May 2025 arXiv papers — page 88

Weichang Wu, Xiaolu Zhang, Jun Zhou, Yuchen Li

User Behavior Sequence (UBS) modeling is crucial in industrial applications. As data scale and task diversity grow, UBS pretraining methods have become increasingly pivotal. State-of-the-art UBS pretraining methods rely on predicting behavior distributions. The key step in these methods is constructing a selected behavior vocabulary. However, this manual ste

Jiacong Chen, Qingyu Mao, Youneng Bao, Xiandong Meng

3D Gaussian Splatting (3DGS) has emerged as a high-fidelity and efficient paradigm for online free-viewpoint video (FVV) reconstruction, offering viewers rapid responsiveness and immersive experiences. However, existing online methods face challenge in prohibitive storage requirements primarily due to point-wise modeling that fails to exploit the motion prop

Alejandro Moreno Arcas, Albert Sanchis, Jorge Civera, Alfons Juan

Adaptation of foundation models using low-rank methods is a widespread approach. Another way to adapt these models is to employ orthogonal fine-tuning methods, which are less time and memory efficient despite their good generalization properties. In this work, we propose Householder Orthogonal Fine-tuning (HOFT), a novel orthogonal fine-tuning method that ai

Yuliang Yan, Haochun Tang, Shuo Yan, Enyan Dai

Large language models (LLMs) are considered valuable Intellectual Properties (IP) for legitimate owners due to the enormous computational cost of training. It is crucial to protect the IP of LLMs from malicious stealing or unauthorized deployment. Despite existing efforts in watermarking and fingerprinting LLMs, these methods either impact the text generatio

Luís Dieulefait, Josep González, Joan-C. Lario

We say that a normalized modular form is of CM type modulo $\ell$ by an imaginary quadratic field $K$ if its Fourier coefficients $a_p$ are congruent to $0$ modulo a prime $\mathcal L\mid \ell$ for every prime $p$ that is inert in $K$. In this paper, we address the following question. Let $f$ be a weight~$2$ cuspidal Hecke eigenform without complex multiplic

Junichi Baba

We present a chemo-dynamical $N$-body/hydrodynamic simulation of an isolated Milky Way-like galaxy to investigate how bar formation influences star formation rates, stellar migration, and the resulting age and metallicity distributions of disk stars. Focusing on the transient epoch of bar formation, a phase that triggers gas inflows, enhances local star form

Mohamed Amine Ketata, David Lüdke, Leo Schwinn, Stephan Günnemann

Building generative models for relational databases (RDBs) is important for many applications, such as privacy-preserving data release and augmenting real datasets. However, most prior works either focus on single-table generation or adapt single-table models to the multi-table setting by relying on autoregressive factorizations and sequential generation. Th

Heejae Suh, Yejin Jeon, Deokhyung Kang, Taehee Park

Small large language models (sLLMs) offer the advantage of being lightweight and efficient, which makes them suitable for resource-constrained environments. However, sLLMs often struggle to maintain topic consistency in task-oriented dialogue systems, which is critical for scenarios such as service chatbots. Specifically, it is important to ensure that the m

Tanay Pathak, Masaki Tezuka

It is often expected (and assumed) for a quantum chaotic system that the presence of correlated eigenvalues implies that all the other properties as dictated by random matrix theory are satisfied. We demonstrate using the spin-$1/2$ kicked field Ising model that this is not necessarily true. We study the properties of eigenvalues of the reduced density matri

Huitong Yang, Zhuoxiao Chen, Fengyi Zhang, Zi Huang

Maintaining robust 3D perception under dynamic and unpredictable test-time conditions remains a critical challenge for autonomous driving systems. Existing test-time adaptation (TTA) methods often fail in high-variance tasks like 3D object detection due to unstable optimization and sharp minima. While recent model merging strategies based on linear mode conn

Thomas J. Elliott

Standard formulations of quantum theory are based on complex numbers: Quantum states can be in superpositions, with weights given by complex probability amplitudes. Motivated by quantum theory promising a range of practical advantages over classical for a multitude of tasks, we investigate how the presence of complex amplitudes in quantum theory can yield op

Zhouhao Sun, Zhiyuan Kan, Xiao Ding, Li Du

With the evolution of large language models (LLMs), their robustness against individual simple biases has been enhanced. However, we observe that the ensemble of multiple simple biases still exerts a significant adverse impact on LLMs. Given that real-world data samples are typically confounded by a wide range of biases, LLMs tend to exhibit unstable perform

Pankaj Borah, Pradipta Ghosh

In this work, we extend the scalar sector of the conventional hyperchargeless inert triplet model (ITM) to include a second dark matter (DM) candidate, which appears to be a pseudo-Nambu-Goldstone boson (pNGB). The usual ITM with an extended scalar sector offers a DM candidate along with novel signatures at different experiments, e.g., colliders, gravitation

Giovanni Servedio, Alessandro De Bellis, Dario Di Palma, Vito Walter Anelli

Factual hallucinations are a major challenge for Large Language Models (LLMs). They undermine reliability and user trust by generating inaccurate or fabricated content. Recent studies suggest that when generating false statements, the internal states of LLMs encode information about truthfulness. However, these studies often rely on synthetic datasets that l

Ayush Pandey, Rohail Asim, Jean Louis K. E. Fendji, Talal Rahwan

Over 2.6 billion people remain without access to the Internet in 2025. This phenomenon is especially pronounced in developing regions, where cost and infrastructure limitations are major barriers to connectivity. In response, we design SONIC, a low-cost, scalable data delivery system that builds on existing infrastructures: FM radio for downlink broadcasting

Lovisa Hagström, Youna Kim, Haeun Yu, Sang-goo Lee

Incorporating external knowledge is crucial for knowledge-intensive tasks, such as question answering and fact checking. However, language models (LMs) may ignore relevant information that contradicts outdated parametric memory or be distracted by irrelevant contexts. While many context utilisation manipulation techniques (CMTs) have recently been proposed t

Zirui Song, Guangxian Ouyang, Mingzhe Li, Yuheng Ji

Large Vision-Language Models (LVLMs) have recently advanced robotic manipulation by leveraging vision for scene perception and language for instruction following. However, existing methods rely heavily on costly human-annotated training datasets, which limits their generalization and causes them to struggle in out-of-domain (OOD) scenarios, reducing real-wor

Sergiu I. Vacaru, Elşen Veli Veliev

We construct new classes of solutions describing generic off-diagonal deformations of regular Schwarzschild black holes (BHs) in general relativity (GR). Examples of such (primary) diagonal metrics reducing the Einstein equations to integrable systems of nonlinear ordinary differential equations were studied in a recent work by R. Casadio, A. Kamenshchik and

Majid Mohammadi, Siu Lun Chau, Krikamol Muandet

Kernel methods are widely used in machine learning and statistics for their flexibility and expressive power, yet their black-box nature limits adoption in high-stakes applications. Shapley value-based attribution methods such as SHAP, and kernel-specific adaptations including RKHS-SHAP, provide a principled framework for explainability -- but exact computat

M. Scheck, R. Chapman, K. Mashtakov, R. Meeten

The relative ordering of $J^-$ levels of multiplets resulting from two-body excitations, which include a $J^{\pi}=3^-$ state that can contribute to the octupole phonon, are investigated in a simplistic shell-model approach. To calculate the relative level ordering, harmonic oscillator wavefunctions and a residual $\delta$ interaction are used. The simplistic

Yuting Huang, Meitong Guo, Yiquan Wu, Ang Li

Recent advances in LegalAI have primarily focused on individual case judgment analysis, often overlooking the critical appellate process within the judicial system. Appeals serve as a core mechanism for error correction and ensuring fair trials, making them highly significant both in practice and in research. To address this gap, we present the AppealCase da

Vaishali Maheshkar, Aadarsh Anantha Ramakrishnan, Charuvahan Adhivarahan, Karthik Dantu

According to the EPA, only 25% of waste is recycled, and just 60% of U.S. municipalities offer curbside recycling. Plastics fare worse, with a recycling rate of only 8%; an additional 16% is incinerated, while the remaining 76% ends up in landfills. The low plastic recycling rate stems from contamination, poor economic incentives, and technical difficulties,

Jiaxin Liu, Jia Wang, Saihui Hou, Min Ren

In recent years, the explosive advancement of deepfake technology has posed a critical and escalating threat to public security: diffusion-based digital human generation. Unlike traditional face manipulation methods, such models can generate highly realistic videos with consistency via multimodal control signals. Their flexibility and covertness pose severe

Hao Yin, Claudio De Persis, Bayu Jayawardhana, Santiago Sanchez Escalonilla Plaza

This paper studies the design of neural network (NN)-based controllers for unknown nonlinear systems, using contraction analysis. A Neural Ordinary Differential Equation (NODE) system is constructed by approximating the unknown draft dynamics with a feedforward NN. Incremental sector bounds and contraction theory are applied to the activation functions and t

Rosamaria Rescigno, Lubomira Softova

We consider the Cauchy-Dirichlet problem for second-order quasilinear non-divergence form operators of parabolic type. The data are Cara\-th\'e\-o\-dory functions, and the principal part is of $VMO_x$-type with respect to the variables $ (x,t).$ Assuming the existence of a strong solution $u_0,$ we apply the Implicit Function Theorem in a small domain of thi

Sogol Sanjaripour, Archana Aravindan, Gabriela Canalizo, Shoubaneh Hemmati

Identifying AGNs in dwarf galaxies is critical for understanding black hole formation but remains challenging due to their low luminosities, low metallicities, and star formation-driven emission that can obscure AGN signatures. Machine learning (ML) techniques, particularly unsupervised methods, offer new ways to address these challenges by uncovering patter

SiYoung Jang, Roberto Morabito

The widespread adoption of Language Models (LMs) across industries is driving interest in deploying these services across the computing continuum, from the cloud to the network edge. This shift aims to reduce costs, lower latency, and improve reliability and privacy. Small Language Models (SLMs), enabled by advances in model compression, are central to this

Anshu Xiong, Songmao Zhang

The notion of relevance was proposed for stability of justification status of a single argument in incomplete argumentation frameworks (IAFs) in 2024 by Odekerken et al. To extend the notion, we study the relevance for stability of verification status of a set of arguments in this paper, i.e., the uncertainties in an IAF that have to be resolved in some situ

Włodzimierz Lewoniewski, Krzysztof Węcel, Witold Abramowicz

This study presents a comparative analysis of 55 Wikipedia language editions employing a citation index alongside a synthetic quality measure. Specifically, we identified the most significant Wikipedia articles within distinct topical areas, selecting the top 10, top 25, and top 100 most cited articles in each topic and language version. This index was built

Runcong Zhao, Chengyu Cao, Qinglin Zhu, Xiucheng Lv

Complex narrative contexts often challenge language models' ability to follow instructions, and existing benchmarks fail to capture these difficulties. To address this, we propose Concise-SAE, a training-free framework that improves instruction following by identifying and editing instruction-relevant neurons using only natural language instructions, without

Hongyu Li, Matteo Nerini, Shanpu Shen, Bruno Clerckx

Written by its inventors, this first tutorial on Beyond-Diagonal Reconfigurable Intelligent Surfaces (BD-RISs) provides the readers with the basics and fundamental tools necessary to appreciate, understand, and contribute to this emerging and disruptive technology. Conventional (Diagonal) RISs (D-RISs) are characterized by a diagonal scattering matrix $\math

Zhiyuan Wu, Sheng Sun, Yuwei Wang, Min Liu

Heterogeneous device-edge-cloud computing infrastructures have become widely adopted in telecommunication operators and Wide Area Networks (WANs), offering multi-tier computational support for emerging intelligent services. With the rapid proliferation of Large Language Model (LLM) services, efficiently coordinating inference tasks and reducing communication

Antonio Martínez Ibarra, Julian James Stephen, Aurora González Vidal, K. R. Jayaram

This work examines latency, throughput, and other metrics when performing inference on confidential GPUs. We explore different traffic patterns and scheduling strategies using a single Virtual Machine with one NVIDIA H100 GPU, to perform relaxed batch inferences on multiple Large Language Models (LLMs), operating under the constraint of swapping models in an

Minghui Du, Pengcheng Wang, Ziren Luo, Wen-Biao Han

Taiji, a Chinese space-based gravitational wave (GW) detection project, aims to explore the millihertz GW universe with unprecedented sensitivity. By observing astrophysical and cosmological sources, including Galactic binaries, massive black hole binaries, extreme mass-ratio inspirals, and stochastic gravitational wave backgrounds, etc., Taiji is expected t

Roberto Morabito, SiYoung Jang

The rapid adoption of generative AI (GenAI), particularly Large Language Models (LLMs), has exposed critical limitations of cloud-centric deployments, including latency, cost, and privacy concerns. Meanwhile, Small Language Models (SLMs) are emerging as viable alternatives for resource-constrained edge environments, though they often lack the capabilities of

Aditya Bhattacharya, Katrien Verbert

During job recruitment, traditional applicant selection methods often lack transparency. Candidates are rarely given sufficient justifications for recruiting decisions, whether they are made manually by human recruiters or through the use of black-box Applicant Tracking Systems (ATS). To address this problem, our work introduces a multi-agent AI system that

Augusto Luis Ballardini, Miguel Ángel Sotelo

Achieving full automation in self-driving vehicles remains a challenge, especially in dynamic urban environments where navigation requires real-time adaptability. Existing systems struggle to handle navigation plans when faced with unpredictable changes in road layouts, spontaneous detours, or missing map data, due to their heavy reliance on predefined carto

Alex Degtyarev, Igor Dolgachev, Shigeyuki Kondo

We study K3 surfaces of degree 6 containing two sets of 12 skew lines such that each line from a set intersects exactly six lines from the other set. These surfaces arise as hyperplane sections of the cubic line complex associated with the pencil of desmic quartic surfaces introduced by George Humbert and recently studied by the second and third authors. We

Suvarthi Sarkar, Dhanesh V, Ketan Singh, Aryabartta Sahu

With the increasing prevalence of computationally intensive workflows in cloud environments, it has become crucial for cloud platforms to optimize energy consumption while ensuring the feasibility of user workflow schedules with respect to strict deadlines and reliability constraints. The key challenges faced when cloud systems provide virtual machines of va

Lingfeng Wang, Hualing Lin, Senda Chen, Tao Wang

While humans effortlessly draw visual objects and shapes by adaptively allocating attention based on their complexity, existing multimodal large language models (MLLMs) remain constrained by rigid token representations. Bridging this gap, we propose ALTo, an adaptive length tokenizer for autoregressive mask generation. To achieve this, a novel token length p

Noga Amit, Omer Reingold, Guy N. Rothblum

We revisit the foundations of fairness and its interplay with utility and efficiency in settings where the training data contain richer labels, such as individual types, rankings, or risk estimates, rather than just binary outcomes. In this context, we propose algorithms that achieve stronger notions of evidence-based fairness than are possible in standard s

Seyedeh Fatemeh Ebrahimi, Jaakko Peltonen

Topic models often fail to capture low-prevalence, domain-critical themes, so-called minority topics, such as mental health themes in online comments. While some existing methods can incorporate domain knowledge, such as expected topical content, methods allowing guidance may require overly detailed expected topics, hindering the discovery of topic divisions

Qi Jiang, Lei Zhang, Longquan Li, Lin Wan

Time-series single-cell RNA-sequencing (scRNA-seq) datasets offer unprecedented insights into the dynamics and heterogeneity of cellular systems. These systems exhibit multiscale collective behaviors driven by intricate intracellular gene regulatory networks and intercellular interactions of molecules. However, inferring interacting cell population dynamics

Dario Di Palma, Alessandro De Bellis, Giovanni Servedio, Vito Walter Anelli

Large Language Models (LLMs) have rapidly become central to NLP, demonstrating their ability to adapt to various tasks through prompting techniques, including sentiment analysis. However, we still have a limited understanding of how these models capture sentiment-related information. This study probes the hidden layers of Llama models to pinpoint where senti

David Krongauz, Hido Pinto, Sarah Kohn, Yanir Marmor

Human speech contains paralinguistic cues that reflect a speaker's physiological and neurological state, potentially enabling non-invasive detection of various medical phenotypes. We introduce the Human Phenotype Project Voice corpus (HPP-Voice): a dataset of 7,188 recordings in which Hebrew-speaking adults count for 30 seconds, with each speaker linked to u

Jiří Klimeš, Pavel Soldán

Homonuclear spin-polarized alkali-metal trimers in their lowest-lying electronic state are investigated theoretically. Their equilibrium geometries and binding energies are determined with the state-of-the-art quantum chemical methods at three levels of approximation. The equilibrium geometries obtained $R_{\rm eq}({\rm Li}_3) = 3.100$ \r{A}, $R_{\rm eq}({\r

Zishuo Bao, Yibo Liu, Changyutao Qiu

With the rise of different language model architecture, fine-tuning is becoming even more important for down stream tasks Model gets messy, finding proper hyperparameters for fine-tuning. Although BO has been tried for hyperparameter tuning, most of the existing methods are oblivious to the fact that BO relies on careful choices of acquisition functions, whi

Andrey Dymov

We consider a damped/driven cubic NLS equation on a torus under the limit when first the amplitude of solutions goes to zero and then the period of the torus goes to infinity. We suggest another proof of the kinetic approximation for the energy spectrum under a subcritical scaling, extending to the exact solutions result obtained in [Dymov, Kuksin, Maiocchi,

Haibo Liu, Junqing Chen, Guang Lin

Reconstructing complex 3D interfaces from indirect measurements remains a grand challenge in scientific computing, particularly for ill-posed inverse problems like Electrical Impedance Tomography (EIT). Traditional shape optimization struggles with topological changes and regularization tuning, while emerging deep learning approaches often compromise physica

Giorgio Consigli, Darinka Dentcheva, Francesca Maggioni, Giovanni Micheli

We consider a financial intermediary managing assets and liabilities exposed to several risk sources and seeking an optimal portfolio strategy to minimise the initial capital invested and the total risk associated with investment losses and financial debt. We formulate the problem as a multistage stochastic programming model, with a time-consistent dynamic r

Yao Wei, Muhammad Usman, Hazrat Bilal

The problems that tobacco workshops encounter include poor curing, inconsistencies in supplies, irregular scheduling, and a lack of oversight, all of which drive up expenses and worse quality. Large quantities make manual examination costly, sluggish, and unreliable. Deep convolutional neural networks have recently made strides in capabilities that transcend

Jing Li, Yanqi Song, Sujuan Qin, Fei Gao

Kernel methods serve as powerful tools to capture nonlinear patterns behind data in machine learning. The quantum kernel, integrating kernel theory with quantum computing, has attracted widespread attention. However, existing studies encounter performance bottlenecks when processing complex data with localized structural patterns, stemming from the limitatio

Shuzheng Si, Haozhe Zhao, Cheng Gao, Yuzhuo Bai

Teaching large language models (LLMs) to be faithful in the provided context is crucial for building reliable information-seeking systems. Therefore, we propose a systematic framework, CANOE, to reduce faithfulness hallucinations of LLMs across different downstream tasks without human annotations. Specifically, we first synthesize short-form question-answeri

Huynh Thi Thanh Binh, Le Van Cuong, Dang Hai Dang, Le Trong Vinh

Recently, Wireless Rechargeable Sensor Networks (WRSNs) that leveraged the advantage of wireless energy transfer technology have opened a promising opportunity in solving the limited energy issue. However, an ineffective charging strategy may reduce the charging performance. Although many practical charging algorithms have been introduced, these studies main

Xinxing Shi, Xiaoyu Jiang, Mauricio A. Álvarez

Gaussian Process (GP) Variational Autoencoders (VAEs) extend standard VAEs by replacing the fully factorised Gaussian prior with a GP prior, thereby capturing richer correlations among latent variables. However, performing exact GP inference in large-scale GPVAEs is computationally prohibitive, often forcing existing approaches to rely on restrictive kernel

Yuetong Liu, Yunqiu Xu, Yang Wei, Xiuli Bi

Restoring nighttime images affected by multiple adverse weather conditions is a practical yet under-explored research problem, as multiple weather conditions often coexist in the real world alongside various lighting effects at night. This paper first explores the challenging multi-weather nighttime image restoration task, where various types of weather degr

Davide Gorbani, Giuseppe L'Erario, Hosameldin Awadalla Omer Mohamed, Daniele Pucci

We propose a novel Model Predictive Control (MPC) framework for a jet-powered flying humanoid robot. The controller is based on a linearised centroidal momentum model to represent the flight dynamics, augmented with a second-order nonlinear model to explicitly account for the slow and nonlinear dynamics of jet propulsion. A key contribution is the introducti

Yanbo Zhang, Sumeer A. Khan, Adnan Mahmud, Huck Yang

With recent Nobel Prizes recognising AI contributions to science, Large Language Models (LLMs) are transforming scientific research by enhancing productivity and reshaping the scientific method. LLMs are now involved in experimental design, data analysis, and workflows, particularly in chemistry and biology. However, challenges such as hallucinations and rel

Patryk Tajs, Mateusz Skarupski, Jakub Rydzewski

Unsupervised machine learning has recently gained much attention in the field of molecular dynamics (MD). Particularly, dimensionality reduction techniques have been regularly employed to analyze large volumes of high-dimensional MD data to gain insight into hidden information encoded in MD trajectories. Among many such techniques, t-distributed stochastic n

Jiaqi Li, Xinyi Dong, Yang Liu, Zhizhuo Yang

We present a novel pipeline, ReflectEvo, to demonstrate that small language models (SLMs) can enhance meta introspection through reflection learning. This process iteratively generates self-reflection for self-training, fostering a continuous and self-evolving process. Leveraging this pipeline, we construct ReflectEvo-460k, a large-scale, comprehensive, self

Yu Zhang, Xingzhuo Guo, Haoran Xu, Jialong Wu

Diffusion and flow-based models have enabled significant progress in generation tasks across various modalities and have recently found applications in predictive learning. However, unlike typical generation tasks that encourage sample diversity, predictive learning entails different sources of stochasticity and requires sampling consistency aligned with the

Yubin He

Under a reasonable decay assumption on the approximating function, we establish a zero-full law for the Hausdorff measure of sets of inhomogeneous Dirichlet non-improvable affine forms with weights, thereby answering a question posed by Kim and Kim (\S 5.3, Adv. Math., 2022).

Joerg F. Schneider

The presented approach aims at estimating the lateral variation of seismic velocities for a seismic timelapse survey. The monitor survey is depth migrated with the known velocity model of the base survey. An analysis is performed to determine whether the moveout to be expected for the monitor survey after residual prestack depth migration (RPSM) can be used

Robbert Reijnen, Yaoxin Wu, Zaharah Bukhsh, Yingqian Zhang

Deep reinforcement learning (DRL) has been widely used for dynamic algorithm configuration, particularly in evolutionary computation, which benefits from the adaptive update of parameters during the algorithmic execution. However, applying DRL to algorithm configuration for multi-objective combinatorial optimization (MOCO) problems remains relatively unexplo

Kuicai Dong, Yujing Chang, Shijie Huang, Yasheng Wang

Document Visual Question Answering (DocVQA) faces dual challenges in processing lengthy multimodal documents (text, images, tables) and performing cross-modal reasoning. Current document retrieval-augmented generation (DocRAG) methods remain limited by their text-centric approaches, frequently missing critical visual information. The field also lacks robust

Noémie Bergues, Arthur Carré, Paul Join-Lambert, Brice Hoffmann

Predicting the 3D conformation of small molecules within protein binding sites is a key challenge in drug design. When a crystallized reference ligand (template) is available, it provides geometric priors that can guide 3D pose prediction. We present a two-stage method for ligand conformation generation guided by such templates. In the first stage, we introd

Y. Zhou, M. Li, Y. Guan, Y. Wang

We experimentally and numerically investigate the clogging behavior of granular materials in a two-dimensional vertical pipe. The nonmonotonicity of clogging probability found in a cylindrical vertical pipe [L\'opez et al., Phys. Rev. E 102, 010902 (2020)] is also observed in the two-dimensional case. We numerically demonstrate that the clogging probability

Yangfan Luo, Jindong Wang, Shuonan Wu

We devise local projection stabilization (LPS) methods for advection problems in the $\boldsymbol{H}$(curl) and $\boldsymbol{H}$(div) spaces, employing conforming finite element spaces of arbitrary order within a unified framework. The key ingredient is a local inf-sup condition, enabled by enriching the approximation space with appropriate $\boldsymbol{H}$(

Vera Neplenbroek, Arianna Bisazza, Raquel Fernández

Generative Large Language Models (LLMs) infer user's demographic information from subtle cues in the conversation -- a phenomenon called implicit personalization. Prior work has shown that such inferences can lead to lower quality responses for users assumed to be from minority groups, even when no demographic information is explicitly provided. In this work

Meng Yan, Cai Xu, Xujing Wang, Ziyu Guan

Recommender systems based on graph neural networks perform well in tasks such as rating and ranking. However, in real-world recommendation scenarios, noise such as user misuse and malicious advertisement gradually accumulates through the message propagation mechanism. Even if existing studies mitigate their effects by reducing the noise propagation weights,

N. Saradha, Divyum Sharma

Let $F(x,y)$ be an irreducible form of degree $r\geq 3$ and having $s+1$ non-zero coefficients. Let $h\geq 1$ be an integer and consider the Thue inequality $$|F(x,y)|\leq h.$$ Following the seminal work of Thue in 1909, several papers were written giving an upper bound for the number of solutions of the above inequality as $\ll c(r,s,h)$ where $c(r,s,h)$ is

Jiquan Shan, Junxiao Wang, Lifeng Zhao, Liang Cai

Recently, vision transformers (ViTs) have achieved excellent performance on vision tasks by measuring the global self-attention among the image patches. Given $n$ patches, they will have quadratic complexity such as $\mathcal{O}(n^2)$ and the time cost is high when splitting the input image with a small granularity. Meanwhile, the pivotal information is ofte

Sagar Deep Deb, Suvakash Dey, Deepak K. Agrawal

Ventilator dyssynchrony (VD) is often described as a mismatch between a patient breathing effort and the ventilator support during mechanical ventilation. This mismatch is often associated with an increased risk of lung injury and longer hospital stays. The manual VD detection method is unreliable and requires considerable effort from medical professionals.

Sekhar Ghosh, R. Lakshmi, Chao Zhang

This work focuses on the nonhomogeneous nonlocal double phase problem \begin{align*} L_au(x)=f(x,u,D_s^p u, D_{a,t}^q u) \text{ in } \Omega, \end{align*} where $\Omega\subset\mathbb{R}^N$ is a bounded domain with Lipschitz boundary, $0<s,t<1<p\leq q<\infty$ with $tq\leq sp$ and the operator $L_a$ is defined as \begin{align*} L_a u(x)&=2\operatorname{P.V.}\in

Ikhlasul Akmal Hanif, Eryawan Presma Yulianrifat, Jaycent Gunawan Ongris, Eduardus Tjitrahardja

This paper presents our approach for SemEval 2025 Task 11 Track A, focusing on multilabel emotion classification across 28 languages. We explore two main strategies: fully fine-tuning transformer models and classifier-only training, evaluating different settings such as fine-tuning strategies, model architectures, loss functions, encoders, and classifiers. O

Guiyao Tie, Xueyang Zhou, Tianhe Gu, Ruihang Zhang

Recent advances in Multi-Modal Large Language Models (MLLMs) have enabled unified processing of language, vision, and structured inputs, opening the door to complex tasks such as logical deduction, spatial reasoning, and scientific analysis. Despite their promise, the reasoning capabilities of MLLMs, particularly those augmented with intermediate thinking tr

G. González-Torà, A. A. C. Sander, N. Moens, J. O. Sundqvist

Classical Wolf-Rayet (cWR) stars are evolved massive stars that have lost most of their H envelope and exhibit dense, extended atmospheres with strong, line-driven winds. Accurately modeling wind launching from optically thick layers remains a challenge. Two main approaches have advanced our understanding: 1D stationary atmosphere models with consistent hydr

Atsuya Hasegawa, François Le Gall, Augusto Modanese

We present relation problems whose input size is $n$ such that they can be solved with no communication for entanglement-assisted quantum communication models, but require $\Omega(n)$ qubit communication for $2$-way quantum communication models without prior shared entanglement. This is the maximum separation of quantum communication complexity with and with

Mengzhu Liu, Zhengqiu Zhu, Chuan Ai, Chen Gao

During sudden disaster events, accurately predicting public panic sentiment on social media is crucial for proactive governance and crisis management. Current efforts on this problem face three main challenges: lack of finely annotated data hinders emotion prediction studies, unmodeled risk perception causes prediction inaccuracies, and insufficient interpre

Shuhei Yoshida

The ghost imaging (GI) technique, which has attracted attention as a highly sensitive and noise-resistant technique, employs a spatially modulated illuminating light and a single-pixel detector. Generally, the information acquired by GI is the transmittance or reflectance distribution of an object. A method has also been proposed to measure the complex ampli

Qu He, Weikun Li, Guangmin Dai, Hao Chen

Fish have endured millions of years of evolution, and their distinct rigid-flexible body structures offer inspiration for overcoming challenges in underwater robotics, such as limited mobility, high energy consumption, and adaptability. This paper introduces SpineWave, a biomimetic robotic fish featuring a fish-spine-like rigid-flexible transition structure.

Mohamed S. Elmahdy, Marius Staring, Patrick J. H. de Koning, Samer Alabed

Accurate and efficient quantification of cardiac function is essential for the estimation of prognosis of cardiovascular diseases (CVDs). One of the most commonly used metrics for evaluating cardiac pumping performance is left ventricular ejection fraction (LVEF). However, LVEF can be affected by factors such as inter-observer variability and varying pre-loa

Gianmarco Perrupato, Tommaso Rizzo

We analytically solve the critical dynamics of the Kob-Andersen kinetically constrained model of supercooled liquids on the Bethe lattice, employing a combinatorial argument based on the cavity method. For arbitrary values of graph connectivity z and facilitation parameter m, we demonstrate that the critical behavior of the order parameter is governed by equ

Gilles Courtois, Pablo Lessa, Emiliano Sequeira

We study the volume growth of horospheres in a Heintze group of the form R ___ A R d with A a diagonal derivation. We conclude that the isometry and quasi-isometry classes of horospheres (with their intrinsic geometry) coincide. Furthermore, if A is not a scalar multiple of the identity, then there are exactly two such classes, characterized by their volume

Gianmarco Del Sarto, Matthias Hieber, Tarek Zöchling

This article investigates an energy balance model coupled to the primitive equations by a dynamic boundary condition with and without noise on the boundary. It is shown that this system is globally strongly well-posed both in the deterministic setting for arbitrary large data in $W^{2(1-\frac{1}{p}),p}$ for $p \in [2,\infty)$ and in the stochastic setting fo

Aditya Bhattacharya, Simone Stumpf, Katrien Verbert

As Artificial Intelligence (AI) becomes increasingly integrated into high-stakes domains like healthcare, effective collaboration between healthcare experts and AI systems is critical. Data-centric steering, which involves fine-tuning prediction models by improving training data quality, plays a key role in this process. However, little research has explored

Renfei Dang, Zhening Li, Shujian Huang, Jiajun Chen

Reasoning models often exhibit overthinking, characterized by redundant reasoning steps. We identify \emph{internal bias} elicited by the input question as a key trigger of such behavior. Upon encountering a problem, the model immediately forms a preliminary guess about the answer, which we term an internal bias since it may not be explicitly generated, and

Oliver Grainge, Michael Milford, Indu Bodala, Sarvapali D. Ramchurn

TAT-VPR is a ternary-quantized transformer that brings dynamic accuracy-efficiency trade-offs to visual SLAM loop-closure. By fusing ternary weights with a learned activation-sparsity gate, the model can control computation by up to 40% at run-time without degrading performance (Recall@1). The proposed two-stage distillation pipeline preserves descriptor qua

Zhaoxin Wang, Handing Wang, Cong Tian, Yaochu Jin

Multimodal large language models (MLLMs) enable powerful cross-modal reasoning capabilities. However, the expanded input space introduces new attack surfaces. Previous jailbreak attacks often inject malicious instructions from text into less aligned modalities, such as vision. As MLLMs increasingly incorporate cross-modal consistency and alignment mechanisms

Henrik Abgaryan, Tristan Cazenave, Ararat Harutyunyan

Large Language Models (LLMs) have demonstrated impressive reasoning capabilities, yet their direct application to NP-hard combinatorial problems (CPs) remains underexplored. In this work, we systematically investigate the reasoning abilities of LLMs on a variety of NP-hard combinatorial optimization tasks and introduce ACCORD: Autoregressive Constraint-satis

Xiaotian Zhao, Zixuan Li, Yichen Cai, Tianju Wang

Dataflow is a critical yet underexplored factor in automatic macro placement, which is becoming increasingly important for developing intelligent design automation techniques that minimize reliance on manual adjustments and reduce design iterations. Existing macro or mixed-size placers with dataflow awareness primarily focus on intrinsic relationships among

Jonas Stein, Jannis Lutz, Moritz Sölderer, Maximilian Adler

Quantum Simulation-based Optimization (QuSO) is a recently proposed class of optimization problems that entails industrially relevant problems characterized by cost functions or constraints that depend on summary statistic information about the simulation of a physical system or process. This work extends initial theoretical results that proved an up-to-expo

Daniele Avitabile, Francesca Cavallini, Svetlana Dubinkina, Gabriel J. Lord

We develop and analyse numerical schemes for uncertainty quantification in neural field equations subject to random parametric data in the synaptic kernel, firing rate, external stimulus, and initial conditions. The schemes combine a generic projection method for spatial discretisation to a stochastic collocation scheme for the random variables. We study the

Yichen Li, Qiankun Liu, Zhenchao Jin, Jiuzhe Wei

Small object detection in intricate environments has consistently represented a major challenge in the field of object detection. In this paper, we identify that this difficulty stems from the detectors' inability to effectively learn discriminative features for objects of small size, compounded by the complexity of selecting high-quality small object sample

Jisu Han, Jaemin Na, Wonjun Hwang

Test-time adaptation aims to adapt to realistic environments in an online manner by learning during test time. Entropy minimization has emerged as a principal strategy for test-time adaptation due to its efficiency and adaptability. Nevertheless, it remains underexplored in continual test-time adaptation, where stability is more important. We observe that th

S. Paul, K. P. Arunbabu, M. Chakraborty, S. K. Gupta

The GRAPES-3 large area muon telescope with its sixteen independent modules records the high energy (>1 GeV) muons continuously over 2.3 sr of the sky. However, the recorded muon rates are contaminated by instrumental effects and instabilities spanning both short- and long-timescales, such as variations in the efficiency of the detector. We present an automa

Jiazhi Mo, Hailu Kuang, Xiaoqi Li

As network security issues continue gaining prominence, password security has become crucial in safeguarding personal information and network systems. This study first introduces various methods for system password cracking, outlines password defense strategies, and discusses the application of machine learning in the realm of password security. Subsequently

Sangita Goswami, Sudipta Das

In the present work, we have studied the dynamics of accelerating universe considering a simple parametrization of the equation of state parameter in an interacting scenario. In this toy model, the dark energy component is allowed to interact with the dark matter component through a source term. The expressions for various relevant cosmological parameters fo

Heide Narnhofer

Generalizing the relation between spin-systems and Fermi-systems on the lattice we construct for a spin-system with dimension d an algebra for which quasifree time-evolutions exist. With appropriate assumptions the gauge invariant subalgebra common for both algebras is invariant under this time-evolution and on this subalgebra is norm-asymptotically abelian.

Bin Chen, Zezhou Hu, Xin-Cheng Mao

In this paper, we investigate the quantum field theory in Klein space that has two time directions. To study the canonical quantization, we select the ``length of time" $q$ as the evolution direction of the system. In our novel construction, some additional modes beyond the plane wave modes are crucial in the canonical quantization and the later derivation o