May 2025 arXiv papers — page 3

Basudha Pal, Rama Chellappa, Muhammad Umair

While echocardiography and MRI are clinical standards for evaluating cardiac structure, their use is limited by cost and accessibility.We introduce a direct classification framework that predicts severe left ventricular hypertrophy from chest X-rays, without relying on anatomical measurements or demographic inputs. Our approach achieves high AUROC and AUPRC,

Josh Givens, Song Liu, Henry W J Reeve

Score matching is a vital tool for learning the distribution of data with applications across many areas including diffusion processes, energy based modelling, and graphical model estimation. Despite all these applications, little work explores its use when data is incomplete. We address this by adapting score matching (and its major extensions) to work with

Ayon Mukherjee, Oleksandr Sverdlov, Ngoc-Thuy Ha, Yu Deng

Randomized discontinuation design (RDD) is an enrichment strategy commonly used to address limitations of traditional placebo-controlled trials, particularly the ethical concern of prolonged placebo exposure. RDD consists of two phases: an initial open-label phase in which all eligible patients receive the investigational medicinal product (IMP), followed by

Peng Xia, Jinglu Wang, Yibo Peng, Kaide Zeng

Medical Large Vision-Language Models (Med-LVLMs) have shown strong potential in multimodal diagnostic tasks. However, existing single-agent models struggle to generalize across diverse medical specialties, limiting their performance. Recent efforts introduce multi-agent collaboration frameworks inspired by clinical workflows, where general practitioners (GPs

Hadi Hosseini, Grzegorz Lisowski, Shraddha Pathak

The Deferred Acceptance (DA) algorithm is an elegant procedure for finding a stable matching in two-sided matching markets. It ensures that no pair of agents prefers each other to their matched partners. In this work, we initiate the study of two-sided manipulations in matching markets as non-cooperative games. We introduce the accomplice manipulation game,

Xintong Jiang

In this paper, we prove the canonical bundle formula for Fano type fibrations and Shokurov's conjecture on boundedness of complements for Fano type threefold pairs $(X,B)$ with fibration structures in large characteristics. In particular, we prove the conjecture when $-(K_X+B)\not\equiv 0$ is nef and not big in large characteristics.

Pingping Zeng, Gongqiu Zhang, Weinan Zhang

Drawdown risk, an important metric in financial risk management, poses significant computational challenges due to its highly path-dependent nature. This paper proposes a unified framework for computing five important drawdown quantities introduced in Landriault et al. (2015) and Zhang (2015) under general Markov models. We first establish linear systems and

Ming Wang, Peidong Wang, Lin Wu, Xiaocui Yang

Constrained by the cost and ethical concerns of involving real seekers in AI-driven mental health, researchers develop LLM-based conversational agents (CAs) with tailored configurations, such as profiles, symptoms, and scenarios, to simulate seekers. While these efforts advance AI in mental health, achieving more realistic seeker simulation remains hindered

Márcio Ferreira, Constança Providência

The slope of the neutron star mass-radius curve, $dM/dR$, is studied to understand the information it may carry about the composition of neutron stars, particularly with regard to the presence of non-nucleonic degrees of freedom. This study uses two large sets of relativistic mean-field equations of state with either nucleonic or nucleonic and hyperonic degr

Jiahui Geng, Thy Thy Tran, Preslav Nakov, Iryna Gurevych

Existing attacks against multimodal language models (MLLMs) primarily communicate instructions through text accompanied by adversarial images. In contrast, we exploit the capabilities of MLLMs to interpret non-textual instructions, specifically, adversarial images or audio generated by our novel method, Con Instruction. We optimize these adversarial examples

Jonathan B. Hill

We establish a generic symmetrization property for dependent random variables $\{x_{t}\}_{t=1}^{n}$ on $\mathbb{R}^{p}$, where $p$ $>>$ $n$ is allowed. We link $\mathbb{E}\psi (\max_{1\leq i\leq p}|1/n\sum_{t=1}^{n}(x_{i,t}$ $-$ $\mathbb{E}x_{i,t})|)$ to $\mathbb{E}\psi (\max_{1\leq i\leq p}|1/n$ $\sum_{t=1}^{n}\eta _{t}(x_{i,t}$ $-$ $\mathbb{E}% x_{i,t})|)$

Zhaoying Wang, Xingxing Zuo, Wei Dong

Lightweight long-range mapping is critical for safe navigation of UAV swarms in large-scale unknown environments. Traditional stereo vision systems with fixed short baselines face limited perception ranges. To address this, we propose Flying Co-Stereo, a cross-agent collaborative stereo vision system that leverages the wide-baseline spatial configuration of

Mehdi Bejani, Guillermo Perez-de-Arenaza-Pozo, Julián D. Arias-Londoño, Juan I. Godino-LLorente

Missing data is a relevant issue in time series, especially in biomedical sequences such as those corresponding to smooth pursuit eye movements, which often contain gaps due to eye blinks and track losses, complicating the analysis and extraction of meaningful biomarkers. In this paper, a novel imputation framework is proposed using Self-Attention-based Impu

Levin Maier

In this article, we combine V. Arnold's celebrated approach via the Euler-Arnold equation -- describing the geodesic flow on a Lie group equipped with a right-invariant metric \cite{Arnold66} -- with his formulation of the motion of a charged particle in a magnetic field \cite{ar61}. We introduce the \emph{magnetic Euler-Arnold equation}, which is the Euleri

Mattias Ergon, Maximilian Stritzinger, Francesco Taddia, Jesper Sollerman

We present HYDE, a new one-dimensional hydrodynamical code, and use it to construct a grid of supernova (SN) models based on solar-metallicity bare helium-core models evolved to the verge of core-collapse with MESA STAR. This grid is suited to model Type IIb SNe, which progenitor stars are thought to have lost all but a tiny fraction of their hydrogen envelo

Kaustabh Mondal, Gunja Sachdeva

Let $G$ be a connected semisimple simply connected Lie group with a compact Cartan subgroup and let $\Gamma$ be a uniform lattice in $G$. Let $\widehat{G}_d$ denote the set of equivalence classes of unitary discrete series representations of $G$. We prove that for any finite subset of $\widehat{G}_d$ satisfying a certain condition, the associated finite set

Huayu Huang, Banglei Guan, Yang Shang, Qifeng Yu

Photomechanics is a crucial branch of solid mechanics. The localization of point targets constitutes a fundamental problem in optical experimental mechanics, with extensive applications in various missions of UAVs. Localizing moving targets is crucial for analyzing their motion characteristics and dynamic properties. Reconstructing the trajectories of points

Gogulakrishnan Thiyagarajan, Prabhudarshi Nayak

Containerization, driven by Docker, has transformed application development and deployment by enhancing efficiency and scalability. However, the rapid adoption of container technologies introduces significant security challenges that require careful management. This paper investigates key areas of container security, including runtime protection, network saf

Alexandre Bouras, Audrey Durand, Richard Khoury

This study explores strategies for optimizing news headline recommendations through preference-based learning. Using real-world data of user interactions with French-language online news posts, we learn a headline recommender agent under a contextual bandit setting. This allows us to explore the impact of translation on engagement predictions, as well as the

Wenzhang Liu, Muqaddar Abbas, Pei Zhang, Jiawei Lai

We present a theoretical study demonstrating enhanced tunability of the photonic spin Hall effect (PSHE) using a strongly interacting Rydberg atomic medium under electromagnetically induced transparency (EIT) conditions. In contrast to conventional approaches that rely on static refractiveindex profiles or metamaterials, here the PSHE is controlled via a non

Ruihan Yang, Yikai Zhang, Aili Chen, Xintao Wang

Large language models (LLMs) have enabled agents to perform complex reasoning and decision-making through free-form language interactions. However, in open-ended language action environments (e.g., negotiation or question-asking games), the action space can be formulated as a joint distribution over tokens, resulting in an exponentially large action space. S

Daniel Jezierski, Wojciech Grochala

Based on theoretical calculations, we introduce a new AgIIAgIIIF5 monoclinic polymorph with a rare high spin AgIII. Our analysis of the experimental xray diffraction data available in the literature reveals that this polymorph was likely prepared in the past in a mixture with the triclinic form of the same compound. Theoretical calculations reproduce very we

Shunyi Lan, Xiangcun Meng

Millisecond pulsars are rapidly rotating neutron stars, and it is now widely accepted that their extremely short rotation periods result from the accretion of material from a companion star. Binary evolution theory predicts that millisecond pulsars can have various types of companion stars. However, in observations, binary pulsars with giant companions, refe

Changyue Wang, Weihang Su, Qingyao Ai, Yujia Zhou

Knowledge editing aims to efficiently update Large Language Models (LLMs) by modifying specific knowledge without retraining the entire model. Among knowledge editing approaches, in-context editing (ICE) offers a lightweight solution by injecting new knowledge directly into the input context, leaving model parameters unchanged. However, existing ICE approach

L. Nađđerđ, J. Milošević, D. Devetak, F. Wang

The method of Q-cumulants has been shown as a powerful tool to study the fine details of the azimuthal anisotropies in high-energy nucleus-nucleus collisions. A new method for the fast calculation of arbitrary order Q-cumulant $v_{n}\{2k\}$ values, based on the partition of a non-negative integer l $\le$ m for calculation of the 2m-particle azimuthal correla

Yudong Zhang, Ruobing Xie, Xingwu Sun, Jiansheng Chen

The choice of a suitable visual language projector (VLP) is critical to the successful training of large visual language models (LVLMs). Mainstream VLPs can be broadly categorized into compressed and uncompressed projectors, and each offers distinct advantages in performance and computational efficiency. However, their security implications have not been tho

Junquan Huang, Zong-Gan Chen, Yuncheng Jiang, Zhi-Hui Zhan

GCN-based traveling salesman problem (TSP) solvers face two critical challenges: poor cross-scale generalization for TSPs and high training costs. To address these challenges, we propose a Subgraph-Based Rescaling Graph Convolutional Network (RsGCN). Focusing on the scale-dependent features (i.e., features varied with problem scales) related to nodes and edg

Fasheng Xu, Jing Hou, Wei Chen, Karen Xie

Generative AI (GenAI) is rapidly transforming knowledge work, yet its implications for organizational hierarchies remain poorly understood. Unlike earlier automation technologies, GenAI can both perform tasks autonomously and assist human workers, while its intrinsic fallibility, the tendency to produce confident but incorrect outputs, demands continuous hum

Mohammad Alakhrass

In this paper, we investigate the generalized numerical radius $\omega_N$, associated with a matrix norm $N$ defined by $\omega_N(X) = \sup_{\theta \in \mathbb{R}} N(\operatorname{Re}(e^{i\theta}X))$. We focus on matrices whose numerical ranges are contained in sectors of the complex plane (sectorial matrices) and derive upper bounds for $\omega_N(XY)$ and $

Hang Fana, Mingxuan Lib, Zuhan Zhanga, Long Chengc

The integration of wind energy into power grids necessitates accurate ultra-short-term wind power forecasting to ensure grid stability and optimize resource allocation. This study introduces M2WLLM, an innovative model that leverages the capabilities of Large Language Models (LLMs) for predicting wind power output at granular time intervals. M2WLLM overcomes

Tianhui Liu, Hetian Pang, Xin Zhang, Tianjian Ouyang

Understanding urban socioeconomic conditions through visual data is a challenging yet essential task for sustainable urban development and policy planning. In this work, we introduce \textit{CityLens}, a comprehensive benchmark designed to evaluate the capabilities of Large Vision-Language Models (LVLMs) in predicting socioeconomic indicators from satellite

Souvik Dey, Dipankar Ghosh, Siddhartha Pramanik, Tony J. Puthenpurakal

Let $R$ be a Noetherian ring, $I_1,\ldots,I_r$ be ideals of $R$, and $N\subseteq M$ be finitely generated $R$-modules. Let $S = \bigoplus_{\underline{n} \in \mathbb{N}^r} S_{\underline{n}}$ be a Noetherian standard $\mathbb{N}^r$-graded ring with $S_{\underline{0}} = R$, and $\mathcal{M} $ be a finitely generated $\mathbb{Z}^r$-graded $S$-module. For $ \unde

Richard Connor, Alan Dearle, Ben Claydon

Many modern search domains comprise high-dimensional vectors of floating point numbers derived from neural networks, in the form of embeddings. Typical embeddings range in size from hundreds to thousands of dimensions, making the size of the embeddings, and the speed of comparison, a significant issue. Quantisation is a class of mechanism which replaces the

Runtao Ren, Jian Ma, Jianxi Luo

Retrieval-Augmented Generation (RAG) systems in the Intellectual Property (IP) field often struggle with diverse user queries, including colloquial expressions, spelling errors, and ambiguous terminology, leading to inaccurate retrieval and suboptimal responses. To address this challenge, we propose Multi-Angle Question Generation and Retrieval Fine-Tuning M

Chenhao Wu, Qingbo Wu, Haoran Wei, Shuai Chen

With the help of powerful generative models, Semantic Image Compression (SIC) has achieved impressive performance at ultra-low bitrate. However, due to coarse-grained visual-semantic alignment and inherent randomness, the reliability of SIC is seriously concerned for reconstructing completely different object instances, even they are semantically consistent

Mehran Z-Abyaneh, Mehrdad Farhoudi

We reduce the Taub-NUT metric dimensionally to three spatial dimensions by treating time as an extra curled dimension, and end up with the 3-dimensional Einstein field equations plus a corresponding Maxwell type equations for a gravitomagnetic field, associated with the NUT charge, which also acts as a source for the Einstein field equations. In this approac

Hui Li, Jie Xie, Hyukjoon Kwon, Yixin Zhao

Fluctuation theorems have elevated the second law of thermodynamics to a statistical realm by establishing a connection between time-forward and time-reversal probabilities, providing invaluable insight into nonequilibrium dynamics. While well established in classical systems, their quantum generalization, incorporating coherence and the diversity of quantum

Xingtong Ge, Xin Zhang, Tongda Xu, Yi Zhang

The Distribution Matching Distillation (DMD) has been successfully applied to text-to-image diffusion models such as Stable Diffusion (SD) 1.5. However, vanilla DMD suffers from convergence difficulties on large-scale flow-based text-to-image models, such as SD 3.5 and FLUX. In this paper, we first analyze the issues when applying vanilla DMD on large-scale

Yinchao Yang, Zhaohui Yang, Chongwen Huang, Wei Xu

This paper introduces a novel framework for integrated sensing, computing, and semantic communication (ISCSC) within vehicular networks comprising a roadside unit (RSU) and multiple autonomous vehicles. Both the RSU and the vehicles are equipped with local knowledge bases to facilitate semantic communication. The framework incorporates a secure communication

Shida Wang, Jalal Fadili, Peter Ochs

In this paper, we study convergence rates of the cubic regularized proximal quasi-Newton method (\csr) for solving non-smooth additive composite problems that satisfy the so-called Kurdyka-\L ojasiewicz (K\L ) property with respect to some desingularization function $\phi$ rather than strong convexity. After a number of iterations $k_0$, Cubic SR1 PQN exhibi

Chenxu Liu, Zhiyu Gu, Guoquan Wu, Ying Zhang

Quality assurance of web applications is critical, as web applications play an essential role in people's daily lives. To reduce labor costs, automated web GUI testing (AWGT) is widely adopted, exploring web applications via GUI actions such as clicks and text inputs. However, these approaches face limitations in generating continuous and meaningful action s

Yuxi Sun, Aoqi Zuo, Wei Gao, Jing Ma

Large Language Models (LLMs) often exhibit knowledge disparities across languages. Encouraging LLMs to \textit{abstain} when faced with knowledge gaps is a promising strategy to reduce hallucinations in multilingual settings. Current abstention strategies for multilingual scenarios primarily rely on generating feedback in various languages using LLMs and per

Tzu-Shen Wang, Jimmy Dani, Juan Garay, Soamar Homsi

In this work, we present an efficient secure multi-party computation MPC protocol that provides strong security guarantees in settings with dishonest majority of participants who may behave arbitrarily. Unlike the popular MPC implementation known as SPDZ [Crypto '12], which only ensures security with abort, our protocol achieves both complete identifiability

Alexey Tikhonov, Sergei Shteiner, Anna Bykova, Ivan P. Yamshchikov

Russenorsk, a pidgin language historically used in trade interactions between Russian and Norwegian speakers, represents a unique linguistic phenomenon. In this paper, we attempt to analyze its lexicon using modern large language models (LLMs), based on surviving literary sources. We construct a structured dictionary of the language, grouped by synonyms and

A. V. Glushkov, K. G. Lebedev, L. T. Ksenofontov, A. V. Saburov

The two most powerful extensive air showers (EAS) with energies of about $10^{20}$ eV, registered at the Yakutsk EAS array during the entire observation period of 1974-2024, are considered. Both showers hit the array near the center and triggered all surface detectors and underground muon detectors with a threshold energy of $E_{\mu} = 1.0 \times \cos \theta

Ketan Vinayak Warghat, Yogesh Biswal, Sukesh Sharma, Pankaj Sharadchandra Kolhe

We experimentally investigate the breakup dynamics of viscoelastic jets composed of carboxymethyl cellulose (CMC) solutions, focusing on the dripping and Rayleigh regimes at low flow rates. By varying the CMC concentration, needle diameter ($D_n$), and flow rate ($Q$), we analyze the effects of elasticity, viscosity, and flow conditions on jet stability and

Yuxiang Zhang, Fan Zhong

Phase transitions, as one of the most intriguing phenomena in nature, are divided into first-order phase transitions (FOPTs) and continuous ones in current classification. While the latter shows striking phenomena of scaling and universality, the former has recently also been demonstrated to exhibit scaling and universal behavior within a mesoscopic, coarse-

Tianhui Zhang, Bei Peng, Danushka Bollegala

In commonsense generation, given a set of input concepts, a model must generate a response that is not only commonsense bearing, but also capturing multiple diverse viewpoints. Numerous evaluation metrics based on form- and content-level overlap have been proposed in prior work for evaluating the diversity of a commonsense generation model. However, it remai

Yaxiong Wang, Zhenqiang Zhang, Lechao Cheng, Zhun Zhong

Test-time adaption (TTA) has witnessed important progress in recent years, the prevailing methods typically first encode the image and the text and design strategies to model the association between them. Meanwhile, the image encoder is usually frozen due to the absence of explicit supervision in TTA scenarios. We identify a critical limitation in this parad

Yang Zheng, Mengqi Huang, Nan Chen, Zhendong Mao

Text-guided 3D editing aims to precisely edit semantically relevant local 3D regions, which has significant potential for various practical applications ranging from 3D games to film production. Existing methods typically follow a view-indiscriminate paradigm: editing 2D views indiscriminately and projecting them back into 3D space. However, they overlook th

A. M. Shikin, N. L. Zaitsev, A. V. Eryzhenkov, R. V. Makeev

Magnetic topological systems based on MnBi2Te4 have recently attracted significant attention due to their rich interplay between magnetism and topological electronic states. In this work, using density functional theory (DFT), we investigate topological phase transitions (TPTs) in Mn1-xGexBi2Te4 compounds with both ferromagnetic (FM) and antiferromagnetic (A

Sujit Janaa, Vikash Kumar Ojha

We investigate the quark Wigner distribution in a frame-independent, three-dimensional position space within the framework of the dressed quark model. It is observed that the distributions are concentrated near the center of the target and gradually diminish as one moves away in both the longitudinal and transverse directions. The distribution exhibits symme

Zherui Li, Yan Mi, Zhenhong Zhou, Houcheng Jiang

Large Language Model-based Multi-Agent Systems (MASs) have demonstrated strong advantages in addressing complex real-world tasks. However, due to the introduction of additional attack surfaces, MASs are particularly vulnerable to misinformation injection. To facilitate a deeper understanding of misinformation propagation dynamics within these systems, we int

Muhammad Islam, Tao Huang, Euijoon Ahn, Usman Naseem

This paper presents an in-depth survey on the use of multimodal Generative Artificial Intelligence (GenAI) and autoregressive Large Language Models (LLMs) for human motion understanding and generation, offering insights into emerging methods, architectures, and their potential to advance realistic and versatile motion synthesis. Focusing exclusively on text

Andrea Civilini, Fabrizio de Vico Fallani, Vito Latora

Identifying patterns of relations among the units of a complex system from measurements of their activities in time is a fundamental problem with many practical applications. Here, we introduce a method that detects dependencies of any order in multivariate time series data. The method first transforms a multivariate time series into a symbolic sequence, and

Dohyun Lee, Seungil Chad Lee, Chanwoo Yang, Yujin Baek

Large language models (LLMs) have demonstrated strong performance across various tasks, leveraging their exceptional in-context learning ability with only a few examples. Accordingly, the selection of optimal in-context examples has been actively studied in the field of machine translation. However, these studies presuppose the presence of a demonstration po

Marie Kunešová, Aleš Pražák, Jan Lehečka

We present a system for non-intrusive prediction of speech quality in noisy and enhanced speech, developed for Track 3 of the VoiceMOS 2024 Challenge. The task required estimating the ITU-T P.835 metrics SIG, BAK, and OVRL without reference signals and with only 100 subjectively labeled utterances for training. Our approach uses wav2vec 2.0 with a two-stage

Hanxiao Qu, Krzysztof Gogol, Florian Groetschla, Claudio Tessone

Decentralized Finance (DeFi) lending enables permissionless borrowing via smart contracts. However, it faces challenges in optimizing interest rates, mitigating bad debt, and improving capital efficiency. Rule-based interest-rate models struggle to adapt to dynamic market conditions, leading to inefficiencies. This work applies Offline Reinforcement Learning

M. S. Guimaraes, I. Roditi, S. P. Sorella

The violation of the Bell-CHSH inequality in a relativistic scalar Quantum Field Theory is analysed by means of a set of bounded Hermitian operators constructed out of the unitary Weyl operators. These operators allow for both analytic and numerical approaches. While the former relies on the modular theory of Tomita-Takesaki, the latter is devised through an

Viola Hegedűs, Szabolcs Mészáros, Blanka Világos, Marco Pignatari

We obtain two-infall galactic chemical evolution (GCE) models simulating the chemical evolution of the Milky Way as constrained by a golden sample of $394,000$ stellar abundances of the Milky Way Mapper survey from the 19th data release of SDSS-V. The separation between the chemical thin and thick disks is defined using [Mg/M]. We use the chemical evolution

Mingxin Wei, Jinrui Zhang, Peter Groot Koerkamp, Andre Aarnink

The management of thermal environments in pig farming is crucial for optimizing animal health, productivity, and operational energy efficiency. This study introduces a novel thermal ventilation model (TVM) based on enthalpy balance, which integrates both temperature and humidity control to address the specific thermal regulation requirements of pig housing i

Vassilis Apidopoulos, Cesare Molinari, Juan Peypouquet, Silvia Villa

We introduce and analyze a continuous primal-dual dynamical system in the context of the minimization problem $f(x)+g(Ax)$, where $f$ and $g$ are convex functions and $A$ is a linear operator. In this setting, the trajectories of the Arrow-Hurwicz continuous flow may not converge, accumulating at points that are not solutions. Our proposal is inspired by the

Krzysztof Gogol, Szczepan Gurgul, Faizan Nehal Siddiqui, David Branes

Ethereum's scalability limitations pose significant challenges for the adoption of decentralized applications (dApps). Zero-Knowledge Rollups (ZK Rollups) present a promising solution, bundling transactions off-chain and submitting validity proofs on-chain to enhance throughput and efficiency. In this work, we examine the technical underpinnings of ZK Rollup

Diogo Landau, Ingeborg de Pater, Mihaela Mitici, Nishant Saurabh

Complex systems such as aircraft engines are continuously monitored by sensors. In predictive aircraft maintenance, the collected sensor measurements are used to estimate the health condition and the Remaining Useful Life (RUL) of such systems. However, a major challenge when developing prognostics is the limited number of run-to-failure data samples. This c

Raghav Mehta, Karthik Gopinath, Ben Glocker, Juan Eugenio Iglesias

We propose UNSURF, a novel uncertainty measure for cortical surface reconstruction of clinical brain MRI scans of any orientation, resolution, and contrast. It relies on the discrepancy between predicted voxel-wise signed distance functions (SDFs) and the actual SDFs of the fitted surfaces. Our experiments on real clinical scans show that traditional uncerta

Anders Malthe Westerkam, Alba Spliid Damkjær, Rasmus Erik Villadsen, Magnus Ørum Bastrup Poulsen

We investigate the second-order characteristics of the radar return signal from a swarm of rotor drones. We consider the case of a swarm of identical drones, with each a number of rotors comprised of a number of rotor blades. By considering the orientation and speed of each rotor as stochastic variables, we derive expressions for the autocorrelation function

Ashutosh Gupta, Thomas A. Henzinger, Konstantin Kueffner, Kaushik Mallik

Input-output robustness appears in various different forms in the literature, such as robustness of AI models to adversarial or semantic perturbations and individual fairness of AI models that make decisions about humans. We propose runtime monitoring of input-output robustness of deployed, black-box AI models, where the goal is to design monitors that would

Xinyi Wang, Lirong Gao, Haobo Wang, Yiming Zhang

Parameter-Efficient Fine-Tuning (PEFT) methods have emerged as a widely adopted strategy for adapting pre-trained Large Language Models (LLMs) to downstream tasks, significantly reducing memory and computational costs. However, most existing PEFT techniques uniformly deploy LoRA adapters across all layers, disregarding the intrinsic heterogeneity of layer co

Ali Ghanizadeh, Ali Ahmadi, Arash Bahrami

The internal structure of the Fin-Ray fingers plays a significant role in their adaptability and grasping performance. However, modeling the grasp force and deformation behavior for design purposes is challenging. When the Fin-Ray finger becomes more rigid and capable of exerting higher forces, it becomes less delicate in handling objects. The contrast betwe

Hans Kjeldsen, Timothy R. Bedding, Yaguang Li, Frank Grundahl

We present time-series radial velocities of the G8 subgiant star beta Aql obtained in 2022 and 2023 using SONG-Tenerife and, for the first time, SONG-Australia. We also analyse a sector of TESS photometry that overlapped with the 2022 SONG data. The resulting power spectrum clearly shows solar-like oscillations centred at 430 muHz. The TESS light curve shows

Ankita Negi, Leon Merten Lohse, Sven Velten, Ilya Sergeev

Phase retrieval is at the heart of adaptive optics and modern high-resolution imaging. Without phase information, optical systems are limited to intensity-only measurements, hindering full reconstruction of object structures and wavefront dynamics essential for advanced applications. Here, we address a one-dimensional phase problem linking energy and time, w

Linhao Ye, Lang Yu, Zhikai Lei, Qin Chen

Retrieval-augmented generation (RAG) is usually integrated into large language models (LLMs) to mitigate hallucinations and knowledge obsolescence. Whereas,conventional one-step retrieve-and-read methods are insufficient for multi-hop question answering, facing challenges of retrieval semantic mismatching and the high cost in handling interdependent subquest

Shaofeng Zhang, Shengcai Liu, Ning Lu, Jiahao Wu

Combinatorial optimization problems are widely encountered in real-world applications. A critical research challenge lies in designing high-quality heuristic algorithms that efficiently approximate optimal solutions within a reasonable time. In recent years, many works have explored integrating Large Language Models (LLMs) with Evolutionary Algorithms to aut

Valeria Giunta, Thomas Hillen, Mark A. Lewis, Jonathan R. Potts

From tumour invasion to cell sorting and animal territoriality, many biological systems rely on nonlocal interactions that drive complex spatial organisation. Partial differential equations (PDEs) with nonlocal advection are increasingly recognised as powerful tools for capturing such phenomena. However, most research has focused on one-dimensional domains,

Shuguo Hu, Jun Hu, Huaiwen Zhang

Large Language Models (LLMs) can assist multimodal fake news detection by predicting pseudo labels. However, LLM-generated pseudo labels alone demonstrate poor performance compared to traditional detection methods, making their effective integration non-trivial. In this paper, we propose Global Label Propagation Network with LLM-based Pseudo Labeling (GLPN-L

Carlisle Casuga, Henri Hänninen, Heikki Mäntysaari

We determine the initial condition of the Balitsky-Kovchegov evolution equation at next-to-leading order (NLO) accuracy using HERA deep inelastic scattering data. Posterior distributions characterizing the initial condition are extracted using Bayesian inference. The total cross section and charm quark production data from HERA are found to provide stringent

Jun Wu, Patrick Huang, Jiangtao Wen, Yuxing Han

Despite rapid progress in large language models (LLMs), the statistical structure of their weights, activations, and gradients-and its implications for initialization, training dynamics, and efficiency-remains largely unexplored. We empirically show that these quantities in LLMs are well modeled by generalized Gaussian (GG) distributions, and introduce a uni

Levin Maier

In this paper, we introduce \emph{$\ell^p$-information geometry}, an infinite dimensional framework that shares key features with the geometry of the space of probability densities \( \mathrm{Dens}(M) \) on a closed manifold, while also incorporating aspects of measure-valued information geometry. We define the \emph{$\ell^2$-probability simplex} with a nonc

Denise Moerel, Tijl Grootswagers

In cognitive neuroscience research, Representational Dissimilarity Matrices (RDMs) are often incomplete because pairwise similarity judgments cannot always be exhaustively collected as the number of pairs rapidly increases with the number of conditions. Existing methods to fill these missing values, such as deep neural network imputation, are powerful but co

Aviv Jan, Dean Tahory, Omer Talmi, Omar Abo Mokh

Multi-hop questions still stump large language models (LLMs), which struggle to link information across multiple reasoning steps. We introduce Auto-Patch, a novel method that dynamically patches hidden states during inference to enhance multi-hop reasoning in LLMs. Building on the PatchScopes framework, Auto-Patch selectively modifies internal representation

Eunsu Kim, Haneul Yoo, Guijin Son, Hitesh Patel

As large language models (LLMs) continue to advance, the need for up-to-date and well-organized benchmarks becomes increasingly critical. However, many existing datasets are scattered, difficult to manage, and make it challenging to perform evaluations tailored to specific needs or domains, despite the growing importance of domain-specific models in areas su

Junseo Kim, Jongwook Han, Dongmin Choi, Jongwook Yoon

Visual persuasion, which uses visual elements to influence cognition and behaviors, is crucial in fields such as advertising and political communication. With recent advancements in artificial intelligence, there is growing potential to develop persuasive systems that automatically generate persuasive images tailored to individuals. However, a significant bo

Yameng Liu, Jianhua Zhang, Yuxiang Zhang, Hongbo Xing

Integrated Sensing And Communication (ISAC) has been identified as a key 6G application by ITU and 3GPP, with standardization efforts already underway. Sensing tasks, such as target localization, demand more precise characterization of the sensing target (ST) in ISAC channel modeling. The ST couples complexly with environmental scatterers, potentially blocki

Mohit Arora, Pratyush Shukla, Shivali Chopra

Unmanned Aerial Vehicles (UAVs) are one of the most revolutionary inventions of 21st century. At the core of a UAV lies the central processing system that uses wireless signals to control their movement. The most popular UAVs are quadcopters that use a set of four motors, arranged as two on either side with opposite spin. An autonomous UAV is called a drone.

Zekun Wang, Minghua Ma, Zexin Wang, Rongchuan Mu

Large Vision-Language Models (LVLMs) have achieved remarkable success, yet their significant computational demands hinder practical deployment. While efforts to improve LVLM efficiency are growing, existing methods lack comprehensive evaluation across diverse backbones, benchmarks, and metrics. In this work, we systematically evaluate mainstream acceleration

Hongjie Zhu, Zezheng Zhang, Zeyu Zhang, Yu Bai

Alternating Current Optimal Power Flow (AC-OPF) aims to optimize generator power outputs by utilizing the non-linear relationships between voltage magnitudes and phase angles in a power system. However, current AC-OPF solvers struggle to effectively represent the complex relationship between variable distributions in the constraint space and their correspond

Leila Mahmoodi, Peyman Moghadam, Munawar Hayat, Christian Simon

We introduce Flashback Learning (FL), a novel method designed to harmonize the stability and plasticity of models in Continual Learning (CL). Unlike prior approaches that primarily focus on regularizing model updates to preserve old information while learning new concepts, FL explicitly balances this trade-off through a bidirectional form of regularization.

Duzhen Zhang, Yong Ren, Zhong-Zhi Li, Yahan Yu

Multimodal Continual Instruction Tuning (MCIT) aims to finetune Multimodal Large Language Models (MLLMs) to continually align with human intent across sequential tasks. Existing approaches often rely on the Mixture-of-Experts (MoE) LoRA framework to preserve previous instruction alignments. However, these methods are prone to Catastrophic Forgetting (CF), as

Abhisek Ray, Lukas Esterle

Federated Learning (FL) has emerged as a powerful paradigm for training machine learning models across distributed data sources while preserving data locality. However, the privacy of local data is always a pivotal concern and has received a lot of attention in recent research on the FL regime. Moreover, the lack of domain heterogeneity and client-specific s

Wei Tao, Xiaoyang Qu, Kai Lu, Jiguang Wan

Since the point cloud data is inherently irregular and unstructured, point cloud semantic segmentation has always been a challenging task. The graph-based method attempts to model the irregular point cloud by representing it as a graph; however, this approach incurs substantial computational cost due to the necessity of constructing a graph for every point w

Yang Zhang, Buqing Nie, Zhanxiang Cao, Yangqing Fu

Reinforcement Learning (RL)-based methods have significantly improved the locomotion performance of legged robots. However, these motion policies face significant challenges when deployed in the real world. Robots operating in uncertain environments struggle to adapt to payload variations and external disturbances, resulting in severe degradation of motion p

Shijun Cheng, Tariq Alkhalifah

Physics-informed neural networks (PINNs) offer a powerful framework for seismic wavefield modeling, yet they typically require time-consuming retraining when applied to different velocity models. Moreover, their training can suffer from slow convergence due to the complexity of of the wavefield solution. To address these challenges, we introduce a latent dif

A. A. Vidotto

There are several physical processes that mediate the interaction between an exoplanet and its host star, with the four main ones being due to magnetic, particle (stellar outflow), radiative and tidal interactions. These interactions can be observed at different wavelengths, from X-ray to radio. Their strengths depend on the architecture of planetary systems

Shaoxiong Ji, Zihao Li, Jaakko Paavola, Hengyu Luo

This paper investigates a critical design decision in the practice of massively multilingual continual pre-training -- the inclusion of parallel data. Specifically, we study the impact of bilingual translation data for massively multilingual language adaptation of the Llama3 family of models to 500 languages. To this end, we construct the MaLA bilingual tran

Jintai Chen, Fangqing Liu, Xueming Yan, Han Huang

This study aims to optimize the evaluation metric of multimodal multi-objective optimization problems using a Regionalized Metric Framework, which provides a certain boost to research in this field. Existing evaluation metrics usually use the reference set as the evaluation basis, which inevitably leads to reference set dependence. To optimize this problem,

Shuai Zhao, Heyan Huang, Xinge Li, Xiaokang Chen

Neural networks have demonstrated exceptional performance in supervised learning, benefiting from abundant high-quality annotated data. However, obtaining such data in real-world scenarios is costly and labor-intensive. Semi-supervised learning (SSL) offers a solution to this problem. Recent studies, such as Semi-ViT and Noisy Student, which employ consisten

Cunhang Fan, Ying Chen, Jian Zhou, Zexu Pan

The brain-assisted target speaker extraction (TSE) aims to extract the attended speech from mixed speech by utilizing the brain neural activities, for example Electroencephalography (EEG). However, existing models overlook the issue of temporal misalignment between speech and EEG modalities, which hampers TSE performance. In addition, the speech encoder in c

Andreas Themelis, Ziyuan Wang

The Bregman proximal mapping and Bregman-Moreau envelope are traditionally studied for functions defined on the entire space $\mathbb{R}^n$, even though these constructions depend only on the values of the function within (the interior of) the domain of the distance-generating function (dgf). While this convention is largely harmless in the convex setting, i

Donghao Qu, Jiajun Zhang, Yajun Wu, Zhang Zhao

This paper presents the design and validation of a digital receiver system developed for the next-generation radio interferometer projects. The receiver supports 8 analog inputs with 12-bit, 4GHz sampling and performs real-time signal processing using FPGA-based channelization. Field experiments were conducted to observe the Sun, a satellite beacon, and Cass

Karl Kunisch, Jesper Schröder

The Kalman(-Bucy) filter is the natural choice for the state reconstruction of disturbed, linear dynamical systems based on flawed and incomplete measurements. Taking a deterministic viewpoint this work investigates possible extensions of the concept to systems with uncertain dynamics and noise covariances. In a theoretical analysis error bounds in terms of