May 2025 arXiv papers — page 76

Yossi Bokor Bleile, Lisbeth Fajstrup, Teresa Heiss, Anne Marie Svane

Motivated by applications in chemistry, we give a homlogical definition of tunnels, or more generally cobordisms, connecting disjoint parts of a cell complex. For a filtered complex, this defines a persistence module. We give a method for identifying birth and death times using kernel persistence and a matrix reduction algorithm for pairing birth and death t

Seth Siriya, Julian D. Schiller, Victor G. Lopez, Matthias A. Müller

In many sampled-data applications, observers are designed based on approximately discretized models of continuous-time systems, where usually only the discretized system is analyzed in terms of its detectability. In this paper, we show that if the continuous-time system satisfies certain linear matrix inequality (LMI) conditions, and the sampling period of t

Moule Lin, Shuhao Guan, Weipeng Jing, Goetz Botterweck

While offering a principled framework for uncertainty quantification in deep learning, the employment of Bayesian Neural Networks (BNNs) is still constrained by their increased computational requirements and the convergence difficulties when training very deep, state-of-the-art architectures. In this work, we reinterpret weight-sharing quantization technique

Jingyi Sun, Greta Warren, Irina Shklovski, Isabelle Augenstein

Understanding sources of a model's uncertainty regarding its predictions is crucial for effective human-AI collaboration. Prior work proposes using numerical uncertainty or hedges ("I'm not sure, but ..."), which do not explain uncertainty that arises from conflicting evidence, leaving users unable to resolve disagreements or rely on the output. We introduce

Lukas Koller, Tobias Ladner, Matthias Althoff

Neural networks are ubiquitous. However, they are often sensitive to small input changes. Hence, to prevent unexpected behavior in safety-critical applications, their formal verification -- a notoriously hard problem -- is necessary. Many state-of-the-art verification algorithms use reachability analysis or abstract interpretation to enclose the set of possi

Barry Minemyer

In this paper we prove that, at least in even complex dimensions, the ratio of Chern numbers for a closed complex hyperbolic branched cover manifold are not all equal to the corresponding ratio of Chern numbers for a closed complex hyperbolic manifold. This leads to an answer for a question posed by Deraux and Seshadri, and proves that an almost $1/4$-pinche

Francois Chaubard, Mykel Kochenderfer

During inference, Recurrent Neural Networks (RNNs) scale constant in both FLOPs and GPU memory with increasing context length, as they compress all prior tokens into a fixed-size memory. In contrast, transformers scale linearly in FLOPs and, at best, linearly in memory during generation, since they must attend to all previous tokens explicitly. Despite this

Yuanya She

In this paper, we propose a robust and reinforcement-learning-enhanced network intrusion detection system (NIDS) designed for class-imbalanced and few-shot attack scenarios in Industrial Internet of Things (IIoT) environments. Our model integrates a TabTransformer for effective tabular feature representation with Proximal Policy Optimization (PPO) to optimiz

Yanglei Song, Berkan Dulek, Sinan Gezici

The composite binary hypothesis testing problem within the Neyman-Pearson framework is considered. The goal is to maximize the expectation of a nonlinear function of the detection probability, integrated with respect to a given probability measure, subject to a false-alarm constraint. It is shown that each power function can be realized by a generalized Baye

CMS Collaboration

A measurement of event-shape variables is presented, using a data sample produced in a special run with approximately one inelastic proton-proton collision per bunch crossing. The data were collected with the CMS detector at a center-of-mass energy of 13 TeV, corresponding to an integrated luminosity of 64 $\mu$b$^{-1}$. A number of observables related to th

Daniel VandenHeuvel, Sheehan Olver

We develop an efficient $hp$-finite element method for piecewise-smooth differential equations with periodic boundary conditions, using orthogonal polynomials defined on circular arcs. The operators derived from this basis are banded and achieve optimal complexity regardless of $h$ or $p$, both for building the discretisation and solving the resulting linear

NEXT Collaboration, C. Adams, H. Almazán, V. Álvarez

The NEXT collaboration is dedicated to the study of double beta decays of $^{136}$Xe using a high-pressure gas electroluminescent time projection chamber. This advanced technology combines exceptional energy resolution ($\leq 1\%$ FWHM at the $Q_{\beta\beta}$ value of the neutrinoless double beta decay) and powerful topological event discrimination. Building

Hao Wang, Licheng Pan, Zhichao Chen, Xu Chen

Training time-series forecast models presents unique challenges in designing effective learning objectives. Existing methods predominantly utilize the temporal mean squared error, which faces two critical challenges: (1) label autocorrelation, which leads to bias from the label sequence likelihood; (2) excessive amount of tasks, which increases with the fore

Yu-cheng Chen, Ronin Wu, M. H. Cheng, Min-Hsiu Hsieh

Quantum computing promises to revolutionize many-body simulations for quantum chemistry, but its potential is constrained by limited qubits and noise in current devices. In this work, we introduce the Lossy Quantum Selected Configuration Interaction (Lossy-QSCI) framework, which combines a lossy subspace Hamiltonian preparation pipeline with a generic QSCI s

Riccardo Ontani

We consider genus zero quasimap invariants of smooth projective targets of the form $V/\!/G$, where $V$ is a representation of a reductive group $G$. In particular we consider integrals of cohomology classes arising as characteristic classes of the universal quasimap. In this setting, we provide a way to express the invariants of $V/\!/G$ in terms of invaria

Fabian Deuser, Philipp Hausenblas, Hannah Schieber, Daniel Roth

Contrastive learning is a representational learning paradigm in which a neural network maps data elements to feature vectors. It improves the feature space by forming lots with an anchor and examples that are either positive or negative based on class similarity. Hard negative examples, which are close to the anchor in the feature space but from a different

PeiZhen Cheng, XingQian Chen, GuiLin Liao, Ying Gu

In this manuscript, an improved Seyfert-LINER classification line (= S-L line) is proposed in the [N~{\sc ii}] BPT diagram, based on a sample of 47,968 low redshift narrow emission line galaxies from SDSS DR16, motivated by different S-L lines reported in the [N~{\sc ii}] BPT diagram through different methods. The method proposed by Kewley et al. in 2006 is

Juan C. Boschero, Niels M. P. Neumann, Ward van der Schoot, Frank Phillipson

Distributed quantum computing (DQC) is crucial for high-volume quantum processing in the NISQ era. Many different technologies are utilized to implement a quantum computer, each with a different advantages and disadvantages. Various research is performed on how to implement DQC within a certain technology, but research on DQC between different technologies i

Wiebke Hutiri, Mircea Cimpoi, Morgan Scheuerman, Victoria Matthews

Dataset transparency is a key enabler of responsible AI, but insights into multimodal dataset attributes that impact trustworthy and ethical aspects of AI applications remain scarce and are difficult to compare across datasets. To address this challenge, we introduce Trustworthy and Ethical Dataset Indicators (TEDI) that facilitate the systematic, empirical

Martin Panholzer, Michael Haring, Thomas Wallek, Robert E. Zillich

Properties of classical molecular systems can be calculated with integral equation theories based on the Ornstein-Zernike (OZ) equation and a complementing closure relation. One such closure relation is the hyper netted chain (HNC) approximation, which neglects the so-called bridge function. We present a new way to use machine learning to train a deep operat

Tim Rosenbach, David Heidrich, Alexander Weinert

One important step in software development is testing the finished product with actual users. These tests aim, among other goals, at determining unintuitive behavior of the software as it is presented to the end-user. Moreover, they aim to determine inconsistencies in the user-facing interface. They provide valuable feedback for the development of the softwa

Abhiti Mishra, Yash Patel, Ambuj Tewari

Transformers robustly exhibit the ability to perform in-context learning, whereby their predictive accuracy on a task can increase not by parameter updates but merely with the placement of training samples in their context windows. Recent works have shown that transformers achieve this by implementing gradient descent in their forward passes. Such results, h

Vincent Wüst, Erdem Eray Cil, Laurent Schmalen

Forward error correcting (FEC) codes are used in many communication standards with a wide range of re quirements. FEC codes should work close to capacity, achieve low error floors, and have low decoding complexity. In this paper, we propose a novel category of low-density parity-check (LDPC) codes, based on protograph codes with local irregularity. This new

Anna Van Elst, Igor Colin, Stephan Clémençon

This paper addresses the problem of robust estimation in gossip algorithms over arbitrary communication graphs. Gossip algorithms are fully decentralized, relying only on local neighbor-to-neighbor communication, making them well-suited for situations where communication is constrained. A fundamental challenge in existing mean-based gossip algorithms is thei

Marc Lalonde, Hamed Ghodrati

The task of grading atopic dermatitis (or AD, a form of eczema) from patient images is difficult even for trained dermatologists. Research on automating this task has progressed in recent years with the development of deep learning solutions; however, the rapid evolution of multimodal models and more specifically vision-language models (VLMs) opens the door

Shy-el Cohen, Yoni Choukroun, Eliya Nachmani

We introduce a novel deep learning method for decoding error correction codes based on the Mamba architecture, enhanced with Transformer layers. Our approach proposes a hybrid decoder that leverages Mamba's efficient sequential modeling while maintaining the global context capabilities of Transformers. To further improve performance, we design a novel layer-

Kalle Lahtinen, Einari Vaaras, Liisa Mustanoja, Okko Räsänen

Study of affect in speech requires suitable data, as emotional expression and perception vary across languages. Until now, no corpus has existed for natural expression of affect in spontaneous Finnish, existing data being acted or from a very specific communicative setting. This paper presents the first such corpus, created by annotating 12,000 utterances fo

Giacomo Magnifico, Eduard Barbu

Clear and effective explanations are essential for human understanding and knowledge dissemination. The scope of scientific research aiming to understand the essence of explanations has recently expanded from the social sciences to machine learning and artificial intelligence. Explanations for machine learning decisions must be impactful and human-like, and

Shaghayegh Hosseinpour, Sanchari Das

Smishing, or SMS-based phishing, poses an increasing threat to mobile users by mimicking legitimate communications through culturally adapted, concise, and deceptive messages, which can result in the loss of sensitive data or financial resources. In such, we present a multi-channel smishing detection model that combines country-specific semantic tagging, str

Alessandro Ramponi, M. Elisabetta Tessitore

In this paper, we present a mathematical model to describe the temporal evolution of delinquent behavior, treating it as a socially transmitted phenomenon influenced by peer interactions, thus similar to an epidemic. We consider a compartmental framework involving three ordinary differential equations to describe the dynamics among the three population group

Nicolas Castanet, Olivier Sigaud, Sylvain Lamprier

Goal-Conditioned Reinforcement Learning (GCRL) enables agents to autonomously acquire diverse behaviors, but faces major challenges in visual environments due to high-dimensional, semantically sparse observations. In the online setting, where agents learn representations while exploring, the latent space evolves with the agent's policy, to capture newly disc

Zezhong Wang, Xingshan Zeng, Weiwen Liu, Yufei Wang

Mathematical reasoning through Chain-of-Thought (CoT) has emerged as a powerful capability of Large Language Models (LLMs), which can be further enhanced through Test-Time Scaling (TTS) methods like Beam Search and DVTS. However, these methods, despite improving accuracy by allocating more computational resources during inference, often suffer from path homo

Shingo Hirano, Naoki Yoshida

Ambient dark matter (DM) around binary black holes can imprint characteristic signatures on gravitational waves emitted from their merger. The exact signature depends sensitively on the DM density profile around the black holes. We run very high resolution cosmological hydrodynamics simulations of first star formation that follow the collapse of a $3\times10

Hang Yuan, Bin Yu, Haotian Li, Shijun Yang

Modern reasoning models, such as OpenAI's o1 and DeepSeek-R1, exhibit impressive problem-solving capabilities but suffer from critical inefficiencies: high inference latency, excessive computational resource consumption, and a tendency toward overthinking -- generating verbose chains of thought (CoT) laden with redundant tokens that contribute minimally to t

Xuchen Pan, Yanxi Chen, Yushuo Chen, Yuchang Sun

Trinity-RFT is a general-purpose, unified and easy-to-use framework designed for reinforcement fine-tuning (RFT) of large language models. It is built with a modular and decoupled design, consisting of (1) an RFT-core that unifies and generalizes synchronous/asynchronous, on-policy/off-policy, and online/offline modes of RFT; (2) seamless integration for age

Mingkuan Feng, Jinyang Wu, Siyuan Liu, Shuai Zhang

The deployment of large language models (LLMs) is largely hindered by their large number of parameters. Structural pruning has emerged as a promising solution. Prior structured pruning methods directly remove unimportant parameters based on certain metrics, which often causes knowledge loss and necessitates extensive retraining. To overcome this, we introduc

José Correia, Mark Hindmarsh, Kari Rummukainen, David J. Weir

We study gravitational wave production at strong first order phase transitions, with large-scale, long-running simulations of a system with a scalar order parameter and a relativistic fluid. One transition proceeds by detonations with asymptotic wall speed $v_\text{w}=0.92$ and transition strength $\alpha_n=0.67$, and the other by deflagrations, with a nomin

Gerardo Roa-Dabike, Trevor J. Cox, Jon P. Barker, Michael A. Akeroyd

Musical (MSS) source separation of western popular music using non-causal deep learning can be very effective. In contrast, MSS for classical music is an unsolved problem. Classical ensembles are harder to separate than popular music because of issues such as the inherent greater variation in the music; the sparsity of recordings with ground truth for superv

Martin Poblet, Christian Vinther Bertelsen, David Alonso-Tomas, Rahul Singh

One-dimensional photonic crystal (1D-PhC) pillar cavities allow transducing mechanical pillar vibrations to the optical domain, thereby relaxing the requirements typically associated with mechanical motion detection. In this study, we integrate these geometries into a silicon-on-insulator photonics platform and explore their optical and mechanical properties

Shihao Li, Chenglong Li, Aihua Zheng, Jin Tang

Multi-spectral object re-identification (ReID) brings a new perception perspective for smart city and intelligent transportation applications, effectively addressing challenges from complex illumination and adverse weather. However, complex modal differences between heterogeneous spectra pose challenges to efficiently utilizing complementary and discrepancy

Jingyun Du

Bluetooth backscatter systems, as a crucial technology for low-power communication in the Internet of Things (IoT), have witnessed remarkable development in recent years. This article comprehensively analyzes multiple related papers, including the latest advancements in RF-Transformer and B2Loc systems, summarizes their research progress, challenges faced, a

Carlos Duaso Pueyo, Harry Goodhew, Ciaran McCulloch, Enrico Pajer

Much of our intuition about Effective Field Theories (EFTs) stems from their formulation in flat spacetime, yet EFTs have become indispensable tools in cosmology, where time-dependent backgrounds are the norm. In this work, we demonstrate that in spacetimes undergoing significant expansion-such as accelerated FLRW and de Sitter backgrounds-the contributions

Jürgen Dölz, Jolanda Weygandt

Spectral clustering is a popular unsupervised learning technique which is able to partition unlabelled data into disjoint clusters of distinct shapes. However, the data under consideration are often experimental data, implying that the data is subject to measurement errors and measurements may even be lost or invalid. These uncertainties in the corrupted inp

Daeun Kyung, Hyunseung Chung, Seongsu Bae, Jiho Kim

Doctor-patient consultations require multi-turn, context-aware communication tailored to diverse patient personas. Training or evaluating doctor LLMs in such settings requires realistic patient interaction systems. However, existing simulators often fail to reflect the full range of personas seen in clinical practice. To address this, we introduce PatientSim

Theodore D. Drivas, Daniel Ginsberg, Marc Nualart

Inviscid laminar flow is a stationary solution of the incompressible Euler equations whose streamlines foliate the fluid domain. Their structure on symmetric domains is rigid: all laminar flows occupying straight periodic channels are shear and on regular annuli they are circular. Laminarity can persist to slight deformations of these domains provided the ba

Hei Yi Mak, Tan Lee

The majority of inhabitants in Hong Kong are able to read and write in standard Chinese but use Cantonese as the primary spoken language in daily life. Spoken Cantonese can be transcribed into Chinese characters, which constitute the so-called written Cantonese. Written Cantonese exhibits significant lexical and grammatical differences from standard written

Yihe Fan, Wenqi Zhang, Xudong Pan, Min Yang

As foundation models grow increasingly more intelligent, reliable and trustworthy safety evaluation becomes more indispensable than ever. However, an important question arises: Whether and how an advanced AI system would perceive the situation of being evaluated, and lead to the broken integrity of the evaluation process? During standard safety tests on a ma

Stefan H. Strub, Lorenzo Speri, Domenico Giardini

The Laser Interferometer Space Antenna (LISA) is a planned space-based observatory designed to detect gravitational waves (GWs) within the millihertz frequency range. LISA is anticipated to observe the inspiral of compact objects into black holes at the centers of galaxies, so called extreme-mass-ratio inspirals (EMRIs). However, the extraction of these long

Michael Hassid, Gabriel Synnaeve, Yossi Adi, Roy Schwartz

Reasoning large language models (LLMs) heavily rely on scaling test-time compute to perform complex reasoning tasks by generating extensive "thinking" chains. While demonstrating impressive results, this approach incurs significant computational costs and inference time. In this work, we challenge the assumption that long thinking chains results in better re

Boxu Chen, Ziwei Zheng, Le Yang, Zeyu Geng

Large Vision-Language Models (LVLMs) have achieved remarkable success but continue to struggle with object hallucination (OH), generating outputs inconsistent with visual inputs. While previous work has proposed methods to reduce OH, the visual decision-making mechanisms that lead to hallucinations remain poorly understood. In this paper, we propose VaLSe, a

I. V. Anikin

For the reconstruction problem, the universal representation of inverse Radon transforms implies the needed complexity of the direct Radon transforms which leads to the additional contributions. In the standard theory of generalized functions, if the outset (origin) function which generates the Radon image is a pure-real function, as a rule, the complexity o

Devika Dass, Amol Delmade, Agastya Raj, Eoin Kenny

We demonstrate the end-to-end transmission of digital coherent and analog radio-over-fiber signals, at mmWave and sub-THz frequencies, over the HEAnet live production metro network using Optical Spectrum-as-a-Service (OSaaS), transparently connected to a passive optical network.

Ramanathan Swaminathan

This research work reveals the strengths of intertwining a deep custom convolutional neural network with a disruptive Vision Transformer, both fused together with a radical Cross-Attention module. Here, two high-yielding datasets for artificial intelligence models in detecting glaucoma, namely ACRIMA and Drishti, are utilized. The Cross-Attention mechanism f

Ping Li, Jianan Ni, Bo Pang

Action recognition models using deep learning are vulnerable to adversarial examples, which are transferable across other models trained on the same data modality. Existing transferable attack methods face two major challenges: 1) they heavily rely on the assumption that the decision boundaries of the surrogate (a.k.a., source) model and the target model are

Hang Yang, Dexue Zhang

We present a Stone duality for bitopological spaces in analogy to the duality between Stone spaces and Boolean algebras, in the same vein as the duality between d-sober bitopological spaces and spatial d-frames established by Jung and Moshier. Precisely, we introduce the notion of d-Boolean algebras and prove that the category of such algebras is dually equi

Buyan Li, Jie Xiao

Based on the construction of simple Lie algebras via root category and following Chevalley's results, we construct Chevalley groups from the root category. Then we prove the Bruhat decomposition and the simplicity of the Chevalley groups, and calculate the orders of finite Chevalley groups.

Jonas Seng, Fabrizio Ventola, Zhongjie Yu, Kristian Kersting

Despite the growing interest in designing truly interactive hyperparameter optimization (HPO) methods, to date, only a few allow to include human feedback. Existing interactive Bayesian optimization (BO) methods incorporate human beliefs by weighting the acquisition function with a user-defined prior distribution. However, in light of the non-trivial inner o

Friederike Preusse

We propose a method that combines the closed testing framework with the concept of safe anytime-valid inference (SAVI) to compute lower confidence bounds for the true discovery proportion in a multiple testing setting. The proposed procedure provides confidence bounds that are valid at every observation time point and that are simultaneous for all possible s

Iffat Fatima, Patricia Lago

In the international software engineering research community, the premier conference (ICSE) features since a decade a special track on the role of SE In Society (or SEIS track). In this work, we want to use the articles published in this track as a proxy or example of the research in this field, in terms of covered topics, trends, and gaps. Also, since SEIS

Jonathan B. Hill

We derive an $\mathcal{L}_{q}$-maximal inequality for zero mean dependent random variables $\{x_{t}\}_{t=1}^{n}$ on $\mathbb{R}^{p}$, where $p$ $>>$ $% n $ is allowed. The upper bound is a familiar multiple of $\ln (p)$ and an $% l_{\infty }$ moment, as well as Kolmogorov distances based on Gaussian approximations $(\rho _{n},\tilde{\rho}_{n})$, derived with

Brian B. Moser, Arundhati S. Shanbhag, Stanislav Frolov, Federico Raue

Coreset selection targets the challenge of finding a small, representative subset of a large dataset that preserves essential patterns for effective machine learning. Although several surveys have examined data reduction strategies before, most focus narrowly on either classical geometry-based methods or active learning techniques. In contrast, this survey p

Ana Petrovska, Guan Erjiage, Stefan Kugele

In the last two decades, the popularity of self-adaptive systems in the field of software and systems engineering has drastically increased. However, despite the extensive work on self-adaptive systems, the literature still lacks a common agreement on the definition of these systems. To this day, the notion of self-adaptive systems is mainly used intuitively

Louis Béthune, David Vigouroux, Yilun Du, Rufin VanRullen

What is the shortest path between two data points lying in a high-dimensional space? While the answer is trivial in Euclidean geometry, it becomes significantly more complex when the data lies on a curved manifold -- requiring a Riemannian metric to describe the space's local curvature. Estimating such a metric, however, remains a major challenge in high dim

Manuel Morante, Naveed ur Rehman

We introduce Variational Latent Mode Decomposition (VLMD), a new algorithm for extracting oscillatory modes and associated connectivity structures from multivariate signals. VLMD addresses key limitations of existing Multivariate Mode Decomposition (MMD) techniques -including high computational cost, sensitivity to parameter choices, and weak modeling of int

Yuxin Yang, Yinan Zhou, Yuxin Chen, Ziqi Zhang

Composed Image Retrieval (CIR) aims to retrieve target images from a gallery based on a reference image and modification text as a combined query. Recent approaches focus on balancing global information from two modalities and encode the query into a unified feature for retrieval. However, due to insufficient attention to fine-grained details, these coarse f

Mingning Guo, Mengwei Wu, Jiarun He, Shaoxian Li

With the rapid advancement of low-altitude remote sensing and Vision-Language Models (VLMs), Embodied Agents based on Unmanned Aerial Vehicles (UAVs) have shown significant potential in autonomous tasks. However, current evaluation methods for UAV-Embodied Agents (UAV-EAs) remain constrained by the lack of standardized benchmarks, diverse testing scenarios a

Wang Cai, Hsiu-Yuan Huang, Zhixiang Wang, Yunfang Wu

In-Context derived Vector (ICV) methods extract task-relevant representations from large language models (LLMs) and reinject them during inference, achieving comparable performance to few-shot In-Context Learning (ICL) without repeated demonstration processing. However, existing ICV methods remain sensitive to ICL-specific factors, often use coarse or semant

Timotheus Kampik

This chapter gives an introduction to agent-oriented programming in JavaScript. It provides an example-based walk-through of how to implement abstractions for reasoning loop agents in vanilla JavaScript. The initial example is used as a stepping stone for explaining how to implement slightly more advanced agents and multi-agent systems using JS-son, a JavaSc

Tazeek Bin Abdur Rakib, Ambuj Mehrish, Lay-Ki Soon, Wern Han Lim

Large-language-model (LLM) agents excel at reactive dialogue but struggle with proactive, goal-driven interactions due to myopic decoding and costly planning. We introduce DialogXpert, which leverages a frozen LLM to propose a small, high-quality set of candidate actions per turn and employs a compact Q-network over fixed BERT embeddings trained via temporal

Ömer Faruk Akgül, Feiyu Zhu, Yuxin Yang, Rajgopal Kannan

Temporal Knowledge Graphs (TKGs) represent dynamic facts as timestamped relations between entities. TKG completion involves forecasting missing or future links, requiring models to reason over time-evolving structure. While LLMs show promise for this task, existing approaches often overemphasize supervised fine-tuning and struggle particularly when historica

Jianxiang Zang, Meiling Ning, Yongda Wei, Shihan Dou

Recently, the concept of ``compression as intelligence'' has provided a novel informatics metric perspective for language models (LMs), emphasizing that highly structured representations signify the intelligence level of LMs. However, from a geometric standpoint, the word representation space of highly compressed LMs tends to degenerate into a highly anisotr

Can Kutlu Yüksel, Tomáš Vyhlídal, Silviu-Iulian Niculescu

The paper proposes an alternative way to achieve the Internal Model Principle (IMP) in contrast to the standard way, where a model of the signal one wishes to track/reject is directly substituted into the closed-loop. The proposed alternative approach relies on an already-existing stabilizing controller, which can be further augmented with a Youla-Ku\v{c}era

Luca Fehlings, Bojian Zhang, Paolo Gibertini, Martin A. Nicholson

Recent efforts to improve the efficiency of neuromorphic and machine learning systems have centred on developing of specialised hardware for neural networks. These systems typically feature architectures that go beyond the von Neumann model employed in general-purpose hardware such as GPUs, offering potential efficiency and performance gains. However, neural

Florian Kalinke, Shakeel Gavioli-Akilagun

This article studies the problem of online non-parametric change point detection in multivariate data streams. We approach the problem through the lens of kernel-based two-sample testing and introduce a sequential testing procedure based on random Fourier features, running with logarithmic time complexity per observation and with overall logarithmic space co

Hakan Osman Caldag, Martin Alan Bees

The combined impact of axial stretching and cross-stream diffusion on the downstream transport of solute is termed Taylor dispersion. The dispersion of active suspensions is qualitatively distinct: viscous and external torques can establish non-uniform concentration fields with weighted access to shear, modifying mean drift and effective diffusivity. It woul

Peilin Chen, Xiaoxuan Yang

Large language models (LLMs) have gained great success in various domains. Existing systems cache Key and Value within the attention block to avoid redundant computations. However, the size of key-value cache (KV cache) is unpredictable and can even be tens of times larger than the weights in the long context length scenario. In this work, we propose Titanus

Devika Dass, Dan Kilper, Liam Barry, Marco Ruffini

We experimentally investigate the transparent coexistence of heterogeneous Analog Radio-over-Fiber (ARoF) and Digital Coherent Optical (DCO) signals in a converged metro/PON network. Our streamlined setup employs RF generation via optical heterodyning, so that both carrier and modulated signals can be generated centrally and transmitted to the antenna site,

Lukas Edman, Helmut Schmid, Alexander Fraser

The CUTE benchmark showed that LLMs struggle with character understanding in English. We extend it to more languages with diverse scripts and writing systems, introducing EXECUTE. Our simplified framework allows easy expansion to any language. Tests across multiple LLMs reveal that challenges in other languages are not always on the character level as in Eng

Dekai Zhu, Stefan Gavranovic, Flavien Boussuge, Benjamin Busam

Data augmentation is widely used to train deep learning models to address data scarcity. However, traditional data augmentation (TDA) typically relies on simple geometric transformation, such as random rotation and rescaling, resulting in minimal data diversity enrichment and limited model performance improvement. State-of-the-art generative models for 3D sh

Valentin Kladov, Johan Messchendorp, James Ritman

The real-time data processing of the next generation of experiments conducted at FAIR requires a reliable reconstruction of event topologies and, therefore, will depend heavily on in-situ calibration procedures. A neural network-based approach can provide fast real-time calibrations based on continuously available environmental data. We applied this approach

Sambaran Banerjee

Young massive clusters (YMC) and open clusters (OC) are widely considered as potential environments for assembling merging binary stellar-remnant black holes (BBH) via dynamical interactions. However, such moderate mass systems are susceptible to being disrupted by the external tidal field of their host galaxies, potentially limiting their effectiveness as G

Anjie Le, Henan Liu, Yue Wang, Zhenyu Liu

Ultrasound is a widely-used imaging modality critical to global healthcare, yet its interpretation remains challenging due to its varying image quality on operators, noises, and anatomical structures. Although large vision-language models (LVLMs) have demonstrated impressive multimodal capabilities across natural and medical domains, their performance on ult

Yu Xie, Jielei Zhang, Pengyu Chen, Weihang Wang

Diffusion-based scene text synthesis has progressed rapidly, yet existing methods commonly rely on additional visual conditioning modules and require large-scale annotated data to support multilingual generation. In this work, we revisit the necessity of complex auxiliary modules and further explore an approach that simultaneously ensures glyph accuracy and

Harish G. Ramaswamy, L. A. Prashanth

We consider the problem of estimating and optimizing utility-based shortfall risk (UBSR) of a loss, say $(Y - \hat Y)^2$, in the context of a regression problem. Empirical risk minimization with a UBSR objective is challenging since UBSR is a non-linear function of the underlying distribution. We first derive a concentration bound for UBSR estimation using i

Ildi Alla, Valeria Loscri

Emerging 5G millimeter-wave and sub-6 GHz networks enable high-accuracy indoor localization, but security and privacy vulnerabilities pose serious challenges. In this paper, we identify and address threats including location spoofing and adversarial signal manipulation against 5G-based indoor localization. We formalize a threat model encompassing attackers w

Haoran Dou, Heng Yu

Context. The environment plays a crucial role in galaxy evolution, particularly for galaxies infalling into clusters. Accurately estimating the infall times of galaxies from observations can significantly enhance our understanding of the environmental effects on galaxy evolution. Aims. This paper aims to evaluate existing methods for estimating infall times

Taro Fukazawa, Haruki Okumura, Tetsuya Fukushima, Hisazumi Akai

We propose an approach for exploring magnetic structures by using Liechtenstein's method for exchange couplings from the results of first-principles calculations. Our method enables efficient and accurate exploration of stable magnetic structures by greatly reducing the number of firstprinciples calculations required. We apply our method to the magnetic stru

Abdul Basit, Maha Nawaz, Muhammad Shafique

Non-invasive brain-computer interfaces (BCIs) have the potential to enable intuitive control of prosthetic limbs for individuals with upper limb amputations. However, existing EEG-based control systems face challenges related to signal noise, classification accuracy, and real-time adaptability. In this work, we present BRAVE, a hybrid EEG and voice-controlle

Peiliang Gong, Yucheng Wang, Min Wu, Zhenghua Chen

Source-Free Domain Adaptation (SFDA) aims to adapt a pre-trained model from an annotated source domain to an unlabelled target domain without accessing the source data, thereby preserving data privacy. While existing SFDA methods have proven effective in reducing reliance on source data, they struggle to perform well on multivariate time series (MTS) due to

Amir Hossein Rahmati, Sanket Jantre, Weifeng Zhang, Yucheng Wang

Low-Rank Adaptation (LoRA) offers a cost-effective solution for fine-tuning large language models (LLMs), but it often produces overconfident predictions in data-scarce few-shot settings. To address this issue, several classical statistical learning approaches have been repurposed for scalable uncertainty-aware LoRA fine-tuning. However, these approaches neg

Nils Caci, Dominik Chudy, Pablo Daniel Mendez Mariscal, Daniel Ueltschi

Using a combined analysis from Poisson-Dirichlet and symmetry-breaking calculations as well as quantum Monte Carlo simulations, we examine the ordered phase and the thermal phase transition of the three-dimensional spin-1 quantum magnet on the simple cubic lattice with bilinear and biquadratic interactions and SU(3) internal symmetry. We obtain exact results

Yanping Fu, Xinyuan Liu, Tianyu Li, Yike Ma

Topology reasoning, which unifies perception and structured reasoning, plays a vital role in understanding intersections for autonomous driving. However, its performance heavily relies on the accuracy of lane detection, particularly at connected lane endpoints. Existing methods often suffer from lane endpoints deviation, leading to incorrect topology constru

Sourabh Biswas, Srijaya Nandi, Tuhin Subhra Pal, Aesha Lahiri

Interspecific communication plays a critical role in mediating human-animal interactions, particularly in contexts involving access to anthropogenic resources. This study investigates the influence of human gazing on the begging strategies of free-ranging dogs in urban and peri-urban environments. Begging behaviour, commonly observed in dogs seeking food fro

Patrick Leask, Neel Nanda, Noura Al Moubayed

Sparse autoencoders (SAEs) are a popular method for decomposing Large Langage Models (LLM) activations into interpretable latents. However, due to their substantial training cost, most academic research uses open-source SAEs which are only available for a restricted set of models of up to 27B parameters. SAE latents are also learned from a dataset of activat

Dong Zhang, Lingfeng He, Rui Yan, Fei Shen

While recent advances in image editing have enabled impressive visual synthesis capabilities, current methods remain constrained by explicit textual instructions and limited editing operations, lacking deep comprehension of implicit user intentions and contextual reasoning. In this work, we introduce a new image editing paradigm: reasoning-guided generative

Weiwen Liu, Jiarui Qin, Xu Huang, Xingshan Zeng

Large Language Model (LLM) agents represent a promising shift in human-AI interaction, moving beyond passive prompt-response systems to autonomous agents capable of reasoning, planning, and goal-directed action. While LLM agents are technically capable of performing a broad range of tasks, not all of these capabilities translate into meaningful usability. Th

Reda Chhaibi, Serge Gratton, Samuel Vaiter

In this paper, we propose an accelerated version for the Sinkhorn algorithm, which is the reference method for computing the solution to Entropic Optimal Transport. Its main draw-back is the exponential slow-down of convergence as the regularization weakens $\varepsilon \rightarrow 0$. Thanks to spectral insights on the behavior of the Hessian, we propose to

Xiaoyu Zhang, Tai Wang, Yi Qin Gao, Yunlong Xiao

We recently proposed a scheme to generalize collinear functionals to the noncollinear regime, termed the multicollinear approach. The resulting noncollinear functionals preserve spin symmetry while providing numerically stable higher-order functional derivatives. This scheme has already been applied to noncollinear spin-flip TDDFT and its analytic gradient c

Junhong Zhang, Zhihui Lai

Kernel methods are powerful tools for nonlinear learning with well-established theory. The scalability issue has been their long-standing challenge. Despite the existing success, there are two limitations in large-scale kernel methods: (i) The memory overhead is too high for users to afford; (ii) existing efforts mainly focus on kernel ridge regression (KRR)

Aleksandar Tomčić, Miloš Savić, Dušan Simić, Miloš Radovanović

The role of high-degree nodes, or hubs, in shaping graph dynamics and structure is well-recognized in network science, yet their influence remains underexplored in the context of dynamic graph embedding. Recent advances in representation learning for graphs have shown that random walk-based methods can capture both structural and temporal patterns, but often