October 2025 arXiv papers — page 202

Sai Karthikeya Vemuri, Adithya Ashok Chalain Valapil, Tim Büchner, Joachim Denzler

Transferring the recent advancements in deep learning into scientific disciplines is hindered by the lack of the required large-scale datasets for training. We argue that in these knowledge-rich domains, the established body of scientific theory provides reliable inductive biases in the form of governing physical laws. We address the ill-posed inverse proble

Marius Bozga, Radu Iosif, Florian Zuleger

An aggregative composition is a binary operation obeying the principle that the whole is determined by the sum of its parts. The development of graph algebras, on which the theory of formal graph languages is built, relies on aggregative compositions that behave like disjoint union, except for a set of well-marked interface vertices from both sides, that are

Timothy Pistotti, Jason Brown, Michael Witbrock

Recent studies employing Large Language Models (LLMs) to test the Argument from the Poverty of the Stimulus (APS) have yielded contrasting results across syntactic phenomena. This paper investigates the hypothesis that characteristics of the stimuli used in recent studies, including lexical ambiguities and structural complexities, may confound model performa

Sushmita Agarwal, Vaidehi S. Paliya

Very-high energy (VHE; $>$100 GeV) $\gamma$-ray emission originates via some of the most extreme particle acceleration processes in the universe. Considering beamed active galactic nuclei, i.e., blazars, only a small fraction, mainly high synchrotron peak BL Lacs, have been detected in the VHE band with the ground-based Cherenkov telescopes. We utilized $\si

Ju Gao, Fang Shen

We present an angular--momentum--resolved energetic formulation of the Aharonov--Bohm (AB) response for a confined Dirac electron based on two gauge--invariant interaction functionals: a magnetization--field functional and a current--potential functional. Using exact Dirac eigenmodes in a cylindrical cavity threaded by a solenoidal flux, we show that the mag

Luke Stevenson, Sanchari Das

Mobile healthcare (mHealth) applications promise convenient, continuous patient-provider interaction but also introduce severe and often underexamined security and privacy risks. We present an end-to-end audit of 272 Android mHealth apps from Google Play, combining permission forensics, static vulnerability analysis, and user review mining. Our multi-tool as

Zhangyue Yin, Qiushi Sun, Zhiyuan Zeng, Zhiyuan Yu

Test-time scaling has emerged as a transformative paradigm for enhancing the performance of large reasoning models, enabling dynamic allocation of computational resources during inference. However, as the landscape of reasoning models rapidly expands, a critical question remains: how can we systematically compare and evaluate the test-time scaling capabiliti

Arjun Agarwal, Rachel Chen, Rohan Garg, Jared Kettinger

Given a finite abelian group $G$ and elements $x, y \in G$, we prove that there exists $\phi \in \text{Aut}(G)$ such that $\phi(x) = y$ if and only if $G/\langle x \rangle \cong G/\langle y \rangle$. This result leads to our development of the two fastest known algorithms to determine if two elements of a finite abelian group are automorphic images of one an

Fabian Tschofenig, Douglas Guilbeault

An enduring challenge in contagion theory is that the pathways contagions follow through social networks exhibit emergent complexities that are difficult to predict using network structure. Here, we address this challenge by developing a causal modeling framework that (i) simulates the possible network pathways that emerge as contagions spread and (ii) ident

Jacob Calvert, Andréa W. Richa, Dana Randall

From the formation of ice in small clusters of water molecules to the mass raids of army ant colonies, the emergent behavior of collectives depends critically on their size. At the same time, common wisdom holds that such behaviors are robust to the loss of individuals. This tension points to the need for a more systematic study of how number influences coll

Aueaphum Aueawatthanaphisut, Nyi Wunna Tun

The comparative evaluation between classical and quantum reinforcement learning (QRL) paradigms was conducted to investigate their convergence behavior, robustness under observational noise, and computational efficiency in a benchmark control environment. The study employed a multilayer perceptron (MLP) agent as a classical baseline and a parameterized varia

Bertram Taetz, Gal Bordelius

Generating accurate and coherent image captions in a continual learning setting remains a major challenge due to catastrophic forgetting and the difficulty of aligning evolving visual concepts with language over time. In this work, we propose a novel multi-loss framework for continual image captioning that integrates semantic guidance through prompt-based co

Moritz Alker, David C. Schedl, Andreas Stöckl

This study examines the use of social media and news images to detect and measure hailstones, utilizing pre-trained multimodal large language models. The dataset for this study comprises 474 crowdsourced images of hailstones from documented hail events in Austria, which occurred between January 2022 and September 2024. These hailstones have maximum diameters

Hans Weytjens, Wouter Verbeke

This book chapter introduces the principles and practical applications of uncertainty quantification in machine learning. It explains how to identify and distinguish between different types of uncertainty and presents methods for quantifying uncertainty in predictive models, including linear regression, random forests, and neural networks. The chapter also c

Sergio M. Rezende, Byron Freelon, Roberto L. Rodríguez-Suárez

The magnetic properties of two-dimensional (2D) materials have been attracting increasing attention in recent years due to their unique behavior and possible applications in new devices. One material of great interest is the 2D van der Waals (vdW) crystal CrSBr, that exhibits antiferromagnetic (AF) order at low temperatures due to an interlayer AF exchange i

Qin Dong, Yuntian Tang, Heming Jia, Yunhang Shen

Low-Rank Adaptation (LoRA) has emerged as a dominant method in Parameter-Efficient Fine-Tuning (PEFT) for large language models, which augments the transformer layer with one down-projection $A$ and one up-projection $B$. However, LoRA's reliance on a single down-projection matrix ($A$) creates a representational bottleneck, as this solitary feature extracto

Rawnak Sultana, Mojtaba Taghipour Kaffash, Gianluca Gubbiotti, Yi Ji

We report a combined experimental and numerical investigation of spin-wave dynamics in a hybrid magnonic crystal consisting of a CoFeB artificial spin ice (ASI) of stadium-shaped nanoelements patterned atop a continuous NiFe film, separated by a 5 nm Al2O3 spacer. Using Brillouin light scattering spectroscopy, we probe the frequency dependence of thermal spi

S. K. Maurya, Abdul Aziz, Ksh. Newton Singh, G. Mustafa

In this investigation we examine the astrophysical consequences of the influence of pressure anisotropy on the physical properties of observed pulsars within the background of $f(Q,T)$ gravity by choosing a specific form $f(Q, T)=\psi_1\, Q + \psi_2 T$, where $\psi_1$ and $\psi_2$ are the model parameters. Initially, we solve the modified field equations for

Hudson de Martim

In high-stakes legal domains, retrieval must preserve not only semantic relevance, but also the hierarchy, temporality, and causal provenance of legal norms. Standard Retrieval-Augmented Generation (RAG), based mainly on semantic similarity over text fragments, cannot reliably provide this level of control. Prior work on SAT-Graph RAG addressed the represent

Timothy Pistotti, Jason Brown, Michael Witbrock

Recent studies probing the Argument from the Poverty of the Stimulus (APS) have applied Large Language Models (LLMs) to test the learnability of complex syntax through surprisal-based metrics. However, divergent conclusions raise questions concerning the insights these metrics offer. While Wilcox et al. (2024) used direct minimal pair comparisons (the "wh-ef

Daniel Otten, Trevor Stalnaker, Nathan Wintersgill, Oscar Chaparro

The use of generative AI (GenAI) tools has fundamentally transformed software development. Central to this shift is prompt engineering, the practice of crafting textual prompts to guide GenAI tools in generating useful content. Although prompt engineering has emerged as a critical skill, prior research has focused primarily on cataloging of prompting techniq

J. M. Do Ó, R. F. Freire, J. Giacomoni, E. S. Medeiros

In this paper we study the problem $-\mathrm{div}(\rho(x_N)\nabla u)=a|u|^{p-2}u$ in $\mathbb{R}^N_+$, $-\partial u/\partial x_N=b|u|^{q-2}u$ in $\mathbb{R}^{N-1}$ where $a,b \in \mathbb{R}$, $p,q\in (1,\infty)$ and $\rho$ is a positive weight. We establish regularity results for weak solutions and, using a variational approach combined with a new Pohozaev-t

Jana Hartenstein, Maximilian Stegemeyer

The string topology coproduct on the homology of the free loop space of a closed manifold induces a string cobracket on $S^1$-equivariant homology. We give a complete computation of the string topology coproduct for surfaces of higher genus by describing an algorithm which computes the coproduct of a cyclic word in terms of generators of the fundamental grou

Moritz Schneider, Robert Krug, Narunas Vaskevicius, Luigi Palmieri

Empowerment, an information-theoretic measure of an agent's potential influence on its environment, has emerged as a powerful intrinsic motivation and exploration framework for reinforcement learning (RL). Besides for unsupervised RL and skill learning algorithms, the specific use of empowerment as a pre-training signal has received limited attention in the

Weiheng Zhong, Qibang Liu, Diab Abueidda, Seid Koric

Neural operators have emerged as powerful tools for learning nonlinear mappings between function spaces, enabling real-time prediction of complex dynamics in diverse scientific and engineering applications. With their growing adoption in engineering design evaluation, a wide range of neural operator architectures have been proposed for various problem settin

Zhiliang Deng, Zhiyuan Wang, Xiaomei Yang, Xiaofei Guan

We present a novel Bayesian framework for inverse problems in which the pos terior distribution is interpreted as the intensity measure of a Poisson point process (PPP). The posterior density is approximated using kernel density estimation, and the superposition property of PPPs is then exploited to enable efficient sampling from each kernel component. This

Xuemin Tu, Jinjin Zhang

Stochastic balancing domain decomposition by constraints (BDDC) algorithms are developed and analyzed for the sampling of the solutions of linear stochastic elliptic equations with random coefficients. Different from the deterministic BDDC algorithms, the stochastic BDDC algorithms have online and offline stages. At the offline stage, the Polynomial Chaos (P

Tien-Dat Nguyen, Thien-Minh Nguyen, Vinh-Hao Nguyen

Onboard simultaneous localization and mapping (SLAM) methods are commonly used to provide accurate localization information for autonomous robots. However, the coordinate origin of SLAM estimate often resets for each run. On the other hand, UWB-based localization with fixed anchors can ensure a consistent coordinate reference across sessions; however, it req

Francesca Crispo, Angelica Pia Di Feola, Carlo Romano Grisanti

The paper is concerned with the 3D-initial value problem for power-law fluids with shear dependent viscosity in a spatially periodic domain. The goal is the construction of a weak solution enjoying an energy equality. The results hold assuming an initial data $v_0\in J^2(\Omega)$ and for $p\in \left(\frac 95,2\right)$. It is interesting to observe that the r

Suzan Kagan, Sankar Sudhir, Ben Hamilton, Ramya Dwivedi

Peoples Water Data proves that a people-powered, open-data model can deliver reliable household water evidence at scale, driving equitable decisions, accountability, and safer water where it is most needed.

Xueyan Li, Guinan Su, Mrinmaya Sachan, Jonas Geiping

Large Language Models (LLMs) are increasingly applied to complex tasks that require extended reasoning. In such settings, models often benefit from diverse chains-of-thought to arrive at multiple candidate solutions. This requires two competing objectives: to inject enough stochasticity to explore multiple reasoning chains, and to ensure sufficient accuracy

Yotam Gafni

Transaction Fee Mechanisms (TFMs) study auction design in the Blockchain context, and emphasize robustness against miner and user collusion, moreso than traditional auction theory. \cite{chung2023foundations} introduce the notion of a mechanism being $c$-Side-Contract-Proof ($c$-SCP), i.e., robust to a collusion of the miner and $c$ users. Later work \cite{c

Cristina Luna, Robert Field, Steven Kay

Current planetary rovers operate at traverse speeds of approximately 10 cm/s, fundamentally limiting exploration efficiency. This work presents integrated AI systems which significantly improve autonomy through three components: (i) the FASTNAV Far Obstacle Detector (FOD), capable of facilitating sustained 1.0 m/s speeds via computer vision-based obstacle de

Imranur Rahman, Md Rayhanur Rahman

Code completion can help developers improve efficiency and ease the development lifecycle. Although code completion is available in modern integrated development environments (IDEs), research lacks in determining what makes a good context for code completion based on the information available to the IDEs for the large language models (LLMs) to perform better

Tao Zhu, Yinfeng Yu, Liejun Wang, Fuchun Sun

Diffusion models have demonstrated remarkable performance in speech synthesis, but typically require multi-step sampling, resulting in low inference efficiency. Recent studies address this issue by distilling diffusion models into consistency models, enabling efficient one-step generation. However, these approaches introduce additional training costs and rel

Lucas Barrault, Lisa Bugnet, Stéphane Mathis, Joey S. G. Mombarg

Gamma Dor stars are ideal targets for studies of the innermost dynamical properties of stars, due to their rich frequency spectrum of gravito-inertial modes propagating in the radiative envelope. Recent studies found that these modes could couple at the core-to-envelope interface with pure inertial modes in their sub-inertial regime, forming the so-called in

Johanna Müller-Horn, Hans-Walter Rix, Kareem El-Badry, Ben Pennell

We present a rigorous identification of candidates for dormant black holes (BHs) and neutron stars (NSs) in binaries using summary statistics from Gaia DR3, rather than full orbital solutions. Although Gaia astrometric orbits have already revealed a small sample of compact object binaries, many systems remain undetected due to stringent quality cuts imposed

Cristina Luna, Alba Guerra, Almudena Moreno, Manuel Esquer

Planetary exploration missions require robust locomotion systems capable of operating in extreme environments over extended periods. This paper presents the DISTANT (Distant Transmission and Steering Systems) design, a novel approach for relocating rover traction and steering actuators from wheel-mounted positions to a thermally protected warm box within the

Asiye Arif, Tugba Yurdakadim

Here, we introduce three kinds of neural network operators of convolution type which are activated by q-deformed and \b{eta}-parametrized half hyperbolic tangent function. We obtain quantitative convergence results to the identity operator with the use of modulus of continuity. Global smoothness preservation of our operators are also presented and the iterat

Karen Jia-Hui Li, Simone Balloccu, Ondrej Dusek, Ehud Reiter

The increasing trust in large language models (LLMs), especially in the form of chatbots, is often undermined by the lack of their extrinsic evaluation. This holds particularly true in nutrition, where randomised controlled trials (RCTs) are the gold standard, and experts demand them for evidence-based deployment. LLMs have shown promising results in this fi

Jingjun Bao, Lijun Ji

Two families of sets \(\mathcal{A}\) and \(\mathcal{B}\) are called \emph{cross-\(t\)-intersecting} if \(|A \cap B| \geq t\) for all \(A \in \mathcal{A}\) and \(B \in \mathcal{B}\). Determining the maximum product of sizes for such cross-\(t\)-intersecting families is an active problem in extremal set theory. In this paper, we verify the following cross-\(t\

I. Kontogiannis, Y. Zhu, K. Barczynski, M. Z. Stiefel

Magnetic flux emergence and decay in the Sun span from days to months. However, their tracking is typically limited to about half a solar rotation when relying on single-vantage-point observations. Combining observations from both the Earth-facing and far side of the Sun, we monitored the magnetic and coronal evolution and characterised the non-potentiality

Yunyi Ni, Finn Carter, Ze Niu, Emily Davis

Robust invisible watermarking aims to embed hidden information into images such that the watermark can survive various image manipulations. However, the rise of powerful diffusion-based image generation and editing techniques poses a new threat to these watermarking schemes. In this paper, we present a theoretical study and method demonstrating that diffusio

Owen T. Huber, Raghu G. Raj, Tianyu Chen, Zacharie I. Idriss

This paper introduces a novel methodology of adapting the representation of videos based on the dynamics of their scene content variation. In particular, we demonstrate how the clustering of dynamic mode decomposition eigenvalues can be leveraged to learn an adaptive video representation for separating structurally distinct morphologies of a video. We extend

Eashan Adhikarla, Yixin Liu, Brian D. Davison

Low-light image enhancement (LLIE) is vital for safety-critical applications such as surveillance, autonomous navigation, and medical imaging, where visibility degradation can impair downstream task performance. Recently, diffusion models have emerged as a promising generative paradigm for LLIE due to their capacity to model complex image distributions via i

Binhong Li, Xiao Yan, Shangqi Lu

Approximate nearest neighbor (ANN) search in high-dimensional metric spaces is a fundamental problem with many applications. Over the past decade, proximity graph (PG)-based indexes have demonstrated superior empirical performance over alternatives. However, these methods often lack theoretical guarantees regarding the quality of query results, especially in

Simon Rousseau, Gabriel Seyfarth, Georg Knebel, Dai Aoki

We report resistivity and Hall effect measurements in the chiral antiferromagnet EuPtSi. Depending on the magnetic field orientation with respect to the crystallographic axes, EuPtSi presents different topological magnetic phases below the N\'eel temperature $T_N=4.05$K. In particular, for a field $H \parallel $ [111], it exhibits the well known skyrmion lat

Tobias J. Bauer

Subject of this thesis is the implementation of an AI-based Gaze Tracking system using RGBD images that contain both color (RGB) and depth (D) information. To fuse the features extracted from the images, a module based on the Transformer architecture is used. The combination of RGBD input images and Transformers was chosen because it has not yet been investi

Periklis Mantenoglou, Rishi Hazra, Pedro Zuidberg Dos Martires, Luc De Raedt

Owing to their reasoning capabilities, large language models (LLMs) have been evaluated on planning tasks described in natural language. However, LLMs have largely been tested on planning domains without constraints. In order to deploy them in real-world settings where adherence to constraints, in particular safety constraints, is critical, we need to evalua

Ron Keuth, Paul Kaftan, Mattias P. Heinrich

The generalization of the Transformer architecture via MetaFormer has reshaped our understanding of its success in computer vision. By replacing self-attention with simpler token mixers, MetaFormer provides strong baselines for vision tasks. However, while extensively studied on natural image datasets, its use in medical imaging remains scarce, and existing

Ezgi Hatipoğlu, Ayman A. El-Zoka, Yujun Zhao, Stanislav Mráz

Nanoporous metals allow for tailoring composition and surface-to-volume ratio, both aspects critical for applications in catalysis. Here, Ag70Au30 (2 at. %) films with a face-centered cubic structure were deposited at 400{\deg}C, either {111}-textured or randomly oriented. Upon chemical dealloying, atom probe tomography of the nanoporous structure reveals th

Scott Frees

Large language models translate natural language into database queries, yet context window limitations prevent direct deployment in reporting systems where complete datasets exhaust available tokens. The Model Context Protocol specification defines ResourceLink for referencing external resources, but practical patterns for implementing scalable reporting arc

Michele Benaco, Dimitrios Karamitros, Sami Nurmi, Kimmo Tuominen

We investigate scalar-induced stochastic gravitational waves from adiabatic curvature perturbations sourced by a spectator field via the modulated reheating mechanism. We consider a spectator scalar with Higgs-like couplings and inflaton decay via shift symmetric dimension-five operators. The spectator is assumed to be in the Sitter vacuum and it sources blu

Markus Hirvensalo

We analyze the Fr\'echet derivative $F$, that maps a perturbation in conductivity to the linearized change in boundary measurements governed by the conductivity equation. The domain is taken to be the unit ball $B \subset \mathbb{R}^d$ with $d \geq 2$, and we choose perturbations $\eta$ from the Hilbert space $L^2(B)$. Under the condition that the perturbati

Aaron Nagel, Johannes Löwe

To simulate highly-resolved flow fields, we extend the Quantum Lattice Boltzmann Method (QLBM) to be able to simulate multiple time steps without state extraction or reinitialization. We adjust and extend given QLBM approaches from the literature to completely remove the need to measure or reinitialize the flow field in between the simulation time steps. The

W. Dietrich, J. Wicht

Hot Jupiters are Jupiter-sized exoplanets with close-in orbits, characterized by extreme day-night temperature contrasts due to synchronous rotation. These planets offer unique observational opportunities through transit photometry, transmission spectroscopy, and infrared (IR) phase curve analysis, which reveal information about heat redistribution and atmos

Katharina Hafner, Sara Shamekh, Guillaume Bertoli, Axel Lauer

Improvements of Machine Learning (ML)-based radiation emulators remain constrained by the underlying assumptions to represent horizontal and vertical subgrid-scale cloud distributions, which continue to introduce substantial uncertainties. In this study, we introduce a method to represent the impact of subgrid-scale clouds by applying ML to learn processes f

Dayyán O'Brien, Barry Haddow, Emily Allaway, Pinzhen Chen

Conducting contamination-free evaluation of mathematical capabilities can be difficult for two reasons: models may memorize a test set once it is made public, and current mathematical benchmarks are prone to overfitting due to having limited diversity of symbols and rules, coupled with closed-ended answers. This paper proposes a method to leverage these shor

Florian Buchner, Johannes Schörghuber, Nico Unglert, Jesús Carrete

We present msmJAX, a Python package implementing the multilevel summation method with B-spline interpolation, a linear-scaling algorithm for efficiently evaluating electrostatic and other long-range interactions in particle-based simulations. Built on the JAX framework, msmJAX integrates naturally with the machine-learning methods that are transforming chemi

Andrea Mecchina, Roberta Pappadà, Nicola Torelli

Understanding the dependence structure of asset returns is fundamental in risk assessment and is particularly relevant in a portfolio diversification strategy. We propose a clustering approach where evidence accumulated in a multiplicity of classifications is achieved using classical hierarchical procedures and multiple copula-based dissimilarity measures. A

Kaixiang Zhang, Zhaojian Li, Wei Lin

Distributed control of connected and automated vehicles has attracted considerable interest for its potential to improve traffic efficiency and safety. However, such control schemes require sharing privacy-sensitive vehicle data, which introduces risks of information leakage and potential malicious activities. This paper investigates the stability and privac

Yifeng Chen, J. W. Sanders

A model of consciousness is proposed which, having a logical basis, lends itself to simulation using a simple mathematical model called Consciousness as Entropy Reduction (CER). The approach has been inspired by previous models such as GWT, IIT and an earlier less mainstream model called "Feature Map" in Psychology. CER considers the contents of consciousnes

Félix Rebotier

We study the phenomenon of coming down from infinity - that is, when the process starts from infinity and never returns to it - for continuous-state branching processes with generalized drift. We provide sufficient conditions on the drift term and the branching mechanism to ensure both non-explosion and coming down from infinity, without requiring the associ

Vaughn Gzenda, Robin Chhabra

Soft robotic crawlers are mobile robots that utilize soft body deformability and compliance to achieve locomotion through surface contact. Designing control strategies for such systems is challenging due to model inaccuracies, sensor noise, and the need to discover locomotor gaits. In this work, we present a model-based reinforcement learning (MB-RL) framewo

Alexander Oestreicher, Chris Clarkson, Julian Adamek, Sofie Marie Koksbang

The era of real-time cosmology has begun. It is now possible to directly measure the apparent drift of high-redshift astronomical sources across the sky $\textit{in real time}$. This so-called $\textit{position drift}$ provides a valuable probe of the peculiar velocity field and cosmic structure formation by giving direct access to the transverse velocity, w

Weihua Zheng, Zhengyuan Liu, Tanmoy Chakraborty, Weiwen Xu

Large language models (LLMs) are now used worldwide, yet their multimodal understanding and reasoning often degrade outside Western, high-resource settings. We propose MMA-ASIA, a comprehensive framework to evaluate LLMs' cultural awareness with a focus on Asian contexts. MMA-ASIA centers on a human-curated, multilingual, and multimodally aligned multiple-ch

Sándor P. Fekete, Phillip Keldenich, Dominik Krupke, Michael Perk

We consider a class of optimization problems that are fundamental to testing in modern configurable software systems, e.g., in automotive industries. In pairwise interaction sampling, we are given a (potentially very large) configuration space, in which each dimension corresponds to a possible Boolean feature of a software system; valid configurations are th

Luca Barbieri, Pascal Démoulin

We develop a three-dimensional kinetic model of the solar transition region and corona in which the plasma above the chromosphere is collisionless and embedded in a uniform magnetic field. Heating occurs intermittently at discrete locations on the chromospheric surface, modeled through a surface coarse-graining procedure that produces non-thermal boundary co

Fritz Peter Hessberger

In the present study we want to give an overview on low lying isomeric states in the heaviest nuclei. After a short report on the early history on the discovery of nuclear isomerism and attempts to understand their physical nature, decay probabilities and structure of all low lying isomeric states in heaviest nuclei with half-lives typically longer than one

Max van Drunen, Sanne Vrijenhoek

News recommender systems increasingly determine what news individuals see online. Over the past decade, researchers have extensively critiqued recommender systems that prioritise news based on user engagement. To offer an alternative, researchers have analysed how recommender systems could support the media's ability to fulfil its role in democratic society

Simon Hackl, Simon Hubmer, Ronny Ramlau

Ultrasound imaging is a widely used, non-invasive diagnostic tool in modern medicine. A crucial assumption is a constant sound speed in the observed medium. For large scale sound speed variations, this assumption leads to blurred and distorted images. In this paper, we present a Geometrical Acoustics based Focusing Algorithm (GOAT) which is able to correct f

Songyuan Sui, Zihang Xu, Xia Hu

Time series classification (TSC) spans diverse application scenarios, yet labeled data are often scarce, making task-specific training costly and inflexible. Recent reasoning-oriented large language models (LLMs) show promise in understanding temporal patterns, but purely zero-shot usage remains suboptimal. We propose FETA, a multi-agent framework for traini

Randall Balestriero, Nicolas Ballas, Mike Rabbat, Yann LeCun

Joint Embedding Predictive Architectures (JEPAs) learn representations able to solve numerous downstream tasks out-of-the-box. JEPAs combine two objectives: (i) a latent-space prediction term, i.e., the representation of a slightly perturbed sample must be predictable from the original sample's representation, and (ii) an anti-collapse term, i.e., not all sa

Vinícius dos Passos de Souza, Antonio Alvaro Ranha Neves

This work applies Mie scattering theory to provide a new perspective on the propagation of light through a spherical obstacle, offering a novel explanation for the formation of the Poisson spot (also known as the Arago or Fresnel spot). We demonstrate that the diffraction patterns generated by a sphere and by a circular disk can be understood as complementar

R. A. Konoplya, D. Ovchinnikov, J. Schee

We study the optical properties of black holes endowed with primary Proca hair, focusing on the distinctive double-peak structure generated in the effective potential by the massive vector field. This novel feature drastically modifies the geodesic motion of both photons and massive particles, leading to qualitatively new dynamical and observational signatur

Xenia Heilmann, Ernst Althaus, Mattia Cerrato, Nick Johannes Peter Rassau

A sum-product network (SPN) is a graphical model that allows several types of probabilistic inference to be performed efficiently. In this paper, we propose a privacy-preserving protocol which tackles structure generation and parameter learning of SPNs. Additionally, we provide a protocol for private inference on SPNs, subsequent to training. To preserve the

Sebastian Hartweg, Dusan k. Bozanic, Gustavo A. Garcia-Macias, Laurent Nahon

Spectroscopic techniques that are sensitive to molecular chirality are important analytical tools to quantitatively determine enantiomeric excess and purity of chiral molecular samples. Many chiroptical processes however produce weak enantio-specific asymmetries due to their origin relying on weak magnetic dipole or electric quadrupole effects. Photoelectron

Mishal Assif P K, Yuliy Baryshnikov

A function on a topological space is called unimodal if all of its super-level sets are contractible. A minimal unimodal decomposition of a function $f$ is the smallest number of unimodal functions that sum up to $f$. The problem of decomposing a given density function into its minimal unimodal components is fundamental in topological statistics. We show tha

Zheyue Tan, Mustapha Abdullahi, Tuo Shi, Huining Yuan

Reinforcement learning (RL) has become a pivotal component of large language model (LLM) post-training, and agentic RL extends this paradigm to operate as agents through multi-turn interaction and tool use. Scaling such systems exposes two practical bottlenecks: (1) context length grows rapidly during training, inflating memory usage and latency, and trigger

Hoang Van Quyet

We investigate tricritical phase transitions in a holographic model of topological superconductivity using Einstein-Maxwell gravity coupled with a charged scalar field in Anti-de Sitter spacetime. By incorporating both gravitational backreaction and quartic self-interaction $V(\phi) = \lambda \phi^4$, we demonstrate that the system exhibits both second-order

Vladislav Egorov, Stepan Osipov

In many applications of spin models, the fast estimation of their critical temperatures and other physical properties is of great importance. In this work, we present the analytical expressions estimating the critical properties of inhomogeneous Ising models with ferromagnetic interactions. The expressions were obtained within the framework of the m-vicinity

Jakub Skowronski, Riccardo Maria Gesuè, Axel Boeltzig, Giovanni Francesco Ciani

Studies of charged-particle reactions for low-energy nuclear astrophysics require high sensitivity, which can be achieved by means of detection setups with high efficiency and low backgrounds, to obtain precise measurements in the energy region of interest for stellar scenarios. High-efficiency total absorption spectroscopy is an established and powerful too

Andy S. Anker, John L. A. Gardner, Louise A. M. Rosset, Andrew L. Goodwin

Materials with bespoke properties have long been identified by computational searches, and their experimental realisation is now coming within reach through autonomous laboratories. Scattering experiments are central to verifying the atomic structures of autonomously synthesised materials. Yet, interpreting these measurements typically requires user expertis

Longkun Guo, Zeyu Lin, Chaoqi Jia, Chao Chen

Many real-world applications pose challenges in incorporating fairness constraints into the $k$-center clustering problem, where the dataset consists of $m$ demographic groups, each with a specified upper bound on the number of centers to ensure fairness. Focusing on big data scenarios, this paper addresses the problem in a streaming setting, where data poin

Ludwig Stage, Mirela Riveni, Raimundas Matulevičius, Dimka Karastoyanova

Provenance in scientific workflows is essential for understand- ing and reproducing processes, while in business processes, it can ensure compliance and correctness and facilitates process mining. However, the provenance of process adaptations, especially modifications during execu- tion, remains insufficiently addressed. A review of the literature reveals a

Mohamed Bal-Ghaoui, Fayssal Sabri

High-dimensional data remains a pervasive challenge in machine learning, often undermining model interpretability and computational efficiency. While Large Language Models (LLMs) have shown promise for dimensionality reduction through feature selection, existing LLM-based approaches frequently lack structured reasoning and transparent justification for their

Huang-Cheng Chou, Chi-Chun Lee

Over the past two decades, speech emotion recognition (SER) has received growing attention. To train SER systems, researchers collect emotional speech databases annotated by crowdsourced or in-house raters who select emotions from predefined categories. However, disagreements among raters are common. Conventional methods treat these disagreements as noise, a

Ian W. Stephens, Simon Coude, Philip C. Myers, Catherine Zucker

Stars primarily form in galactic spiral arms within dense, filamentary molecular clouds. The largest and most elongated of these molecular clouds are referred to as ``bones," which are massive, velocity-coherent filaments (lengths ~20 to >100 pc, widths ~1-2 pc) that run approximately parallel and in close proximity to the Galactic plane. While these bones h

Thanh Tran, Phuong Vo, Thomas Sheppard, Timothy Grotjohn

This work studies the correlation between mean retardance and thickness of diamond substrates grown homoepitaxially via microwave plasma-enhanced chemical vapor deposition (MPCVD). We measure the retardance of a diamond substrate in two orientations: perpendicular and parallel to the growth direction. Our experimental results demonstrate that the correlation

Mai AlKhamissi, Yunze Xiao, Badr AlKhamissi, Mona Diab

Cultural evaluation of large language models has become increasingly important, yet current benchmarks often reduce culture to static facts or homogeneous values. This view conflicts with anthropological accounts that emphasize culture as dynamic, historically situated, and enacted in practice. To analyze this gap, we introduce a four-part framework that cat

Jacob Bamberger, Iolo Jones, Dennis Duncan, Michael M. Bronstein

Deep generative models often face a fundamental tradeoff: high sample quality can come at the cost of memorisation, where the model reproduces training data rather than generalising across the underlying data geometry. We introduce Carr\'e du champ flow matching (CDC-FM), a generalisation of flow matching (FM), that improves the quality-generalisation tradeo

M. P. Thejitha, Anusree Anand, S. N. Fathima

Motivated by the recent work of several authors on vanishing coefficients of the arithmetic progression in certain $q$-series expansion, we study some variants of these $q$-series and prove some comparable results. For instance, if $\sum_{n=0}^{\infty}c_1(n)q^n=\left(\pm q^2,\pm q^3; q^5\right)_\infty^2 \left( q, q^{14}; q^{15}\right)_\infty$, then $c_1(5n+3

Pavel Souček, Stanislava Debnárová, Šárka Zuzjaková, Shuyao Lin

High- and medium-entropy nitride coatings from the Cr-Hf-Mo-Ta-W-N system were studied using ab initio calculations and experiments to clarify the role of entropy and individual elements in phase stability, microstructure, and high-temperature behaviour. Formation energy calculations indicated that nitrogen vacancies stabilise the cubic (fcc) phase, with haf

Renato Ferreira Pinto, Diptaksho Palit, Sofya Raskhodnikova

We initiate a systematic study of the computational complexity of property testing, focusing on the relationship between query and time complexity. While traditional work in property testing has emphasized query complexity, relatively little is known about the computational hardness of property testers. Our goal is to chart the landscape of time-query interp

Ruchi Guo, Jiahua Jiang, Bangti Jin, Wuwei Ren

Fluorescence Molecular Tomography (FMT) is a widely used non-invasive optical imaging technology in biomedical research. It usually faces significant accuracy challenges in depth reconstruction, and conventional iterative methods struggle with poor $z$-resolution even with advanced regularization. Supervised learning approaches can improve recovery accuracy

Merlin Christ

We construct relative $3$-Calabi--Yau categories related with higher Teichm\"uller theory. We further study their corresponding cosingularity categories and the additive categorification of the corresponding cluster algebras. The input for our constructions is a marked surface with boundary and a Dynkin quiver $I$. In the case of the triangle, these categori

Xiong Liu, Wenhua Wang

Let $\alpha\in\mathbb{R}$, $p\in[1,\infty)$, $q\in(0,\infty]$, $\mathbf{W}$ be a matrix weight, and $A$ be an expansive dilation on $\mathbb{R}^d$. In this paper, the authors firstly investigate and develop some aspects of homogeneous anisotropic Besov spaces $\dot{B}^{\alpha,q}_{p,A}(\mathbb{R}^d,\mathbf{W})$ and inhomogeneous anisotropic Besov spaces $B^{\

Burkhard Ringlein, Jan van Lunteren, Radu Stoica, Thomas Parnell

A long-standing goal in both industry and academia is to develop an LLM inference platform that is portable across hardware architectures, eliminates the need for low-level hand-tuning, and still delivers best-in-class efficiency. In this work, we demonstrate that portable, efficient cross-platform LLM inference is indeed possible and share our experience. W

Aman Singh, Aastha Mishra, Deepak Kapa, Suryank Joshi

A monoped's jump height and energy consumption depend on both, its mechanical design and control strategy. Existing co-design frameworks typically optimize for either maximum height or minimum energy, neglecting their trade-off. They also often omit gearbox parameter optimization and use oversimplified actuator mass models, producing designs difficult to rep

Vitor Magno de O. S. Bezerra, Gabriel F. A. Bastos, Jugurta Montalvão

Mel-frequency cepstral coefficients (MFCCs) are an important feature in speech processing. A deeper understanding of their properties can contribute to the work that is being done with both classical and deep learning models. This study challenges the long-held assumption that MFCCs lack relevant temporal information by investigating their relationship with