March 2025 arXiv papers — page 176

Hiroshi Nakano, Krzysztof Blachut, Kamil Jeziorek, Piotr Wzorek

As the quantities of data recorded by embedded edge sensors grow, so too does the need for intelligent local processing. Such data often comes in the form of time-series signals, based on which real-time predictions can be made locally using an AI model. However, a hardware-software approach capable of making low-latency predictions with low power consumptio

Artem Vasin, Valery Krivchenko, Dmitry Kovalev, Fedyor Stonyakin

First-order methods for minimization and saddle point (min-max) problems are widely used for solving large-scale problems, in particular arising in machine learning. The majority of works obtain favorable complexity guarantees of such methods, assuming that exact gradient information is available. At the same time, even the use of floating-point representati

Jinmyeong An, Sangwon Ryu, Heejin Do, Yunsu Kim

Online grooming is a severe social threat where sexual predators gradually entrap child victims with subtle and gradual manipulation. Therefore, timely intervention for online grooming is critical for proactive protection. However, previous methods fail to determine the optimal intervention points (i.e., jump to conclusions) as they rely on chat-level risk l

Hasan Abed Al Kader Hammoud, Bernard Ghanem

We propose DiffCLIP, a novel vision-language model that extends the differential attention mechanism to CLIP architectures. Differential attention was originally developed for large language models to amplify relevant context while canceling out noisy information. In this work, we integrate this mechanism into CLIP's dual encoder (image and text) framework.

Chaocan Xue, Bineng Zhong, Qihua Liang, Yaozong Zheng

Vision transformers (ViTs) have emerged as a popular backbone for visual tracking. However, complete ViT architectures are too cumbersome to deploy for unmanned aerial vehicle (UAV) tracking which extremely emphasizes efficiency. In this study, we discover that many layers within lightweight ViT-based trackers tend to learn relatively redundant and repetitiv

Xingming Liao, Meiyu Zeng, Canyu Chen, Nankai Lin

The proliferation of AI-Generated Content (AIGC), especially deepfake videos, poses a severe threat to social trust by enabling fraud, privacy violations and disinformation. Existing AI-generated video detection (AGVD) benchmarks focus on open-source model generated videos, yet commercial closed-source models produce more realistic, temporally coherent video

Sijie Zhao, Feng Liu, Xueliang Zhang, Hao Chen

The increasing impact of climate change and extreme weather events has spurred growing interest in deep learning for weather research. However, existing studies often rely on weather data in pixel space, which presents several challenges such as smooth outputs in model outputs, limited applicability to a single pressure-variable subset (PVS), and high data s

Peter K. Friz, Khoa Le, Huilin Zhang

Rough stochastic differential equations (RSDEs) are common generalisations of Ito SDEs and Lyons RDEs and have emerged as new tool in several areas of applied probability, including non-linear stochastic filtering, pathwise stochastic optimal control, volatility modelling in finance and mean-fields analysis of common noise system. We here take a unified pers

Xiaohai Li, Bineng Zhong, Qihua Liang, Zhiyi Mo

The consistency between the semantic information provided by the multi-modal reference and the tracked object is crucial for visual-language (VL) tracking. However, existing VL tracking frameworks rely on static multi-modal references to locate dynamic objects, which can lead to semantic discrepancies and reduce the robustness of the tracker. To address this

Nachiket U. Bapat, Randy C. Paffenroth, Raghvendra V. Cowlagi

We study the problem of synthetic generation of samples of environmental features for autonomous vehicle navigation. These features are described by a spatiotemporally varying scalar field that we refer to as a threat field. The threat field is known to have some underlying dynamics subject to process noise. Some "real-world" data of observations of various

Jiusi Yu, Haitao Li, Shijie Kang, Dongyi Wang

Metasurfaces have offered unprecedented control over electromagnetic (EM) waves across a wide range of frequency spectrum by manipulating their phase, amplitude, and polarization at subwavelength scales. Full wavefront control using metasurfaces requires 2{\pi} phase modulation, which is essential for advanced optical and photonic engineering. Common approac

Long Peng, Anran Wu, Wenbo Li, Peizhe Xia

Arbitrary-scale super-resolution (ASSR) aims to reconstruct high-resolution (HR) images from low-resolution (LR) inputs with arbitrary upsampling factors using a single model, addressing the limitations of traditional SR methods constrained to fixed-scale factors (\textit{e.g.}, $\times$ 2). Recent advances leveraging implicit neural representation (INR) hav

Pengxiang Xue, Yuankui Ma, Taekyun Kim, Dae San Kim

Let Y be a random variable whose moment generating function exists in a neighborhood of the origin. The aim of this paper is to study the probabilistic degenerate poly-Bell polynomials associated with the random variable Y, arising from the degenerate polyexponential functions, which are probabilistic extensions of degenerate versions of the poly-Bell polyno

Xiangdi Fu, Kunyu Guo, Dilong Li

In this paper, we investigate contractive projections, conditional expectations, and idempotent coefficient multipliers on the Hardy spaces $H^p(\mathbb{T})$ for $0<p<1$. For such values of $p$, we first establish a general extension theorem for contractive projections in a probability $L^p$-space. Combining this theorem with the study of conditional expecta

Qian Zeng, Xin Lin, Jingyi Gao, Yang Yu

Previous studies have demonstrated the strong performance of Graph Neural Networks (GNNs) in node classification. However, most existing GNNs adopt a node-centric perspective and rely on global message passing, leading to high computational and memory costs that hinder scalability. To mitigate these challenges, subgraph-based methods have been introduced, le

Kousuke Ishihara, Fumitaka Nakamura, Patricio Sanhueza, Masao Saito

Stars form from dense cores in turbulent molecular clouds. According to the standard scenario of star formation, dense cores are created by cloud fragmentation. However, the physical mechanisms driving this process are still not fully understood from an observational standpoint. However, the physical mechanisms driving this process are still not fully unders

Junyao Peng

We study special $\mathbb{G}_m$-equivariant degenerations of a smooth del Pezzo surface $X$ induced by valuations that are log canonical places of $(X,C)$ for a nodal anti-canonical curve $C$. We show that the space of special valuations in the dual complex of $(X,C)$ is connected and admits a locally finite partition into sub-intervals, each associated to a

Alexandra E. Ballentine, Raghvendra V. Cowlagi

Performance and reliability analyses of autonomous vehicles (AVs) can benefit from tools that ``amplify'' small datasets to synthesize larger volumes of plausible samples of the AV's behavior. We consider a specific instance of this data synthesis problem that addresses minimizing the AV's exposure to adverse environmental conditions during travel to a fixed

Guilherme N. C. Amaral, Mahmoud Sedahmed, Margarida M. Telo da Gama, Rodrigo C. V. Coelho

Simulations of nematohydrodynamics on graphics processing units (GPUs) are typically performed using double precision, which ensures accuracy but significantly increases computational cost. However, consumer-grade GPUs are optimized for single-precision calculations, making double-precision simulations inefficient on widely available hardware. In this work,

Nhat A. Nghiem

It is shown that quantum computer can detect the existence of root of a function almost exponentially more efficient than the classical counterpart. It is also shown that a quantum computer can produce quantum state corresponding to the solution of nonlinear algebraic equations quadratically faster than the best known classical approach. Various applications

Ruchi Bhatt, Shreya Bansal, Amanpreet Chander, Rupinder Kaur

Understanding plant growth dynamics is essential for applications in agriculture and plant phenotyping. We present the Growth Modelling (GroMo) challenge, which is designed for two primary tasks: (1) plant age prediction and (2) leaf count estimation, both essential for crop monitoring and precision agriculture. For this challenge, we introduce GroMo25, a da

Pranoy Panda, Kancheti Sai Srinivas, Vineeth N Balasubramanian, Gaurav Sinha

Data in the real world often has an evolving distribution. Thus, machine learning models trained on such data get outdated over time. This phenomenon is called model drift. Knowledge of this drift serves two purposes: (i) Retain an accurate model and (ii) Discovery of knowledge or insights about change in the relationship between input features and output va

Peigen Cao

$F$-invariant for a pair of good elements (e.g. cluster monomials) in cluster algebras is introduced by the author in a previous work. A key feature of $F$-invariant is that it is a coordinate-free invariant, that is, it is mutation invariant under the initial seed mutations. $E$-invariant for a pair of decorated representations of quivers with potentials is

Renhao Lu

Semantic segmentation is a core task in computer vision with applications in biomedical imaging, remote sensing, and autonomous driving. While standard loss functions such as cross-entropy and Dice loss perform well in general cases, they often struggle with fine structures, particularly in tasks involving thin structures or closely packed objects. Various w

Ting Zhang, Hai-Chuan Xu, Wei-Xing Zhou

We investigate the extreme return connectedness between the food, fossil energy, and clean energy markets using the quantile connectedness approach, which combines the traditional spillover index with quantile regression. Our results show that return connectedness at the tails (57.91% for the right tail and 61.47% for the left tail) is significantly higher t

Pranoy Panda, Siddharth Tandon, Vineeth N Balasubramanian

Fair credit assignment is essential in various machine learning (ML) applications, and Shapley values have emerged as a valuable tool for this purpose. However, in critical ML applications such as data valuation and feature attribution, the uniform weighting of Shapley values across subset cardinalities leads to unintuitive credit assignments. To address thi

Yanqing Shen, Sanping Zhou, Jingwen Fu, Ruotong Wang

Visual place recognition is a challenging task for autonomous driving and robotics, which is usually considered as an image retrieval problem. A commonly used two-stage strategy involves global retrieval followed by re-ranking using patch-level descriptors. Most deep learning-based methods in an end-to-end manner cannot extract global features with sufficien

Stephen D. Cohen, Peter V. Danchev, Tomás Oliveira e Silva

We classify those finite fields $\mathbb{F}_q$, for $q$ a power of some fixed prime number, whose members are the sum of an $n$-potent element with $n>1$ and a 4-potent element. It is shown that there are precisely ten non-trivial pairs $(q,n)$ for which this is the case. This continues a recent publication by Cohen-Danchev et al. in Turk. J. Math. (2024) in

Ting Zhang, Peng-Fei Li, Wei-Xing Zhou

The study examines the return connectedness between climate policy uncertainty (CPU), clean energy, fossil energy, and food markets. Using the time-domain method of Diebold and Yilmaz (2012) and frequency-domain methods of Barun{\'{i}}k and K{\v{r}}hl{\'{i}}k (2018), we find substantial spillover effects between these markets. Furthermore, high frequency dom

Hao Xu, Tengfei Xue, Dongnan Liu, Yuqian Chen

3D neuroimages provide a comprehensive view of brain structure and function, aiding in precise localization and functional connectivity analysis. Segmentation of white matter (WM) tracts using 3D neuroimages is vital for understanding the brain's structural connectivity in both healthy and diseased states. One-shot Class Incremental Semantic Segmentation (OC

Souradeep Ghosh, L. Yashvanth, Chandra R. Murthy

Intelligent reflecting surfaces (IRSs) are envisioned to enhance the performance of mmWave wireless systems. In practice, multiple mobile operators (MO) coexist in an area and provide simultaneous and independent services to user-equipments (UEs) on different frequency bands. Then, if each MO deploys an IRS to enhance its performance, the IRSs also alter the

Yu-Jia Gao, La-Tai You, Gong-Ping Zheng

A self-bound dipolar quantum droplet modulated by an shallow optical latticeare studied. It is found that the behaviors of the droplets in an optical lattice are similar with those without the optical lattice. In the shallow enough limit the properties of the periodically-modulated dipolar droplets are exactly the same with those without the optical lattice,

Yingfeng Luo, Tong Zheng, Yongyu Mu, Bei Li

The field of neural machine translation (NMT) has changed with the advent of large language models (LLMs). Much of the recent emphasis in natural language processing (NLP) has been on modeling machine translation and many other problems using a single pre-trained Transformer decoder, while encoder-decoder architectures, which were the standard in earlier NMT

Yun Wang, Minxing Li, Ping He

Both simulations and observations indicate that the so-called missing baryons reside in the intergalactic medium known as the warm-hot intergalactic medium (WHIM). In this study we employed the IllustrisTNG50-1 simulation to demonstrate that knowledge of the turbulence in the cosmic baryonic fluid is crucial for correctly understanding both the spatial distr

Banglong Liu, Niuniu Qi, Xia Zeng, Lydia Dehbi

Polynomial inequality proving is fundamental to many mathematical disciplines and finds wide applications in diverse fields. Current traditional algebraic methods are based on searching for a polynomial positive definite representation over a set of basis. However, these methods are limited by truncation degree. To address this issue, this paper proposes an

Xianglai Yuan, Yichao Yao, Han Wu, Minyu Feng

Since Granell et al. proposed a multiplex network for information and epidemic propagation, researchers have explored how information propagation affects epidemic dynamics. However, the role of individuals acquiring information through physical interactions has received relatively less attention. In this work, we introduce a novel source of information: phys

Harukuni Ikeda

We show that norm-conserving spin models driven by temporally hyperuniform noise exhibit a sharp ergodicity-breaking transition in the absence of interactions. In the nonergodic phase, the dynamics freeze into configurations determined by the initial condition. Our analysis demonstrates that such interaction-free ergodicity breaking arises generically whenev

Dmitri Gal'tsov, Rostom Karsanov

We perform full integration of the stationary axisymmetric Einstein-Maxwell-dilaton-axion (EMDA) theory with and without potential using a recently proposed generalization of Carter's approach to spacetimes beyond type D, allowing the Killing tensor. Crucial to our construction is a new parametrization of the dilaton and axion fields based on the analyticity

Yaxuan Li, Han Jiang, Yifei Ma, Shihua Qin

Dynamic Magnetic Resonance Imaging (MRI) of the vocal tract has become an increasingly adopted imaging modality for speech motor studies. Beyond image signals, systematic data loss, noise pollution, and audio file corruption can occur due to the unpredictability of the MRI acquisition environment. In such cases, generating audio from images is critical for d

Yixin Yang, Yang Zhou, Hui Huang

Recently, 2D Gaussian Splatting (2DGS) has demonstrated superior geometry reconstruction quality than the popular 3DGS by using 2D surfels to approximate thin surfaces. However, it falls short when dealing with glossy surfaces, resulting in visible holes in these areas. We find that the reflection discontinuity causes the issue. To fit the jump from diffuse

Christopher J. N. Coveney, David P. Tew

We present the diagrammatic theory of the irreducible self-energy and Bethe-Salpeter kernel that naturally arises within the Green's function formalism for a general $N$-body non-hermitian interaction. In this work, we focus specifically on the coupled-cluster self-energy generated by the similarity transformation of the electronic structure Hamiltonian. We

Diogo da Silva Machado

We consider one dimensional holomorphic foliations with isolated singularities that leave invariant a local complete intersection. We establish explicit formulas for the total GSV index of such foliations and obtain bounds for this index. As applications, we derive several consequences related to Poincar\'e's problem for foliations on projective spaces.

Yijie Mo, Xiao-Jiao Wang, Zhongbo Yan

Odd-parity pairings offer a natural pathway for realizing topological superconductivity. When two identical even-parity superconductors form a $\pi$-junction, the metallic material sandwiched between them experiences an effective odd-parity pairing, facilitating the emergence of topological superconductivity in the intermediate region. In this work, we consi

Andronick Arutyunov, Oleg Muravev

We construct a description of graded derivations in group algebras. Using this result for arbitrary graduation of the group algebra, we describe all possible structures of DG algebras. The corresponding examples are given. The description is given in terms of characters on a groupoid analogous to the groupoid of inner action.

Xuyao Zhang, Milan Ilić, Beat Signer

Dynamic data physicalisation is an emerging field of research, investigating the representation and exploration of data via multiple modalities, beyond traditional visual methods. Despite the development of various data physicalisation applications in recent years, the integration of diverse hardware components remains both time-consuming and costly. Further

Tiffany Ding, Dominique Perrault-Joncas, Orit Ronen, Michael I. Jordan

As e-commerce marketplaces continue to grow in popularity, it has become increasingly important to understand the role and impact of marketplace operators on competition and social welfare. We model a marketplace operator as an entity that not only facilitates third-party sales but can also choose to directly participate in the market as a competing seller.

Xiaodong Liu, Qingxiang Shi

This work is dedicated to a novel sampling method for accurately reconstructing elastic and electromagnetic sources from the far field patterns. We show that the proposed indicators in the form of integrals with full far field patterns are exactly the source functions. These facts not only give constructive uniqueness proofs of the inverse source problems, b

Yuxiang Zhang, Yuqi Yang, Jiangming Shu, Xinyan Wen

Traditional agentic workflows rely on external prompts to manage interactions with tools and the environment, which limits the autonomy of reasoning models. We position \emph{Large Agent Models (LAMs)} that internalize the generation of \emph{Chain-of-Action (CoA)}, enabling the model to autonomously decide when and how to use external tools. Our proposed Au

George Chacko, Minhyuk Park, Vikram Ramavarapu, Ananth Grama

Whether citations can be objectively and reliably used to measure productivity and scientific quality of articles and researchers can, and should, be vigorously questioned. However, citations are widely used to estimate the productivity of researchers and institutions, effectively creating a 'grubby' motivation to be well-cited. We model citation growth, and

Julia Ramos González, Enrico Vitale

In the first part of the paper, we establish an homotopical version of the snail lemma (which is a generalization of the classical snake lemma). In the second part, we introduce the category $\mathbf{Seq}(\mathcal A)$ of sequentiable families of arrows in a category $\mathcal A$ and we compare it with the category of chain complexes in $\mathcal A.$ We apply

Fan Fei, Md Shumon Mia, Ahmed E. Elbanna, Jinhyun Choo

Computational modeling of faulting processes is an essential tool for understanding earthquake mechanics but remains challenging due to the structural and material complexities of fault zones. The phase-field method has recently enabled unified modeling of fault propagation and off-fault damage; however, its capability has been restricted to simplified anti-

Zhenglu Jiang, Zanchen Zhuo

A new potential is presented for spherical galaxies. The technique of the construction of our model is similar to that given by An and Evans. In a special case, its mass density becomes a special one of the Hernquist model. Another special model is primarily discussed, and its intrinsic properties, such as velocity dispersions and surface densities, can be s

Farzam Tajdari, Amin Rezasoltani

Recent research has paid little attention to complex driving behaviors, namely merging car-following and lane-changing behavior, and how lane-changing affects algorithms designed to model and control a car-following vehicle. During the merging behavior, the Follower Vehicle (FV) might significantly diverge from typical car-following models. Thus, this paper

Jinjie Shi, Hongchen Chu, Aurelien Merkel, Chenkai Liu

While parity transformation represents a fundamental symmetry operation in physics, its implications remain underexplored in metamaterial science. Here, we introduce a framework leveraging parity transformation to construct parity-inverted counterparts of arbitrary three-dimensional meta-atoms, enabling the creation of parity-engineered metamaterial slabs. W

Xuan-May Le, Ling Luo, Uwe Aickelin, Minh-Tuan Tran

Patient-ventilator asynchrony (PVA) is a common and critical issue during mechanical ventilation, affecting up to 85% of patients. PVA can result in clinical complications such as discomfort, sleep disruption, and potentially more severe conditions like ventilator-induced lung injury and diaphragm dysfunction. Traditional PVA management, which relies on manu

ChunYin Hau

In this paper, we propose a condition on the coefficients of a cohomology-valued power series, which we call ``asymptotically Mittag-Leffler''. We show that if the $J$-function of a Fano manifold is asymptotically Mittag-Leffler, then it has the exponential growth as $t\to +\infty$. This provides an alternative method to compute the principal asymptotic clas

Shijie Li, Zhongyao Cheng, Rong Li, Shuai Li

Monocular Semantic Scene Completion (MonoSSC) reconstructs and interprets 3D environments from a single image, enabling diverse real-world applications. However, existing methods are often constrained by the local receptive field of Convolutional Neural Networks (CNNs), making it challenging to handle the non-uniform distribution of projected points (Fig. \r

Qiyuan He, Angela Yao

Personalized image generation with text-to-image diffusion models generates unseen images based on reference image content. Zero-shot adapter methods such as IP-Adapter and OminiControl are especially interesting because they do not require test-time fine-tuning. However, they struggle to balance preserving personalized content and adherence to the text prom

Yao Cheng, Yibo Zhao, Jiapeng Zhu, Yao Liu

Large Language Models (LLMs) have demonstrated significant potential across various domains. However, they often struggle with integrating external knowledge and performing complex reasoning, leading to hallucinations and unreliable outputs. Retrieval Augmented Generation (RAG) has emerged as a promising paradigm to mitigate these issues by incorporating ext

Shijie Li, Xun Xu, Si Yong Yeo, Xulei Yang

Motion forecasting is a crucial component of autonomous driving systems, enabling the generation of accurate and smooth future trajectories to ensure safe navigation to the destination. In previous methods, potential future trajectories are often absent in the scene encoding stage, which may lead to suboptimal outcomes. Additionally, prior approaches typical

Yihua Shao, Deyang Lin, Fanhu Zeng, Minxi Yan

Diffusion models have been widely adopted in image and video generation. However, their complex network architecture leads to high inference overhead for its generation process. Existing diffusion quantization methods primarily focus on the quantization of the model structure while ignoring the impact of time-steps variation during sampling. At the same time

Jiangdong Cai, Haotian Jiang, Zhenrong Shen, Yonghao Li

The deployment of computer-aided diagnosis systems for cervical cancer screening using whole slide images (WSIs) faces critical challenges due to domain shifts caused by staining variations across different scanners and imaging environments. While existing stain augmentation methods improve patch-level robustness, they fail to scale to WSIs due to two key li

Chenxin Wei, Dawod Muhamed, Wenting Lu, Haikai Zou

Thermoelectric (Bi1-x Sbx)2Te3 (BST-x) compounds with x=0.2, 0.7 and 0.9 have been studied using synchrotron angle-dispersive powder x-ray diffraction in a diamond anvil cell up to 25 GPa (at room temperature). The results clearly indicate that all compounds of this study follow a similar structural evolution with the one of pure Bi2Te3 and Sb2Te3 under pres

Jimmy Xuekai Li, Thomas Flottmann, Max Millen, Shuai Chen

Understanding the response of coal mechanical properties to dewatering and gas depletion is critical for estimating borehole stability and designing infill coal seam gas (CSG) wells. Despite its importance, the full impact of these processes on coal strength remains little explored. This study aims to quantify these effects through a combination of results f

Rustem Khasanov, Igor Plokhikh, Thomas J. Hicken, Hubertus Luetkens

High-pressure studies reveal a stark contrast between the superconducting properties of double-layer Ruddlesden-Popper (RP) nickelates La$_2$PrNi$_2$O$_7$ and La$_3$Ni$_2$O$_7$. While La$_2$PrNi$_2$O$_7$ exhibits bulk superconductivity, La$_3$Ni$_2$O$_7$ displays filamentary behavior, suggesting that superconductivity is confined to phase interfaces rather t

Yuzheng Wang, Zhaoyu Chen, Dingkang Yang, Yuanhang Wang

Adversarial Robustness Distillation (ARD) is a promising task to boost the robustness of small-capacity models with the guidance of the pre-trained robust teacher. The ARD can be summarized as a min-max optimization process, i.e., synthesizing adversarial examples (inner) & training the student (outer). Although competitive robustness performance, existing A

Adrian Baule

Score-based diffusion models generate samples from an unknown target distribution using a time-reversed diffusion process. While such models represent state-of-the-art approaches in industrial applications such as artificial image generation, it has recently been noted that their performance can be further improved by considering injection noise with heavy t

Thierry Valet, Kei Yamamoto, Benjamin Pigeau, Grégoire de Loubens

In the context of an ever-expanding experimental and theoretical interest in the magnetization dynamics of mesoscopic magnetic structures, both in the classical and quantum regimes, we formulate a low energy field theory for the linear spin-waves in finite and textured ferromagnets and we perform its constrained canonical quantization. The introduction of a

Thierry Valet, Kei Yamamoto, Benjamin Pigeau, Grégoire de Loubens

In the context of a growing interdisciplinary interest in the angular momentum of wave fields, the spin-wave case has yet to be fully explored, with the extensively studied notion of spin transport being only part of the broader picture. Here we report experimental evidence for magnon orbital angular momentum, demonstrating that the mode exhibits rotation ra

Mohammad S. Ahmad, Zan A. Naeem, Michaël Aupetit, Ahmed Elmagarmid

Tabular data embedded in PDF files, web pages, and other types of documents is prevalent in various domains. These tables, which we call human-centric tables (HCTs for short), are dense in information but often exhibit complex structural and semantic layouts. To query these HCTs, some existing solutions focus on transforming them into relational formats. How

Shaohong Du, Qianqian Hou, Xiaoping Xie

In this paper, we develop a modified nonlinear dynamic diffusion (DD) finite element method for convection-diffusion-reaction equations. This method is free of stabilization parameters and is capable of precluding spurious oscillations. We prove existence and, under an assumption of small mesh size, uniqueness of the discrete solution, and derive the optimal

Chengcheng Zhu, Jiale Zhang, Di Wu, Guodong Long

Federated learning is a distributed learning paradigm that facilitates the collaborative training of a global model across multiple clients while preserving the privacy of local datasets. To address inherent challenges related to data heterogeneity and satisfy personalized needs, a new direction within FL, known as personalized Federated Learning (pFL), has

Jiaxin Ai, Pengfei Zhou, Zhaopan Xu, Ming Li

As multi-modal large language models (MLLMs) frequently exhibit errors when solving scientific problems, evaluating the validity of their reasoning processes is critical for ensuring reliability and uncovering fine-grained model weaknesses. Since human evaluation is laborious and costly, prompting MLLMs as automated process judges has become a common practic

Rajan Das Gupta, Md. Tanzib Hosain, M. F. Mridha, Salah Uddin Ahmed

LLM chatbot interfaces allow students to get instant, interactive assistance with homework, but doing so carelessly may not advance educational objectives. In this study, an interactive homework help system based on DeepSeek R1 is developed and first implemented for students enrolled in a large computer science beginning programming course. In addition to an

Fan Yin, Philippe Laban, Xiangyu Peng, Yilun Zhou

Malicious content generated by large language models (LLMs) can pose varying degrees of harm. Although existing LLM-based moderators can detect harmful content, they struggle to assess risk levels and may miss lower-risk outputs. Accurate risk assessment allows platforms with different safety thresholds to tailor content filtering and rejection. In this pape

Zhigang Bao, Giorgio Cipolloni, László Erdős, Joscha Henheik

We consider the Wigner minor process, i.e. the eigenvalues of an $N\times N$ Wigner matrix $H^{(N)}$ together with the eigenvalues of all its $n\times n$ minors, $H^{(n)}$, $n\le N$. The top eigenvalues of $H^{(N)}$ and those of its immediate minor $H^{(N-1)}$ are very strongly correlated, but this correlation becomes weaker for smaller minors $H^{(N-k)}$ as

Bruno Pinheiro, Ignacio Ponce, Daniel Dotta, Federico Milano

This paper introduces a novel formulation to evaluate the local synchronization of power system devices, namely Synchronization Energy (SE). The formulation is derived based on the complex frequency concept and the Teager Energy Operator applied to the complex power. This formulation offers valuable insights into the relationship between complex frequency of

Rasul Dent, Pedro Ortiz Suarez, Thibault Clérice, Benoît Sagot

Automatic language identification is frequently framed as a multi-class classification problem. However, when creating digital corpora for less commonly written languages, it may be more appropriate to consider it a data mining problem. For these varieties, one knows ahead of time that the vast majority of documents are of little interest. By minimizing reso

Abdessatar Souissi, Amenallah Andolsi

This paper introduces a Heisenberg picture approach to Matrix Product States (MPS), offering a rigorous yet intuitive framework to explore their structure and classification. MPS efficiently represent ground states of quantum many-body systems, with infinite MPS (iMPS) capturing long-range correlations and thermodynamic behavior. We classify MPS into project

Junyi Wu, Zhiteng Li, Zheng Hui, Yulun Zhang

Recently, Diffusion Transformers (DiTs) have emerged as a dominant architecture in video generation, surpassing U-Net-based models in terms of performance. However, the enhanced capabilities of DiTs come with significant drawbacks, including increased computational and memory costs, which hinder their deployment on resource-constrained devices. Current accel

Yue Cai, Jie Liu, Kang-Jie Ma, Lei Tan

Nonreciprocal devices, such as isolator or circulator, are crucial for information routing and processing in quantum networks. Traditional nonreciprocal devices, which rely on the application of bias magnetic fields to break time-reversal symmetry and Lorentz reciprocity, tend to be bulky and require strong static magnetic fields. This makes them challenging

Aritra Banerjee, Abir Ghosh

We revisit the general analytic solution space for relativistic $(1+1)$-dimensional hydrodynamics for a perfect fluid flowing along the longitudinal direction. We work out the explicit one-parameter family of interpolating flows between boost-invariant and boost-non-invariant regimes, where a direct and simple dialing of the parameter at the level of solutio

Jianwen Sun, Yukang Feng, Chuanhao Li, Fanrui Zhang

Unified multimodal understanding and generation have recently received much attention in the area of vision and language. Existing UniMs are designed to simultaneously learn both multimodal understanding and generation capabilities, demanding substantial computational resources, and often struggle to generate interleaved text-image. We present ARMOR, a resou

Yuqi Wang, Ruifeng Hu

Predicting particle-laden flows requires accurate fluid force models. However, a reliable particle force model for finite-size particles in turbulent flows remains lacking. In the present work, a fluid force model for a finite-size spherical particle in turbulence is developed by simulating turbulent flow past a fixed spherical particle using particle-resolv

S. Mohammedzadeh, R. de Lamare

Robust adaptive beamforming (RAB) based on interference-plus-noise covariance (IPNC) matrix reconstruction can experience serious performance degradation in the presence of look direction and array geometry mismatches, particularly when the input signal-to-noise ratio (SNR) is large. In this work, we present a RAB technique to address covariance matrix recon

Z. Xie, A. K. Lei, H. Zheng, W. C. Zhang

The pseudorapidity density distributions of charged particles and the transverse momentum spectra of identified particles in proton-proton (pp) collisions at the center-of-mass energies ranging from $\sqrt{s}=200$ GeV to 13 TeV have been systematically studied using the newly released parton and cascade model PACIAE 4.0 based on PYTHIA 8.3. The available exp

Wataru Urasaki, Kouji Tahata, Sadao Tomizawa

In two-way contingency tables under an asymmetric situation, where the row and column variables are defined as explanatory and response variables, respectively, quantifying the extent to which the explanatory variable contributes to predicting the response variable is important. One quantification method is the association measure, which indicates the degree

Xiaoyang Liu, Yuquan Wang, Zheng Chen, Jiezhang Cao

Currently, methods for single-image deblurring based on CNNs and transformers have demonstrated promising performance. However, these methods often suffer from perceptual limitations, poor generalization ability, and struggle with heavy or complex blur. While diffusion-based methods can partially address these shortcomings, their multi-step denoising process

Sebastian Engelke, Nicola Gnecco, Frank Röttger

The behavior of extreme observations is well-understood for time series or spatial data, but little is known if the data generating process is a structural causal model (SCM). We study the behavior of extremes in this model class, both for the observational distribution and under extremal interventions. We show that under suitable regularity conditions on th

Huilin Wang

This paper systematically investigates the thermodynamic properties of classical oscillators under different statistical distributions, focusing on the behavior of uniform distribution, two-level distribution, gamma distribution, log-normal distribution, and F-distribution as the nonequilibrium parameter q varies. By calculating Helmholtz free energy and ent

Xingwei Tan, Chen Lyu, Hafiz Muhammad Umer, Sahrish Khan

Detecting toxic language including sexism, harassment and abusive behaviour, remains a critical challenge, particularly in its subtle and context-dependent forms. Existing approaches largely focus on isolated message-level classification, overlooking toxicity that emerges across conversational contexts. To promote and enable future research in this direction

Long Guo, Ying Zhang, Qi Qin, Guanjun Liu

Over the past decades, the performance design of closed-chain legged mechanisms (CLMs) has not been adequately addressed. Most existing design methodologies have predominantly relied on trajectory synthesis, which inadvertently prioritizes less critical performance aspects. This study proposes a hierarchical multi-objective optimization strategy to address t

Chuang Li, Lanfang Huang, Dian He, Yanhua Wen

The serverless platform aims to facilitate cloud applications' straightforward deployment, scaling, and management. Unfortunately, the distributed nature of serverless computing makes it difficult to port traditional security tools directly. The existing serverless solutions primarily identify potential threats or performance bottlenecks through post-analysi

Jie He, Yu Fu

Commonsense reasoning (CR) has been studied in many pieces of domain and has achieved great progress with the aid of large datasets. Unfortunately, most existing CR datasets are built in English, so most previous work focus on English. Furthermore, as the annotation of commonsense reasoning is costly, it is impossible to build a large dataset for every novel

ZhiGuang Tan, YouNeng Guo, ShengJie Wang, Hua Zheng

Quantum Chromodynamics (QCD) is the fundamental theory describing quark interactions, and various quark models based on QCD have been widely used to study the properties of hadrons, including their structures and mass spectra. However, unlike Quantum Electrodynamics (QED) and the Bohr model of the hydrogen atom, there is no direct classical analogy for hadro

Jialin Lu, Junjie Shan, Ziqi Zhao, Ka-Ho Chow

As object detection becomes integral to many safety-critical applications, understanding its vulnerabilities is essential. Backdoor attacks, in particular, pose a serious threat by implanting hidden triggers in victim models, which adversaries can later exploit to induce malicious behaviors during inference. However, current understanding is limited to singl

Leia Greenberg, Haim Avron

Reduced Rank Regression (RRR) is a widely used method for multi-response regression. However, RRR assumes a linear relationship between features and responses. While linear models are useful and often provide a good approximation, many real-world problems involve more complex relationships that cannot be adequately captured by simple linear interactions. One

Oscar Enajite, Onyekachuku Mike Osiele, Samuel Emovokerayec Omoyibo, Kingsley Imoni-Ogbe

The technological world is in search of environmentally friendly methods of generating energy for the use of the ever growing needs of power for sustainable development of visually all facet of life. Solar energy form an environmentally friendly solution for the reduction of global warming. For usage in photovoltaic applications, this work explores the produ

Chen-Lin Zhang, Lin Sui, Shuming Liu, Fangzhou Mu

Temporal localization in untrimmed videos, which aims to identify specific timestamps, is crucial for video understanding but remains challenging. This task encompasses several subtasks, including temporal action localization, temporal video grounding, moment retrieval, and generic event boundary detection. Existing methods in each subfield are typically des

Xiaodong Liu, Qingxiang Shi, Jing Wang

This work is dedicated to uniqueness and numerical algorithms for determining the point sources of the biharmonic wave equation using scattered fields at sparse sensors. We first show that the point sources in both $\mathbb{R}^2$ and $\mathbb{R}^3$ can be uniquely determined from the multifrequency sparse scattered fields. In particular, to deal with the cha