May 2025 arXiv papers — page 109

Danish Furekh Dar, Stephan Fritzsche

We present a detailed comparison of theoretical approaches for modeling strong-field ionization by few-cycle laser pulses. The dipole approximation is shown to accurately capture interference structures in photoelectron spectra, while non-dipole effects introduce significant momentum shifts along the propagation direction. Two complementary analytical method

Radu Ioan Boţ, Enis Chenchene, Ernö Robert Csetnek, David Alexander Hulett

We analyze fast diagonal methods for simple bilevel programs. Guided by the analysis of the corresponding continuous-time dynamics, we provide a unified convergence analysis under general geometric conditions, including H\"olderian growth and the Attouch-Czarnecki condition. Our results yield explicit convergence rates and guarantee weak convergence to a sol

Prasanna Parasurama, Panos Ipeirotis

Algorithmic tools are increasingly used in hiring to improve fairness and diversity, often by enforcing constraints such as gender-balanced candidate shortlists. However, we show theoretically and empirically that enforcing equal representation at the shortlist stage does not necessarily translate into more diverse final hires, even when there is no gender b

Tye Lidman, Lisa Piccirillo

One approach to produce a pair of homeomorphic-but-not-diffeomophic closed 4-manifolds is to find a knot which is smoothly slice in one but not the other. This approach has never been run successfully. We give the first examples of a pair of closed 4-manifolds with the same integer cohomology ring where the diffeomorphism type is distinguished by this approa

Dániel Gerbner, Shujing Miao

For a family of graphs $\cF$, a graph is called $\cF$-free if it does not contain any member of $\cF$ as a subgraph. Given a collection of graphs $(G_1,\ldots,G_t)$ on the same vertex set $V$ of size $n$, a rainbow graph on $V$ is obtained by taking at most one edge from each $G_i$. We say that a collection is rainbow $\cF$-free if it contains no rainbow cop

Alexander Wendt, Matthew J. Storey, Michael Miller, Dalton Anderson

Surface acoustic waves (SAWs) enable a wide array of technologies including RF filters, chemical and biological sensors, acousto-optic devices, acoustic control of microfluidic flow in lab-on-a-chip systems, and quantum phononics. While numerous methods exist for generating SAWs, they each have intrinsic limitations that inhibit performance, operation at hig

Quan Xie, Jiajia Liu, Robert Erdélyi, Yuming Wang

Vortices have been observed at various heights within the solar atmosphere and are suggested to potentially play great roles in heating the solar upper atmosphere. Multiple automated vortex detection methods have been developed and applied to detect vortices. We aim to improve the $\Gamma$-functions method for vortex identification by optimizing the value of

Jian-Wei Liu, Gui-Geng Liu, Bo Zhang, Hao-Chang Mo

Valleytronics has emerged as a promising paradigm, enabling comprehensive control of the valley degree of freedom (DoF) for energy-efficient and high-speed information processing. However, backscattering-induced valley depolarization remains a fundamental limitation, stemming from the weak topological protection of the valley Hall phase. Here, we propose and

Vinícius Pereira da S. Oliveira, Danilo S. Borges, Erick M. Franklin, Jorge Peixinho

Unstable systems of fluidized grains in a very-narrow vertical tube can auto-defluidize after some time, the settling particles forming either a glass- or crystal-like structure. We carried out experiments using different polymer spheres, of known friction and roughness, fluidized in water. A diagram was obtained for the \RefereeA{shell-settled} particles wh

Nobuhiro Ueda, Yuyang Dong, Krisztián Boros, Daiki Ito

With the increasing adoption of Large Language Models (LLMs) and Vision-Language Models (VLMs), rich document analysis technologies for applications like Retrieval-Augmented Generation (RAG) and visual RAG are gaining significant attention. Recent research indicates that using VLMs yields better RAG performance, but processing rich documents remains a challe

Sita Vriend, David Hägele, Daniel Weiskopf

There exists limited theoretical guidance on integrating visualization and sonification. In this paper, we address this gap by investigating audiovisual semiotics for uncertainty representation: joining uncertainty visualization and sonification to combine audiovisual channels for enhancing users' perception of uncertainty. We conducted two preregistered cro

Alexeï Molin, Simon Dupourqué, Nicolas Clerc, Étienne Pointecouteau

Context. The NewAthena mission planned for launch in the late 2030s will carry X-IFU, an integral field unit spectrometer that will obtain unique insight into the X-ray hot universe through its combination of spectral and spatial capabilities. Its high spectral resolution will allow a mapping of turbulent velocities of the hot gas in galaxy clusters, providi

Ulrike Kuhl, Annika Bush

Although the integration of artificial intelligence (AI) into everyday tasks improves efficiency and objectivity, it also risks transmitting bias to human decision-making. In this study, we conducted a controlled experiment that simulated hiring decisions to examine how biased AI recommendations - augmented with or without counterfactual explanations - influ

Maitreya Prafulla Chitale, Ketaki Mangesh Shetye, Harshit Gupta, Manav Chaudhary

Obtaining high-quality, pre-submission feedback is a critical bottleneck in the academic publication lifecycle for researchers. We introduce AutoRev, an automated author-centric feedback system that generates structured, actionable guidance prior to formal peer review. AutoRev employs a graph-based retrieval-augmented generation framework that models each pa

Mauro Ballicchia, Clemens Etl, Mihail Nedjalkov, David K. Ferry

The electric interaction between two nearby evolving electrons triggers the correlation between their waves and governs the operation of logical devices called Coulomb entanglers. Of technological interest in the presence of magnetic fields are multi-spatial evolution scenarios beyond pure state descriptions. The two-electron density matrix becomes eight-dim

Ziyue Liu, Meredith L. Carr, Norberto C. Nadal-Caraballo, Madison C. Yawn

Bayesian networks (BN) have advantages in visualizing causal relationships and performing probabilistic inference analysis, making them ideal tools for coastal hazard analysis and characterizing the compound mechanisms of coastal hazards. Meanwhile, the Joint Probability Method (JPM) has served as the primary probabilistic assessment approach used to develop

Gianluca Rizzi, Angela Madeo

Microstructured materials, such as architected metamaterials and phononic crystals, exhibit complex wave propagation phenomena due to their internal structure. While full-scale numerical simulations can capture these effects, they are computationally demanding, especially in time-domain analyses. To overcome this limitation, effective continuum models have b

Colin Gilgenbach, Menglin Zhu, James M. LeBeau

We present \code{phaser}, an open-source Python package that provides a unified interface to both conventional and gradient descent-based ptychographic algorithms. Features such as mixed-state probe, probe position correction, and multislice ptychography make experimental reconstructions practical and robust. Reconstructions are specified in a declarative fo

Anh Duc Nguyen, Ilia Markov, Frank Zhengqing Wu, Ali Ramezani-Kebrya

Modern deep neural networks exhibit heterogeneity across numerous layers of various types such as residuals, multi-head attention, etc., due to varying structures (dimensions, activation functions, etc.), distinct representation characteristics, which impact predictions. We develop a general layer-wise quantization framework with tight variance and code-leng

Ava Elizabeth Scott, Lev Tankelevitch, Payod Panda, Rishi Vanukuru

Despite decades of HCI and Meeting Science research, complaints about ineffective meetings are still pervasive. We argue that meeting technologies lack support for prospective reflection, that is, thinking about why a meeting is needed and what might happen. To explore this, we designed a Meeting Purpose Assistant (MPA) technology probe to coach users to art

Amruta Lambe, Shambhu Nath Sharma

The Koopmanization embeds the bilinearization via the action of the infinitesimal stochastic Koopman operator on the observables associated with the controlled nonlinear It\^o stochastic differential system without explicit linearizations. The stochastic evolutions of controlled Markov processes assume the structure of controlled nonlinear It\^o stochastic d

Jialong Han, Si Zhang, Ke Zhang

Parameter-Efficient Fine-Tuning (PEFT) has emerged as a critical paradigm for adapting Large Language Models (LLMs) to downstream tasks, among which Low-rank Adaptation (LoRA) represents one of the most widely adopted methodologies. However, existing LoRA-based approaches exhibit two fundamental limitations: unstable training dynamics and inefficient knowled

L. Bandiera, R. Cai, S. Carsi, S. Cesare

This study investigates the performance of bent silicon crystals intended to channel hadrons in a fixed-target experiment at the Large Hadron Collider (LHC). The phenomenon of planar channelling in bent crystals enables extremely high effective bending fields for positively charged hadrons within compact volumes. Particles trapped in the potential well of hi

J. Schroeder, S. Howard, C. Eberle, J. Esslinger

All measurements of continuous signals rely on taking discrete snapshots, with the Nyquist-Shannon theorem dictating sampling paradigms. We present a broader framework of information-optimal measurement, showing that traditional sampling is optimal only when we are entirely ignorant about the system under investigation. This insight unlocks methods that effi

Mengyao Guo, Jinda Han, Ze Gao, Yuan Zhuang

This paper explores the evolving landscape of human-machine co-creation, focusing on its development in the context of the ACM Conference on Human Factors in Computing Systems (CHI) from 2014 to 2024. We employ co-word analysis to identify emerging trends, central themes, and the intellectual trajectory of this field. The study highlights the shift from view

Ziwei Zheng, Michael Yang, Jack Hong, Chenxiao Zhao

Large Vision-Language Models excel at multimodal understanding but struggle to deeply integrate visual information into their predominantly text-based reasoning processes, a key challenge in mirroring human cognition. To address this, we introduce DeepEyes, a model that learns to "think with images", trained end-to-end with reinforcement learning without req

Xingxing Weng, Chao Pang, Gui-Song Xia

Vision-language modeling (VLM) aims to bridge the information gap between images and natural language. Under the new paradigm of first pre-training on massive image-text pairs and then fine-tuning on task-specific data, VLM in the remote sensing domain has made significant progress. The resulting models benefit from the absorption of extensive general knowle

Antoine Castagnède, Laura Filion, Frank Smallenburg

In experimental systems, colloidal particles are virtually always at least somewhat polydisperse, which can have profound effects on their ability to crystallize. Unfortunately, accurately predicting the effects of polydispersity on phase behavior using computer simulations remains a challenging task. As a result, our understanding of the equilibrium phase b

Ruoxin Chen, Junwei Xi, Zhiyuan Yan, Ke-Yue Zhang

Existing detectors are often trained on biased datasets, leading to the possibility of overfitting on non-causal image attributes that are spuriously correlated with real/synthetic labels. While these biased features enhance performance on the training data, they result in substantial performance degradation when applied to unbiased datasets. One common solu

Bhavana Vannarth Shobhana, Yen-lin Chien, Jonathan Diamant, Badri Nath

Latency is a key indicator of Internet service performance. Continuously tracking the latency of client requests enables service operators to quickly identify bottlenecks, perform adaptive resource allocation or routing, and mitigate attacks. Passively measuring the response latency at intermediate vantage points is attractive since it provides insight into

Siqiao Huang, Jialong Wu, Qixing Zhou, Shangchen Miao

World models, which predict future transitions from past observation and action sequences, have shown great promise for improving data efficiency in sequential decision-making. However, existing world models often require extensive domain-specific training and still produce low-fidelity, coarse predictions, limiting their usefulness in complex environments.

Sho Inoue, Shai Wang, Haizhou Li

Despite significant progress in neural spoken dialog systems, personality-aware conversation agents -- capable of adapting behavior based on personalities -- remain underexplored due to the absence of personality annotations in speech datasets. We propose a pipeline that preprocesses raw audio recordings to create a dialogue dataset annotated with timestamps

Nicolas Brunel, Vincent Hakim, Jean-Pierre Nadal

The perceptron has served as a prototypical neuronal learning machine in the physics community interested in neural networks and artificial intelligence, which included G\'erard Toulouse as one of its prominent figures. It has also been used as a model of Purkinje cells of the cerebellum, a brain structure involved in motor learning, in the early influential

Xin Li, Mengbing Liu, Li Wei, Jiancheng An

Large Language Models (LLMs) have achieved impressive results across a broad array of tasks, yet their capacity for complex, domain-specific mathematical reasoning-particularly in wireless communications-remains underexplored. In this work, we introduce WirelessMathBench, a novel benchmark specifically designed to evaluate LLMs on mathematical modeling chall

Souvik Dey, Jian Liu, Liran Shaul

Given a compactly generated triangulated category $\mathcal{T}$ equipped with an action of a graded-commutative Noetherian ring $R$, generalizing results of Letz, we prove a general result concerning the openness with respect to levels of compact objects in $\mathcal{T}$. Applications are given to derived categories of commutative Noetherian rings, derived c

Bartosz Cywiński, Emil Ryd, Senthooran Rajamanoharan, Neel Nanda

As language models become more powerful and sophisticated, it is crucial that they remain trustworthy and reliable. There is concerning preliminary evidence that models may attempt to deceive or keep secrets from their operators. To explore the ability of current techniques to elicit such hidden knowledge, we train a Taboo model: a language model that descri

Yutong Liu, Ziyue Zhang, Ban Ma-bao, Yuqing Cai

Tibetan is a low-resource language with minimal parallel speech corpora spanning its three major dialects-\"U-Tsang, Amdo, and Kham-limiting progress in speech modeling. To address this issue, we propose FMSD-TTS, a few-shot, multi-speaker, multi-dialect text-to-speech framework that synthesizes parallel dialectal speech from limited reference audio and expl

Jialong Han, Si Zhang, Ke Zhang

Fine-tuning Large Language Models (LLMs) has become increasingly challenging due to their massive scale and associated computational costs. Parameter-Efficient Fine-Tuning (PEFT) methodologies have been proposed as computational alternatives; however, their implementations still require significant resources. In this paper, we present OSoRA (Output-Dimension

Evangelos Pournaras, Srijoni Majumdar, Thomas Wellings, Joshua C. Yang

Voting methods are instrumental design elements of democracies. Citizens use them to express and aggregate their preferences to reach a collective decision. However, voting outcomes can be as sensitive to voting rules as they are to people's voting choices. Despite significance and interdisciplinary scientific progress, several democracies keep relying on ou

Denis Mazzucato, Abdalrhman Mohamed, Juneyoung Lee, Clark Barrett

Many security- and performance-critical domains, such as cryptography, rely on low-level verification to minimize the trusted computing surface and allow code to be written directly in assembly. However, verifying assembly code against a realistic machine model is a challenging task. Furthermore, certain security properties -- such as constant-time behavior

Neelabh Sinha

Language Models (LMs) have revolutionized natural language processing, enabling high-quality text generation through prompting and in-context learning. However, models often struggle with long-context summarization due to positional biases, leading to suboptimal extraction of critical information. There are techniques to improve this with fine-tuning, pipeli

Mingfang Zhang, Ryo Yonetani, Yifei Huang, Liangyang Ouyang

This paper presents a novel inertial localization framework named Egocentric Action-aware Inertial Localization (EAIL), which leverages egocentric action cues from head-mounted IMU signals to localize the target individual within a 3D point cloud. Human inertial localization is challenging due to IMU sensor noise that causes trajectory drift over time. The d

Aydin Abedinia, Shima Tabakhi, Vahid Seydi

Recent advancements in semi-supervised deep learning have introduced effective strategies for leveraging both labeled and unlabeled data to improve classification performance. This work proposes a semi-supervised framework that utilizes a distance-based weighting mechanism to prioritize critical training samples based on their proximity to test data. By focu

K. Dolgikh, A. Korochkin, G. Rubtsov, D. Semikoz

The standard assumption about the influence of the turbulent intergalactic magnetic field (IGMF) on the images of ultra-high-energy cosmic rays (UHECR) sources is that the latter are formed in a random walk mode in the deflection angle. As a result, the images are symmetrically broadened to angular scales proportional to the IGMF strength and the square root

Clàudia Soriano-Guerrero, Daniele Viganò, Rosalba Perna, Albert Elias-López

In Hot Jupiters (HJs), atmospherically induced magnetic fields are expected to play an important role in controlling the wind circulation and in determining their inflated radii. Here we perform 1D plane-parallel magnetohydrodynamic (MHD) simulations of HJ atmospheric columns, using the wind and thermodynamic profiles generated by global circulation models o

Sifan Li, Ming Tao, Hao Zhao, Ling Shao

Text-to-Image (T2I) has been prevalent in recent years, with most common condition tasks having been optimized nicely. Besides, counterfactual Text-to-Image is obstructing us from a more versatile AIGC experience. For those scenes that are impossible to happen in real world and anti-physics, we should spare no efforts in increasing the factual feel, which me

Seunghyuk Cho, Zhenyue Qin, Yang Liu, Youngbin Choi

Plane geometry problem solving (PGPS) has recently gained significant attention as a benchmark to assess the multi-modal reasoning capabilities of large vision-language models. Despite the growing interest in PGPS, the research community still lacks a comprehensive overview that systematically synthesizes recent work in PGPS. To fill this gap, we present a s

Vedanshi Chetan Shah, Ab Mosca

With the introduction of the Visualization for Communication workshop (VisComm) at IEEE VIS and in light of the COVID-19 pandemic, there has been renewed interest in studying visualization as a medium of communication. However the characteristics and definition of this line of study tend to vary from paper to paper and person to person. In this work, we exam

Egor Bakaev, Florestan Brunck, Christoph Hertrich, Jack Stade

This work studies the expressivity of ReLU neural networks with a focus on their depth. A sequence of previous works showed that $\lceil \log_2(n+1) \rceil$ hidden layers are sufficient to compute all continuous piecewise linear (CPWL) functions on $\mathbb{R}^n$. Hertrich, Basu, Di Summa, and Skutella (NeurIPS'21 / SIDMA'23) conjectured that this result is

Seung Hun Lee, Wonse Jo, Lionel P. Robert, Dawn M. Tilbury

Path planning is crucial for the navigation of autonomous vehicles, yet these vehicles face challenges in complex and real-world environments. Although a global view may be provided, it is often outdated, necessitating the reliance of Unmanned Ground Vehicles (UGVs) on real-time local information. This reliance on partial information, without considering the

Umberto Cappellazzo, Minsu Kim, Stavros Petridis, Daniele Falavigna

Audio-Visual Speech Recognition (AVSR) enhances robustness in noisy environments by integrating visual cues. While recent advances integrate Large Language Models (LLMs) into AVSR, their high computational cost hinders deployment in resource-constrained settings. To address this, we propose Llama-SMoP, an efficient Multimodal LLM that employs a Sparse Mixtur

Mikael Chala, Luis Gil, Zhe Ren

We perform the first computation of phase-transition parameters to cubic order in $\lambda\sim m^2/T^2$, where $m$ is the scalar mass and $T$ is the temperature, in a simple model resembling the Higgs sector of the SMEFT. We use dimensional reduction, including 1-loop matching corrections for terms of dimension 6 (in 4-dimensional units), 2-loop contribution

Christiaan T. van Campenhout, Romane Le Dizès Castell, Tess Heeremans, Sander Woutersen

Amorphous and glassy materials are important for many advanced applications, from flexible solar cells to drug delivery systems. To this end, new glasses are in high demand, but precise chemical design of amorphous materials remains challenging. By studying the crystallization of mixed salt solutions, we have discovered an entirely new type of amorphous mate

Inder Pal Singh, Enjie Ghorbel, Anis Kacem, Djamila Aouada

This paper introduces a discriminator-free adversarial-based approach termed DDA-MLIC for Unsupervised Domain Adaptation (UDA) in the context of Multi-Label Image Classification (MLIC). While recent efforts have explored adversarial-based UDA methods for MLIC, they typically include an additional discriminator subnet. Nevertheless, decoupling the classificat

Maria Petropavlova

We present new constraints on neutrino secret interactions ($\nu$SI) by studying high-energy neutrinos from well-known astrophysical sources, such as SN1987A, the blazars TXS $0506+056$ and PKS $0735+178$, the active galaxy NGC 1068, and the KM3-230213A neutrino event. We expand existing limits by probing a previously unconstrained region of the mediator mas

Debarshi Basu, Ashish Chandra, Qiang Wen

These notes present a comprehensive analysis of shockwave geometries in holographic settings, focusing on $\textrm{T}\overline{\textrm{T}}$-deformed BTZ black holes and their extensions. By constructing deformed metrics and employing Kruskal coordinates, we examine out-of-time-ordered correlators (OTOCs) as probes of quantum chaos. We also study localized sh

Rajat Kanti Bhattacharjee, Meghali Nandi, Amrit Jha, Gunajit Kalita

This paper proposes deep learning techniques of generating designs for clothing, focused on handloom fabric and discusses the associated challenges along with its application. The capability of generative neural network models in understanding artistic designs and synthesizing those is not yet explored well. In this work, multiple methods are employed incorp

Xiang Li, Xianfu Cheng, Dezhuang Miao, Xiaoming Zhang

Multimodal Sentiment Analysis (MSA) with missing modalities has attracted increasing attention recently. While current Transformer-based methods leverage dense text information to maintain model robustness, their quadratic complexity hinders efficient long-range modeling and multimodal fusion. To this end, we propose a novel and efficient Text-enhanced Fusio

Maheak Dave, Aniket Kumar Singh, Aryan Pareek, Harshita Jha

Deep Neural Networks have achieved remarkable achievements across various domains, however balancing performance and generalization still remains a challenge while training these networks. In this paper, we propose a novel framework that uses a cyclic optimization strategy to concurrently optimize the model and its input data for better training, rethinking

Sebastian Barzaghi, Simona Colitti, Arianna Moretti, Giulia Renda

This paper introduces a pipeline for integrating semantic metadata, 3D models, and storytelling, enhancing cultural heritage digitization. Using the Aldrovandi Digital Twin case study, it outlines a reusable workflow combining RDF-driven narratives and data visualization for creating interactive experiences to facilitate access to cultural heritage.

Virgile Guemard

We present a topological approach to lifting a quantum CSS code. In previous work, we proposed lifting a CSS code by constructing covering spaces over its 2D simplicial complex representation, known as the Tanner cone-complex. This idea was inspired by the work of Freedman and Hastings, which associates CSS codes with handlebodies. In this paper, we show how

Simona Bisiani, Agnes Gulyas, John Wihbey, Bahareh Heravi

In this paper, we present UKTwitNewsCor, a comprehensive dataset for understanding the content production, dissemination, and audience engagement dynamics of online local media in the UK. It comprises over 2.5 million online news articles published between January 2020 and December 2022 from 360 local outlets. The corpus represents all articles shared on Twi

Roosa Risto, Mohit Sethi, Mika Katara

The Cyber Resilience Act (CRA) is a new European Union (EU) regulation aimed at enhancing the security of digital products and services by ensuring they meet stringent cybersecurity requirements. This paper investigates the challenges that industrial equipment manufacturing companies anticipate while preparing for compliance with CRA through a comprehensive

Saulo Diles, Miguel Angel Martin Contreras, Alfredo Vega

Holographic models of QCD provide the spectrum of heavy vector meson masses and electromagnetic decay constants through bulk computations of the current-current correlation function. Conversely, the phenomenology of heavy vector mesons is articulated by the constituent heavy quark model utilizing a non-relativistic approximation. By applying the Segre formul

Robert Allison, Tomasz Maciążek, Henry Bourne

The growing body of literature on training-data reconstruction attacks raises significant concerns about deploying neural network classifiers trained on sensitive data. However, differentially private (DP) training (e.g. using DP-SGD) can defend against such attacks with large training datasets causing only minimal loss of network utility. Folklore, heuristi

Jan De Beule, Philipp Heering, Sam Mattheus, Klaus Metsch

The investigation into large families of non-opposite flags in finite spherical buildings has been a recent addition to a long line of research in extremal combinatorics, extending classical results in vector and polar spaces. This line of research falls under the umbrella of Erdős-Ko-Rado (EKR) problems, but poses some extra difficulty on the algebraic leve

Bo Feng, Zhengfeng Lai, Shiyu Li, Zizhen Wang

Existing video understanding benchmarks often conflate knowledge-based and purely image-based questions, rather than clearly isolating a model's temporal reasoning ability, which is the key aspect that distinguishes video understanding from other modalities. We identify two major limitations that obscure whether higher scores truly indicate stronger understa

Maria Cuellar, Hon Kiu, To, Arush Mehrotra

Facial recognition technology (FRT) is increasingly used in criminal investigations, yet most evaluations of its accuracy rely on high-quality images, unlike those often encountered by law enforcement. This study examines how five common forms of image degradation--contrast, brightness, motion blur, pose shift, and resolution--affect FRT accuracy and fairnes

Weihao Xia, Chenliang Zhou, Cengiz Oztireli

Tactile perception is profoundly influenced by the surface properties of objects in contact. However, despite their crucial role in shaping tactile experiences, these material characteristics have been largely neglected in existing tactile representation learning methods. Most approaches primarily focus on aligning tactile data with visual or textual informa

Wenjun Hou, Yi Cheng, Kaishuai Xu, Heng Li

Large language models (LLMs) have demonstrated remarkable capabilities in various domains, including radiology report generation. Previous approaches have attempted to utilize multimodal LLMs for this task, enhancing their performance through the integration of domain-specific knowledge retrieval. However, these approaches often overlook the knowledge alread

E. Fontanet, S. Udry, D. Ségransan, P. Figueira

With around 200 detections of exoplanets around giant stars to date, our knowledge of the population of exoplanets orbiting evolved hosts more massive than the Sun remains limited. The CORALIE radial-velocity search for companions around evolved stars (CASCADES) was launched in 2006 with the aim of improving our understanding of the demographics of exoplanet

Tiehan Cui, Yanxu Mao, Peipei Liu, Congying Liu

Although large language models (LLMs) have achieved remarkable advancements, their security remains a pressing concern. One major threat is jailbreak attacks, where adversarial prompts bypass model safeguards to generate harmful or objectionable content. Researchers study jailbreak attacks to understand security and robustness of LLMs. However, existing jail

Rafael Corsi Ferrao, Igor dos Santos Montagner, Rodolfo Azevedo

This paper investigates code quality education by analyzing how errors are introduced and corrected in group projects within an embedded systems course. We identify who introduces errors, who fixes them, and when these actions occur. Students learn code quality rules for C and embedded systems. We address three questions: RQ1: What is the impact of group for

Kosmas Alexandridis, Vasileios Titopoulos, Giorgos Dimitrakopoulos

Attention mechanisms, particularly within Transformer architectures and large language models (LLMs), have revolutionized sequence modeling in machine learning and artificial intelligence applications. To compute attention for increasingly long sequences, specialized accelerators have been proposed to execute key attention steps directly in hardware. Among t

Leonardo Bertolazzi, Manuel Vargas Guzmán, Raffaella Bernardi, Maciej Malicki

Large language models (LLMs) are increasingly evaluated on reasoning tasks, yet their logical abilities remain contested. To address this, we study LLMs' reasoning in a well-defined fragment of logic: syllogistic reasoning. We cast the problem as premise selection and construct controlled datasets to isolate logical competence. Beyond evaluation, an open cha

Kyungeun Lee, Moonjung Eo, Hye-Seung Cho, Dongmin Kim

Despite the widespread use of tabular data in real-world applications, most benchmarks rely on average-case metrics, which fail to reveal how model behavior varies across diverse data regimes. To address this, we propose MultiTab, a benchmark suite and evaluation framework for multi-dimensional, data-aware analysis of tabular learning algorithms. Rather than

Shamsuddeen Hassan Muhammad, Ibrahim Said Ahmad, Idris Abdulmumin, Falalu Ibrahim Lawan

Hausa Natural Language Processing (NLP) has gained increasing attention in recent years, yet remains understudied as a low-resource language despite having over 120 million first-language (L1) and 80 million second-language (L2) speakers worldwide. While significant advances have been made in high-resource languages, Hausa NLP faces persistent challenges, in

Shiyin Tan, Dongyuan Li, Renhe Jiang, Zhen Wang

Popularity bias occurs when popular items are recommended far more frequently than they should be, negatively impacting both user experience and recommendation accuracy. Existing debiasing methods mitigate popularity bias often uniformly across all users and only partially consider the time evolution of users or items. However, users have different levels of

Ehsan Doostmohammadi, Marco Kuhlmann

Retrieval-augmented language models have demonstrated performance comparable to much larger models while requiring fewer computational resources. The effectiveness of these models crucially depends on the overlap between query and retrieved context, but the optimal degree of this overlap remains unexplored. In this paper, we systematically investigate how va

Lin Wang, Guido Burkard

We study the spin relaxation in a single-electron bilayer graphene quantum dot due to the spin-orbit coupling. The spin relaxation is assisted by the emission of acoustic phonons via the bond-length change and deformation potential mechanisms and $1/f$ charge noise. In the perpendicular magnetic-field dependence of the spin relaxation rate $T_1^{-1}$, we pre

Mengqiu Zhou, Meng Zhang, Howard H. Yang, Roy D. Yates

The proliferation of mobile devices and real-time status updating applications has motivated the optimization of data freshness in the context of age of information (AoI). Meanwhile, increasing computational demands have inspired research on CPU scheduling. Since prior CPU scheduling strategies have ignored data freshness and prior age-minimization strategie

Yue Fei, Jingjing Liu, Yuyou Yao, Yusheng Peng

CVT (Centroidal Voronoi Tessellation)-based remeshing optimizes mesh quality by leveraging the Voronoi-Delaunay framework to optimize vertex distribution and produce uniformly distributed vertices with regular triangles. Current CVT-based approaches can be classified into two categories: (1) exact methods (e.g., Geodesic CVT, Restricted Voronoi Diagrams) tha

Jinwang Song, Hongying Zan, Kunli Zhang, Lingling Mu

Text-to-SQL, which maps natural language to SQL queries, has benefited greatly from recent advances in Large Language Models (LLMs). While LLMs offer various paradigms for this task, including prompting and supervised fine-tuning (SFT), SFT approaches still face challenges such as complex multi-stage pipelines and poor robustness to noisy schema information.

Mengzhao Chen, Chaoyi Zhang, Jing Liu, Yutao Zeng

Large language models (LLMs) demand substantial computational and memory resources, creating deployment challenges. Quantization-aware training (QAT) addresses these challenges by reducing model precision while maintaining performance. However, the scaling behavior of QAT, especially at 4-bit precision (W4A4), is not well understood. Existing QAT scaling law

Thilagaraj Ravi, Rajnandan Choudhury Das, Heramb Vivek Bhusane, Samrat Roy

We report the direct loading of Yb atoms in the magneto-optical trap (MOT) using the intercombination narrow optical transition 6s$^2$ $^1$S$_0$ $\rightarrow$ 6s6p $^3$P$_1$ at 556 nm (green), known as green MOT with limited power of green laser, 10 mW. Direct loading of the green MOT is achieved by superimposing the green laser beam, inside a hollow core of

Zihao Feng, Xiaoxue Wang, Bowen Wu, Weihong Zhong

Task-oriented dialogue systems based on Large Language Models (LLMs) have gained increasing attention across various industries and achieved significant results. Current approaches condense complex procedural workflows into a single agent to achieve satisfactory performance on large-scale LLMs. However, these approaches face challenges to achieve comparable

Fulong Yao, Wenju Zhou, Huosheng Hu

Visually impaired individuals rely heavily on accurate and timely information about obstacles and their surrounding environments to achieve independent living. In recent years, significant progress has been made in the development of assistive technologies, particularly vision-based systems, that enhance mobility and facilitate interaction with the external

Jungseob Lee, Seongtae Hong, Hyeonseok Moon, Heuiseok Lim

Adapting large language models to other languages typically employs supervised fine-tuning (SFT) as a standard approach. However, it often suffers from an overemphasis on English performance, a phenomenon that is especially pronounced in data-constrained environments. To overcome these challenges, we propose \textbf{Cross-Lingual Optimization (CLO)} that eff

Abdul-Kazeem Shamba

This paper explores the use of contrastive learning and generative adversarial networks for generating realistic underwater images from synthetic images with uniform lighting. We investigate the performance of image translation models for generating realistic underwater images using the VAROS dataset. Two key evaluation metrics, Fr\'echet Inception Distance

Orlane Zang, Grégoire Barrué, Tony Quertier

Data encoding plays a fundamental and distinctive role in Quantum Machine Learning (QML). While classical approaches process data directly as vectors, QML may require transforming classical data into quantum states through encoding circuits, known as quantum feature maps or quantum embeddings. This step leverages the inherently high-dimensional and non-linea

Filip Vaverka, Ondrej Vysocky, Lubomir Riha

We present a lightweight tool for the analysis and tuning of application data placement in systems with heterogeneous memory pools. The tool allows non-intrusively identifying, analyzing, and controlling the placement of individual allocations of the application. We use the tool to analyze a set of benchmarks running on the Intel Sapphire Rapids platform wit

Fuga Komori, Akio K. Inoue, Ken Mawatari, Yuma Sugahara

The H~{\sc i} gas distribution in damped Lyman $\alpha$ absorbers (DLAs) has remained elusive due to the point-source nature of background quasar emission. Observing DLAs against spatially extended background galaxies provides a new method for constraining their size and structure. Using the Keck Cosmic Web Imager, we present the first ``silhouette'' image o

Yu. M. Andreev, A. Antonov, M. A. Ayala Torres, D. Banerjee

Since its approval in 2016, NA64 has pioneered light dark matter (LDM) searches with electron, positron, muon, and hadron beams. The experiment has successfully met its primary objectives, as outlined in the EPPS input (2018), and even exceeded them, producing results that demonstrate its ability to operate in a near-background-free environment. The Physics

Ali Taheri, Vahideh Vahidifar

We present new gradient estimates and Harnack inequalities for positive solutions to nonlinear slow diffusion equations. The framework is that of a smooth metric measure space $(\mathscr M,g,d\mu)$ with invariant weighted measure $d\mu=e^{-\phi} dv_g$ and diffusion operator $\Delta_\phi=e^\phi {\rm div} (e^{-\phi} \nabla)$ -- the $\phi$-Laplacian. The nonlin

Abderrahman Skiredj, Ferdaous Azhari, Houdaifa Atou, Nouamane Tazi

Open-source large language models (LLMs) still marginalise Moroccan Arabic (Darija), forcing practitioners either to bolt on heavyweight Arabic adapters or to sacrifice the very reasoning skills that make LLMs useful. We show that a rigorously quality-over-quantity alignment strategy can surface fluent Darija while safeguarding the backbone s cross-lingual r

Alexandre Popier, Laurent Denis, Dorian Cacitti-Holland

We prove that the solution of the backward stochastic differential equation with terminal singularity has a Malliavin derivative, which is the limit of the derivative of the approximating sequence. We also provide the asymptotic behavior of this derivative close to the terminal time. We apply this result to the regularity of the related partial differential

Rao Ma, Mengjie Qian, Vyas Raina, Mark Gales

The combination of pre-trained speech encoders with large language models has enabled the development of speech LLMs that can handle a wide range of spoken language processing tasks. While these models are powerful and flexible, this very flexibility may make them more vulnerable to adversarial attacks. To examine the extent of this problem, in this work we

Eirini Panteli, Paulo E. Santos, Nabil Humphrey

This paper presents AquaSignal, a modular and scalable pipeline for preprocessing, denoising, classification, and novelty detection of underwater acoustic signals. Designed to operate effectively in noisy and dynamic marine environments, AquaSignal integrates state-of-the-art deep learning architectures to enhance the reliability and accuracy of acoustic sig

Sergei M. Butorin

The degrees of the localization and hybridization of the valence $f$ states/covalency of the chemical bonding in actinide and lanthanide dioxides were investigated using the atomic, crystal-field multiplet and Anderson impurity model (AIM) approaches to calculate actinide $5d$ and lanthanide $3d$ x-ray photoemission spectra (XPS). The actinide $5d$ XPS can b

Hongyu Zhao, Yangyang Xu, Chenglin Liao

A specific failure mode designated as transient micro-short circuit (TMSC) has been identified in practical battery systems, exhibiting subtle and latent characteristics with measurable voltage deviations. To further improve the safe use of lithium-ion batteries (LIBs), this letter introduces a novel method for the precise detection of this TMSC faults withi