistently improve the effectiveness of memorization audits without introducing additional computational costs. Our work hence serves as an important stepping stone toward reliable memorization audits for large sequence models.

cs.CR↗

arXiv2025

The Tug of War Within: Mitigating the Fairness-Privacy Conflicts in Large Language Models

Ensuring awareness of fairness and privacy in Large Language Models (LLMs) is critical. Interestingly, we discover a counter-intuitive trade-off phenomenon that enhancing an LLM's privacy awareness through Supervised Fine-Tuning (SFT) methods significantly decreases its fairness awareness with thousands of samples. To address this issue, inspired by the information theory, we introduce a training-free method to \textbf{S}uppress the \textbf{P}rivacy and fa\textbf{I}rness coupled \textbf{N}eurons (\textbf{SPIN}), which theoretically and empirically decrease the mutual information between fairness and privacy awareness. Extensive experimental results demonstrate that SPIN eliminates the trade-off phenomenon and significantly improves LLMs' fairness and privacy awareness simultaneously without compromising general capabilities, \eg improving Qwen-2-7B-Instruct's fairness awareness by 12.2\% and privacy awareness by 14.0\%. More crucially, SPIN remains robust and effective with limited annotated data or even when only malicious fine-tuning data is available, whereas SFT methods may fail to perform properly in such scenarios. Furthermore, we show that SPIN could generalize to other potential trade-off dimensions. We hope this study provides valuable insights into concurrently addressing fairness and privacy concerns in LLMs and can be integrated into comprehensive frameworks to develop more ethical and responsible AI systems. Our code is available at https://github.com/ChnQ/SPIN.

cs.AI↗

arXiv2025

Light as Deception: GPT-driven Natural Relighting Against Vision-Language Pre-training Models

While adversarial attacks on vision-and-language pretraining (VLP) models have been explored, generating natural adversarial samples crafted through realistic and semantically meaningful perturbations remains an open challenge. Existing methods, primarily designed for classification tasks, struggle when adapted to VLP models due to their restricted optimization spaces, leading to ineffective attacks or unnatural artifacts. To address this, we propose \textbf{LightD}, a novel framework that generates natural adversarial samples for VLP models via semantically guided relighting. Specifically, LightD leverages ChatGPT to propose context-aware initial lighting parameters and integrates a pretrained relighting model (IC-light) to enable diverse lighting adjustments. LightD expands the optimization space while ensuring perturbations align with scene semantics. Additionally, gradient-based optimization is applied to the reference lighting image to further enhance attack effectiveness while maintaining visual naturalness. The effectiveness and superiority of the proposed LightD have been demonstrated across various VLP models in tasks such as image captioning and visual question answering.

cs.CV↗

arXiv2025

A Composite Predictive-Generative Approach to Monaural Universal Speech Enhancement

It is promising to design a single model that can suppress various distortions and improve speech quality, i.e., universal speech enhancement (USE). Compared to supervised learning-based predictive methods, diffusion-based generative models have shown greater potential due to the generative capacities from degraded speech with severely damaged information. However, artifacts may be introduced in highly adverse conditions, and diffusion models often suffer from a heavy computational burden due to many steps for inference. In order to jointly leverage the superiority of prediction and generation and overcome the respective defects, in this work we propose a universal speech enhancement model called PGUSE by combining predictive and generative modeling. Our model consists of two branches: the predictive branch directly predicts clean samples from degraded signals, while the generative branch optimizes the denoising objective of diffusion models. We utilize the output fusion and truncated diffusion scheme to effectively integrate predictive and generative modeling, where the former directly combines results from both branches and the latter modifies the reverse diffusion process with initial estimates from the predictive branch. Extensive experiments on several datasets verify the superiority of the proposed model over state-of-the-art baselines, demonstrating the complementarity and benefits of combining predictive and generative modeling.

eess.AS↗

arXiv2025

On the Distortion of Multi-winner Election Using Single-Candidate Ballots

In this paper, we study the distortion bounds for voting mechanisms in multi-winner elections in general metric spaces. Our study pertains to the case in which each voter only reports her favorite candidate amongst $m$ possible choices. Given that candidates' locations are undisclosed to the mechanism, the mechanism has to form a $w-$winner committee based solely on the number of votes received by candidates. We establish distortion bounds for both truthful and non-truthful mechanisms. Our research highlights the significance of the $σ$ parameter, which represents the ratio between maximum and minimum distances among all candidate pairs. We show that the distortion is linear in $σ$. First, we demonstrate that all mechanisms possess a distortion greater than $1+\frac{w-1}{w+1}(σ-1)$. To give an upper bound, we study the Single Non-Transferable Vote (SNTV) mechanism, whose distortion is at most $1+2σ$. Second, we retrieve the upper bounds for strategyproof mechanisms. In particular, we infer an upper bound by examining the Random Sequential Dictator mechanism that achieves a distortion less than $1+4σ$ when $w=2$.

cs.GT↗

arXiv2025

Jailbreaking Prompt Attack: A Controllable Adversarial Attack against Diffusion Models

Text-to-image (T2I) models can be maliciously used to generate harmful content such as sexually explicit, unfaithful, and misleading or Not-Safe-for-Work (NSFW) images. Previous attacks largely depend on the availability of the diffusion model or involve a lengthy optimization process. In this work, we investigate a more practical and universal attack that does not require the presence of a target model and demonstrate that the high-dimensional text embedding space inherently contains NSFW concepts that can be exploited to generate harmful images. We present the Jailbreaking Prompt Attack (JPA). JPA first searches for the target malicious concepts in the text embedding space using a group of antonyms generated by ChatGPT. Subsequently, a prefix prompt is optimized in the discrete vocabulary space to align malicious concepts semantically in the text embedding space. We further introduce a soft assignment with gradient masking technique that allows us to perform gradient ascent in the discrete vocabulary space. We perform extensive experiments with open-sourced T2I models, e.g. stable-diffusion-v1-4 and closed-sourced online services, e.g. DALLE2, Midjourney with black-box safety checkers. Results show that (1) JPA bypasses both text and image safety checkers (2) while preserving high semantic alignment with the target prompt. (3) JPA demonstrates a much faster speed than previous methods and can be executed in a fully automated manner. These merits render it a valuable tool for robustness evaluation in future text-to-image generation research.

cs.CR↗

arXiv2025

Distilling Textual Priors from LLM to Efficient Image Fusion

Multi-modality image fusion aims to synthesize a single, comprehensive image from multiple source inputs. Traditional approaches, such as CNNs and GANs, offer efficiency but struggle to handle low-quality or complex inputs. Recent advances in text-guided methods leverage large model priors to overcome these limitations, but at the cost of significant computational overhead, both in memory and inference time. To address this challenge, we propose a novel framework for distilling large model priors, eliminating the need for text guidance during inference while dramatically reducing model size. Our framework utilizes a teacher-student architecture, where the teacher network incorporates large model priors and transfers this knowledge to a smaller student network via a tailored distillation process. Additionally, we introduce spatial-channel cross-fusion module to enhance the model's ability to leverage textual priors across both spatial and channel dimensions. Our method achieves a favorable trade-off between computational efficiency and fusion quality. The distilled network, requiring only 10% of the parameters and inference time of the teacher network, retains 90% of its performance and outperforms existing SOTA methods. Extensive experiments demonstrate the effectiveness of our approach. The implementation will be made publicly available as an open-source resource.

cs.CV↗