October 2025 arXiv papers — page 49

Zhichao Chen, Hongyuan Sheng, Hao Wang, Jiaze Ma

Green hydrogen production via electrolysis is crucial for decarbonization but faces significant economic hurdles primarily due to the high cost of the electricity. However, current electrolyzer-based hydrogen production processes predominantly rely on the single-scale Day-Ahead Market (DAM) for electricity procurement, failing to fully exploit the economic b

Xiaofeng Zhu, Yunshen Zhou

Microsoft Copilot suites serve as the universal entry point for various agents skilled in handling important tasks, ranging from assisting a customer with product purchases to detecting vulnerabilities in corporate programming code. Each agent can be powered by language models, software engineering operations, such as database retrieval, and internal \& exte

Zora Zhiruo Wang, Yijia Shao, Omar Shaikh, Daniel Fried

AI agents are continually optimized for tasks related to human work, such as software engineering and professional writing, signaling a pressing trend with significant impacts on the human workforce. However, these agent developments have often not been grounded in a clear understanding of how humans execute work, to reveal what expertise agents possess and

Ayushi Chhipa, M. Vivek, Nayana A. J., P. Kharb

SDSSJ1539+3954 ($z\approx 1.935$), a radio-quiet weak-line quasar (WLQ), exhibited exceptional X-ray variability in 2019$-$2020, with its X-ray flux increasing by over 20 times from 2013 to 2019 and subsequently dropping by at least a factor of nine in 2020. Motivated by the empirical correlations between X-ray and radio emission in AGN cores, we carried out

Wenrui Cai, Defa Zhu, Qingjie Liu, Qiyang Min

Normalization layer constitutes an essential component in neural networks. In transformers, the predominantly used RMSNorm constrains vectors to a unit hypersphere, followed by dimension-wise rescaling through a learnable scaling coefficient $\gamma$ to maintain the representational capacity of the model. However, RMSNorm discards the input norm information

Zhen Xie, Zhipeng Zhang, Ruizhi Yang

We investigate the observational capabilities of the upcoming LACT Cherenkov telescope array for the microquasar SS 433 through detailed simulations. Our results indicate that a detection significance of 5 sigma can be achieved with approximately 30 hours of observation. This exposure, coupled with LACT's excellent angular resolution, enables the spatial sep

Canxiang Yan, Chunxiang Jin, Dawei Huang, Haibing Yu

Existing speech models suffer from competing requirements on token representations by understanding and generation tasks. This discrepancy in representation prevents speech language models from performing instruction-based free-form editing. To solve this challenge, we introduce a novel framework that unifies speech understanding, generation, and editing. Th

Junjielong Xu, Boyin Tan, Xiaoyuan Liu, Chao Peng

While language model agents have advanced software engineering, existing test-based supervision is limiting its scalability on real-world issues. The reason is twofold: (1) high-coverage tests are naturally rare in the wild, and (2) building and running test sandbox is heavy and fragile. To unlock supervision scaling, we propose R4P, a reasoning-based method

Junbang Liu

In this note, we give a proof of the uniform log-continuity of the solution to complex Monge-Amp\`ere equations on compact Hermitian manifolds, which is a generalization of the result of Guo-Phong-Tong-Wang in the K\"ahler case.

Simon Scherrer, Adrian Perrig, Stefan Schmid

To understand the fairness properties of the BBR congestion-control algorithm (CCA), previous research has analyzed BBR behavior with a variety of models. However, previous model-based work suffers from a trade-off between accuracy and interpretability: While dynamic fluid models generate highly accurate predictions through simulation, the causes of their pr

Yizhuo Wu, Francesco Fioranelli, Chang Gao

Radar-based human activity recognition (HAR) is attractive for unobtrusive and privacy-preserving monitoring, yet many CNN/RNN solutions remain too heavy for edge deployment, and even lightweight ViT/SSM variants often exceed practical compute and memory budgets. We introduce Neural-HAR, a dimension-gated CNN accelerator tailored for real-time radar HAR on r

Ping Lin, Hatem Zaag

This paper is concerned with a feedback approximate controllability problem of blowup points for the heat equation. We show that the system is approximately controllable for blowup points with feedback controls and the feedback operator is bounded at any time before blowup. It is also proved that the blowup profile for feedback controllability of blowup poin

Chaoming Song

We construct a supersymmetric extension of the Fock-Goncharov cluster ensemble associated with a split basic classical Lie supergroup $G$ and a marked bordered surface $S$. The resulting structure defines a super higher-Teichm\"uller geometry: a split super--thickening of $(\mathscr A_{G,S}, \mathscr X_{G,S})$ equipped with a mutation atlas preserving a cano

Omar Naim, Krish Sharma, Niyar R Barman, Nicholas Asher

Large Language Models (LLMs) typically come with a fixed architecture, despite growing evidence that not all layers contribute equally to every downstream task. We introduce TALE (Task-Aware Layer Elimination), an inference-time method that improves task performance by selectively removing layers that are irrelevant or detrimental for a given task. TALE opti

Massimo Fascinari, Vincent English

This paper investigates how artificial intelligence (AI) can be effectively integrated into Strategic Technology Management (STM) practices to enhance the strategic alignment and effectiveness of technology investments. Through a mixed-methods approach combining quantitative survey data (n=230) and qualitative expert interviews (n=14), this study addresses t

Binxiao Xu, Junyu Feng, Shaolin Lu, Yulin Luo

The rapid development of Vision-language models (VLMs) enables open-ended perception and reasoning. Recent works have started to investigate how to adapt general-purpose VLMs into personalized assistants. Even commercial models such as ChatGPT now support model personalization by incorporating user-specific information. However, existing methods either learn

Maksym Luz, Mykhailo Moklyachuk

The problem of optimal estimation of linear functional ${{A}_{N}}\xi =\sum\limits_{k=0}^{N}{a(k)\xi (k)}\,$ depending on the unknown values of a stochastic sequence $\xi (m)$ with stationary $n$-th increments from observations of the sequence $\xi (k)$ at points $k=-1,-2,\ldots $ and of the sequence $\xi (k)+\eta (k)$ at points of time $k=N+1,N+2,\ldots $ is

Yasmin Moslem, Muhammad Hazim Al Farouq, John D. Kelleher

Large language models (LLMs) have transformed many areas of natural language processing, including machine translation. However, efficient deployment of LLMs remains challenging due to their intensive computational requirements. In this paper, we address this challenge and present our submissions to the Model Compression track at the Conference on Machine Tr

Erin Pierce

We consider the rationality of the Fourier coefficients of a particular paramodular Siegel Eisenstein series of level $N^2$ with weight $k\geq 4$. We show that the coefficients lie in a number field.

Tehreem Zahra, Oreeda Shabbir, Bushra Majeed, Mubasher Jamil

In this article, we investigate the orbital dynamics and quasi-periodic oscillations (QPOs) surrounding a static, spherically symmetric geometry of an Einstein-nonlinear Maxwell-Yukawa (ENMY) black hole (BH). Using the Hamiltonian formalism, we derive equations of motion and analyze the effective potential. We determine the innermost stable circular orbits (

Kai Ye, Bowen Liu, Jianghang Lin, Jiayi Ji

Referring Remote Sensing Image Segmentation (RRSIS) aims to segment instances in remote sensing images according to referring expressions. Unlike Referring Image Segmentation on general images, acquiring high-quality referring expressions in the remote sensing domain is particularly challenging due to the prevalence of small, densely distributed objects and

Naoya Kitajima

We study the dynamics of isolated closed domain walls with 3+1 numerical relativity. A closed wall shrinks due to its own surface tension, and its surface energy is converted to the kinetic energy, leading to implosion. Then, it can result in the formation of a black hole. First, we focus on spherically symmetric closed domain walls and clarify whether they

Li Zhou, Lutong Yu, You Lyu, Yihang Lin

Speech Language Models (SLMs) have made significant progress in spoken language understanding. Yet it remains unclear whether they can fully perceive non lexical vocal cues alongside spoken words, and respond with empathy that aligns with both emotional and contextual factors. Existing benchmarks typically evaluate linguistic, acoustic, reasoning, or dialogu

Jiaqi Wen, Jianyi Yang

This paper studies Distributionally Robust Optimization (DRO), a fundamental framework for enhancing the robustness and generalization of statistical learning and optimization. An effective ambiguity set for DRO must involve distributions that remain consistent with the nominal distribution while being diverse enough to account for a variety of potential sce

Arunima Bhattacharya, W. Jacob Ogden

In this paper, we establish optimal regularity for H\"older continuous Hamiltonian stationary Lagrangian graphs in $\mathbb{C}^n$. We prove that such a graph is smooth whenever its H\"older exponent is strictly larger than $\frac{1}{3}$ and the Lagrangian phase is supercritical, which yields semi-convexity of the potential. We establish the optimality of our

Kshitish Kumar Pradhan, Dushmanta Sahu, Raghunath Sahoo

The behavior of vector mesons in extreme environments provides a unique probe of non-perturbative Quantum Chromodynamics. We investigate the conditions for Bose-Einstein condensation (BEC) of spin-1 $\rho$ mesons in dense rotating hadronic matter, a regime relevant to the peripheral heavy-ion collisions and the interiors of rapidly rotating neutron stars. Wh

Chunyu Li, Shoubin Chen, Dong Li, Weixing Xue

Multi-agent cooperative SLAM often encounters challenges in similar indoor environments characterized by repetitive structures, such as corridors and rooms. These challenges can lead to significant inaccuracies in shared location identification when employing point cloud-based techniques. To mitigate these issues, we introduce TWC-SLAM, a multi-agent coopera

Immanuel Ben Porat, Antoine Gagnebin, Mikaela Iacobelli, Jonathan Junné

We prove that polynomial velocity moments of solutions to the 2D magnetized Vlasov-Poisson system and the 3D magnetized screened Vlasov-Poisson equation remain finite for all times, provided they are finite initially, even when the external magnetic field $B=B(t,x)$ is space-time dependent. We deduce propagation of regularity, thereby implying the existence

Anooshka Bajaj, Deven Mahesh Mistry, Sahaj Singh Maini, Yash Aggarwal

In-context learning is governed by both temporal and semantic relationships, shaping how Large Language Models (LLMs) retrieve contextual information. Analogous to human episodic memory, where the retrieval of specific events is enabled by separating events that happened at different times, this work probes the ability of various pretrained LLMs, including t

Piyushkumar Patel

While Large Language Models have transformed how we interact with AI systems, they suffer from a critical flaw: they confidently generate false information that sounds entirely plausible. This hallucination problem has become a major barrier to deploying these models in real-world applications where accuracy matters. We developed a fact verification framewor

Kaibalyapati Mishra

In this paper, I develop a refinement of stability for matching markets with incomplete information. I introduce Information-Credible Pairwise Stability (ICPS), a solution concept in which deviating pairs can use credible, costly tests to reveal match-relevant information before deciding whether to block. By leveraging the option value of information, ICPS s

Zakaria El Fatouaki, El Mustapha Hrida, Abderahhim Jabar, Abdellah Tahiri

A theoretical study of a number of properties of A2CrH6 (where A = Ba, Sr, and Ca) hydride perovskites with the Cambridge Serial Total Energy Package (CASTEP). These include structural, hydrogen storage, mechanical, phonon, thermodynamic, electronic, and optical properties. The lattice constants of the compounds studied are in the range from 7.220 {\AA} to 8

Travis Askham, Tristan Goodwill, Jeremy G Hoskins, Peter Nekrasov

The dynamics of surface waves traveling along the boundary of a liquid medium are changed by the presence of floating plates and membranes, contributing to a number of important phenomena in a wide range of applications. Mathematically, if the fluid is only partly covered by a plate or membrane, the order of derivatives of the surface-boundary conditions jum

Eeham Khan, Firas Saidani, Owen Van Esbroeck, Richard Khoury

Despite the widespread adoption of Large Language Models (LLMs), their strongest capabilities remain largely confined to a small number of high-resource languages for which there is abundant training data. Recently, continual pre-training (CPT) has emerged as a means to fine-tune these models to low-resource regional dialects. In this paper, we study the use

L. Pagani, R. Saldanha, B. M. Loer, G. S. Ortega

Liquid nitrogen is commonly used in cryogenic applications and is a promising medium for the direct immersion cooling of sensors used for nuclear and particle physics experiments. The scintillation properties of gaseous nitrogen are well-documented, but little is known about the scintillation of liquid nitrogen. If present, scintillation light from interacti

Xinyuan Jiang, Jian Wu, Weiyi Pan

Chiral magnetic textures have attracted considerable attention owing to their topological properties and potential applications in spintronic devices. Here, we employ first-principles calculations together with atomic spin dynamics simulations to explore the switching between skyrmions and Yoshimori-type spin spirals induced by Li adsorption in Janus two-dim

Kanad Pardeshi, Bryan Wilder, Aarti Singh

Online learning algorithms continually update their models as data arrive, making it essential to accurately estimate the expected loss at the current time step. The prequential method is an effective estimation approach which can be practically deployed in various ways. However, theoretical guarantees have previously been established under strong conditions

Raihan Ahamed Rifat, Fuyad Hasan Bhoyan, Md Humaion Kabir Mehedi, Md Kaviul Hossain

Diabetic foot ulcer (DFU) detection is a clinically significant yet challenging task due to the scarcity and variability of publicly available datasets. To solve these problems, we propose ConMatFormer, a new hybrid deep learning architecture that combines ConvNeXt blocks, multiple attention mechanisms convolutional block attention module (CBAM) and dual att

Arunima Bhattacharya, Ravi Shankar, Jeremy Wall, Diego Yepez

We establish interior estimates for singularities of the Lagrangian mean curvature flow when the Lagrangian phase is critical, i.e., $|\Theta|\geq (n-2)\tfrac{\pi}{2}$, and extend our results to the broader class of Lagrangian mean curvature type equations. Our gradient estimates require certain structural conditions, and we construct $C^{\alpha}$ singular v

Sai Krishna Ghanta, Ramviyas Parasuraman

We consider the distributed pose-graph optimization (PGO) problem, which is fundamental in accurate trajectory estimation in multi-robot simultaneous localization and mapping (SLAM). Conventional iterative approaches linearize a highly non-convex optimization objective, requiring repeated solving of normal equations, which often converge to local minima and

Yiwen Tang, Qiuyu Zhao, Zenghui Sun, Jinsong Lan

In Taobao e-commerce visual search, user behavior analysis reveals a large proportion of no-click requests, suggesting diverse and implicit user intents. These intents are expressed in various forms and are difficult to mine and discover, thereby leading to the limited adaptability and lag in platform strategies. This greatly restricts users' ability to expr

Wentao Guo, Wenzeng Zhang

This paper presents environment-adaptive pinch-lifting built on a slot-constrained adaptive linkage (SCAL) and instantiated in two complementary fingers: SCAL-R, a rotational-drive design with an active fingertip that folds inward after contact to form an envelope, and SCAL-L, a linear-drive design that passively opens on contact to span wide or weak-feature

Jacek Gosciniak

In this paper, we propose the Wheatstone bridge configuration for enabling real-time and closed-loop stabilization and calibration of photonic devices integrated on chip. The measurement of the optical power propagating in a waveguide is achieved by leveraging the photo-thermal resistance variation of one of the resistors that comprise a bridge, which is eit

Ningli Xu, Rongjun Qin

Cross-view localization and synthesis are two fundamental tasks in cross-view visual understanding, which deals with cross-view datasets: overhead (satellite or aerial) and ground-level imagery. These tasks have gained increasing attention due to their broad applications in autonomous navigation, urban planning, and augmented reality. Cross-view localization

Christian Klein, Christof Sparber

We study the nonlinear Schr\"odinger equation with a competing cubic-quintic power law nonlinearity on the waveguide domain $\mathbb R_x \times \mathbb T_{L_y}$. This model is globally well-posed and admits line solitary wave solutions, whose transverse (in-)stability is numerically investigated. We consider both spatially localized perturbations and periodi

Yuanhao Lai, Pengfei Zheng, Chenpeng Ji, Yan Li

Gaussian-Process-based Bayesian optimization (GP-BO), is a prevailing model-based framework for DBMS auto-tuning. However, recent work shows GP-BO-based DBMS auto-tuners significantly outperformed auto-tuners based on SMAC, which features random forest surrogate models; such results motivate us to rethink and investigate the limitations of GP-BO in auto-tune

Qi Liu, Yanzhao Zhang, Mingxin Li, Dingkun Long

Text embedding models deliver competitive retrieval performance with high efficiency, but their ranking fidelity remains limited compared to LLM-based listwise rerankers, which capture fine-grained query-document and document-document interactions at high computational cost. We propose E2Rank (Efficient Embedding-based Ranking), a unified framework that exte

Jiali Cheng, Anjishnu Kumar, Roshan Lal, Rishi Rajasekaran

Large Language Model (LLM) web agents often struggle with long-horizon web navigation and web task completion in new websites, producing inefficient action sequences unless fine-tuned on environment-specific data. We show that experience-driven memory, combined with look-ahead action simulation, is sufficient for LLM agents to adapt to unseen web environment

Yong Huang, Wenjing Wang, Dalong Zhang, Junjie Wang

Radio frequency (RF) fingerprinting techniques provide a promising supplement to cryptography-based approaches but rely on dedicated equipment to capture in-phase and quadrature (IQ) samples, hindering their wide adoption. Recent advances advocate easily obtainable channel state information (CSI) by commercial WiFi devices for lightweight RF fingerprinting,

Rongfu Guo, Yantao Yang

This work presents a unified numerical framework for simulating incompressible flows within the coupled fluid-porous-medium system and involving heat and solute transport and phase-changing process. A complete set of governing equations is established based on the Darcy-Brinkman equation, the advection-diffusion equations for heat and solute, and a phase fie

Urja Kohli, Aditi Singh, Arun Sharma

Big Language Models (LLMs) are changing the way businesses use software, the way people live their lives and the way industries work. Companies like Google, High-Flyer, Anthropic, OpenAI and Meta are making better LLMs. So, it's crucial to look at how each model is different in terms of performance, moral behaviour and usability, as these differences are bas

Khai Le-Duc, Duy M. H. Nguyen, Phuong T. H. Trinh, Tien-Phat Nguyen

Faithful reasoning in medical vision-language models (VLMs) requires not only accurate predictions but also transparent alignment between textual rationales and visual evidence. While Chain-of-Thought (CoT) prompting has shown promise in medical visual question answering (VQA), no large-scale expert-level dataset has captured stepwise reasoning with precise

Shmuel Lorber, Yonatan Dubi

We introduce Qlustering, a quantum-inspired algorithm for unsupervised learning that leverages network-based quantum transport to perform data clustering. In contrast to traditional distance-based methods, Qlustering treats the steady-state dynamics of quantum particles propagating through a network as a computational resource. Data are encoded as input stat

Fujiang Yuan, Yangrui Fan, Xiaohuan Bing, Zhen Tian

Accurate traffic flow forecasting is essential for intelligent transportation systems and urban traffic management. However, single model approaches often fail to capture the complex, nonlinear, and multi scale temporal patterns in traffic flow data. This study proposes a decomposition driven hybrid framework that integrates Seasonal Trend decomposition usin

Van Le, Tan Le

SpoofTrackBench is a reproducible, modular benchmark for evaluating adversarial robustness in real-time localization and tracking (RTLS) systems under radar spoofing. Leveraging the Hampton University Skyler Radar Sensor dataset, we simulate drift, ghost, and mirror-type spoofing attacks and evaluate tracker performance using both Joint Probabilistic Data As

Chon-Teng Belmiro Chu, Hao-Chung Chen, Ting Hsu, Hsiang-Yu Lo

Electric-field-induced ion heating is a major obstacle in scalable trapped-ion quantum computing. We present a theoretical study of a novel 3D-printed ion trap with a skeleton electrode structure, designed to reduce heating by minimizing surface area near the ion. Compared to a conventional blade trap with identical confinement parameters, the skeleton trap

Changwon Lee, Tak Hur, Daniel K. Park

Real-time, scalable, and accurate decoding is a critical component for realizing a fault-tolerant quantum computer. While Transformer-based neural decoders such as \textit{AlphaQubit} have demonstrated high accuracy, the computational complexity of their core attention mechanism, which scales as $\mathcal{O}(d^4)$ with code distance $d$, results in decoding

Riddhik Basu, Arkaprava Roy

Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by cognitive decline, structural brain changes, and genetic predispositions. This study leverages machine-learning and statistical techniques to investigate the mechanistic relationships between cognitive function, genetic markers, and neuroimaging biomarkers in AD progression

Jian Ma

Testing copula hypothesis is of fundamental importance in the applications of copula theory. In this paper we proposed a copula hypothesis testing with copula entropy. Since copula entropy is a unified theory in probability and therefore testing copula hypothesis based on it can be applied to any types of copula function. The test statistic is defined as the

Jason Li, Connor Mowry, Satish Rao

We present a faster algorithm for low-diameter decompositions on directed graphs, matching the $O(\log n\log\log n)$ loss factor from Bringmann, Fischer, Haeupler, and Latypov (ICALP 2025) and improving the running time to $O((m+n\log\log n)\log n\log\log n)$ in expectation. We then apply our faster low-diameter decomposition to obtain an algorithm for negat

Kiri Kawato, Yukio Hayashi

Although the analysis of loops is not so much because of the complications, it has already been found that heuristically enhancing loops decreases the variance of degree distributions for improving the robustness of connectivity. While many real scale-free networks are known to contain shorter loops such as triangles, it remains to investigate the distributi

Oskar Grocholski, Piotr H. Chankowski

Applying the previously developed systematic thermal (imaginary time) perturbative expansion to the relevant effective field theory we compute the free energy $F$ of the diluted gas of (nonrelativistic) spin $1/2$ fermions interacting through a spin-independent repulsive two-body potential as a function of the numbers $N_+$ and $N_-$ of spin up and spin down

Yujie Wan, Chenxuan Liu, Shuai Wang, Tong Zhang

Gaussian splatting (GS) struggles with degraded rendering quality on low-cost devices. To address this issue, we present edge collaborative GS (ECO-GS), where each user can switch between a local small GS model to guarantee timeliness and a remote large GS model to guarantee fidelity. However, deciding how to engage the large GS model is nontrivial, due to t

Petr Klenovský

We compute the electronic and emission properties of Coulomb-correlated multi-particle states (X$^0$, X$^\pm$, XX) in weakly confining GaAs/AlGaAs quantum dots using an 8-band $\mathbf{k}\!\cdot\!\mathbf{p}$ model coupled to continuum elasticity and configuration interaction (CI). We evaluate polarization-resolved oscillator strengths and radiative rates bot

Zahra Taghizadeh, Mohammad Shahverdikondori, Arian Noori, Alireza Dadgarnia

Lipreading has emerged as an increasingly important research area for developing robust speech recognition systems and assistive technologies for the hearing-impaired. However, non-English resources for visual speech recognition remain limited. We introduce LRW-Persian, the largest in-the-wild Persian word-level lipreading dataset, comprising $743$ target wo

Radu Ioan Bot, Jalal Fadili, Dang-Khoa Nguyen

In this paper, we prove that the iterates of the accelerated Nesterov's algorithm in the critical regime do converge in the weak topology to a global minimizer of an $L$-smooth function in a real Hilbert space, hence answering positively a conjecture posed by H. Attouch and co-authors a decade ago. This result is the algorithmic case of a very recent result

Jean-Marie Dufour, Abderrahim Taamouti, Meilin Tong

We provide pairwise-difference (Gini-type) representations of higher-order central moments for both general random variables and empirical moments. Such representations do not require a measure of location. For third and fourth moments, this yields pairwise-difference representations of skewness and kurtosis coefficients. We show that all central moments pos

Mengshen Wang, Zuocheng Zhang, Hua Xu

Recent pulsar timing array results, including the NANOGrav 15-year data set, show evidence for a stochastic gravitational-wave background (GWB) in the nanohertz band. We present a Bayesian framework to compare three possible origins: (i) a background from supermassive black hole binary mergers, (ii) a first-order phase transition in the early Universe, and (

Jose Luis Ponton, Eduardo Alvarado, Lin Geng Foo, Nuria Pelechano

Human motion is fundamentally driven by continuous physical interaction with the environment. Whether walking, running, or simply standing, the forces exchanged between our feet and the ground provide crucial insights for understanding and reconstructing human movement. Recent advances in wearable insole devices offer a compelling solution for capturing thes

Amin Heyrani Nobari, Lyle Regenwetter, Cyril Picard, Ligong Han

Structural topology optimization (TO) is central to engineering design but remains computationally intensive due to complex physics and hard constraints. Existing deep-learning methods are limited to fixed square grids, a few hand-coded boundary conditions, and post-hoc optimization, preventing general deployment. We introduce Optimize Any Topology (OAT), a

Wei Chen, Linjun Peng, Zhiyi Huang, Haoyue Dai

Recovering causal structure in the presence of latent variables is an important but challenging task. While many methods have been proposed to handle it, most of them require strict and/or untestable assumptions on the causal structure. In real-world scenarios, observed variables may be affected by multiple latent variables simultaneously, which, generally s

Kaitong Cai, Jusheng Zhang, Yijia Fan, Jing Yang

Retrieval-Augmented Generation (RAG) faces a core bottleneck with knowledge-sparse and semantically ambiguous long-tail queries, where retrieval noise distorts reasoning and necessitates costly post-processing. To tackle this, we propose RaCoT (Retrieval-aware Contrastive-of-Thought), a novel framework that shifts contrastive thinking to the pre-retrieval st

Xi Fang, Zhiqiang Cao, Fan Li

Composite endpoints are increasingly used in clinical trials to capture treatment effects across multiple or hierarchically ordered outcomes. Although inference procedures based on win statistics, such as the win ratio, win odds, and win difference (net benefit) have gained traction in individually randomized trials, their methodological development for clus

Hongzhi Huang, Xian-Tao Huang

In \cite{NPZ24}, Navarro-Pan-Zhu proved that the fundamental group of an open manifold with nonnegative Ricci curvature and linear volume growth contains a subgroup isomorphic to $\mathbb{Z}^k$ with finite index. They further asked whether the existence of a torsion-free element in the fundamental group forces the universal cover to split off an isometric $\

Ivan N. Mikhailov

In this note we construct a geodesic line in the Gromov-Hausdorff class for which the real line with a natural metric is an interior point.

Hao Li, Zhengyu Zou, Fangfu Liu, Xuanyang Zhang

Humans naturally perceive the geometric structure and semantic content of a 3D world as intertwined dimensions, enabling coherent and accurate understanding of complex scenes. However, most prior approaches prioritize training large geometry models for low-level 3D reconstruction and treat high-level spatial understanding in isolation, overlooking the crucia

Lazare Osmanov, Nilanjan Banerjee

This study dives into the applicability of using automated discovery of conserved quantities in dynamical systems relevant to accelerator physics. Specifically, we explore the performance of AI Poincar\'e in analyzing numerical trajectory data obtained using the McMillan system of non-linear integrable optics. A comprehensive evaluation of the algorithm's pe

Ahmad Al-Badawi

This work explores the shadow of a black hole within the framework of $F(R)$-ModMax gravity coupled with a cloud of strings. The Einstein field equations are solved for a nonlinear ModMax electromagnetic source in the context of $F(R)$ gravity and a string cloud. From this solution, we obtain analytical expressions for the photon sphere and shadow radii. Our

Weiwei Hu, Ming-Jun Lai, Hao-Ning Wu

We consider a novel approach for the enhancement of fluid mixing via pure stirring strategies building upon the Least Action Principle (LAP) for incompressible flows. The LAP is formally analogous to the Benamou--Brenier formulation of optimal transport, but imposes an incompressibility constraint. Our objective is to find a velocity field, generated by Hami

Mithul Chander, Sai Pragnya Ranga, Prathamesh Mayekar

We introduce the {\em Atlas Urban Index} (AUI), a metric for measuring urban development computed using Sentinel-2 \citep{spoto2012sentinel2} satellite imagery. Existing approaches, such as the {\em Normalized Difference Built-up Index} (NDBI), often struggle to accurately capture urban development due to factors like atmospheric noise, seasonal variation, a

Daniel Ahlberg, Maria Deijfen, Tiffany Y. Y. Lo

Two sets of objects of size $n$ are to be matched to each other based on i.i.d. costs associated to every pair of objects. Objects prefer to be matched as cheaply as possible, and a matching is said to be stable if there is no pair of objects that would prefer to match to each other rather than to their current partners. Properties of such matchings are anal

R. V. Romanik, O. A. Dobush, M. P. Kozlovskii, I. V. Pylyuk

Entropy of the cell fluid model with Curie-Weiss interaction is obtained in analytical form as a function of temperature and chemical potential. A parametric equation is derived representing the entropy as a function of density. Features of both the entropy per particle and the entropy per cell are investigated at the entropy-density and entropy-chemical pot

Matteo El-Hariry, Andrej Orsula, Matthieu Geist, Miguel Olivares-Mendez

The growing need for autonomous on-orbit services such as inspection, maintenance, and situational awareness calls for intelligent spacecraft capable of complex maneuvers around large orbital targets. Traditional control systems often fall short in adaptability, especially under model uncertainties, multi-spacecraft configurations, or dynamically evolving mi

Vittorio Bernuzzi, Leonardo Rossi, Tomaso Fontanini, Massimo Bertozzi

Self-supervised learning (SSL) has recently emerged as a key strategy for building foundation models in remote sensing, where the scarcity of annotated data limits the applicability of fully supervised approaches. In this work, we introduce WaveMAE, a masked autoencoding framework tailored for multispectral satellite imagery. Unlike conventional pixel-based

Junpeng Gong, Chunkai Wang, Hao Li, Jinyong Ma

A/B testing has become the cornerstone of decision-making in online markets, guiding how platforms launch new features, optimize pricing strategies, and improve user experience. In practice, we typically employ the pairwise $t$-test to compare outcomes between the treatment and control groups, thereby assessing the effectiveness of a given strategy. To be tr

Ivan N. Mikhailov

In this note we show that the Lipschitz distance between the classes of metric spaces at finite Gromov-Hausdorff distances from the one-point metric space and the real line with the natural metric, respectively, is positive.

Ivan Nasonov, Gaiane Panina

We prove that for $n>3$ each generic simple polytope in $\mathbb{R}^n$ contains a point with at least $2n+4$ emanating normals to the boundary. This result is a piecewise-linear counterpart of a long-standing problem about normals to smooth convex bodies.

Shu Zhao, Tianyi Shen, Nilesh Ahuja, Omesh Tickoo

Multimodal Retrieval-Augmented Generation (MRAG) has emerged as a promising method to generate factual and up-to-date responses of Multimodal Large Language Models (MLLMs) by incorporating non-parametric knowledge from external knowledge bases. However, existing MRAG approaches suffer from static retrieval strategies, inflexible modality selection, and subop

Wenlong Li, Yifei Xu, Yuan Rao, Zhenhua Wang

Video anomaly detection (VAD) focuses on identifying anomalies in videos. Supervised methods demand substantial in-domain training data and fail to deliver clear explanations for anomalies. In contrast, training-free methods leverage the knowledge reserves and language interactivity of large pre-trained models to detect anomalies. However, the current fixed-

Huaxin Lin, Hang Wang

We show that Mumford's Approximately Macroscopically Unique (AMU) states exist for quantum systems consisting of unbounded self-adjoint operators when the commutators are small. In particular, AMU states always exist in position and momentum systems when the Planck constant $|\hbar|$ is sufficiently small. However, we show that these standard quantum mechani

Ruslan Berdichevsky, Shai Nahum-Gefen, Elad Ben Zaken

Despite their impressive generalization capabilities, instruction-tuned Large Language Models often underperform on text classification benchmarks. We introduce SALSA, a coherent pipeline that combines structured prompting, class-to-token mapping, and parameter-efficient fine-tuning, thereby avoiding cold-start training. Each class label is mapped to a disti

Jiezhong Wu, Reiichiro Kawai

We establish a practical and easy-to-implement sequential stopping rule for the martingale central limit theorem, focusing on Monte Carlo methods for estimating the mean of a non-iid sequence of martingale difference type. Starting with an impractical scheme based on the standard martingale central limit theorem, we progressively address its limitations from

Joel Rorseth, Parke Godfrey, Lukasz Golab, Divesh Srivastava

If-then rules are widely used to explain machine learning models; e.g., "if employed = no, then loan application = rejected." We present the first proposal to apply rules to explain the emerging class of large language models (LLMs) with retrieval-augmented generation (RAG). Since RAG enables LLM systems to incorporate retrieved information sources at infere

Teresa Arias-Marco, Zdenek Dusek

A new geometrical definition of naturally reductive Finsler manifold using geodeic graph is proposed, with a possible generalization. Based on a construction from a recent paper by the authors, Finsler metrics based on naturally reductive Riemannian metrics $g_i$ are studied. Explicit examples of purely Finsler naturally reductive $\alpha_i$-type metrics are

Shan Zhong, Shutong Ding, He Diao, Xiangyu Wang

Reliable value estimation serves as the cornerstone of reinforcement learning (RL) by evaluating long-term returns and guiding policy improvement, significantly influencing the convergence speed and final performance. Existing works improve the reliability of value function estimation via multi-critic ensembles and distributional RL, yet the former merely co

Ollie Olby, Rory Baggott, Namid Stillman

The recent application of deep learning models to financial trading has heightened the need for high fidelity financial time series data. This synthetic data can be used to supplement historical data to train large trading models. The state-of-the-art models for the generative application often rely on huge amounts of historical data and large, complicated m

Jiuniu Wang, Gongjie Zhang, Quanhao Qian, Junlong Gao

Scalable Vector Graphics (SVGs) are fundamental to digital design and robot control, encoding not only visual structure but also motion paths in interactive drawings. In this work, we introduce RoboSVG, a unified multimodal framework for generating interactive SVGs guided by textual, visual, and numerical signals. Given an input query, the RoboSVG model firs

Hibiki Ayabe, Kazushi Okamoto, Koki Karube, Atsushi Shibata

Structural fireproof classification is vital for disaster risk assessment and insurance pricing in Japan. However, key building metadata such as construction year and structure type are often missing or outdated, particularly in the second-hand housing market. This study proposes a multi-task learning model that predicts these attributes from facade images.

Bar Shaybet, Vladimir Tourbabin, Boaz Rafaely

Accurate Direction-of-Arrival (DOA) estimation in reverberant environments remains a fundamental challenge for spatial audio applications. While deep learning methods have shown strong performance in such conditions, they typically lack a mechanism to assess the reliability of their predictions - an essential feature for real-world deployment. In this work,

Rikiya Takehi, Fernando Diaz, Tetsuya Sakai

Users tend to remember failures of a search session more than its many successes. This observation has led to work on search robustness, where systems are penalized if they perform very poorly on some queries. However, this principle of robustness has been overlooked within a single query. An ambiguous or underspecified query (e.g., ``jaguar'') can have seve