October 2025 arXiv papers — page 58

Zehua Liu, Tilo Burghardt

As most ''in the wild'' data collections of the natural world, the North America Camera Trap Images (NACTI) dataset shows severe long-tailed class imbalance, noting that the largest 'Head' class alone covers >50% of the 3.7M images in the corpus. Building on the PyTorch Wildlife model, we present a systematic study of Long-Tail Recognition methodologies for

Marta Contreiras Silva, Daniel Faria, Catia Pesquita

Constructing comprehensive knowledge graphs requires the use of multiple ontologies in order to fully contextualize data into a domain. Ontology matching finds equivalences between concepts interconnecting ontologies and creating a cohesive semantic layer. While the simple pairwise state of the art is well established, simple equivalence mappings cannot prov

Michał Czakon, Kilian Erhard Minguez, Felix Eschment

We provide the last missing ingredient necessary to approximate one-loop amplitudes in QCD with massive quarks in the limit of vanishing energy of a single gluon up to terms suppressed by this energy. Our main result is a soft operator acting in color and spin space that manipulates the momenta of the hard partons while keeping them on-shell and respecting m

Ying Xue, Jiaxi Jiang, Rayan Armani, Dominik Hollidt

Tracking human full-body motion using sparse wearable inertial measurement units (IMUs) overcomes the limitations of occlusion and instrumentation of the environment inherent in vision-based approaches. However, purely IMU-based tracking compromises translation estimates and accurate relative positioning between individuals, as inertial cues are inherently s

Xingyu Cheng, Reese Lance, Nikhil Nagabandi, Andrey Smirnov

We consider quantum difference equation (QDE) for equivariant quantum K-theory of the Grassmannian. In this paper we obtain a solution to the QDE and use the solution to asymptotically derive the Bethe ansatz equations. In the limit, we obtain similar results for the cohomological analogue. For both cases, we describe the nonequivariant solutions as well. As

Jonathan Bragg, Mike D'Arcy, Nishant Balepur, Dan Bareket

AI agents hold the potential to revolutionize scientific productivity by automating literature reviews, replicating experiments, analyzing data, and even proposing new directions of inquiry; indeed, there are now many such agents, ranging from general-purpose "deep research" systems to specialized science-specific agents, such as AI Scientist and AIGS. Rigor

Erhan Bayraktar, Bingyan Han, Jingjie Zhang

We study a goal-based portfolio selection problem in which an investor aims to meet multiple financial goals, each with a specific deadline and target amount. Trading the stock incurs a strictly positive transaction cost. Using the stochastic Perron's method, we show that the value function is the unique viscosity solution to a system of quasi-variational in

Han Yang, Guangjun Qin

This paper introduces a novel dynamic knowledge distillation framework, Gompertz-CNN, which integrates the Gompertz growth model into the training process to address the limitations of traditional knowledge distillation. Conventional methods often fail to capture the evolving cognitive capacity of student models, leading to suboptimal knowledge transfer. To

Inbazhagan Ravikumar, Ram Sundhar, Narendhiran Vijayakumar

Micro aerial vehicles are becoming increasingly important in search and rescue operations due to their agility, speed, and ability to access confined spaces or hazardous areas. However, designing lightweight aerial systems presents significant structural, aerodynamic, and computational challenges. This work addresses two key limitations in many low-cost aeri

Mitchell Marfinetz

CoW Protocol batch auctions aggregate user intents and rely on solvers to find optimal execution paths that maximize user surplus across heterogeneous automated market makers (AMMs) under stringent auction deadlines. Deterministic single-objective heuristics that optimize only expected output frequently fail to exploit split-flow opportunities across multipl

Marc Mars, Gabriel Sánchez-Pérez

In this paper, we study the asymptotic structure of the Fefferman-Graham ambient metric. We prove that every straight ambient metric admits a conformal completion with a well-defined null infinity, and that the asymptotic expansion of the metric at infinity can be related to that at the homothetic horizon. Furthermore, in even dimensions, we show that the Fe

P. O. Kazinski, A. A. Sokolov

The Compton process with the initial states of photons and neutrons described by the density matrices of a general form is studied for low energies of photons. The coherent contribution to the inclusive probability to record a photon is investigated in detail. This contribution gives the hologram of the neutron one-particle density matrix. The evolution of t

Zheng Liu

We construct Hida families of theta lifts from definite orthogonal and unitary groups. A major ingredient of the construction is the choice of test Schwartz functions at places dividing $p$. We select a special type of Schwartz functions endowed with the equivariance property for the action of $\mathbb{U}_p$-operators.

Gabriele Fiore, Ciaran Williams

We present the electroweak corrections for the production of a photon pair through gluon fusion, focusing on the contribution from the first two generations of quarks. The two-loop amplitude is calculated using a series of projection operators which define scalar form factors. In order to evaluate the Master Integrals which appear in this process we employ b

N. S. Martorana, G. D'Agata, A. Barbon, G. Cardella

The development of new detectors based on Silicon Carbide (SiC) is currently a topic of interest within the scientific community. The significant features of SiC make it highly promising for detecting charged particles, neutrons, and $\gamma$/X radiation. In this framework, within the SAMOTHRACE (Sicilian Micro and Nano Technology Research and Innovation Cen

Saak Gabriyelyan, Evgenii Reznichenko

Being motivated by the notions of $\kappa$-Fr\'{e}chet--Urysohn spaces and $k'$-spaces introduced by Arhangel'skii, the notion of sequential spaces and the study of Ascoli spaces, we introduce three new classes of compact-type spaces. They are defined by the possibility to attain each or some of boundary points $x$ of an open set $U$ by a sequence in $U$ con

CMS Collaboration

A search is presented for narrow resonances, with a mass between 0.6 and 1.8 TeV, decaying to pairs of jets, in proton-proton collisions at $\sqrt{s}$ = 13 TeV. The search is performed using dijets that are reconstructed, selected, and recorded in a compact form by the high-level trigger in a technique referred to as "data scouting", from data collected in 2

Rina G. Rast, Carol E. Jones, Mark W. Suffak, Jonathan Labadie-Bartz

We conduct a systematic study on the effects of rapid rotation on predicted Be star observables. We use the three-dimensional Monte Carlo radiative transfer code, \textsc{hdust}, to model a comprehensive range of Be star subtypes at varying rotation rates. Using these models, we predict $V$ magnitude and photometric color, H$\alpha$ line profiles, and polari

Pankaj K. Agarwal, Benjamin Holmgren, Alex Steiger

Let $W \subset \mathbb{R}^2$ be a planar polygonal environment with $n$ vertices, and let $[k] = \{1,\ldots,k\}$ denote $k$ unit-square robots translating in $W$. Given source and target placements $s_1, t_1, \ldots, s_k, t_k \in W$ for each robot, we wish to compute a collision-free motion plan $\mathbf{\pi}$, i.e., a coordinated motion for each robot $i$ a

Olusola Babalola, Bolanle Ojokoh, Olutayo Boyinbode

This research examines the potential of datasets generated by Large Language Models (LLMs) to support Natural Language Processing (NLP) tasks, aiming to overcome challenges related to data acquisition and privacy concerns associated with real-world data. Focusing on negative valence text, a critical component of sentiment analysis, we explore the use of LLM-

Tala Aljaafari, Varun Kanade, Philip Torr, Christian Schroeder de Witt

Deploying reinforcement learning (RL) in safety-critical settings is constrained by brittleness under distribution shift. We study out-of-distribution (OOD) detection for RL time series and introduce DEEDEE, a two-statistic detector that revisits representation-heavy pipelines with a minimal alternative. DEEDEE uses only an episodewise mean and an RBF kernel

Yoana R. Chorbadzhiyska, Peter A. Ivanov, Charlie Nation

The time-dependence of multi-point observable correlation functions are essential quantities in analysis and simulation of quantum dynamics. Open quantum systems approaches utilize two-point correlations to describe the influence of an environment on a system of interest, and in studies of chaotic quantum system, the out-of-time-ordered correlator (OTOC) is

Ziqi Gao, Qiufu Li, Linlin Shen

Compared to 2D data, the scale of point cloud data in different domains available for training, is quite limited. Researchers have been trying to combine these data of different domains for masked autoencoder (MAE) pre-training to leverage such a data scarcity issue. However, the prior knowledge learned from mixed domains may not align well with the downstre

L. T. Khusainova, K. S. Kolegov

This paper is devoted to the problem of identification of colloidal assemblies using the example of two-dimensional coatings (monolayer assemblies). Colloidal systems are used in various fields of science and technology, for example, in applications for photonics and functional coatings. The physical properties depend on the morphology of the structure of th

Jie Zhao, Kang Hao Cheong

Evolutionary algorithms (EAs) have proven effective in exploring the vast solution spaces typical of graph-structured combinatorial problems. However, traditional encoding schemes, such as binary or numerical representations, often fail to straightforwardly capture the intricate structural properties of networks. Through employing the image-based encoding to

Debarun Paul, Sourav Pal, Deepthi Moorkanat, Antara Dey

The redshifted 21-cm signal from the dark ages offers a powerful probe of cosmological models and the underlying dark matter (DM) microphysics. We investigate deviations from the standard $\Lambda$CDM prediction, an absorption trough of approximately $-40.6\,\mathrm{mK}$ at redshift $z\simeq85.6$, in the context of co-SIMP (strongly interacting massive parti

Wonki Lee, Annalisa Pillepich, Dylan Nelson, Myungkook James Jee

Double radio relics, pairs of diffuse radio features located on opposite sides of merging galaxy clusters, are a rare subclass of radio relics that are believed to trace merger shocks and provide valuable constraints on plasma acceleration models and merger history. With the number of known double relics growing in recent and upcoming radio surveys, statisti

Faisal Hamman, Pasan Dissanayake, Yanjun Fu, Sanghamitra Dutta

Knowledge distillation is a promising approach to transfer capabilities from complex teacher models to smaller, resource-efficient student models that can be deployed easily, particularly in task-aware scenarios. However, existing methods of task-aware distillation typically require substantial quantities of data which may be unavailable or expensive to obta

Forough Mahpouya, Sabrina Casucci, Suzanne Sullivan, Christopher Barrick

Informal caregiving often carries a significant emotional, physical, and financial toll, yet caregiver burden is often underrepresented in healthcare research and methods. Existing caregiver burden instruments, while valuable in clinical research, often lack compatibility with observational datasets regularly used in health services research and planning. Th

Lucas Rouhi, Christophe Droz

Metamaterials derive their unconventional properties from engineered microstructures, with periodic lattices providing a versatile framework for modeling wave propagation. Dispersion relations, obtained from Bloch-Floquet theory, govern how waves propagate, attenuate, or localize within such systems. Extending interactions beyond nearest neighbors, through n

Benjamin N. Miller, David H. Meyer, Carter A. Montag, Omar Nagib

Rydberg atomic radio-frequency (rf) sensors are an emerging technology platform that relies on vaporous atoms, interrogated with laser beams and nearly ionized, to receive rf signals. Rydberg rf sensors have a number of interesting fundamental distinctions from traditional receiver technologies, such as those based on metallic antennas, since they are govern

Swarnavo Basu, Karen Alim

Flows are essential to transport resources over large distances. As soon as diffusion becomes time-limiting, flows are needed. Flows are key for the function of multiple human organs, from the blood vasculature to the lungs, the digestive tract, the lymphatic system, and many more. While physics governs the flow dynamics, biology's response to flows governs

George T. Hookway, Martin B. Nielsen, Guy R. Davies, Mikkel N. Lund

The pattern of individual mode frequencies in solar-like oscillators provides valuable insight into their properties and interior structures. The identification and characterisation of these modes requires high signal-to-noise and frequency resolution. The KEYSTONE project unlocks the asteroseismic potential of the K2 mission by providing individually reduce

Christoph Schlegel, Xinyuan Sun

In the neon-lit nights of 2026, Johnson \& Johnson unveiled X. A pill, not larger than a snowflake, that promised a tempest of change. This miraculous drug didn't just allow people to cherry-pick memories to erase from their minds, it could also leave a reminder of this erasure in the minds of those who ingested it. Amidst the iconic red-bricked walls of Har

K. K. Ernazarov

We consider the scalar-Einstein-Gauss-Bonnet (sEGB) $4d$ gravitational model with a scalar field $\varphi\left(u\right)$, Einstein and Gauss-Bonnet terms. The model action contains a potential term $U\left(\varphi\right)$, a Gauss-Bonnet coupling function $f\left(\varphi\right)$ and a parameter $\varepsilon = \pm 1$, where $\varepsilon = 1$ corresponds to th

F. Herzog, B. Ruijl, T. Ueda, J. Vermaseren

At the end of 2016, we computed the five-loop (N$^4$LO) contributions to the beta function in perturbative Quantum Chromodynamics (QCD), its generalization to non-Abelian gauge theories with a simple compact Lie group, and for Quantum Electrodynamics (QED). Here we recall main tools used in and specifically developed for this computation and its main analyti

Qiguang Chen, Jinhao Liu, Libo Qin, Yimeng Zhang

Understanding how information is dynamically accumulated and transformed in human reasoning has long challenged cognitive psychology, philosophy, and artificial intelligence. Existing accounts, from classical logic to probabilistic models, illuminate aspects of output or individual modelling, but do not offer a unified, quantitative description of general hu

Fabian Becker, Sudip KC, Lorenz J. J. Sauerzopf, Tim Schneider

We present polarization-resolved optical absorption measurements on Er3+ ions in CaWO4 under vector magnetic fields, focusing on charge-compensated sites arising from local Ca2+ vacancies. While the known axial Er3+ site displays a single symmetric Zeeman-split transition pattern consistent with S4 symmetry, two additional sites exhibit more complex spectral

Helge Dietert, Lukas Niebel

We prove Nash's $G$ bound for the Kolmogorov equation with rough coefficients. Our proof is inspired by the treatment of the parabolic problem by Nash (1958) and Fabes and Stroock (1986). To transfer their ideas to the kinetic setting, we employ critical kinetic trajectories. From Nash's $G$ bound, we recover the sharp lower bound on the fundamental solution

Markus Exner, Rohan Srikumar, Richard Blättner, Peter Schmelcher

We report the observation of spin-singlet ultra-long range Rydberg butterfly molecules consisting of a ground-state atom bound to a Rydberg atom by $P$-wave scattering of $^{87}$Rb Rydberg electrons from $^{87}$Rb(5s) atoms. A three-photon excitation scheme enables the photoassociation of these molecules by weakly admixing Rb($18f_{7/2}$) states. The measure

Ilia Chernobrovkin, Maurice Debray, Frederik Holst Knudsen, Thibault Capelle

Imaging spatial mode profiles is important for understanding the behavior of mechanical resonators. The recent development of phononic circuits has increased the demand for a fast imaging method based on principles of coherent detection. However, it becomes complicated to perform measurements on a large surface area. Here, we present a frequency-detuned coll

Xiaoxi Li, Wenxiang Jiao, Jiarui Jin, Guanting Dong

Large reasoning models have demonstrated strong problem-solving abilities, yet real-world tasks often require external tools and long-horizon interactions. Existing agent frameworks typically follow predefined workflows, which limit autonomous and global task completion. In this paper, we introduce DeepAgent, an end-to-end deep reasoning agent that performs

Alan Luner, Benjamin Grimmer

We present a performant gradient method for smooth convex optimization, drawing inspiration from several recent advances in the field. Our algorithm, the Adaptive Subgame Perfect Gradient Method (ASPGM) is based on the notion of subgame perfection, attaining a dynamic strengthening of minimax optimality. At each iteration, ASPGM makes a momentum-type update,

Wenyi Wang, Piotr Piękos, Li Nanbo, Firas Laakom

Recent studies operationalize self-improvement through coding agents that edit their own codebases. They grow a tree of self-modifications through expansion strategies that favor higher software engineering benchmark performance, assuming that this implies more promising subsequent self-modifications. However, we identify a mismatch between the agent's self-

Yorie Nakahira, Fangzhou Xiao, Victoria Kostina, John C. Doyle

This paper focuses on rate-limited control of the generalized Ornstein-Uhlenbeck process where the control action can be either multiplicative or additive, and the noise variance can depend on the control action. We derive a lower bound on the data rate necessary to achieve the desired control cost. The lower bound is attained with equality if the control is

Vishnu Varma, Bernhard Mueller, Raphael Hirschi

Core-collapse supernovae (CCSNe) are among the most energetic and complex astrophysical phenomena, requiring threedimensional (3D) simulations to capture their intricate explosion mechanisms. One of the key ingredients for such simulations is the 3D pre-collapse structure, which can impact the development and geometry of the subsequent explosion. While stell

Jens E. d'Hondt, Wieger R. Punter, Odysseas Papapetrou

The increasing need for data privacy and the demand for robust machine learning models have fueled the development of synthetic data generation techniques. However, current methods often succeed in replicating simple summary statistics but fail to preserve both the pairwise and higher-order correlation structure of the data that define the complex, multi-var

Elle Miller, Trevor McInroe, David Abel, Oisin Mac Aodha

Achieving safe, reliable real-world robotic manipulation requires agents to evolve beyond vision and incorporate tactile sensing to overcome sensory deficits and reliance on idealised state information. Despite its potential, the efficacy of tactile sensing in reinforcement learning (RL) remains inconsistent. We address this by developing self-supervised lea

Oscar Davis, Michael S. Albergo, Nicholas M. Boffi, Michael M. Bronstein

Geometric data and purpose-built generative models on them have become ubiquitous in high-impact deep learning application domains, ranging from protein backbone generation and computational chemistry to geospatial data. Current geometric generative models remain computationally expensive at inference -- requiring many steps of complex numerical simulation -

Yuan Zhong

Motivated by applications to the dynamic control of queueing networks, we develop a simulation-based scheme, the so-called multilevel Picard (MLP) approximation, for solving high-dimensional drift control problems whose states are constrained to stay within the nonnegative orthant, over a finite time horizon. We prove that under suitable conditions, the MLP

Jiaxiang Liu, Yuan Wang, Jiawei Du, Joey Tianyi Zhou

Cross-modal alignment aims to map heterogeneous modalities into a shared latent space, as exemplified by models like CLIP, which benefit from large-scale image-text pretraining for strong recognition capabilities. However, when operating in resource-constrained settings with limited or low-quality data, these models often suffer from overconfidence and degra

Xuhui Zhou, Valerie Chen, Zora Zhiruo Wang, Graham Neubig

Recent advances in coding agents have made them capable of planning, editing, running, and testing complex code bases. Despite their growing ability in coding tasks, these systems still struggle to infer and track user intent, especially when instructions are underspecified or context-dependent. To bridge this gap, we introduce ToM-SWE, a dual-agent architec

Xueyuan Lin, Cehao Yang, Ye Ma, Ming Li

Recently, large language models (LLMs) have demonstrated outstanding reasoning capabilities on mathematical and coding tasks. However, their application to financial tasks-especially the most fundamental task of stock movement prediction-remains underexplored. We study a three-class classification problem (up, hold, down) and, by analyzing existing reasoning

Kuicai Dong, Shurui Huang, Fangda Ye, Wei Han

Deep Research systems have revolutionized how LLMs solve complex questions through iterative reasoning and evidence gathering. However, current systems remain fundamentally constrained to textual web data, overlooking the vast knowledge embedded in multimodal documents Processing such documents demands sophisticated parsing to preserve visual semantics (figu

Sera Cremonini, Li Li, Xiao-Long Liu, Jun Nian

We revisit the computation of the shear viscosity to entropy ratio $\eta/s$ at finite chemical potential in a holographic model that takes into account the quantum fluctuations in the IR region of near-extremal black branes. Such quantum corrections can be computed from JT gravity and generate non-trivial temperature dependence for $\eta/s$, which deviates f

Emmanuel Dare Alalade, Ashraf Matrawy

Previous studies on PTA have focused on analyzing privacy threats based on the potential areas of occurrence and their likelihood of occurrence. However, an in-depth understanding of the threat actors involved, their actions, and the intentions that result in privacy threats is essential. In this paper, we present a novel Privacy Threat Model Framework (PTMF

Timothé Boulet, Xavier Hinaut, Clément Moulin-Frier

Software Engineering Agents (SWE-Agents) have proven effective for traditional software engineering tasks with accessible codebases, but their performance for embodied tasks requiring well-designed information discovery remains unexplored. We present the first extended evaluation of SWE-Agents on controller generation for embodied tasks, adapting Mini-SWE-Ag

Thilo Maurer, Markus Bühler, Michael Kröner, Frank Haverkamp

We introduce a prototype FPGA decoder implementing the recently discovered Relay-BP algorithm and targeting memory experiments on the $[[144,12,12]]$ bivariate bicycle quantum low-density parity check code. The decoder is both fast and accurate, achieving a belief propagation iteration time of 24ns. It matches the logical error performance of a floating-poin

Reda Marzouk, Shahaf Bassan, Guy Katz

Although Shapley additive explanations (SHAP) can be computed in polynomial time for simple models like decision trees, they unfortunately become NP-hard to compute for more expressive black-box models like neural networks - where generating explanations is often most critical. In this work, we analyze the problem of computing SHAP explanations for *Tensor N

Diana Cai, Robert M. Gower, David M. Blei, Lawrence K. Saul

We introduce a highly expressive yet distinctly tractable family for black-box variational inference (BBVI). Each member of this family is a weighted product of experts (PoE), and each weighted expert in the product is proportional to a multivariate $t$-distribution. These products of experts can model distributions with skew, heavy tails, and multiple modes

José Rojas, Enrique Casanova, Melvin Arias

We investigate precise structural relations between the standard Schr\"odinger equation and its Carrollian analogue-the Carroll-Schr\"odinger equation-in 1+1 dimensions, with emphasis on dualities, potential maps, and solution behavior. Our contributions proceed in the order of the paper: (i) we encode both dynamics with operators $H$ and $F$ under external

Pauline Mouches, Julien Jung, Armand Demasson, Agnès Guinard

In drug-resistant epilepsy, presurgical evaluation of epilepsy can be considered. Magnetoencephalography (MEG) has been shown to be an effective exam to inform the localization of the epileptogenic zone through the localization of interictal epileptic spikes. Manual detection of these pathological biomarkers remains a fastidious and error-prone task due to t

Diederik van Engelenburg, Christophe Garban, Romain Panis, Franco Severo

We consider sufficiently spread-out Bernoulli percolation in dimensions ${d>6}$. We present a short and simple proof of the up-to-constants estimate for the one-arm probability in both the full-space and half-space settings. These results were previously established by Kozma and Nachmias and by Chatterjee and Hanson, respectively. Our proof improves upon the

Krishna Kumar Madhavan Vijayalakshmi

The Koras-Russell threefolds are a certain family of smooth, affine contractible threefolds exhibiting "exotic" behavior in the algebro-geometric context. Our goal in this note is to extend its $\mathbb{A}^1$-contractibility from a field to a general base scheme. As a consequence, we also give a general strategy to extend the $\mathbb{A}^1$-contractibility o

Lei Liu, Zhenxin Huang, Hong Wang, huanshuo dong

Data-driven deep learning methods like neural operators have advanced in solving nonlinear temporal partial differential equations (PDEs). However, these methods require large quantities of solution pairs\u2014the solution functions and right-hand sides (RHS) of the equations. These pairs are typically generated via traditional numerical methods, which need

Oleksandr Kosenkov, Ehsan Zabardast, Davide Fucci, Daniel Mendez

Context: Consistent requirements and system specifications are essential for the compliance of software systems towards the General Data Protection Regulation (GDPR). Both artefacts need to be grounded in the original text and conjointly assure the achievement of privacy by design (PbD). Objectives: There is little understanding of the perspectives of practi

Jonathan Amar, Edward Liu, Alessandra Breschi, Liangliang Zhang

This paper introduces an innovative Electronic Health Record (EHR) foundation model that integrates Polygenic Risk Scores (PRS) as a foundational data modality, moving beyond traditional EHR-only approaches to build more holistic health profiles. Leveraging the extensive and diverse data from the All of Us (AoU) Research Program, this multimodal framework ai

Xi Chen, Diptaksho Palit, Kabir Peshawaria, William Pires

The model of relative-error property testing of Boolean functions has been the subject of significant recent research effort [CDH+24][CPPS25a][CPPS25b] In this paper we consider the problem of relative-error testing an unknown and arbitrary $f: \{0,1\}^n \to \{0,1\}$ for the property of being a unate function, i.e. a function that is either monotone non-incr

Farhad Pashakhanloo

Biological and artificial learners are inherently exposed to a stream of data and experience throughout their lifetimes and must constantly adapt to, learn from, or selectively ignore the ongoing input. Recent findings reveal that, even when the performance remains stable, the underlying neural representations can change gradually over time, a phenomenon kno

Bho Matthiesen, Armin Dekorsy, Petar Popovski

5G networks offer exceptional reliability and availability, ensuring consistent performance and user satisfaction. Yet they might still fail when confronted with the unexpected. A resilient system is able to adapt to real-world complexity, including operating conditions completely unanticipated during system design. This makes resilience a vital attribute fo

Xuzhao Li, Xuchen Li, Shiyu Hu

Nighttime UAV tracking faces significant challenges in real-world robotics operations. Low-light conditions not only limit visual perception capabilities, but cluttered backgrounds and frequent viewpoint changes also cause existing trackers to drift or fail during deployment. To address these difficulties, researchers have proposed solutions based on low-lig

Yassine El Ouahidi, Jonathan Lys, Philipp Thölke, Nicolas Farrugia

Foundation models have transformed AI by reducing reliance on task-specific data through large-scale pretraining. While successful in language and vision, their adoption in EEG has lagged due to the heterogeneity of public datasets, which are collected under varying protocols, devices, and electrode configurations. Existing EEG foundation models struggle to

Kellen Parker van Dam, Abishek Stephen

Lexical data collection in language documentation often contains transcription errors and undocumented borrowings that can mislead linguistic analysis. We present unsupervised anomaly detection methods to identify phonotactic inconsistencies in wordlists, applying them to a multilingual dataset of Kokborok varieties with Bangla. Using character-level and syl

Diego Doimo

The goal of this thesis is to improve our understanding of the internal mechanisms by which deep artificial neural networks create meaningful representations and are able to generalize. We focus on the challenge of characterizing the semantic content of the hidden representations with unsupervised learning tools, partially developed by us and described in th

Ciara Rowles, Varun Jampani, Simon Donné, Shimon Vainer

Foley Control is a lightweight approach to video-guided Foley that keeps pretrained single-modality models frozen and learns only a small cross-attention bridge between them. We connect V-JEPA2 video embeddings to a frozen Stable Audio Open DiT text-to-audio (T2A) model by inserting compact video cross-attention after the model's existing text cross-attentio

Tomas Lestayo Martinez, Manuel Fernandez Veiega Veiga

Multicast remains a fundamental mechanism for scalable content distribution, yet existing approaches face critical limitations. Traditional multicast trees suffer from path redundancy and inefficient utilization of network resources, while network coding, although capacity-achieving, incurs significant computational overhead and deployment challenges. In thi

Ken Newman, Shaini Naha, Leah Jackson-Blake, Cairistiona Topp

Global sensitivity analysis (GSA) is a recommended step in the use of computer simulation models. GSA quantifies the relative importance of model inputs on outputs (Factor Ranking), identifies inputs that could be fixed, thus simplifying model calibration (Factor Fixing), and pinpointing areas for future data collection (Factor Prioritization). Given the wid

Jia Li, Yajiang Hao

We analytically investigated the dynamical quantum phase transitions in the Bose-Hubbard model using the Loschmidt echo as an observable, revealing that after a quench, the global Loschmidt echo exhibits cusp singularities with a logarithmically divergent rate function near criticality and a critical exponent of zero. Through extensive calculations across va

M. Carretero-Castrillo, M. Ribó, J. M. Paredes, G. Holgado

Gaia DR3 data have revealed new massive runaway stars, while spectroscopic surveys enable detailed characterization. The relative contributions of binary supernova (BSS) and dynamical ejection (DES) scenarios to explain their runaway origin remain poorly constrained, particularly in the Milky Way. We aim to characterize the largest sample of Galactic O-type

Lorenzo Andreoli, Ronen Weiss, Graham Chambers-Wall, Alex Gnech

We present an approach for including relativistic corrections in lepton-nucleus scattering calculations within the Short-Time Approximation (STA). Previous ab-initio studies employed electromagnetic currents expanded in powers of $q/m$, where $q$ is the momentum transfer and $m$ is the nucleon mass, restricting their validity to low-$q$ kinematics. We adopt

Mojca Brglez, Špela Vintar

Large language models are demonstrating increasing capabilities, excelling at benchmarks once considered very difficult. As their capabilities grow, there is a need for more challenging evaluations that go beyond surface-level linguistic competence. Namely, language competence involves not only syntax and semantics but also pragmatics, i.e., understanding si

Jason Wu, Petar Veličković

Neural networks excel at processing unstructured data but often fail to generalise out-of-distribution, whereas classical algorithms guarantee correctness but lack flexibility. We explore whether pretraining Graph Neural Networks (GNNs) on classical algorithms can improve their performance on molecular property prediction tasks from the Open Graph Benchmark:

Matthew Crawford, Pavan Kartik, Reese Lance

We consider cohomological stable envelopes for a natural torus action $\mathsf{T}$ on $X=T^*Gr(k,n)$, introduced by Maulik-Okounkov. We define the $\mathbb{C}^*_\hbar$-equivariant integral of the stable envelope using equivariant localization over the subtorus $\mathbb{C}^*_\hbar\subset\mathsf{T}$, and compute the integral as a non-equivariant limit of the l

Raktim Mukhopadhyay, Marianthi Markatou

Even though several publicly accessible pharmacovigilance databases are available, extracting data from them is a technically challenging process. Existing tools typically focus on a single database. We present SurVigilance, an open-source tool that streamlines the process of retrieving safety data from seven major pharmacovigilance databases. SurVigilance p

Qixiu Li, Yu Deng, Yaobo Liang, Lin Luo

This paper presents a novel approach for pretraining robotic manipulation Vision-Language-Action (VLA) models using a large corpus of unscripted real-life video recordings of human hand activities. Treating human hand as dexterous robot end-effector, we show that "in-the-wild" egocentric human videos without any annotations can be transformed into data forma

Songyuan Li, Teng Wang, Jinrong Tang, Ruiqi Liu

Fully analogue neural computation requires hardware that can implement both linear and nonlinear transformations without digital assistance. While analogue in-memory computing efficiently realizes matrix-vector multiplication, the absence of learnable analogue nonlinearities remains a central bottleneck. Here we introduce KANalogue, a fully analogue realizat

Tran Quang Hoa, Nguyen Duy Phuoc, Tran Nguyen Thanh Son

In this paper, we investigate the weak Lefschetz property for tensor products of Artinian monomial algebras and complete quadratic monomial algebras. As an application, we classify the weak Lefschetz property of the Artinian algebras $A(L_{m,n})$, which are defined by the edge ideals of the lollipop graphs $L_{m,n}$ together with the squares of the variables

Udomsilp Pinsook

In this work, I employ parabolic cylinder functions to quantitatively describe the ARPES spectra of an over-doped Bi2212 across the temperature range 6 - 140K at the antinode k-point. These functions come from the solutions of a particle moving in a system of random scatterers. The parameters, i.e. the overall amplitude (A), the spectral coherence scale (C),

Youngmin Park, Thomas G. Fai

We study the dynamics of molecular motor-driven transport into dendritic spines, which are bulbous intracellular compartments in neurons that play a key role in transmitting signals between neurons. We further develop a stochastic model of vesicle transport in [Park, Singh, and Fai, SIAM J. Appl. Math. 82.3 (2022), pp. 793--820] by showing that second-order

Fangwen Wu, Zheng Wu, Jihong Wang, Yunku Chen

With the rapid development of (multimodal) large language model-based agents, the landscape of agentic service management has evolved from single-agent systems to multi-agent systems, and now to massive-agent ecosystems. Current massive-agent ecosystems face growing challenges, including impersonal service experiences, a lack of standardization, and untrustw

Mohamad Hossein Idjadi, Haotian Cheng, Farshid Ashtiani, Benjia Li

Narrow-linewidth lasers are vital for a broad range of scientific and technological applications, including atomic clocks and precision sensing. Achieving high frequency stability is often as critical as ensuring scalability, portability, and cost-effectiveness in the development of low noise laser systems. Conventional electro-optic stabilization techniques

Graham Chambers-Wall, Justin Lieffers, Garrett B. King, Emanuele Mereghetti

We derive the three-nucleon neutrinoless double beta decay potential in $\Delta$-full chiral effective field theory through next-to-next-to-next-to leading order in Weinberg's power counting. The matrix elements of the resulting operators are computed in light nuclei using Variational Monte Carlo with wave functions constructed from the Norfolk family of nuc

Michael Hofstetter

Using a stochastic control approach we establish couplings of the Liouville field and the sinh-Gordon field with the Gaussian free field in dimension $d=2$, such that the difference is in a Sobolev space of regularity $\alpha>1$. The analysis covers the entire $L^2$ phase. Our main tools are estimates for the short scales of the minimiser of the variational

Thomas Agugliaro

We prove that the standard conjecture of Hodge type holds for powers of abelian threefolds. Along the way, we also prove the conjecture for powers of simple abelian variety of prime dimension over finite fields, and in other related cases based on the notion of Frobenius rank of Lenstra-Zarhin. The main tool is a result comparing two real fiber functors on t

Sandra Mitrović, David Kletz, Ljiljana Dolamic, Fabio Rinaldi

Large Language Models (LLMs) display notable variation in multilingual behavior, yet the role of genealogical language structure in shaping this variation remains underexplored. In this paper, we investigate whether LLMs exhibit sensitivity to linguistic genera by extending prior analyses on the MultiQ dataset. We first check if models prefer to switch to ge

Yuxuan Yang, Hussein Sibai

Safety is a fundamental requirement for autonomous systems operating in critical domains. Control barrier functions (CBFs) have been used to design safety filters that minimally alter nominal controls for such systems to maintain their safety. Learning neural CBFs has been proposed as a data-driven alternative for their computationally expensive optimization

Dae san Kim, Taekyun Kim

We investigate the representation of arbitrary polynomials using probabilistic Bernoulli and degenerate Bernoulli polynomials associated with a random variable $Y$, whose moment generating function exists in a neighborhood of the origin. In addition, this paper explores the problem of representing arbitrary polynomials in terms of their higher-order counterp

Hongwei Zhang, Ji Lu, Shiqing Jiang, Chenxiang Zhu

Long-horizon reasoning in LLM-based agents often fails not from generative weakness but from insufficient verification of intermediate reasoning. Co-Sight addresses this challenge by turning reasoning into a falsifiable and auditable process through two complementary mechanisms: Conflict-Aware Meta-Verification (CAMV) and Trustworthy Reasoning with Structure

Sophie Hall, Florian Dörfler, Timm Faulwasser

Generalized Nash equilibria are used in multi-agent control applications to model strategic interactions between agents that are coupled in the cost, dynamics, and constraints, and provide the foundations for game-theoretic MPC (Receding Horizon Games). We study properties of finite-horizon dynamic GNE trajectories from a system-theoretic perspective. We sho

Marco A. A. de Paula, Mustapha Azreg-Aïnou

Spacetimes arising from nonlinear electrodynamics (NED) are a good laboratory for studying both the nature of regular black holes (RBH) solutions and the imprints of nonlinear electromagnetic fields within this context. Over the past few decades, NED-sourced black hole (BH) spacetimes have attracted considerable attention, but electrically charged RBHs obtai