October 2025 arXiv papers — page 88

Ahmet Eren Sertbaş, Tufan Kumbasar

Accurate modeling of nonlinear systems is essential for reliable control, yet conventional identification methods often struggle to capture latent dynamics while maintaining stability. We propose a \textit{stable-by-design LPV neural network-based state-space} (NN-SS) model that simultaneously learns latent states and internal scheduling variables directly f

Lianchen Jia, Chaoyang Li, Qian Houde, Tianchi Huang

Control algorithms in production environments typically require domain experts to tune their parameters and logic for specific scenarios. However, existing research predominantly focuses on algorithmic performance under ideal or default configurations, overlooking the critical aspect of Tuning Potential. To bridge this gap, we introduce Crucible, an agent th

Chi Tat Chong, Tin Lok Wong

Let $\mathfrak M=(M,\mathcal X)$ be a model of $\mathsf{RCA}_0+\text{$\Sigma^0_2$-bounding}$ in which $\Sigma^0_2(A)$-induction fails for some $A\in\mathcal X$. We show that (i) if $\mathfrak M$ is a model of the combinatorial principle Ramsey's Theorem for Pairs, the Cohesive Set Theorem or the Tree Theorem, then there is a $\Delta^0_1(A)$-instance of the p

Jinfeng Liu, Lingtong Kong, Mi Zhou, Jinwen Chen

We introduce Mono4DGS-HDR, the first system for reconstructing renderable 4D high dynamic range (HDR) scenes from unposed monocular low dynamic range (LDR) videos captured with alternating exposures. To tackle such a challenging problem, we present a unified framework with two-stage optimization approach based on Gaussian Splatting. The first stage learns a

Ho Fai Leung, Xiaoyan Xi, Fei Zuo

On-device virtual assistants like Siri and Google Assistant are increasingly pivotal, yet their capabilities are hamstrung by a reliance on rigid, developer-dependent APIs. GUI agents offer a powerful, API-independent alternative, but their adoption is hindered by the perception of poor performance, as even the best models (e.g. Qwen3-VL-235B) scores are cap

L. Votta, M. Hoppe, J. Decker, E. Devlaminck

The Tokamak \`a Configuration Variable (TCV) is equipped with an advanced set of diagnostics for studying suprathermal electron dynamics. Among these, the vertical electron cyclotron emission (VECE) diagnostic offers valuable insights into the electron energy distribution by measuring electron cyclotron emission (ECE) along a vertical line-of-sight. However,

M. R. Eslahchi, E. Kokabifar

Consider a given square matrix $\textrm {K}$ with square blocks $A_{11},A_{22},\ldots,A_{nn}$ on the main diagonal. This paper aims to compute an optimal perturbation $\Delta$ of a preassigned block $A_{ii}\in\mathbb{C}^{d_i\times d_k}, \left(1\le i\le n\right)$,with respect to the spectral norm distance, such that the perturbed matrix ${\textrm {K}_X}$ has

Daniel Bethell, Simos Gerasimou, Radu Calinescu, Calum Imrie

Reliable navigation in safety-critical environments requires both accurate hazard perception and principled uncertainty handling to strengthen downstream safety handling. Despite the effectiveness of existing approaches, they assume perfect hazard detection capabilities, while uncertainty-aware perception approaches lack finite-sample guarantees. We present

Arpan Maity, Tamal Ghosh

This article compares the performance of six prominent object detection algorithms, YOLOv11, RetinaNet, Fast R-CNN, YOLOv8, RT-DETR, and DETR, on the NEU-DET surface defect detection dataset, comprising images representing various metal surface defects, a crucial application in industrial quality control. Each model's performance was assessed regarding detec

Oleksandr Adamov, Anders Carlsson

This paper explores the challenges of cyberattack attribution, specifically APTs, applying the case study approach for the WhisperGate cyber operation of January 2022 executed by the Russian military intelligence service (GRU) and targeting Ukrainian government entities. The study provides a detailed review of the threat actor identifiers and taxonomies used

Haoran Zhang, Chenhao Zhu, Sicong Guo, Hanzhe Guo

Human players do more than press buttons: they ground what they see on screen into precise keyboard-mouse actions and, when stuck, they seek information before trying again. We ask whether current vision-language models (VLMs) can do the same. Despite encouraging results under simplified control or tool scaffolds, human-like play in a real client - mapping r

Daniele Barettin, Stella V. Kavokina, Evgeny L. Ivchenko, Alexey V. Kavokin

Freestanding van der Waals crystals made of single-atom carbon chains (carbynes) have been recently realized technologically. Here we investigate their electronic and optical properties experimentally, by continuous-wave and time-resolved photoluminescence spectroscopy, and theoretically. Employing a fully three-dimensional tight-binding formalism benchmarke

Gaëlle Aymeric, Emmanuelle Lavaine, Brice Magdalou

This paper investigates the causal impact of the parental environment on the student's academic performance in mathematics, literature and English (as a foreign language), using a new database covering all children aged 8 to 15 of the Madrid community, from 2016 to 2019. Parental environment refers here to the parents' level of education (i.e. the skills the

Han Peng, Peiyu Liu, Zican Dong, Daixuan Cheng

Diffusion language models (DLMs) have emerged as a promising alternative to the long-dominant autoregressive (AR) paradigm, offering a parallelable decoding process that could yield greater efficiency. Yet, in practice, current open-source DLMs often underperform their AR counterparts in speed, limiting their real-world utility. This work presents a systemat

Daniel Bethell, Simos Gerasimou, Radu Calinescu, Calum Imrie

Ensuring the safe exploration of reinforcement learning (RL) agents is critical for deployment in real-world systems. Yet existing approaches struggle to strike the right balance: methods that tightly enforce safety often cripple task performance, while those that prioritize reward leave safety constraints frequently violated, producing diffuse cost landscap

Haichao Ji, Zibo Wang, Cheng Pan, Meng Han

Large Language Models (LLMs) have shown great promise in automating data analytics tasks by interpreting natural language queries and generating multi-operation execution plans. However, existing LLM-agent-based analytics frameworks operate under the assumption of centralized data access, offering little to no privacy protection. In contrast, federated analy

Feifan Xia, Yuyang Fang, Defang Li, Yantong Xie

We present a probabilistic intent modeling framework for large language model (LLM) agents in multi-turn social dialogue. The framework maintains a belief distribution over a partner's latent intentions, initialized from contextual priors and dynamically updated through likelihood estimation after each utterance. The evolving distribution provides additional

Mariano Barone, Antonio Laudante, Giuseppe Riccio, Antonio Romano

The extraction of pharmacological knowledge from regulatory documents has become a key focus in biomedical natural language processing, with applications ranging from adverse event monitoring to AI-assisted clinical decision support. However, research in this field has predominantly relied on English-language corpora such as DrugBank, leaving a significant g

Yuya Sasaki

Graph neural networks (GNNs) are powerful tools for learning from graph-structured data but often produce biased predictions with respect to sensitive attributes. Fairness-aware GNNs have been actively studied for mitigating biased predictions. However, no prior studies have evaluated fairness-aware GNNs on knowledge graphs, which are one of the most importa

Gian-Luca Oppo

We all know that the first laser device was realised by Theodore Maiman at Hughes Labs in 1960. Less known is that the very first computer simulations of the relaxation oscillations displayed by Maiman's laser were also performed in 1960 on a digital IBM 704 computer. The reason is that lasers and almost all photonic devices are described by nonlinear equati

Arpan Maity, Aviroop Pal, MD. Samiul Islam, Tamal Ghosh

Diabetic Retinopathy (DR), a vision-threatening complication of Dia-betes Mellitus (DM), is a major global concern, particularly in India, which has one of the highest diabetic populations. Prolonged hyperglycemia damages reti-nal microvasculature, leading to DR symptoms like microaneurysms, hemor-rhages, and fluid leakage, which, if undetected, cause irreve

Xue Jiang, Yihong Dong, Mengyang Liu, Hongyi Deng

While Large Language Models (LLMs) excel at code generation by learning from vast code corpora, a fundamental semantic gap remains between their training on textual patterns and the goal of functional correctness, which is governed by formal execution semantics. Reinforcement Learning with Verifiable Rewards (RLVR) approaches attempt to bridge this gap using

Shaobo Wang, Yongliang Miao, Yuancheng Liu, Qianli Ma

Large language models (LLMs) have demonstrated impressive reasoning capabilities, but scaling their performance often relies on massive reasoning datasets that are computationally expensive to train on. Existing data selection methods aim to curate smaller, high-quality subsets but often rely on costly external models or opaque heuristics. In this work, we s

Antonio Romano, Giuseppe Riccio, Mariano Barone, Marco Postiglione

Online medical forums have long served as vital platforms where patients seek professional healthcare advice, generating vast amounts of valuable knowledge. However, the informal nature and linguistic complexity of forum interactions pose significant challenges for automated question answering systems, especially when dealing with non-English languages. We p

Yili Wang, Tairan Huang, Changlong He, Qiutong Li

Heterogeneous temporal graphs (HTGs) are ubiquitous data structures in the real world. Recently, to enhance representation learning on HTGs, numerous attention-based neural networks have been proposed. Despite these successes, existing methods rely on a decoupled temporal and spatial learning paradigm, which weakens interactions of spatio-temporal informatio

Spencer King, Irfan Ozen, Karthika Subramani, Saranyan Senthivel

Web-based behavior-manipulation attacks (BMAs) - such as scareware, fake software downloads, tech support scams, etc. - are a class of social engineering (SE) attacks that exploit human decision-making vulnerabilities. These attacks remain under-studied compared to other attacks such as information harvesting attacks (e.g., phishing) or malware infections. P

Nikhilesh Maity, Shibu Meher, Manoj Dey, Abhishek Kumar Singh

Defects in two-dimensional (2D) semiconductors play a decisive role in determining their electronic, optical, catalytic and quantum properties. Understanding how defect energy levels respond to variations in layer thickness is essential for achieving reproducible and scalable device performance. We report the persistence of layer-tolerant defect levels in rh

Samuele De Gaetano, Monica Crico, Giorgio Ferrari, Marco Sampietro

Joint communication and sensing applications require devices that can analyze multiple electromagnetic waves and process them in real time directly in the analog domain. In optics, the growing maturity of photonic integrated platforms allows the fabrication of complex circuits that can perform such operations, but their large number of sensors and actuators

Xiangyu Hong, Che Jiang, Kai Tian, Biqing Qi

Attributing the behavior of Transformer models to internal computations is a central challenge in mechanistic interpretability. We introduce DePass, a unified framework for feature attribution based on a single decomposed forward pass. DePass decomposes hidden states into customized additive components, then propagates them with attention scores and MLP's ac

Nicos Georgiou, Enrico Scalas, Vladislav Vysotsky

We study a single-server priority queue with a finite number of classes, in which the arrivals follow a fractional Poisson process of index $\alpha \in (0,1]$ and the service completions are triggered by an independent fractional Poisson process of index $\beta \in (0,1]$. Each of the customers arriving is assigned at random to one of the priority classes. T

Christopher von Klitzing, Denis Blessing, Henrik Schopmans, Pascal Friederich

Efficient sampling from high-dimensional and multimodal unnormalized probability distributions is a central challenge in many areas of science and machine learning. We focus on Boltzmann generators (BGs) that aim to sample the Boltzmann distribution of physical systems, such as molecules, at a given temperature. Classical variational approaches that minimize

Tong Liu, Zhiwei Fan, Guanyan Peng, Haodan Zhang

Short video streaming has become a dominant paradigm in digital media, characterized by rapid swiping interactions and diverse media content. A key technical challenge is designing an effective preloading strategy that dynamically selects and prioritizes download tasks from an evolving playlist, balancing Quality of Experience (QoE) and bandwidth efficiency

Marina Echeverria Ferrero, Nicolas Moreno, Marco Ellero

Smoothed dissipative particle dynamics (SDPD) is a widely used particle-based method for modelling soft matter systems at mesoscopic and macroscopic scales, offering thermodynamic consistency and direct control over the fluid's transport properties. Here, we present an SDPD model that incorporates the transport of reactants on scales smaller than the discret

Alejandro H. Artiles, Hiromu Yakura, Levin Brinkmann, Mar Canet Sola

In open-ended domains like art, autonomous agents must generate ideas that are both original and internally coherent, yet current Large Language Models (LLMs) either default to familiar cultural patterns or sacrifice coherence when pushed toward novelty. We address this by introducing the Cultural Alien Sampler (CAS), a concept-selection method that explicit

Tianci Bi, Xiaoyi Zhang, Yan Lu, Nanning Zheng

The performance of Latent Diffusion Models (LDMs) is critically dependent on the quality of their visual tokenizers. While recent works have explored incorporating Vision Foundation Models (VFMs) into the tokenizers training via distillation, we empirically find this approach inevitably weakens the robustness of learnt representation from original VFM. In th

Cristina Martinez Montes, Robert Feldt, Cristina Miguel Martos, Sofia Ouhbi

As artificial intelligence advances, large language models (LLMs) are entering qualitative research workflows, yet no reproducible methods exist for integrating them into established approaches like thematic analysis (TA), one of the most common qualitative methods in software engineering research. Moreover, existing studies lack systematic evaluation of LLM

Liyang He, Yuren Zhang, Ziwei Zhu, Zhenghui Li

Retrieval Augmented Generation (RAG) systems are increasingly vital in dynamic domains like online gaming, yet the lack of a dedicated benchmark has impeded standardized evaluation in this area. The core difficulty lies in Dual Dynamics: the constant interplay between game content updates and the shifting focus of the player community. Furthermore, the neces

Atharvan Dogra, Soumya Suvra Ghosal, Ameet Deshpande, Ashwin Kalyan

Large language models are increasingly used for creative writing and engagement content, raising safety concerns about the outputs. Therefore, casting humor generation as a testbed, this work evaluates how funniness optimization in modern LLM pipelines couples with harmful content by jointly measuring humor, stereotypicality, and toxicity. This is further su

Ana Silva, Eliska Greplova

As quantum technology matures, the efficient benchmarking of quantum devices remains a key challenge. Although sample-efficient, information-theoretic benchmarking techniques have recently been proposed, there is still a gap in adapting these techniques to contemporary experiments. In this work, we re-formulate five of the most common randomized benchmarking

Alexander Bentkamp, Jasmin Blanchette, Matthias Hetzenberger

We introduce $\lambda$KBO and $\lambda$LPO, two variants of the Knuth-Bendix order (KBO) and the lexicographic path order (LPO) designed for use with the $\lambda$-superposition calculus. We establish the desired properties via encodings into the familiar first-order KBO and LPO.

Simon Baker, Benjamin Ward

In this paper we study a quantitative notion of exactness within Diophantine approximation. Given $\Psi:(0,\infty)\to (0,\infty)$ and $\omega:(0,\infty)\to (0,1)$ satisfying $\lim_{q\to\infty}\omega(q)=0$, we study the set of points, which we call $E(\Psi,\omega)$, that are $\Psi$-well approximable but not $\Psi(1-\omega)$-well approximable. We prove results

Sombuddha Bhattacharyya, Tuhin Mondal, Suman Kumar Sahoo

In this article, we study Momentum Light Ray Transform (MLRT) on symmetric tensor fields. MLRT is an integral transform in time-space domain ($(t,x)\in \mathbb{R}^{1+n}$), which integrates a scalar function or a tensor field along the light rays with a polynomial type weight. We explore necessary and sufficient conditions for injectivity of MLRT, over genera

Maximilian Buchta, Felix Hoff, Lucas Bothe, Niklas Penner

Sesqui-chalcogenides serve as a critical bridge between traditional semiconductors and quantum materials, offering significant potential in applications such as thermoelectrics, phase change memory, and topological insulators. While considerable attention has been focused on antimony- and bismuth-based compounds, characterized by substantial property changes

Wenjing Dang, Kaixuan Li, Sen Chen, Zhenwei Zhuo

The Proof-of-Concept (PoC) for a vulnerability is crucial in validating its existence, mitigating false positives, and illustrating the severity of the security threat it poses. However, research on PoCs significantly lags behind studies focusing on vulnerability data. This discrepancy can be directly attributed to several challenges, including the dispersio

J. Q. You

This News & Views article provides a perspective on the 2025 Nobel Prize in Physics, including the groundbreaking discovery of macroscopic quantum tunneling and energy quantization in superconducting circuits, the history and causes giving rise to this breakthrough, and its impact on subsequent progress in quantum computing.

Azam Zoshki, Hamid Arian Zad, Katarina Karlova, Jozef Strecka

We investigate magnetic and thermodynamic properties of a spin-1 Heisenberg diamond chain in a magnetic field using a combination of analytical and numerical methods including the variational approach, exact diagonalization, density-matrix renormalization group, localized-magnon theory, and quantum Monte Carlo simulations. In the unfrustrated regime, the mod

Anna Oliveras, Roger Marí, Rafael Redondo, Oriol Guardià

This work introduces a new latent diffusion model to generate high-quality 3D chest CT scans conditioned on 3D anatomical masks. The method synthesizes volumetric images of size 256x256x256 at 1 mm isotropic resolution using a single mid-range GPU, significantly lowering the computational cost compared to existing approaches. The conditioning masks delineate

Kesavan Sekar, Viney Ghai, Reza Ghanbari, Marko Bek

Complex fluids transition from laminar to transitory flow above a critical control parameter, akin to their Newtonian counterparts. In a continuum mechanics sense, fluid elements follow the ensuing complex trajectories, giving rise to secondary flows in terms of macroscopic vortices and patterns thereof. However, if we replace idealized fluid elements with a

Toby Barter, Zheng Gao, Eva Christodoulaki, Jing Chen

Bond markets respond differently to macroeconomic news compared to equity markets, yet most sentiment models are trained primarily on general financial or equity news data. However, bond prices often move in the opposite direction to economic optimism, making general or equity-based sentiment tools potentially misleading. We introduce BondBERT, a transformer

Rodrigo Carmo Terin

We investigate a unified Landau--gauge fixing that continuously interpolates between the viewpoints of the Serreau--Tissier (ST) copy-averaged formulation and the (Refined) Gribov--Zwanziger (RGZ) restriction to the first Gribov region. By combining the ST weight with a GZ-type horizon term and localizing both through the replica trick and the BRST-invariant

Ziwei Deng, Mian Deng, Chenjing Liang, Zeming Gao

Large Language Models (LLMs) are increasingly being explored across a range of reasoning tasks. However, LLMs sometimes struggle with reasoning tasks under uncertainty that are relatively easy for humans, such as planning actions in stochastic environments. The adoption of LLMs for reasoning is impeded by uncertainty challenges, such as LLM uncertainty and e

Thomas Fischer, Yury Person

A conjecture of Talagrand (2010) states that the so-called expectation and fractional expectation thresholds are always within at most some constant factor from each other. We prove for the unweighted case that this is a.a.s. true when the support is a random hypergraph.

Bach Hung Luu, Samuel Harry Gardner, Sinh Cong Lam, Trong Minh Hoang

Mitigating intercell interference by employing fractional frequency reuse algorithms is one of the important approaches to improving user performance in 5G and Beyond 5G cellular network systems, which typically have a high density of Base Stations (BSs). While most frequency reuse algorithms are based on the downlink Signal-to-Interference-plus-Noise Ratio

Yixuan Liu, Xinlei Li, Yi Li

Phishing attacks in Web3 ecosystems are increasingly sophisticated, exploiting deceptive contract logic, malicious frontend scripts, and token approval patterns. We present DeepTx, a real-time transaction analysis system that detects such threats before user confirmation. DeepTx simulates pending transactions, extracts behavior, context, and UI features, and

Ji Du, Xin Wang, Fangwei Hao, Mingyang Yu

At the core of Camouflaged Object Detection (COD) lies segmenting objects from their highly similar surroundings. Previous efforts navigate this challenge primarily through image-level modeling or annotation-based optimization. Despite advancing considerably, this commonplace practice hardly taps valuable dataset-level contextual information or relies on lab

Haru Kondoh, Asako Kanezaki

The field of multimodal robot navigation in indoor environments has garnered significant attention in recent years. However, as tasks and methods become more advanced, the action decision systems tend to become more complex and operate as black-boxes. For a reliable system, the ability to explain or describe its decisions is crucial; however, there tends to

Alvaro Domínguez, Nagi Khalil

The hydrodynamic stationary states of a granular fluid are addressed theoretically when subject to energy injection and a time-independent, but otherwise arbitrary external potential force. When the latter is not too symmetrical in a well defined sense, we show that a quiescent stationary state does not exist, rather than simply being unstable and, correspon

Brandon Livio Annesi, Dario Bocchi, Chiara Cammarota

High-dimensional non-convex loss landscapes play a central role in the theory of Machine Learning. Gaining insight into how these landscapes interact with gradient-based optimization methods, even in relatively simple models, can shed light on this enigmatic feature of neural networks. In this work, we will focus on a prototypical simple learning problem, wh

Eduard Frankford, Tobias Antensteiner, Michael Vierhauser, Clemens Sauerwein

With the recent rapid increase in digitization across all major industries, acquiring programming skills has increased the demand for introductory programming courses. This has further resulted in universities integrating programming courses into a wide range of curricula, including not only technical studies but also business and management fields of study.

Qingqing Gu, Dan Wang, Yue Zhao, Xiaoyu Wang

Chain-of-Thought (CoT) is widely applied to enhance the LLM capability in math, coding and reasoning tasks. However, its performance is limited for open-domain tasks, when there are no clearly defined reasoning steps or logical transitions. To mitigate such challenges, we propose a new prompt-based paradigm called Chain of Conceptual Thoughts (CoCT), which s

Yuanhe Guo, Linxi Xie, Zhuoran Chen, Kangrui Yu

We introduce ImageGem, a dataset for studying generative models that understand fine-grained individual preferences. We posit that a key challenge hindering the development of such a generative model is the lack of in-the-wild and fine-grained user preference annotations. Our dataset features real-world interaction data from 57K users, who collectively have

Loïc Foissy

Noncommutative multi-indices are noncommutative monomials in a $\mathbb{N}$-indexed family of indeterminates. We define on them a $\mathbb{Z}$-graded operadic structure, with the help of a shifting derivation. Multi-indices of degree 0 are called populated: they form a suboperad, isomorphic to the operad of Novikov algebras. This operadic structure, and the

Zhiwei Hao, Jianyuan Guo, Li Shen, Kai Han

Recent advancements in vision transformers (ViTs) have demonstrated that larger models often achieve superior performance. However, training these models remains computationally intensive and costly. To address this challenge, we introduce ScaleNet, an efficient approach for scaling ViT models. Unlike conventional training from scratch, ScaleNet facilitates

Tasha Settewong, Youmei Fan, Raula Gaikovina Kula, Kenichi Matsumoto

Computational notebooks have become the preferred tool of choice for data scientists and practitioners to perform analyses and share results. Notebooks uniquely combine scripts with documentation. With the emergence of generative AI (GenAI) technologies, it is increasingly important, especially in competitive settings, to distinguish the characteristics of h

Alexander Bentkamp, Jasmin Blanchette, Matthias Hetzenberger, Uwe Waldmann

The $\lambda$-superposition calculus is a successful approach to proving higher-order formulas. However, some parts of the calculus are extremely explosive, notably due to the higher-order unifier enumeration and the functional extensionality axiom. In the present work, we introduce an "optimistic" version of $\lambda$-superposition that addresses these two

Ching-Chih Sung, Shuntaro Suzuki, Francis Pingfan Chien, Komei Sugiura

Clarifying the neural basis of speech intelligibility is critical for computational neuroscience and digital speech processing. Recent neuroimaging studies have shown that intelligibility modulates cortical activity beyond simple acoustics, primarily in the superior temporal and inferior frontal gyri. However, previous studies have been largely confined to c

Jiaao Yu, Mingjie Han, Jinkun Jiang, Junyu Dong

The high cost of data annotation has spurred research on training deep learning models in data-limited scenarios. Existing paradigms, however, fail to balance cross-domain transfer and cross-category generalization, giving rise to the demand for Domain-Adaptive Zero-Shot Learning (DAZSL). Although vision-language models (e.g., CLIP) have inherent advantages

Gad Horovitz, Alexander N. Poddubny

We study theoretically continuous-variable entanglement between the motional degrees of freedom of optically trapped massive particles coupled via the Coulomb interaction, in the presence of a feedback control scheme. We perform a detailed analysis of the parametric resonance induced by temporal modulation of the coupling strength, based on the system's coup

Xiao-jing Gao, Yanfeng Ge, Yan Gao

Nodal-sphere semimetals (NSSMs), featuring spherical band degeneracies in momentum space, constitute a fascinating class of topological materials. However, their realization in real materials is severely hampered by discrete crystallographic symmetry constraints, often resulting in gapped ``pseudo'' nodal spheres. Here, combining first-principles calculation

Chenxu Zhang, Fuxiang Huang, Lei Zhang

With climate change intensifying, urban waterlogging poses an increasingly severe threat to global public safety and infrastructure. However, existing monitoring approaches rely heavily on manual reporting and fail to provide timely and comprehensive assessments. In this study, we present Urban Waterlogging Assessment (UWAssess), a foundation model-driven fr

Guangfu Guo, Xiaoqian Lu, Yue Feng

Visual Language Models (VLMs) achieve promising results in medical reasoning but struggle with hallucinations, vague descriptions, inconsistent logic and poor localization. To address this, we propose a agent framework named Medical Visual Reasoning Agent (\textbf{Med-VRAgent}). The approach is based on Visual Guidance and Self-Reward paradigms and Monte Car

Toranosuke Manabe, Yuchi Ishikawa, Hokuto Munakata, Tatsuya Komatsu

Language-audio joint representation learning frameworks typically depend on deterministic embeddings, assuming a one-to-one correspondence between audio and text. In real-world settings, however, the language-audio relationship is inherently many-to-many: one audio segment can be described by multiple captions and vice versa. To address this, we propose Prob

Yizhen Jia, Siyao Xiao, Wenkai Jia, Hui Chen

The increasing of digital radio frequency memory based electronic countermeasures poses a significant threat to the survivability and effectiveness of radar systems. These jammers can generate a multitude of deceptive false targets, overwhelming the radar's processing capabilities and masking targets. Consequently, the ability to robustly discriminate betwee

Adam Chapman

We revisit the famous theorem of Albert's on the cyclicity of tensor products of cyclic $p$-algebras. In the case of tensor products of cyclic $p$-algebras of prime degree, we provide an explicit computation of the resulting cyclic algebra in symbol algebra terms.

Abdullah Ajasa, Mubarak Badamasi Aremu, Ali Nasir

Permanent Magnet Synchronous Motors (PMSMs) are widely employed in high-performance drive systems due to their high efficiency, power density, and precise dynamic behavior. However, nonlinearities, load disturbances, and parameter uncertainties present persistent challenges to control. Sliding-Mode Control (SMC) remains one of the most reliable strategies fo

Gerardo García-Moreno

This thesis is situated within the context of quantum gravity, broadly understood as any effort to explore the interplay between gravitation and the quantum realm, without necessarily requiring the quantization of the gravitational field itself. We focus on emergent theories, particularly those in which the causal structure and geometric concepts underlying

Nathanaëlle Courant, Xavier Leroy

Convertibility checking - determining whether two lambda-terms are equal up to reductions - is a crucial component of proof assistants and dependently-typed languages. Practical implementations often use heuristics to quickly conclude that two terms are or are not convertible without reducing them to normal form. However, these heuristics can backfire, trigg

Jiaao Yu, Shenwei Li, Mingjie Han, Yifei Yin

Recent breakthroughs in reasoning models have markedly advanced the reasoning capabilities of large language models, particularly via training on tasks with verifiable rewards. Yet, a significant gap persists in their adaptation to real world multimodal scenarios, most notably, vision language tasks, due to a heavy focus on single modal language settings. Wh

Jiaao Yu, Mingjie Han, Tao Gong, Jian Zhang

With the rapid growth of video data, text-video retrieval technology has become increasingly important in numerous application scenarios such as recommendation and search. Early text-video retrieval methods suffer from two critical drawbacks: first, they heavily rely on large-scale annotated video-text pairs, leading to high data acquisition costs; second, t

Rahul Soundrarajan, Claudio Fiandrino, Michele Polese, Salvatore D'Oro

Open RAN introduces a flexible, cloud-based architecture for the Radio Access Network (RAN), enabling Artificial Intelligence (AI)/Machine Learning (ML)-driven automation across heterogeneous, multi-vendor deployments. While EXplainable Artificial Intelligence (XAI) helps mitigate the opacity of AI models, explainability alone does not guarantee reliable net

Amalie F. Davidsen, Kristine M. L. Krighaar, Pascale P. Deen, and Kim Lefmann

We present the instrument concept for KVASIR, a backscattering indirect time-of-flight neutron spectrometer for the European Spallation Source (ESS). KVASIR will probe low lying excitations of single crystal hard condensed matter that many advanced technologies rely upon. The instrument is simultaneously optimized for high resolution of energy-and momentum t

Hideaki Noda

This short note studies the asymptotic behavior of a generating function associated with the decimal expansion of \(2^n\). Our aims are twofold: (i) to present a problem on the best possible upper bound for this behavior, and (ii) to introduce a matrix representation that is useful for its analysis. The representation corresponds to a finite-state transfer o

Siyuan Yan, Guo-Qing Jiang, Yuchen Zhang, Xiaoxing Ma

Large language models (LLMs) now support context windows of hundreds of thousands to millions of tokens, enabling applications such as long-document summarization, large-scale code synthesis, multi-document question answering and persistent multi-turn dialogue. However, such extended contexts exacerbate the quadratic cost of self-attention, leading to severe

Mattia Pujatti, Andrea Di Luca, Nicola Peghini, Federico Monegaglia

In glass bottle manufacturing, precise control of forming machines is critical for ensuring quality and minimizing defects. This study presents a deep learning-based control algorithm designed to optimize the forming process in real production environments. Using real operational data from active manufacturing plants, our neural network predicts the effects

Samuel Bilson, Andrew Thompson, Declan Tucker, Jonathan Pearce

Thermocouples are in widespread use in industry, but they are particularly susceptible to calibration drift in harsh environments. Self-validating thermocouples aim to address this issue by using a miniature phase-change cell (fixed-point) in close proximity to the measurement junction (tip) of the thermocouple. The fixed point is a crucible containing an in

Adeel Safder

Deep neural networks (DNNs) achieve remarkable performance but often suffer from overfitting due to their high capacity. We introduce Momentum-Adaptive Gradient Dropout (MAGDrop), a novel regularization method that dynamically adjusts dropout rates on activations based on current gradients and accumulated momentum, enhancing stability in non-convex optimizat

Yuheng Wu, Thanh-Tung Nguyen, Lucas Liebe, Quang Tau

With the rapid proliferation of the Internet of Things, video analytics has become a cornerstone application in wireless multimedia sensor networks. To support such applications under bandwidth constraints, learning-based adaptive quantization for video compression have demonstrated strong potential in reducing bitrate while maintaining analytical accuracy.

Manjie Xu, Xinyi Yang, Jiayu Zhan, Wei Liang

Self-play constitutes a fundamental paradigm for autonomous skill acquisition, whereby agents iteratively enhance their capabilities through self-directed environmental exploration. Conventional self-play frameworks exploit agent symmetry within zero-sum competitive settings, yet this approach proves inadequate for open-ended learning scenarios characterized

Elisha Falbel

We study self-adjoint extensions of a second order differential operator of Sturm-Liouville type on a graph. We relate self-adjointness of the operator to the existence of non-complete trajectories of the Hamiltonian vector field defined by its principal symbol outside the vertices. We define Kirchhoff conditions at the vertices which guarantee a self-adjoin

Miao Zhang, Junpeng Li, ChangChun HUa, Yana Yang

Weakly supervised learning often operates with coarse aggregate signals rather than instance labels. We study a setting where each training example is an $n$-tuple containing exactly m positives, while only the count m per tuple is observed. This NTMP (N-tuple with M positives) supervision arises in, e.g., image classification with region proposals and multi

Joy Karmoker, Masum Billah, Mst Jannatun Ferdous, Akif Islam

Cricket generates a rich stream of visual and contextual information, yet much of its tactical analysis still depends on slow and subjective manual review. Motivated by the need for a more efficient and data-driven alternative, this paper presents an automated approach for cricket video analysis that identifies wicket-taking deliveries, detects the pitch and

Sungwoo Hong, Hyungyu Kim, Sung Mook Lee, Dongmin Seo

We study non-invertible global symmetries in 4d quantum field theories, aiming to generalize existing discussions to theories with multiple instantons and axions, and to make the subject more accessible to particle phenomenology. Building on both the Adler-Bell-Jackiw (ABJ) anomaly construction and the half-space gauging approach, we identify the 3d topologi

Marco Cicalese, Giuliana Fusco, Giovanni Savaré

We study the discrete-to-continuum evolution of a lattice system consisting of two immiscible phases labelled by -1 and +1 in presence of a surfactant phase labelled by 0. The system's energy is described by the classical Blume-Emery-Griffith model on the lattice epsilon Z^2, and its continuum evolution is obtained as epsilon tends to zero through a minimizi

Matthias Lorenzen, Teodoro Alamo, Martina Mammarella, Fabrizio Dabbene

Motivated by the application of using model predictive control (MPC) for motion planning of autonomous mobile robots, a form of output tracking MPC for non-holonomic systems and with non-convex constraints is studied. Although the advantages of using MPC for motion planning have been demonstrated in several papers, in most of the available fundamental litera

Olivier Zemb, Lauren Jouaron, Estelle Jordi, Anais Cazals

Bioinformatic analysis of microbiota revealed that certain metabolic pathways are associated with low- and high- residual feed intake (HRFI and LRFI), such as the amino-acid biosynthesis pathway and the tRNA-aminoacyl synthesis pathway. The latter is associated with increased propionate production. Yet, in vitro fermentation-profile analyses revealed that LR

Linsong Shan, Zecan Yang, Laurence T. Yang, Changlong Li

Tensor decomposition is a powerful tool for data analysis and has been extensively employed in the field of hyperspectral-multispectral image fusion (HMF). Existing tensor decomposition-based fusion methods typically rely on disruptive data vectorization/reshaping or impose rigid constraints on the arrangement of factor tensors, hindering the preservation of

Mikito Kohno, Yasuo Fukui, Takahiro Hayakawa, Yasuo Doi

We report discovery of two CO clouds which are likely falling down to the Galactic plane at more than $35$ km s$^{-1}$. The clouds show head-tail distributions elongated perpendicular to the Galactic plane at $l=331.6^{\circ}$ and $b=0^{\circ}$ as revealed by an analysis of the Mopra CO $J=$1-0 survey data. We derived the distance of the clouds to be $2.46 \

Oluwaseun A. Ajayi, Ogundepo Odunayo

The recent widespread adoption of Large Language Models (LLMs) and machine learning in general has sparked research interest in exploring the possibilities of deploying these models on smaller devices such as laptops and mobile phones. This creates a need for frameworks and approaches that are capable of taking advantage of on-device hardware. The MLX framew

Gunter Malle, A. A. Schaeffer Fry

Recently, Moret\'o and Rizo proposed a conjecture, known as the Picky Conjecture, proposing new character correspondences extending the McKay Conjecture. We prove the Picky Conjecture for all quasi-simple groups of Lie type for non-defining primes. In favourable situations, we also obtain the stronger version postulating preservation of character values up t

F. Ahmadi, B. Bidabad, H. Nasiri

Background and Objective: In brain imaging, geometric surface models are essential for analyzing the 3D shapes of anatomical structures. Alzheimer's disease (AD) is associated with significant cortical atrophy, making such shape analysis a valuable diagnostic tool. The objective of this study is to introduce and validate a novel local surface representation