May 2025 arXiv papers — page 56

Thrassos K. Oikonomou, Sotiris A. Tegos, Panagiotis D. Diamantoulakis, Yuanwei Liu

Pinching-antenna (PA) systems route millimeter wave (mmWave) signals through a leaky waveguide and radiate them at "pinch" apertures, offering low-cost line-of-sight (LoS) coverage. However, when multiple PAs serve multiple users simultaneously, the downlink channel becomes strongly frequency-selective, creating inter-symbol interference (ISI) that existing

Peng Liu, Xiaoming Ren, Fengkai Liu, Qingsong Xie

Recent advancements in image-to-video (I2V) generation have shown promising performance in conventional scenarios. However, these methods still encounter significant challenges when dealing with complex scenes that require a deep understanding of nuanced motion and intricate object-action relationships. To address these challenges, we present Dynamic-I2V, an

Makoto Suda, Edan Lerner, Eran Bouchbinder

Glassy systems reveal a wide variety of generic behaviors, which lack a unified theoretical description. Here, we study a mean-field model, recently shown to reproduce the universal non-phononic vibrational spectra of glasses, under oscillatory driving forces. The driven mean-field model, featuring a disordered Hamiltonian structure, naturally predicts the s

Ugo Bruzzo, Daniel Hernández Ruipérez

In view of applications to the construction of moduli spaces of objects in algebraic supergeometry, we start a systematic study of stacks in that context. After defining a superstack as a stack over the \'etale site of superschemes, we define quotient superstacks, and, based on previous literature, we see that, in analogy with superschemes, every superstack

Christian Scharrer, Alexander West

We establish an energy quantization for constrained Willmore surfaces, where the constraints are given by area, volume, and total mean curvature, assuming that the underlying conformal structures remain bounded. Furthermore, we show strong compactness of constrained Willmore surfaces under some energy threshold, proving in particular the strong compactness o

Qiushi Sun, Zhoumianze Liu, Chang Ma, Zichen Ding

Large Language Models (LLMs) have extended their impact beyond Natural Language Processing, substantially fostering the development of interdisciplinary research. Recently, various LLM-based agents have been developed to assist scientific discovery progress across multiple aspects and domains. Among these, computer-using agents, capable of interacting with o

Yuan Li, Qi Luo, Xiaonan Li, Bufan Li

Retrieval-Augmented Generation (RAG) integrates external knowledge with Large Language Models (LLMs) to enhance factual correctness and mitigate hallucination. However, dense retrievers often become the bottleneck of RAG systems due to their limited parameters compared to LLMs and their inability to perform step-by-step reasoning. While prompt-based iterativ

Alejandro Carrasco, Victor Rodriguez-Fernandez, Richard Linares

Recent trends are emerging in the use of Large Language Models (LLMs) as autonomous agents that take actions based on the content of the user text prompts. We intend to apply these concepts to the field of Control in space, enabling LLMs to play a significant role in the decision-making process for autonomous satellite operations. As a first step towards thi

Afrah Shaahid, Muzammil Behzad

Underwater images are often affected by complex degradations such as light absorption, scattering, color casts, and artifacts, making enhancement critical for effective object detection, recognition, and scene understanding in aquatic environments. Existing methods, especially diffusion-based approaches, typically rely on synthetic paired datasets due to the

Melis Ilayda Bal, Volkan Cevher, Michael Muehlebach

Large language model pretraining is compute-intensive, yet many tokens contribute marginally to learning, resulting in inefficiency. We introduce Efficient Selective Language Modeling (ESLM), a risk-aware algorithm that improves training efficiency and distributional robustness by performing online token-level batch selection. ESLM leverages per-token statis

Yongxian Wei, Runxi Cheng, Weike Jin, Enneng Yang

Foundation models update slowly due to resource-intensive training, whereas domain-specific models evolve rapidly between releases. Model merging seeks to combine multiple expert models into a single, more capable model, reducing storage and serving costs while supporting decentralized development. Despite its potential, previous studies have primarily focus

Estelle Basset, Gilles Lancien, Antonín Procházka

We show that there are uncountably many mutually non-isomorphic Lipschitz-free spaces over countable, complete, discrete metric spaces. Also there is a countable, complete, discrete metric space whose free space does not embed into the free space of any uniformly discrete metric space. This enhanced diversity is a consequence of the fact that the dentability

Gavril Farkas, Soheyla Feyzbakhsh, Andrés Rojas

We develop a novel approach to the Brill-Noether theory of curves endowed with a degree k cover of the projective line via Bridgeland stability conditions on elliptic K3 surfaces. We first develop the Brill-Noether theory on elliptic K3 surfaces via the notion of Bridgeland stability type for objects in their derived category. As a main application, we show

Anton Tkachenko, Dmitrij Suskevic, Benjamin Adolphi

Large language models (LLMs) have shown promise in software engineering, yet their effectiveness for binary analysis remains unexplored. We present the first comprehensive evaluation of commercial LLMs for assembly code deobfuscation. Testing seven state-of-the-art models against four obfuscation scenarios (bogus control flow, instruction substitution, contr

Bernardo Castro Valerio, Vinícius Albernaz Lacerda, Marc Cheah-Mañe, Pieter Gebraad

Interconnecting price zones and remote renewable energy sources has emerged as a key solution to achieving climate goals. The objective of this work is to present a formulation that extends the base optimal power flow model with price zones constraints to forecast the operations of upcoming offshore wind developments integrated into a multi-terminal DC grid.

Stavros Garoufalidis, Campbell Wheeler

We propose a cycle description of the Habiro cohomology of a smooth variety $X$ over the spectrum $B$ of an \'etale $Z[\lambda]$-algebra and construct explicit nontrivial cycles using either the Picard-Fuchs equation on $X/B$ of a hypergeometric motive, or a push-forward of elements of the Habiro ring of $X/B$. In particular, we give explicit classes for 1-p

Martine Dyring Hansen, Elena Celledoni, Benjamin Kwanen Tapley

We introduce a data-driven method for learning the equations of motion of mechanical systems directly from position measurements, without requiring access to velocity data. This is particularly relevant in system identification tasks where only positional information is available, such as motion capture, pixel data or low-resolution tracking. Our approach ta

Soheil Azarpendar

Liu and Piccirillo developed a combinatorial argument that employs the 10/8-theorem to demonstrate that certain manifolds cannot be obtained via Dehn surgery on a knot. We extend their approach by creating additional examples using chainmail links.

Shintaro Ito, Natsuki Takama, Koichi Ito, Hwann-Tzong Chen

The use of multi-view images acquired by a 360-degree camera can reconstruct a 3D space with a wide area. There are 3D reconstruction methods from equirectangular images based on NeRF and 3DGS, as well as Novel View Synthesis (NVS) methods. On the other hand, it is necessary to overcome the large distortion caused by the projection model of a 360-degree came

Pascal Henkel, Jingrui Li, Patrick Rinke

Quaternary mixed-metal M(II)2M(III)Ch2X3 chalcohalides are an emerging material class for photovoltaic absorbers that combines the beneficial optoelectronic properties of lead-based halide perovskites with the stability of metal chalcogenides. Inspired by the recent discovery of lead-free mixed-metal chalcohalides materials, we utilized a combination of dens

M. J. Neves

The study of effective potential for the scalar Lee-Wick pseudo-electrodynamics in one-loop is presented in this letter. The planar and non-local Lee-Wick pseudo-electrodynamics is so coupled to a complex scalar field sector in 1+2 dimensions, where we achieve the Lee-Wick pseudo-scalar electrodynamics. The effective action formalism is applied such that the

Saman Akbari, Manfred Hauswirth

Serverless computing enables developers to deploy code without managing infrastructure, but suffers from cold start overhead when initializing new function instances. Existing solutions such as "keep-alive" or "pre-warming" are costly and unreliable under bursty workloads. We propose universal workers, which are computational units capable of executing any f

Siyuan Li, Xiangze Meng, Yijian Yang, Yiwen Xu

Human preference research is a significant domain in psychology and psychophysiology, with broad applications in psychiatric evaluation and daily life quality enhancement. This study explores the neural mechanisms of human preference judgments through the analysis of event-related potentials (ERPs), specifically focusing on the early N1 component and the lat

Naoki Ichiji, Takuya Ishida, Ikki Morichika, Daigo Oue

The interaction between circularly polarized (CP) light and matter is governed by two fundamental quantities: spin angular momentum (SAM) and optical chirality (OC). While these quantities are inseparable in free space, they can be selectively enhanced in plasmonic near-field regions through appropriately designed structures. We demonstrate that the excitati

Chao Huang, Benfeng Wang, Jie Wen, Chengliang Liu

Recent advancements in reasoning capability of Multimodal Large Language Models (MLLMs) demonstrate its effectiveness in tackling complex visual tasks. However, existing MLLM-based Video Anomaly Detection (VAD) methods remain limited to shallow anomaly descriptions without deep reasoning. In this paper, we propose a new task named Video Anomaly Reasoning (VA

Bowen Tian, Roel C. G. M. Loonen, Roland Valckenborg, Jan L. M. Hensen

Accurate parameterization of rooftop photovoltaic (PV) installations is critical for effective grid management and strategic large-scale solar deployment. The lack of high-fidelity datasets for PV configuration parameters often compels practitioners to rely on coarse assumptions, undermining both the temporal and numerical accuracy of large-scale PV performa

G. Guiglion

Stellar abundances, coupled with kinematics are a unique way to understand the chemo-dynamical processes that occurred to build the Milky Way and its local volume as we observe today. However, measuring abundances is challenging as one needs to properly address the effect of departure from the Local Thermodynamic Equilibrium (LTE), as well as the commonly us

Yi Wu, Lingting Zhu, Shengju Qian, Lei Liu

In the current research landscape, multimodal autoregressive (AR) models have shown exceptional capabilities across various domains, including visual understanding and generation. However, complex tasks such as style-aligned text-to-image generation present significant challenges, particularly in data acquisition. In analogy to instruction-following tuning f

Yanqi Cheng, Xuxiang Zhao, Tieyong Zeng, Pietro Lio

We introduce the Deep Spectral Prior (DSP), a new framework for unsupervised image reconstruction that operates entirely in the complex frequency domain. Unlike the Deep Image Prior (DIP), which optimises pixel-level errors and is highly sensitive to overfitting, DSP performs joint learning of amplitude and phase to capture the full spectral structure of ima

Stavros Maroulis, Nikos Bikakis, Vassilis Stamatopoulos, George Papastefanatos

Minimizing data-to-analysis time while enabling real-time interaction and efficient analytical computations on large datasets are fundamental objectives of contemporary exploratory systems. Although some of the recent adaptive indexing and on-the-fly processing approaches address most of these needs, there are cases, where they do not always guarantee reliab

Daniel Carter, Nicolas Trotignon

We introduce pathographs as a framework to study graph classes defined by forbidden structures, including forbidding induced subgraphs, minors, etc. Pathographs approximately generalize s-graphs of L\'ev\^eque--Lin--Maffray--Trotignon by the addition of two extra adjacency relations: one between subdivisible edges and vertices called spokes, and one between

Martina Muratore, Jonathan Gair, Olaf Hartwig, Michael L. Katz

Instrumental artefacts, such as glitches, can significantly compromise the scientific output of LISA. Our methodology employs advanced Bayesian techniques, including Reversible Jump Markov Chain Monte Carlo and parallel tempering to find and characterize glitches and astrophysical signals. The robustness of the pipeline is demonstrated through its ability to

Sayan Chakraborty, Pratik Kumar Kundu

We study the Morita equivalence classes of crossed products of rotation algebras $A_\theta$, where $\theta$ is a rational number, by finite and infinite cyclic subgroups of $\mathrm{SL}(2, \mathbb{Z})$. We show that for any such subgroup $F$, the crossed products $A_\theta \rtimes F$ and $A_{\theta'} \rtimes F$ are strongly Morita equivalent, where both $\th

Junhong Lee, Seungwook Kim, Minsu Cho

Recent studies show that simple training-free techniques can dramatically improve the quality of text-to-2D generation outputs, e.g. Classifier-Free Guidance (CFG) or FreeU. However, these training-free techniques have been underexplored in the lens of Score Distillation Sampling (SDS), which is a popular and effective technique to leverage the power of pret

Yavar Taheri Yeganeh, Mohsen Jafari, Andrea Matta

With the recent success of world-model agents, which extend the core idea of model-based reinforcement learning by learning a differentiable model for sample-efficient control across diverse tasks, active inference (AIF) offers a complementary, neuroscience-grounded paradigm that unifies perception, learning, and action within a single probabilistic framewor

Feng Xiong, Hongling Xu, Yifei Wang, Runxi Cheng

Self-taught reasoners (STaRs) enhance the mathematical reasoning abilities of large language models (LLMs) by leveraging self-generated responses for self-training. Recent studies have incorporated reward models to guide response selection or decoding, aiming to obtain higher-quality data. However, they typically allocate a uniform sampling budget across all

Pierce S. Boughton, Yang Yang

Solar sails offer a promising solution for fuel free propulsion, enabling novel mission profiles and deeper space exploration. While reaction wheels are standard for spacecraft attitude control, the large moment of inertia of solar sails often lead to frequent reaction wheel saturation, necessitating momentum offloading via additional control methods. Magnet

Chunyang Li, Junwei Zhang, Anda Cheng, Zhuo Ma

Retrieval-Augmented Generation (RAG) enhances large language models (LLMs) by incorporating external knowledge, but its openness introduces vulnerabilities that can be exploited by poisoning attacks. Existing poisoning methods for RAG systems have limitations, such as poor generalization and lack of fluency in adversarial texts. In this paper, we propose CPA

Lukas Meyer, Andrei-Timotei Ardelean, Tim Weyrich, Marc Stamminger

We introduce FruitNeRF++, a novel fruit-counting approach that combines contrastive learning with neural radiance fields to count fruits from unstructured input photographs of orchards. Our work is based on FruitNeRF, which employs a neural semantic field combined with a fruit-specific clustering approach. The requirement for adaptation for each fruit type l

Hexuan Deng, Wenxiang Jiao, Xuebo Liu, Jun Rao

Large Reasoning Models (LRMs) demonstrate strong performance in complex tasks but often face the challenge of overthinking, leading to substantially high inference costs. Existing approaches synthesize shorter reasoning responses for LRMs to learn, but are inefficient for online usage due to the time-consuming data generation and filtering processes. Meanwhi

Gen Kimura, Aina Mayumi, Haruki Yamashita

We establish the tightest possible Robertson-type preparation uncertainty relation, which explicitly depends on the eigenvalues of the quantum state. The conventional constant $ \tfrac{1}{4} $ is replaced by a state-dependent coefficient $\frac{(\lambda_{\max} + \lambda_{\min})^2}{4(\lambda_{\max} - \lambda_{\min})^2}$, where $ \lambda_{\max} $ and $ \lambda

Roman Gansch, Lina Putze, Tjark Koopmann, Jan Reich

Ensuring safe operation of safety-critical complex systems interacting with their environment poses significant challenges, particularly when the system's world model relies on machine learning algorithms to process the perception input. A comprehensive safety argumentation requires knowledge of how faults or functional insufficiencies propagate through the

Hao Jiang, Zhi-Ming Lu, Yuan Ma, Kai Leong Chong

The modulation of drag through dispersed phases in wall turbulence has been a longstanding focus. This study examines the effects of particle Stokes number ($St$) and Froude number ($Fr$) on drag modulation in turbulent Taylor-Couette (TC) flow, using a two-way coupled Eulerian-Lagrangian approach with Reynolds number $Re_i = r_i \omega_i d/\nu$ fixed at 350

Zixiang Zhao, Haowen Bai, Bingxin Ke, Yukun Cui

The real world is dynamic, yet most image fusion methods process static frames independently, ignoring temporal correlations in videos and leading to flickering and temporal inconsistency. To address this, we propose Unified Video Fusion (UniVF), a novel and unified framework for video fusion that leverages multi-frame learning and optical flow-based feature

Xiu-hui Tan, Yu-feng Zhou

The recently identified \textit{memory burden} effect has the potential to significantly decelerate the evaporation of black holes. Specifically, when approximately half of a black hole's initial mass has been radiated away, the evaporation process is halted. This mechanism allows very light primordial black holes (PBHs) with masses $m_{\rm PBH}<10^{15}$ g t

Louis Lehmann, Maximilian R. Becker, Lucas Tepper, Alexander P. Fellows

The interpretation of sum-frequency-generation (SFG) spectra has been severely limited by the absence of quantitative theoretical predictions of higher-order multipole contributions. Magnetic dipole and electric quadrupole contributions are determined by bulk properties but appear in all experimental SFG spectra, obscuring the connection between measured spe

Zexi Li, Xiangzhu Wang, William F. Shen, Meghdad Kurmanji

Large language Model (LLM) unlearning, i.e., selectively removing information from LLMs, is vital for responsible model deployment. Differently, LLM knowledge editing aims to modify LLM knowledge instead of removing it. Though editing and unlearning seem to be two distinct tasks, we find there is a tight connection between them. In this paper, we conceptuali

Natsuki Takama, Shintaro Ito, Koichi Ito, Hwann-Tzong Chen

Gaussian Splatting (GS) has gained attention as a fast and effective method for novel view synthesis. It has also been applied to 3D reconstruction using multi-view images and can achieve fast and accurate 3D reconstruction. However, GS assumes that the input contains a large number of multi-view images, and therefore, the reconstruction accuracy significant

Miaoyu Li, Qin Chao, Boyang Li

Properly evaluating the ability of Video-Language Models (VLMs) to understand long videos remains a challenge. We propose a long-context video understanding benchmark, Causal2Needles, that assesses two crucial abilities insufficiently addressed by existing benchmarks: (1) extracting information from two separate locations (two needles) in a long video and un

F. Aryasetiawan

The one-particle Green function of a many-electron system is traditionally formulated within the self-energy picture. A different formalism was recently proposed, in which the self-energy is replaced by a dynamical exchange-correlation field, which acts on the Green function locally in both space and time. It was found that there exists a fundamental quantit

Gulfarogh Azam, Mohd Sadique, Saif Ali, Mohammad Nadeem

Transliteration, the process of mapping text from one script to another, plays a crucial role in multilingual natural language processing, especially within linguistically diverse contexts such as India. Despite significant advancements through specialized models like IndicXlit, recent developments in large language models suggest a potential for general-pur

Leander Diaz-Bone, Marco Bagatella, Jonas Hübotter, Andreas Krause

Sparse-reward reinforcement learning (RL) can model a wide range of highly complex tasks. Solving sparse-reward tasks is RL's core premise, requiring efficient exploration coupled with long-horizon credit assignment, and overcoming these challenges is key for building self-improving agents with superhuman ability. Prior work commonly explores with the object

Haoqiang Yang, Congde Yuan, Kun Bai, Mengzhuo Guo

Online display advertising platforms rely on pre-ranking systems to efficiently filter and prioritize candidate ads from large corpora, balancing relevance to users with strict computational constraints. The prevailing two-tower architecture, though highly efficient due to its decoupled design and pre-caching, suffers from cross-domain interaction and coarse

Odunayo Ogundepo, Akintunde Oladipo, Kelechi Ogueji, Esther Adenuga

Researchers working on low-resource languages face persistent challenges due to limited data availability and restricted access to computational resources. Although most large language models (LLMs) are predominantly trained in high-resource languages, adapting them to low-resource contexts, particularly African languages, requires specialized techniques. Se

Wenqing Zhou, Yuxuan Yan, Qianqian Yang

Retrieval-Augmented Generation (RAG) improves factuality by grounding LLMs in external knowledge, yet conventional centralized RAG requires aggregating distributed data, raising privacy risks and incurring high retrieval latency and cost. We present DGRAG, a distributed graph-driven RAG framework for edge-cloud collaborative systems. Each edge device organiz

Nagito Saito, Shintaro Ito, Koichi Ito, Takafumi Aoki

Semantic segmentation is a fundamental task in medical image analysis and autonomous driving and has a problem with the high cost of annotating the labels required in training. To address this problem, semantic segmentation methods based on semi-supervised learning with a small number of labeled data have been proposed. For example, one approach is to train

Guoqing Xia, Pei Xiao, Qu Luo, Bing Ji

In this paper, we investigate integrated sensing and communication (ISAC) in a cell-free (CF) multiple-input multiple-output (MIMO) network, where each access point functions either as an ISAC transmitter or as a sensing receiver. We devote into the ISAC sensing metric using the discrete-time signal-based Cramer-Rao lower bounds (CRLBs) for joint location an

Takayoshi Kusune, Hayata Tsuji, Shinki Oyabu, Hidehiro Kaneda

We carried out the near-infrared ($JHK_{\rm s}$) imaging polarimetric observation with the polarimeter SIRPOL on the Infrared Survey Facility (IRSF) 1.4 m telescope and [CII] line mapping observation with a Fabry-P\'{e}rot spectrometer on board a 100-cm TIFR balloon-borne far-infrared telescope toward NGC 6334, and revealed the relationship between the plane

Emir Aslandogan, Haci Ilhan

Orthogonal Time-Frequency Space modulation stands out as a promising waveform for 6G and beyond wireless communication systems, offering superior performance over conventional methods, particularly in high-mobility scenarios and dispersive channel conditions. Error performance analysis remains crucial for accurately characterizing the reliability of wireless

Shuo Wang, Yun Cheng, Qingye Meng, Olga Saukh

Air quality forecasting (AQF) is critical for public health and environmental management, yet remains challenging due to the complex interplay of emissions, meteorology, and chemical transformations. Traditional numerical models, such as CMAQ and WRF-Chem, provide physically grounded simulations but are computationally expensive and rely on uncertain emissio

Michel Bakni, Abbad Diraneyya, Wael Tellat

Term bases are recognized as one of the most effective components of translation software in time saving and consistency. In spite of the many recent advances in natural language processing (NLP) and large language models (LLMs), major translation platforms have yet to take advantage of these tools to improve their term bases and support scalable content for

Arnaud Vadeboncoeur, Mark Girolami, Andrew M. Stuart

This work is focussed on the inversion task of inferring the distribution over parameters of interest leading to multiple sets of observations. The potential to solve such distributional inversion problems is driven by increasing availability of data, but a major roadblock is blind deconvolution, arising when the observational noise distribution is unknown.

Sel Ly, Anshuman Singh, Petr Vorobev, Yeng Chai Soh

Growing penetration of distributed generation such as solar PV can increase the risk of over-voltage in distribution grids, affecting network security. Therefore, assessment of the so-called, PV hosting capacity (HC) - the maximum amount of PV that a given grid can accommodate becomes an important practical problem. In this paper, we propose a novel chance-c

Pascal Wullschleger, Majid Zarharan, Donnacha Daly, Marc Pouly

We investigate the utility of Large Language Models for automated taxonomy generation and completion specifically applied to taxonomies from the food technology industry. We explore the extent to which taxonomies can be completed from a seed taxonomy or generated without a seed from a set of known concepts, in an iterative fashion using recent prompting tech

Christoph R. Landolt, Christoph Würsch, Roland Meier, Alain Mermoud

Multi-Agent Reinforcement Learning (MARL) has shown great potential as an adaptive solution for addressing modern cybersecurity challenges. MARL enables decentralized, adaptive, and collaborative defense strategies and provides an automated mechanism to combat dynamic, coordinated, and sophisticated threats. This survey investigates the current state of rese

Ayaka Usui, Artur Niezgoda, Manuel Gessner

We demonstrate that spinor Bose-Einstein condensates (BEC) can be operated as an analog simulator of the two-dimensional vibron model. This algebraic model for the description of bending and stretching vibrations of molecules, in the case of a triatomic molecules, exhibits two phases where linear and bent configurations are stabilised. Spinor BECs can be eng

Tomás Caraballo, Francisco Morillas, José Valero

This work proposes a method for modeling and forecasting mortality rates. It constitutes an improvement over previous studies by incorporating both the historical evolution of the mortality phenomenon and its random behavior. In the first part, we introduce the model and analyze mathematical properties such as the existence of solutions and their asymptotic

Matilda Häggblom

We introduce two approximate variants of inclusion dependencies and examine the axiomatization and computational complexity of their implication problems. The approximate variants allow for some imperfection in the database and differ in how this degree is measured. One considers the error relative to the database size, while the other applies a fixed thresh

Lingyu Guo, Victor Zhenyu Guo, Li Lu

The sequence $$ \mathbb{P}^{(c)}=(\lfloor p^c \rfloor)_{p\in \mathbb{P}}\quad (c>0,c\notin \mathbb{N}), $$ is an important subsequence of the well-known Piatetski-Shapiro sequence, where $\mathbb{P}$ is the set of prime numbers and $\lfloor \cdot \rfloor$ is the floor function. We prove that for all $c \in (0, 13/15)$, any large enough integer $N$ can be rep

Sharon E. Meidt, Simon C. O. Glover, Ralf S. Klessen, Adam K. Leroy

New extragalactic measurements of the cloud population-averaged star formation (SF) efficiency per freefall time $\rm\epsilon_{\rm ff}$ from PHANGS show little sign of theoretically predicted dependencies on cloud-scale virial level or velocity dispersion. We explore ways to bring theory into consistency with observations, highlighting systematic variations

Xingzhu Zou, Brajesh Kumar, Rishabh Singh Teja, D. K. Sahu

We present the results of low-resolution spectroscopic and densely sampled multi-band photometric follow-up of supernova (SN) 2024aecx. The SN was discovered in the spiral galaxy NGC 3521 (distance $\sim$11 Mpc) within a day after the explosion. The early spectra of SN 2024aecx show a weak signature of hydrogen lines, which disappeared in $\sim$30 days after

Jean-François Joanny, Joseph O. Indekeu

In these Lecture Notes we aim at clarifying how soft matter physics, and herein notably statistical mechanics and fluid mechanics, can be engaged to understand and manipulate non-equilibrium systems consisting of numerous (microscopic) constituents that convert (chemical) energy to mechanical energy, or vice versa, and that are known as active matter. Hydrod

Jaeyoung Choe, Jihoon Kim, Woohwan Jung

Retrieval-augmented generation (RAG) based large language models (LLMs) are widely used in finance for their excellent performance on knowledge-intensive tasks. However, standardized documents (e.g., SEC filing) share similar formats such as repetitive boilerplate texts, and similar table structures. This similarity forces traditional RAG methods to misident

Kento Asai, Yuki Honda, Hiroki Ishikawa, Joe Sato

We consider symmetry breaking of arbitrary gauge groups on a six-dimensional space-time which consists of a four-dimensional Minkowski space-time $M^4$ and a two-dimensional sphere $S^2$. We expand the gauge fields in the presence of a non-trivial background unique to $S^2$. We analyze Kaluza-Klein(KK) modes of the gauge fields and derive the mass spectrum o

Sen Yan, Fabrizio Gabellieri, Etienne Goffinet, Filippo Castiglione

Nuclear Magnetic Resonance (NMR) spectroscopy leverages nuclear magnetization to probe molecules' chemical environment, structure, and dynamics, with applications spanning from pharmaceuticals to the petroleum industry. Despite its utility, the high cost of NMR instrumentation, operation and the lengthy duration of experiments necessitate the development of

Md Basim Uddin Ahmed, Nima Shiri Harzevili, Jiho Shin, Hung Viet Pham

Large Language Models (LLMs) have shown promise in software engineering tasks, but evaluating their effectiveness in vulnerability detection is challenging due to the lack of high-quality datasets. Most existing datasets are limited to function-level labels, ignoring finer-grained vulnerability patterns and crucial contextual information. Also, poor data qua

Jan Janssen, Janine George, Julian Geiger, Marnik Bercx

Numerous Workflow Management Systems (WfMS) have been developed in the field of computational materials science with different workflow formats, hindering interoperability and reproducibility of workflows in the field. To address this challenge, we introduce here the Python Workflow Definition (PWD) as a workflow exchange format to share workflows between Py

Noga Bar, Mariia Seleznova, Yotam Alexander, Gitta Kutyniok

Proper initialization is critical for Recurrent Neural Networks (RNNs), particularly in long-range reasoning tasks, where repeated application of the same weight matrix can cause vanishing or exploding signals. A common baseline for linear recurrences is Glorot initialization, designed to ensure stable signal propagation--but derived under the infinite-width

T. J. J. M. van Overveld, V. Garbin

Colloidal particles at fluid interfaces can enhance the stability of drops and bubbles. Yet, their effect on mass transfer in these multiphase systems remains ambiguous, with some experiments reporting strongly hindered diffusion, while others show nearly no effect, even at near-complete surface coverage. To resolve this ambiguity, we solve the Fick-Jacobs e

Hua Sun

An $[[n,k,d]]$ quantum maximum-distance-separable code maps $k$ source qudits to $n$ coded qudits such that any $n-(d-1)$ coded qudits may recover all source qudits and $n = k + 2 (d-1)$. The entropy of the joint state of the reference system of $k$ qudits and the $n$ coded qudits is fully characterized - the joint state must be pure, i.e., has entropy zero;

Tassilo Klein, Johannes Hoffart

This position paper argues that foundation models for tabular data face inherent limitations when isolated from operational context - the procedural logic, declarative rules, and domain knowledge that define how data is created and governed. Current approaches focus on single-table generalization or schema-level relationships, fundamentally missing the opera

Sarikul Islam, Nitin Gupta

This paper introduces a novel framework to construct the probability density function (PDF) of non-negative continuous random variables. The proposed framework uses two functions: one is the survival function (SF) of a non-negative continuous random variable, and the other is a weight function, which is an increasing and differentiable function satisfying so

J. Moscatelli, X Benoit Gonin, F. Elias

We introduce a device developed to perform a 3D tracking of passive or active particles under flow, confined in a medium of hundreds micrometers wide. Micro-objects are placed inside a vertical glass capillary and two mirrors are set behind it with a certain angle, making it possible to have the two reflections of the capillary on the same optical plane. A 3

Pengcheng Sun, Erwu Liu, Wei Ni, Rui Wang

Federated Learning (FL) is a distributed machine learning paradigm based on protecting data privacy of devices, which however, can still be broken by gradient leakage attack via parameter inversion techniques. Differential privacy (DP) technology reduces the risk of private data leakage by adding artificial noise to the gradients, but detrimental to the FL u

Fei Wang, Lingda Xu, Zeren Zhang

We address a stability threshold problem of the Couette flow $(y,0,0)$ in a uniform magnetic fleld $\alpha(\sigma,0,1)$ with $\sigma\in\mathbb{Q}$ for the 3D MHD equations on $\mathbb{T}\times\mathbb{R}\times\mathbb{T}$. Previously, the authors in \cite{L20,RZZ25} obtained the threshold $\gamma=1$ for $\sigma\in\mathbb{R}\backslash\mathbb{Q}$ satisfying a ge

Zhou Feng, Jiahao Chen, Chunyi Zhou, Yuwen Pu

Backdoor attacks embed malicious triggers into training data, enabling attackers to manipulate neural network behavior during inference while maintaining high accuracy on benign inputs. However, existing backdoor attacks face limitations manifesting in excessive reliance on training data, poor stealth, and instability, which hinder their effectiveness in rea

Feifei Li, Mi Zhang, Zhaoxiang Wang, Min Yang

Interpretability of point cloud (PC) models becomes imperative given their deployment in safety-critical scenarios such as autonomous vehicles. We focus on attributing PC model outputs to interpretable critical concepts, defined as meaningful subsets of the input point cloud. To enable human-understandable diagnostics of model failures, an ideal critical sub

Dannong Wang, Jaisal Patel, Daochen Zha, Steve Y. Yang

Low-rank adaptation (LoRA) methods show great potential for scaling pre-trained general-purpose Large Language Models (LLMs) to hundreds or thousands of use scenarios. However, their efficacy in high-stakes domains like finance is rarely explored, e.g., passing CFA exams and analyzing SEC filings. In this paper, we present the open-source FinLoRA project tha

Mohamed Masry, Mohamed Amen, Mohamed Elzyat, Mohamed Hamed

Decoding natural language from brain activity using non-invasive electroencephalography (EEG) remains a significant challenge in neuroscience and machine learning, particularly for open-vocabulary scenarios where traditional methods struggle with noise and variability. Previous studies have achieved high accuracy on small-closed vocabularies, but it still st

Luc Dessart

We present nonlocal thermodynamic equilibrium radiative transfer calculations of red supergiant and He-star explosions, extending previous work to focus on the infrared emission from atoms and ions in the ejecta during the nebular-phase (i.e., ~200 to ~500d) -- molecules and dust are ignored. We cover non-rotating solar-metallicity progenitors spanning an in

Alon E. Faraggi

Mirror symmetry is one of the celebrated developments in pure mathematics that arose from an initial observation in worldsheet string constructions. The profound implications of mirror symmetry in the Effective Field Theory (EFT) limit of string compactifications was subsequently understood. In particular, it proved to be an exceptionally useful tool in the

Mohammad Abdulaziz, Thomas Ammer, Shriya Meenakshisundaram, Adem Rimpapa

We present a formal analysis, in Isabelle/HOL, of optimisation algorithms for matroids, which are useful generalisations of combinatorial structures that occur in optimisation, and greedoids, which are a generalisation of matroids. Although some formalisation work has been done earlier on matroids, our work here presents the first formalisation of results on

Junnan Liu, Hongwei Liu, Linchen Xiao, Shudong Liu

We propose a novel framework for comprehending the reasoning capabilities of large language models (LLMs) through the perspective of meta-learning. By conceptualizing reasoning trajectories as pseudo-gradient descent updates to the LLM's parameters, we identify parallels between LLM reasoning and various meta-learning paradigms. We formalize the training pro

You Wang, Li Fang, Hao Zhu, Fei Hu

Neural Radiance Fields (NeRF) have transformed novel view synthesis by modeling scene-specific volumetric representations directly from images. While generalizable NeRF models can generate novel views across unknown scenes by learning latent ray representations, their performance heavily depends on a large number of multi-view observations. However, with lim

Zehong Ma, Shiliang Zhang, Longhui Wei, Qi Tian

Traditional approaches to adapting multi-modal large language models (MLLMs) to new tasks have relied heavily on fine-tuning. This paper introduces Efficient Multi-Modal Long Context Learning (EMLoC), a novel training-free alternative that embeds demonstration examples directly into the model input. EMLoC offers a more efficient, flexible, and scalable solut

Runchen Zhang, Xuke Qiu, Yifei Ma, Zimo Zhao

The Mueller Matrix Polar Decomposition method decomposes a Mueller matrix into a diattenuator, a retarder, and a depolarizer. Among these elements, the retarder, which plays a key role in medical and material characterization, is modelled as a circular retarder followed by a linear retarder when using this approach. However, this model may not accurately ref

X Fan, F Mo, C Chen, H Yamana

Recommender systems presently utilize vast amounts of data and play a pivotal role in enhancing user experiences. Graph Convolution Networks (GCNs) have surfaced as highly efficient models within the realm of recommender systems due to their ability to capture extensive relational information. The continuously expanding volume of data may render the training

Matej Komelj, Vinko Sršan, Kristina Žužek, Sašo Šturm

We diagonalize the quantum two-dimensional spin-1/2 Heisenberg model with Dzyaloshinskii-Moriya interaction (DMI) by applying the variational quantum eigensolver, running on a quantum-computer simulator, which turns out to be a more efficient approach than a classical direct diagonalization for systems with more than 17 sites. The calculated external-magneti

Bariscan Bozkurt, Houssam Zenati, Dimitri Meunier, Liyuan Xu

We study the problem of causal function estimation in the Proxy Causal Learning (PCL) framework, where confounders are not observed but proxies for the confounders are available. Two main approaches have been proposed: outcome bridge-based and treatment bridge-based methods. In this work, we propose two kernel-based doubly robust estimators that combine the

Meiyan Kang, Shizuo Kaji, Sang-Yun Lee, Taegon Kim

Ground Penetrating Radar (GPR) is a widely used Non-Destructive Testing (NDT) technique for subsurface exploration, particularly in infrastructure inspection and maintenance. However, conventional interpretation methods are often limited by noise sensitivity and a lack of structural awareness. This study presents a novel framework that enhances the detection