arXiv subjects

Zeyu Chen

Publications and source records attributed to Zeyu Chen.

At least 55 records · Page 3Linked to original sources

arXiv2026

Lightweight Yet Secure: Secure Scripting Language Generation via Lightweight LLMs

The security of scripting languages such as PowerShell is critical given their powerful automation and administration capabilities, often exercised with elevated privileges. Today, securing these languages still demands substantial human effort to craft and enforce rules, imposing heavy burdens on typical administrators and creating critical production risks (e.g., misoperations that shut down servers).Large language models (LLMs) have demonstrated strong capabilities in code generation, vulnerability detection, and automated repair for languages like Python and JavaScript. However, their ability to assist with generating secure scripting-language code remains largely underexplored. In this paper, we present SecGenEval-PS, a benchmark designed to systematically evaluate LLMs on secure scripting generation, security analysis, and automated repair. Our results show that both proprietary and open-source models fall short in these areas. For instance, over 60% of PowerShell scripts produced by GPT-4o and o3-mini are insecure without structured guidance.To bridge this gap, we propose PSSec, a framework that combines data synthesis with fine-tuning to enhance model security capabilities. We develop a self-debugging agent that integrates static analyzers with the reasoning abilities of advanced LLMs to synthesize large-scale structured triplets of insecure scripts, violation analyses, and corresponding repairs. We then fine-tune lightweight LLMs (as small as 1.7B parameters) using supervised fine-tuning (SFT) and reinforcement learning (RL), enabling security-aware reasoning and the generation of secure PowerShell code.Across multiple LLM families, including GPT and Qwen, \textit{PSSec}-trained models match or surpass general-purpose large models on PowerShell security tasks while reducing inference cost by more than an order of magnitude.

cs.CR↗

arXiv2026

Sortblock: Similarity-Aware Feature Reuse for Diffusion Model

Diffusion Transformers (DiTs) have demonstrated remarkable generative capabilities, particularly benefiting from Transformer architectures that enhance visual and artistic fidelity. However, their inherently sequential denoising process results in high inference latency, limiting their deployment in real-time scenarios. Existing training-free acceleration approaches typically reuse intermediate features at fixed timesteps or layers, overlooking the evolving semantic focus across denoising stages and Transformer blocks.To address this, we propose Sortblock, a training-free inference acceleration framework that dynamically caches block-wise features based on their similarity across adjacent timesteps. By ranking the evolution of residuals, Sortblock adaptively determines a recomputation ratio, selectively skipping redundant computations while preserving generation quality. Furthermore, we incorporate a lightweight linear prediction mechanism to reduce accumulated errors in skipped blocks.Extensive experiments across various tasks and DiT architectures demonstrate that Sortblock achieves over 2$\times$ inference speedup with minimal degradation in output quality, offering an effective and generalizable solution for accelerating diffusion-based generative models.

cs.CV↗

arXiv2025

Personalized Pricing in Social Networks with Individual and Group Fairness Considerations

Personalized pricing assigns different prices to customers for the same product based on customer-specific features to improve retailer revenue. However, this practice often raises concerns about fairness at both the individual and group levels. At the individual level, a customer may perceive unfair treatment if he/she notices being charged a higher price than others. At the group level, pricing disparities can result in discrimination against certain protected groups, such as those defined by gender or race. Existing studies on fair pricing typically address individual and group fairness separately. This paper bridges the gap by introducing a new formulation of the personalized pricing problem that incorporates both dimensions of fairness in social network settings. To solve the problem, we propose FairPricing, a novel framework based on graph neural networks (GNNs) that learns a personalized pricing policy using customer features and network topology. In FairPricing, individual perceived unfairness is captured through a penalty on customer demand, and thus the profit objective, while group-level discrimination is mitigated using adversarial debiasing and a price regularization term. Unlike existing optimization-based personalized pricing, which requires re-optimization whenever the network updates, the pricing policy learned by FairPricing assigns personalized prices to all customers in an updated network based on their features and the new network structure, thereby generalizing to network changes. Extensive experimental results show that FairPricing achieves high profitability while improving individual fairness perceptions and satisfying group fairness requirements.

cs.CY↗

arXiv2025

LLMs as Firmware Experts: A Runtime-Grown Tree-of-Agents Framework

Large Language Models (LLMs) and their agent systems have recently demonstrated strong potential in automating code reasoning and vulnerability detection. However, when applied to large-scale firmware, their performance degrades due to the binary nature of firmware, complex dependency structures, and heterogeneous components. To address this challenge, this paper presents FIRMHIVE, a recursive agent hive that enables LLMs to act as autonomous firmware security analysts. FIRMHIVE introduces two key mechanisms: (1) transforming delegation into a per-agent, executable primitive and (2) constructing a runtime Tree of Agents (ToA) for decentralized coordination. We evaluate FIRMHIVE using real-world firmware images obtained from publicly available datasets, covering five representative security analysis tasks. Compared with existing LLM-agent baselines, FIRMHIVE performs deeper (about 16x more reasoning steps) and broader (about 2.3x more files inspected) cross-file exploration, resulting in about 5.6x more alerts per firmware. Compared to state-of-the-art (SOTA) security tools, FIRMHIVE identifies about 1.5x more vulnerabilities (1,802 total) and achieves 71% precision, representing significant improvements in both yield and fidelity.

cs.CR↗

arXiv2025

MonoGSDF: Exploring Monocular Geometric Cues for Gaussian Splatting-Guided Implicit Surface Reconstruction

Accurate meshing from monocular images remains a key challenge in 3D vision. While state-of-the-art 3D Gaussian Splatting (3DGS) methods excel at synthesizing photorealistic novel views through rasterization-based rendering, their reliance on sparse, explicit primitives severely limits their ability to recover watertight and topologically consistent 3D surfaces.We introduce MonoGSDF, a novel method that couples Gaussian-based primitives with a neural Signed Distance Field (SDF) for high-quality reconstruction. During training, the SDF guides Gaussians' spatial distribution, while at inference, Gaussians serve as priors to reconstruct surfaces, eliminating the need for memory-intensive Marching Cubes. To handle arbitrary-scale scenes, we propose a scaling strategy for robust generalization. A multi-resolution training scheme further refines details and monocular geometric cues from off-the-shelf estimators enhance reconstruction quality. Experiments on real-world datasets show MonoGSDF outperforms prior methods while maintaining efficiency.

cs.CV↗

arXiv2025

Variation Monte Carlo Study on the bilayer $t-J_{\parallel}-J_{\perp}$ model for La$_3$Ni$_2$O$_7$

The discovery of high-temperature superconductivity (HTSC) in La$_3$Ni$_2$O$_7$ has aroused significant interest in exploring the pairing mechanism. Previous studies have proposed an effective d$_{x^2-y^2}$-orbital bilayer $t-J_{\parallel}-J_{\perp}$ model, in which the electrons of the d$_{x^2-y^2}$ orbital are charge carriers, which are subject to the intralayer antiferromagnetic (AFM) superexchange $J_{\parallel}$ and the large interlayer AFM superexchange $J_{\perp}\approx 2J_{\parallel}$, with the latter transferred from the nearly half-filled and hence localized $d_{z^2}$ orbital through the strong Hund's rule coupling. Here we study this model by the variational Monte Carlo (VMC) simulation and find a dominant interlayer s-wave pairing, in which the SC order parameters have a drastic improvement compared with those of the mean field (MF) type of theories. In real materials, the Hund's coupling is finite, leading to reduced $J_{\perp}$, dictating that the MF-type theories have difficulty explaining the HTSC. However, our VMC calculations find that even for effective $J_{\perp}$ as weak as $J_{\perp}=J_{\parallel}$, the interlayer pairing is still considerably large and can be compared with the $T_c$ observed in experiments, which is very weak in MF-type theories. This result indicates the important role of the Gutzwiller projection in improving the $T_c$, which is ignored in the MF-type theories. In addition, our results show that suppressed interlayer hopping can promote interlayer pairing, which is consistent with the fact that the interlayer hopping of the d$_{x^2-y^2}$ orbital in La$_3$Ni$_2$O$_7$ is very weak. Our research offers a new perspective for understanding the pairing mechanism of bilayer nickelates and provides a reference for recent ultra-cold atom experiments in mixed-dimensional systems.

cond-mat.supr-con↗

arXiv2025

Flow of Knowledge: Federated Fine-Tuning of LLMs in Healthcare under Non-IID Conditions

Large language models (LLMs) show great promise in healthcare, but their applications are hindered by data privacy restrictions and the challenges of cross-institution collaboration. Sensitive medical data cannot be centralized, while non-independent and identically distributed (non-IID) characteristics across institutions further complicate convergence and fairness. To address these issues, we present a federated fine-tuning approach based on Low-Rank Adaptation (LoRA), enabling privacy-preserving knowledge flow across institutions. The method iteratively combines local LoRA adaptation with global parameter aggregation, allowing efficient knowledge sharing without exposing raw data. A blockchain identity scheme is used for identifying individual LLM in such a distributed network. We evaluate this approach on heterogeneous and highly non-IID medical text datasets, where experiments demonstrate that federated LoRA not only enhances cross-client generalization but also improves the performance of the weakest client, achieving stable convergence and fairer outcomes. These findings highlight federated LoRA fine-tuning as a practical and effective paradigm for adapting LLMs in healthcare, offering a new path for multi-center medical AI collaboration.

cs.CE↗

arXiv2025

Hierarchical Federated Learning for Social Network with Mobility

Federated Learning (FL) offers a decentralized solution that allows collaborative local model training and global aggregation, thereby protecting data privacy. In conventional FL frameworks, data privacy is typically preserved under the assumption that local data remains absolutely private, whereas the mobility of clients is frequently neglected in explicit modeling. In this paper, we propose a hierarchical federated learning framework based on the social network with mobility namely HFL-SNM that considers both data sharing among clients and their mobility patterns. Under the constraints of limited resources, we formulate a joint optimization problem of resource allocation and client scheduling, which objective is to minimize the energy consumption of clients during the FL process. In social network, we introduce the concepts of Effective Data Coverage Rate and Redundant Data Coverage Rate. We analyze the impact of effective data and redundant data on the model performance through preliminary experiments. We decouple the optimization problem into multiple sub-problems, analyze them based on preliminary experimental results, and propose Dynamic Optimization in Social Network with Mobility (DO-SNM) algorithm. Experimental results demonstrate that our algorithm achieves superior model performance while significantly reducing energy consumption, compared to traditional baseline algorithms.

cs.LG↗

arXiv2025

FedOC: Multi-Server FL with Overlapping Client Relays in Wireless Edge Networks

Multi-server Federated Learning (FL) has emerged as a promising solution to mitigate communication bottlenecks of single-server FL. We focus on a typical multi-server FL architecture, where the regions covered by different edge servers (ESs) may overlap. A key observation of this architecture is that clients located in the overlapping areas can access edge models from multiple ESs. Building on this insight, we propose FedOC (Federated learning with Overlapping Clients), a novel framework designed to fully exploit the potential of these overlapping clients. In FedOC, overlapping clients could serve dual roles: (1) as Relay Overlapping Clients (ROCs), they forward edge models between neighboring ESs in real time to facilitate model sharing among different ESs; and (2) as Normal Overlapping Clients (NOCs), they dynamically select their initial model for local training based on the edge model delivery time, which enables indirect data fusion among different regions of ESs. The overall FedOC workflow proceeds as follows: in every round, each client trains local model based on the earliest received edge model and transmits to the respective ESs for model aggregation. Then each ES transmits the aggregated edge model to neighboring ESs through ROC relaying. Upon receiving the relayed models, each ES performs a second aggregation and subsequently broadcasts the updated model to covered clients. The existence of ROCs enables the model of each ES to be disseminated to the other ESs in a decentralized manner, which indirectly achieves intercell model and speeding up the training process, making it well-suited for latency-sensitive edge environments. Extensive experimental results show remarkable performance gains of our scheme compared to existing methods.

cs.NI↗

arXiv2025

Prototype-Based Pseudo-Label Denoising for Source-Free Domain Adaptation in Remote Sensing Semantic Segmentation

Source-Free Domain Adaptation (SFDA) enables domain adaptation for semantic segmentation of Remote Sensing Images (RSIs) using only a well-trained source model and unlabeled target domain data. However, the lack of ground-truth labels in the target domain often leads to the generation of noisy pseudo-labels. Such noise impedes the effective mitigation of domain shift (DS). To address this challenge, we propose ProSFDA, a prototype-guided SFDA framework. It employs prototype-weighted pseudo-labels to facilitate reliable self-training (ST) under pseudo-labels noise. We, in addition, introduce a prototype-contrast strategy that encourages the aggregation of features belonging to the same class, enabling the model to learn discriminative target domain representations without relying on ground-truth supervision. Extensive experiments show that our approach substantially outperforms existing methods.

cs.CV↗

arXiv2025

Central Concentration and Escape of Ionizing Photons in Galaxies at the Epoch of Reionization

Compact, low-mass galaxies with strong nebular emission are considered promising candidates for efficient ionizing photon production and escape. We present a spatially resolved analysis of 189 galaxies at redshifts $z \sim 6.7-7.6$ in JADES GOODS-N and GOODS-S fields and selected via JWST/NIRCam F410M filter. By employing annular photometry and spectral energy distribution fitting across rest-frame UV to optical wavelengths, we investigate the internal structure of star formation, ionizing photon production and escape, as well as the resolved star formation histories within these galaxies. We find that these galaxies exhibit compact, centrally concentrated, and bursty star formation, especially in lower-mass systems ($\log(M_*/{\rm M_{\odot}}) <9.0$). The central regions of them display extreme [OIII]+H$β$ equivalent widths ($>$1000 Å), high ionizing photon production efficiencies ($ξ_{\text{ion}} \sim 10^{25.6}$ Hz erg$^{-1}$), steep UV slopes ($\sim -2.3$), and elevated escape fractions ($f_{\text{esc}} > 0.08$), with all these properties peaking in the inner regions. These findings reveal outside-in growth and rising star formation histories at $z\sim 7$, with the central regions of them playing a pivotal role in driving cosmic reionization.

astro-ph.GA↗

arXiv2025

PP-DocBee: Improving Multimodal Document Understanding Through a Bag of Tricks

With the rapid advancement of digitalization, various document images are being applied more extensively in production and daily life, and there is an increasingly urgent need for fast and accurate parsing of the content in document images. Therefore, this report presents PP-DocBee, a novel multimodal large language model designed for end-to-end document image understanding. First, we develop a data synthesis strategy tailored to document scenarios in which we build a diverse dataset to improve the model generalization. Then, we apply a few training techniques, including dynamic proportional sampling, data preprocessing, and OCR postprocessing strategies. Extensive evaluations demonstrate the superior performance of PP-DocBee, achieving state-of-the-art results on English document understanding benchmarks and even outperforming existing open source and commercial models in Chinese document understanding. The source code and pre-trained models are publicly available at \href{https://github.com/PaddlePaddle/PaddleMIX}{https://github.com/PaddlePaddle/PaddleMIX}.

cs.CV↗

arXiv2025

3D Gaussian Splatting for Fine-Detailed Surface Reconstruction in Large-Scale Scene

Recent developments in 3D Gaussian Splatting have made significant advances in surface reconstruction. However, scaling these methods to large-scale scenes remains challenging due to high computational demands and the complex dynamic appearances typical of outdoor environments. These challenges hinder the application in aerial surveying and autonomous driving. This paper proposes a novel solution to reconstruct large-scale surfaces with fine details, supervised by full-sized images. Firstly, we introduce a coarse-to-fine strategy to reconstruct a coarse model efficiently, followed by adaptive scene partitioning and sub-scene refining from image segments. Additionally, we integrate a decoupling appearance model to capture global appearance variations and a transient mask model to mitigate interference from moving objects. Finally, we expand the multi-view constraint and introduce a single-view regularization for texture-less areas. Our experiments were conducted on the publicly available dataset GauU-Scene V2, which was captured using unmanned aerial vehicles. To the best of our knowledge, our method outperforms existing NeRF-based and Gaussian-based methods, achieving high-fidelity visual results and accurate surface from full-size image optimization. Open-source code will be available on GitHub.

cs.GR↗

arXiv2025

The Cosmic Evolution and Spatial Distribution of Multiphase Gas associated with QSOs

We investigate the multi-phase gas surrounding QSOs traced by 33 absorption lines (e.g., Ly$α$, C\,\textsc{iv}, Fe\,\textsc{ii}, Mg\,\textsc{ii}, etc.) in the stacked spectra of background sources, using the early data release from the Dark Energy Spectroscopic Instrument. Our analysis reveals that the equivalent width ($ W $) of metal absorption lines decreases with increasing redshift, following an overall trend described by $W \propto (1+z)^{-4.0\pm 2.7}$. Different species that trace multi-phases of QSO-associated gas exhibit distinct evolutionary patterns. Additionally, the $ W $ of these absorption lines decreases with distance ($D$) from QSOs, which can be effectively characterized by a two-halo model. Compared to the projected two point correlation function of galaxies at similar redshifts, low-ionization ions exhibit similar clustering scales, while high-ionization ions show a significantly more extended spatial distribution. We also find that $W_{\text{FeII}}/W_{\text{MgII}}$ increases towards lower redshifts, which can be attributed to evolving star formation histories and/or changes in initial mass function for galaxies. By leveraging multiple absorption tracers, we conduct the first comprehensive investigation of diffuse, multiphase gas from the circumgalactic medium to cosmological scales, offering new insights into baryon cycles and the transport of metals throughout cosmic time.

astro-ph.GA↗

arXiv2025

KDMOS:Knowledge Distillation for Motion Segmentation

Motion Object Segmentation (MOS) is crucial for autonomous driving, as it enhances localization, path planning, map construction, scene flow estimation, and future state prediction. While existing methods achieve strong performance, balancing accuracy and real-time inference remains a challenge. To address this, we propose a logits-based knowledge distillation framework for MOS, aiming to improve accuracy while maintaining real-time efficiency. Specifically, we adopt a Bird's Eye View (BEV) projection-based model as the student and a non-projection model as the teacher. To handle the severe imbalance between moving and non-moving classes, we decouple them and apply tailored distillation strategies, allowing the teacher model to better learn key motion-related features. This approach significantly reduces false positives and false negatives. Additionally, we introduce dynamic upsampling, optimize the network architecture, and achieve a 7.69% reduction in parameter count, mitigating overfitting. Our method achieves a notable IoU of 78.8% on the hidden test set of the SemanticKITTI-MOS dataset and delivers competitive results on the Apollo dataset. The KDMOS implementation is available at https://github.com/SCNU-RISLAB/KDMOS.

cs.CV↗

arXiv2025

Visual Text Processing: A Comprehensive Review and Unified Evaluation

Visual text is a crucial component in both document and scene images, conveying rich semantic information and attracting significant attention in the computer vision community. Beyond traditional tasks such as text detection and recognition, visual text processing has witnessed rapid advancements driven by the emergence of foundation models, including text image reconstruction and text image manipulation. Despite significant progress, challenges remain due to the unique properties that differentiate text from general objects. Effectively capturing and leveraging these distinct textual characteristics is essential for developing robust visual text processing models. In this survey, we present a comprehensive, multi-perspective analysis of recent advancements in visual text processing, focusing on two key questions: (1) What textual features are most suitable for different visual text processing tasks? (2) How can these distinctive text features be effectively incorporated into processing frameworks? Furthermore, we introduce VTPBench, a new benchmark that encompasses a broad range of visual text processing datasets. Leveraging the advanced visual quality assessment capabilities of multimodal large language models (MLLMs), we propose VTPScore, a novel evaluation metric designed to ensure fair and reliable evaluation. Our empirical study with more than 20 specific models reveals substantial room for improvement in the current techniques. Our aim is to establish this work as a fundamental resource that fosters future exploration and innovation in the dynamic field of visual text processing. The relevant repository is available at https://github.com/shuyansy/Visual-Text-Processing-survey.

cs.CV↗

arXiv2025

Potential-Driven Metal Cycling: JADES Census of Gas-Phase Metallicity for galaxies at 1 < z < 7

The gravitational potential is established as a critical determinant of gas-phase metallicity (12+log(O/H)) in low-redshift galaxies, whereas its influence remains unconfirmed at high redshifts. We investigate the correlation between gas-phase metallicity and effective radius ($R_{\rm e}$) for a sample of galaxies with redshifts ranging from 1 to 7, drawn from JADES (JWST Advanced Deep Extragalactic Survey) Data Release 3. We calculate the metallicities using four strong-line methods: ${\rm N2S2Hα}$, ${\rm R23}$, ${\rm N2}$, and ${\rm O3N2}$, respectively. After taking out the evolution of size, we find that the offsets of mass-size relation ($Δ\log R_{\rm e}$) are significantly negatively correlated with the offset of mass-metallicity relation ($Δ\log({\rm O/H})$) for the four metallicity tracers. Regardless of the metallicity tracer used, we obtain Spearman rank $p-$values much less than 0.01, rejecting the null hypothesis that the observed correlation is statistically nonsignificant and attributable to random chance. This is also true for galaxies with $z>3$, with $p-$values less than 0.05 for the four metallicity tracers. We for the first time find evidence of size playing a key role in determining gas-phase metallicity towards cosmic dawn, suggesting that the gravitational potential influences their material-exchange processes with the surrounding environment at very early universe.

astro-ph.GA↗

arXiv2025

Enabling Versatile Controls for Video Diffusion Models

Despite substantial progress in text-to-video generation, achieving precise and flexible control over fine-grained spatiotemporal attributes remains a significant unresolved challenge in video generation research. To address these limitations, we introduce VCtrl (also termed PP-VCtrl), a novel framework designed to enable fine-grained control over pre-trained video diffusion models in a unified manner. VCtrl integrates diverse user-specified control signals-such as Canny edges, segmentation masks, and human keypoints-into pretrained video diffusion models via a generalizable conditional module capable of uniformly encoding multiple types of auxiliary signals without modifying the underlying generator. Additionally, we design a unified control signal encoding pipeline and a sparse residual connection mechanism to efficiently incorporate control representations. Comprehensive experiments and human evaluations demonstrate that VCtrl effectively enhances controllability and generation quality. The source code and pre-trained models are publicly available and implemented using the PaddlePaddle framework at http://github.com/PaddlePaddle/PaddleMIX/tree/develop/ppdiffusers/examples/ppvctrl.

cs.CV↗