December 2025 arXiv papers — page 114

Yuan Gao, Yanshen Li

The interaction of drops floating on liquid surfaces is important for many natural processes and industrial applications. In many of the cases, the system is multicomponent, leading to Marangoni flows on the surface. Here we investigate the competing effect of the attractive ``Cheerios effect'' and the repulsive solutal Marangoni flow by observing the behavi

Anuj Nandi, Manju Sudhakar, Srikar Paavan Tadepalli, Anand Jain

HEL1OS (High Energy L1 Orbiting X-ray Spectrometer) is one of the remote sensing payloads on board Aditya-L1 mission designed to continuously monitor and measure the time-resolved spectra of solar flares between 8 keV and 150 keV. This broad energy range has been covered by using compound semiconductor detectors: cadmium telluride (CdTe: 8 - 70 keV) and cadm

Vitaliy Voytik

The purpose of this article is to extend the applicability of the stationarity principle of the full Jacobi action to non-conservative natural systems and to derive equations of motion corresponding to this extended principle. To this end, in addition to the well-known variation of the Jacobi action with respect to coordinates, we propose independently varia

Fatimah Zohra, Chen Zhao, Hani Itani, Bernard Ghanem

CLIP achieves strong zero-shot image-text retrieval by aligning global vision and text representations, yet it falls behind on fine-grained tasks even when fine-tuned on long, detailed captions. In this work, we propose $\beta$-CLIP, a multi-granular text-conditioned contrastive learning framework designed to achieve hierarchical alignment between multiple t

Jiawei Yan, Ju Liu, Weidong Liu, Jiyuan Tu

We propose a robust and scalable variational Bayes (VB) framework designed to effectively handle contamination and outliers in dataset. Our approach partitions the data into $m$ disjoint subsets and formulates a joint optimization problem based on robust aggregation principles. A key insight is that the full posterior distribution is equivalent to the minimi

Jaehyeok Choi, Seunggyu Kim

We investigate the supercharge cohomology of an $\mathcal{N}=1$ relevant deformation of $\mathcal{N}=4$ super Yang-Mills. By introducing a field redefinition, we integrate out massive fields in a cohomological sense. Then, we construct the monotone cohomologies corresponding to the Kaluza-Klein particles of the dual supergravity solution. Some of the monoton

Yushun Tang, Ziqiong Liu, Jiyuan Jia, Yi Zhang

Online test-time adaptation aims to dynamically adjust a network model in real-time based on sequential input samples during the inference stage. In this work, we find that, when applying a transformer network model to a new target domain, the Query, Key, and Value features of its self-attention module often change significantly from those in the source doma

Mirzabek Alloqulov, Ahmadjon Abdujabbarov, Bobomurat Ahmedov, Chengxun Yuan

Here, we explore the effect of the cloud of strings (CoS) on the gravitational waveforms of extreme mass ratio inspirals (EMRIs). The EMRI system consists of a supermassive black hole (BH) and a compact stellar mass object moving around it. We begin with studying the test particle motion around the Schwarzschild BH surrounded by a CoS by using the Lagrangian

Maria Khilchuk, Vladimir Latypov, Pavel Kleshchev, Alexander Hvatov

Diffusion and Schr\"{o}dinger Bridge models have established state-of-the-art performance in generative modeling but are often hampered by significant computational costs and complex training procedures. While continuous-time bridges promise faster sampling, overparameterized neural networks describe their optimal dynamics, and the underlying stochastic diff

Stelios Stefas, George Zoupanos

Reviving the old proposal of describing gravity as a gauge theory first we describe the construction of the Conformal and the Noncommutative (Fuzzy) Gravities in a gauge-theoretic manner. Then stressing the fact that the tangent group of a curved manifold and the manifold itself do not necessarily have the same dimensions, we show how the above Gravities can

Jiawei Shen, Jia Zhu, Hanghui Guo, Weijie Shi

Temporal Knowledge Graph Reasoning (TKGR) aims to complete missing factual elements along the timeline. Depending on the temporal position of the query, the task is categorized into interpolation and extrapolation. Existing interpolation methods typically embed temporal information into individual facts to complete missing historical knowledge, while extrapo

Long-Zhou Huang, Xu Yang, Min-Qiang Jiang, Yun-Jiang Wang

It was recently shown that vortex-like topological defects with negative winding number in the vibrational modes of a two-dimensional glass under quasistatic shear correlate strongly with plastic events, offering a promising route to predict them. However, many of these vortices, a number that actually grows quadratically with mode frequency, are entirely un

Nardine Osman, Manel Rodriguez-Soto, Jordi Sabater-Mir

In firefighting and other emergency operations, decisions made under pressure carry profound ethical weight and can significantly impact incident outcomes and firefighter safety. Traditional training methods, while foundational, often fall short in adequately preparing firefighters for the complex ethical dilemmas and value conflicts inherent in chaotic emer

Sumanta Paul, Oleksandr Yampolskyy, Zehua Wu, Klaus Müllen

Singlet fission (SF) is a photophysical process where a singlet excitation generates two triplet excited states, enhancing exciton multiplication potentially useful for solar energy conversion. Since SF typically outcompetes radiative decay, single molecule studies of SF have remained elusive. Here, we present single molecule spectroscopy of a terrylenediimi

Haiyang Zheng, Nan Pu, Wenjing Li, Teng Long

The proliferation of synthetic facial imagery has intensified the need for robust Open-World DeepFake Attribution (OW-DFA), which aims to attribute both known and unknown forgeries using labeled data for known types and unlabeled data containing a mixture of known and novel types. However, existing OW-DFA methods face two critical limitations: 1) A confidenc

Maria Khilchuk, Ilya Markov, Alexander Hvatov

In differential equation discovery algorithms, numerical differentiation is usually a fixed preliminary step. Current methods improve robustness with data subsampling and sparsity but often ignore the variability from the differentiation method itself. We show that this choice systematically introduces uncertainty, affecting both equation form and parameter

Hossein Emami, Hadi Sabri

In this work, we investigate the experimental correlation between the pairing gap values and two important observables in the study of nuclear structure (two neutron separation energies and thermal-neutron capture cross-sections). To this aim, we focused on the even-even nuclei in the vicinity of Z=50 and Z = 82 closed proton shells, for which the quantum ph

Sreehari Rajan, Kunal Bhosikar, Charu Sharma

Generating realistic human motions that naturally respond to both spoken language and physical objects is crucial for interactive digital experiences. Current methods, however, address speech-driven gestures or object interactions independently, limiting real-world applicability due to a lack of integrated, comprehensive datasets. To overcome this, we introd

Gelesh G Omathil, Sreeja CS

Deep neural networks possess strong representational capacity yet remain vulnerable to overfitting, primarily because neurons tend to co-adapt in ways that, while capturing complex and fine-grained feature interactions, also reinforce spurious and non-generalizable patterns that inflate training performance but reduce reliability on unseen data. Noise-based

Zdeněk Kasner, Ondřej Dušek

Large language models (LLMs) are becoming central to natural language processing education, yet materials showing their mechanics are sparse. We present AnimatedLLM, an interactive web application that provides step-by-step visualizations of a Transformer language model. AnimatedLLM runs entirely in the browser, using pre-computed traces of open LLMs applied

Abdelsadeq Elfergany, Ammar Adl, Mohammed Kayed

Recommendation systems face challenges in dynamically adapting to evolving user preferences and interactions within complex social networks. Traditional approaches often fail to account for the intricate interactions within cyber-social systems and lack the flexibility to generalize across diverse domains, highlighting the need for more adaptive and versatil

Muhammad Umar Farooq, Abd Ur Rehman, Azka Rehman, Muhammad Usman

Accurate thyroid nodule segmentation in ultrasound images is critical for diagnosis and treatment planning. However, ambiguous boundaries between nodules and surrounding tissues, size variations, and the scarcity of annotated ultrasound data pose significant challenges for automated segmentation. Existing deep learning models struggle to incorporate contextu

A. T. Bajkova, A. A. Smirnov, V. V. Bobylev

The regularity/chaoticity of orbits of 45 globular clusters in the central region of the Galaxy with a radius of 3.5 kpc, which are subject to the greatest influence of the elongated rotating bar, is analyzed. Various methods of analysis are used, namely, the methods of calculating the maximum characteristic Lyapunov exponents (MCLE), MEGNO (Mean Exponential

Wenhao Xu

Quantum synchronisation has recently been proposed as a mechanism for electronic excitation energy transfer in light-harvesting complexes, yet its robustness in driven-dissipative settings remains under active investigation. Here, we revisit this mechanism in cryptophyte photosynthetic antennae using an exciton--vibrational dimer model. By comparing the full

B. Amri, A. Guesmi

In this paper, we investigate the trigonometric Heckman-Opdam polynomials of type $A_1$. We establish connections with ultraspherical polynomials and derive an explicit expression for the associated Poisson kernel. Using the product formula, we introduce a natural convolution structure and develop a theory of fractional integrals associated with these polyno

Qixin Xu, Haozhe Wang, Che Liu, Fangzhen Lin

Current document reasoning paradigms are constrained by a fundamental trade-off between scalability (processing long-context documents) and fidelity (capturing fine-grained, multimodal details). To bridge this gap, we propose CogDoc, a unified coarse-to-fine thinking framework that mimics human cognitive processes: a low-resolution "Fast Reading" phase for s

Hongyang Li, Junyi Tao, Qijie Wei, Ningzhi Yang

Previous work on cross-modal fundus image registration (CMFIR) assumes small cross-modal Field-of-View (FoV) disparity. By contrast, this paper is targeted at a more challenging scenario with large FoV disparity, to which directly applying current methods fails. We propose Crop and Alignment for cross-modal fundus image Registration(CARe), a very simple yet

Wachara Fungwacharakorn, May Myo Zin, Ha-Thanh Nguyen, Yuntao Kong

In this paper, we investigate how language models can perform case-based reasoning (CBR) on non-factorized case bases. We introduce a novel framework, argumentative agentic models for case-based reasoning (AAM-CBR), which extends abstract argumentation for case-based reasoning (AA-CBR). Unlike traditional approaches that require factorization of previous cas

James Drummond, Ömer Gürdoğan, Matthew Rochford, Rowan Wright

It is well-known that the expectation values of null polygonal Wilson loops computed in planar \(\mathcal{N}=4\) super Yang-Mills theory are dual to MHV amplitudes in that theory, and moreover that the duality can be extended to higher helicity sectors through the introduction of super Wilson loops. In this first of a series of papers, we investigate the nat

Hassan Mujtaba, Hamza Naveed, Hanzlah Munir

Authorship analysis has traditionally focused on lexical and stylistic cues within text, while higher-level narrative structure remains underexplored, particularly for low-resource languages such as Urdu. This work proposes a graph-based framework that models Urdu novels as character interaction networks to examine whether authorial style can be inferred fro

Tatsuru Kikuchi

This paper develops a continuous functional framework for treatment effects propagating through geographic space and economic networks. We derive a master equation from three independent economic foundations -- heterogeneous agent aggregation, market equilibrium, and cost minimization -- establishing that the framework rests on fundamental principles rather

Nardine Osman

This paper introduces the concept of value awareness in AI, which goes beyond the traditional value-alignment problem. Our definition of value awareness presents us with a concise and simplified roadmap for engineering value-aware AI. The roadmap is structured around three core pillars: (1) learning and representing human values using formal semantics, (2) e

Vladislav V. Kozhukhov

In the framework of a viscoelastic material model, whose constitutive relation is given by a one-parameter family of Gordon-Schowalter derivatives, the problem of simple shear under acceleration and random velocity motion is considered. For motion with acceleration, the presence of non-zero normal stresses is discovered, which corresponds to the Poynting eff

Jiachang Zhang, Zhiyuan Li, Ziqian Hua, Tong Bao

Based on archival Chandra observations, we present a systematic timing survey of several hundred X-ray sources in M31, M81, and Centaurus A, mostly low-mass X-ray binaries (LXMBs), focusing on searching and characterizing aperiodic and periodic variability within single observation. We identify flares in 24 sources in M31, 5 in M81, and 26 in Cen~A; several

Cameron Jones, Piper Wysocki, MengKe Feng, Gerardo A. Paz-Silva

Practical quantum computers need to continuously exchange data between classical and quantum subsystems during a computation. Mid-circuit measurements of a qubits state are transferred to the classical electronics layer, and their outcome can inform feedforward operations that close the loop back to the quantum layer. These operations are crucial for fault-t

Shiqi Liu, Edward McDonald, Fedor Sukochev, Dmitriy Zanin

On graded Lie groups, we develop a mechanism that transfers the uniformity of maximal hypoellipcity from the frozen coefficients principal part of a differential operator to the full operator. Our approach brings the century-old "freeze-unfreeze" strategy into the hypoelliptic setting, and offers a transparent and flexible framework for lifting symbol-level

Salman Sajad Wani, Saif Al-Kuwari

The perspective-neutral formulation of quantum reference frames (QRFs) treats observers as quantum systems and describes physics relationally from within the composite system. While frame-change maps and frame-invariant resource sums are theoretically understood, their impact on circuit-based quantum information processing has largely remained unexplored. We

Julian Adamek

In two papers, A. Lapi et al. introduce and discuss what they call the $\eta$CDM model, a stochastic framework in which they claim that fluctuations in the density field at the scale of tens of Mpc due to structure formation would effectively drive the accelerated expansion of the Universe. They claim that this qualitative behaviour would emerge from the dyn

Yida Cai, Ranjuexiao Hu, Huiyuan Xie, Chenyang Li

Legal relations serve as an important analytical framework for dispute resolution in civil cases. However, legal relations in Chinese civil cases remain underexplored in the field of legal AI, largely due to the absence of comprehensive schemas. In this work, we first introduce a comprehensive schema for legal relations in civil cases, which contains a hiera

Carlo Abate, Ivan Marisca, Filippo Maria Bianchi

Torch Geometric Pool (tgp) is a pooling library built on top of PyTorch Geometric. Graph pooling methods differ in how they assign nodes to supernodes, how they handle batches, what they return after pooling, and whether they expose auxiliary losses. These differences make it hard to compare methods or reuse the same model code across them. tgp addresses thi

Sander Land, Yuval Pinter

The Unigram tokenization algorithm offers a probabilistic alternative to the greedy heuristics of Byte-Pair Encoding. Despite its theoretical elegance, its implementation in practice is complex, limiting its adoption to the SentencePiece package and adapters thereof. We bridge this gap between theory and practice by providing a clear guide to implementation

Boris Kryzhanovsky, Vladislav Egorov

Expressions for the density of states $D(E)$, where $D(E)$ is the number of states of energy $E$, are well known. The present paper offers the expressions for generalized density of states $D_N(E,m)$, where $D_N(E,m)$ is the number of states with energy $E$ and magnetization $m$ in a one-dimensional $N$-spin chain. The expressions obtained here can be consid

Xiangzhi Cao

In this paper, we are devoted to define p symphonic morphism and characterize it partially as in the case of harmonic morphism.

Rishit Agnihotri, Amit Chaurasia

The transition to electric transportation is a key enabler for intelligent and sustainable cities; however, inadequate charging infrastructure remains a major barrier to large-scale electric vehicle (EV) adoption. This paper presents a scalable Electric Road System (ERS) architecture that enables Dynamic Wireless Charging (DWC) of EVs during motion. The prop

Eung Jin Chun, Hyun Min Lee, Jun-Ho Song

We propose a cogenesis mechanism that unifies the origin of QCD axion dark matter and the baryon asymmetry of the Universe in the framework of Peccei-Quinn pole inflation. The model integrates the Peccei-Quinn symmetry with the seesaw mechanism for neutrino masses. This allows for spontaneous leptogenesis, which generates the required $B-L$ asymmetry around

Enso O. Torres Alegre

Within finite-dimensional generalized probabilistic theories (GPTs), we distinguish between the geometric transition probability tau(psi,phi), defined as the maximum probability of accepting phi when the state is psi, and the predictive probability P(phi|psi) assigned to measurement outcomes. We ask what functional relationship P = Phi(tau) is compatible wit

Jordi Delgado, Marco Linton, Jone Lopez de Gamiz Zearra, Mallika Roy

A group $G$ is said to satisfy the finitely generated intersection property (f.g.i.p.) if the intersection of any two finitely generated subgroups of $G$ is again finitely generated. The aim of this article is to understand when the fundamental group of a graph of groups has the f.g.i.p. Our main results are general criteria for the f.g.i.p. in graphs of gro

Zhou Tao, Shida Wang, Yongxiang Hua, Haoyu Cao

Multimodal Large Language Models have achieved impressive performance on a variety of vision-language tasks, yet their fine-grained visual perception and precise spatial reasoning remain limited. In this work, we introduce DiG (Differential Grounding), a novel proxy task framework where MLLMs learn fine-grained perception by identifying and localizing all di

Rishit Agnihotri, Sandeep Kumar Sharma

Urban Air Mobility (UAM) poses unprecedented traffic coordination challenges, especially with increasing UAV densities in dense urban corridors. This paper introduces a mathematical model using a control algorithm to optimize an Edge AI-driven decentralized swarm architecture for intelligent conflict resolution, enabling real-time decision-making with low la

Martine Hennion, Alexandre Ivanov, Claudine Lacroix, Bernard Hennion

The compound La1-xSrxMnO3 exhibits a charge order (CO) state at $x\approx 1/8$ and $T<T_{co}$, which recalls the CO state with a decrease in the temperature of the superconducting transition, $T_c$, observed in all cuprates at this doping value. Local excitations of lattice and magnetic origins measured in the two-dimensional metallic state of La7/8Sr1/8MnO3

Vaarunay Kaushal, Taranveer Singh

Multi-agent systems where Large Language Models (LLMs) deliberate to form consensus have gained significant attention, yet their practical value over simpler methods remains under-scrutinized. We introduce DELIBERATIONBENCH, a controlled benchmark evaluating three deliberation protocols against a strong baseline of selecting the best response from a pool of

Yuqian Sun, Xingyu Li, Shunyu Yao, Noura Howell

Recent advances in Generative AI (GAI) have led to new opportunities for creativity support. However, this technology has raised ethical concerns in the visual artists community. This paper explores how GAI can assist visual artists in developing original characters (OCs) while respecting their creative agency. We present ORIBA, an AI chatbot leveraging larg

Hao-Ran Yu, Bing-Hang Chen, Kun Xu, Ming-Jie Sheng

$N$-body simulation serves as a critical method for modeling cosmic evolution and poses a significant challenge in high-performance computing. We present CUBE2, an open-source cosmological $N$-body code emphasizing memory efficiency, computational performance, scalability and precision. The core of its algorithm utilizes multi-level Particle-Mesh (PM) method

Alexei H. Sopov, Carlos Tamarit, Raymond R. Volkas

It was previously shown how several explanatory deficiencies of the Standard Model (including the origin of dark matter, matter-antimatter asymmetry, small active neutrino mass, strong CP-conservation and the seeds for large-scale structure formation) may be economically resolved when an experimentally-accessible QCD axion also plays the role of the majoron,

M. Vallinayagam, A. E. Sudheer, A. Kumar, G. Tejaswini

We report a comprehensive investigation of the thermoelectric properties of SbXY (X = Se, Te; Y = Br, I) Janus layers (JL) using spin-polarized first-principles calculations. Ab initio molecular dynamics confirm that the 1T phase ($Pm31$) remains stable up to 1000 K, excluding any phase transitions. The calculated mean-square displacement further evidences t

Borui Xu, Jingzhu Shao, Xiangyu Zhao, Haishan Xu

Recent advances in meta-optics have enabled diverse functionalities in compact optical devices; however, conventional forward design approaches become inadequate as device complexity and scale grow. Inverse design offers a powerful alternative but often requires massive computational resources and neglects mutual coupling effects. Here, we propose and experi

Lankadinee Rathuwadu, Guanli Liu, Christopher Leckie, Renata Borovica-Gajic

Cardinality estimation (CE), the task of predicting the result size of queries is a critical component of query optimization. Accurate estimates are essential for generating efficient query execution plans. Recently, machine learning techniques have been applied to CE, broadly categorized into query-driven and data-driven approaches. Data-driven methods lear

Chengzhi Liu, Yuzhe Yang, Yue Fan, Qingyue Wei

Recent advancements in Multimodal Large Language Models (MLLMs) have significantly enhanced cross-modal understanding and reasoning by incorporating Chain-of-Thought (CoT) reasoning in the semantic space. Building upon this, recent studies extend the CoT mechanism to the visual modality, enabling models to integrate visual information during reasoning throug

Akhil Pandey Akella, Harish Varma Siravuri, Shaurya Rohatgi

Large Language Models are versatile general-task solvers, and their capabilities can truly assist people with scholarly peer review as \textit{pre-review} agents, if not as fully autonomous \textit{peer-review} agents. While incredibly beneficial, automating academic peer-review, as a concept, raises concerns surrounding safety, research integrity, and the v

Zihan Wang, Seungjun Lee, Guangzhao Dai, Gim Hee Lee

Embodied agents face a critical dilemma that end-to-end models lack interpretability and explicit 3D reasoning, while modular systems ignore cross-component interdependencies and synergies. To bridge this gap, we propose the Dynamic 3D Vision-Language-Planning Model (D3D-VLP). Our model introduces two key innovations: 1) A Dynamic 3D Chain-of-Thought (3D CoT

Haizhong Li, Bo Yang

In Euclidean space $\mathbb{R}^n$, the minimization problem of a nonlocal isoperimetric functional with a competition between perimeter and a nonlocal term derived from the negative power of the distance function, has been extensively studied. In this paper, we investigate this nonlocal isoperimetric problem in hyperbolic space $\mathbb{H}^n$, we prove that

Sanjay Das, Swastik Bhattacharya, Shamik Kundu, Arnab Raha

State-space models (SSMs), exemplified by the Mamba architecture, have recently emerged as state-of-the-art sequence-modeling frameworks, offering linear-time scalability together with strong performance in long-context settings. Owing to their unique combination of efficiency, scalability, and expressive capacity, SSMs have become compelling alternatives to

Aheli Poddar, Saptarshi Sahoo, Sujata Ghosh

We study syllogistic reasoning in LLMs from the logical and natural language perspectives. In process, we explore fundamental reasoning capabilities of the LLMs and the direction this research is moving forward. To aid in our studies, we use 14 large language models and investigate their syllogistic reasoning capabilities in terms of symbolic inferences as w

Riju Basak, Daniel Spector

In this paper, we establish Miyachi-Peral-type fixed-time estimates for wave multipliers acting on $\beta$-dimension stable spaces of measures. Our estimates give a refinement of known estimates for the Hardy space. From these bounds, we deduce corresponding estimates for the wave equation with measure data.

Animesh Mishra

Decentralized federated learning (DFL) enables collaborative model training without centralized trust, but it remains vulnerable to Byzantine clients that poison gradients under heterogeneous (Non-IID) data. Existing defenses face a scalability trilemma: distance-based filtering (e.g., Krum) can reject legitimate Non-IID updates, geometric-median methods inc

Yusheng Zheng, Tong Yu, Yiwei Yang, Minghui Jiang

Performance in modern GPU-centric systems increasingly depends on resource management policies, including memory placement, scheduling, and observability. However, uniform policies typically yield suboptimal performance across diverse workloads. Existing approaches present a tradeoff: user-space runtimes provide programmability and flexibility but lack cross

Xiao Zhou, Yuze Sun, Jie Wu, Xiaomeng Huang

Accurately defining the life cycle of the Madden-Julian Oscillation (MJO), the dominant mode of intraseasonal climate variability, remains a foundational challenge due to its propagating nature. The established linear-projection method (RMM index) often conflates mathematical artifacts with physical states, while direct clustering in raw data space is confou

Yucan Guo, Saiping Guan, Miao Su, Jiyao Wei

Sparse Knowledge Graphs (KGs) are commonly encountered in real-world applications, where knowledge is often incomplete or limited. Sparse KG reasoning, the task of inferring missing knowledge over sparse KGs, is inherently challenging due to the scarcity of knowledge and the difficulty of capturing relational patterns in sparse scenarios. Among all sparse KG

Arshyn Altybay, Alibek Yeskermessuly

We investigate linear parabolic equations in divergence form with singular coefficients and non-smooth boundary data. When the diffusion, drift, or potential terms, as well as the initial or boundary conditions, are distributions rather than functions, classical and weak solution concepts become inadequate due to the ill-posedness of products involving distr

Wei Men, Longfei Zhao, Yong Liang Guan, Xiangwang Hou

Integrated sensing and communication (ISAC) technology is crucial for next-generation underwater networks. However, covering multiple users and targets and balancing sensing and communication performance in complex underwater acoustic (UWA) environments remains challenging. This paper proposes an interleaved orthogonal frequency division multiplexing-based M

Wonseok Choi, Sohwi Lim, Nam Hyeon-Woo, Moon Ye-Bin

Instance-level image retrieval aims to find images containing the same object as a given query, despite variations in size, position, or appearance. To address this challenging task, we propose Patchify, a simple yet effective patch-wise retrieval framework that offers high performance, scalability, and interpretability without requiring fine-tuning. Patchif

Zhuang Fan, Yukun Huang, Wenchan Dong, Haodong Yang

Chip-scale all-optical signal broadcasting enables data replication from an optical signal to a large number of wavelength channels, playing a critical role in enabling massive-throughput optical communication and computing systems. The underlying process is four-wave mixing between an optical signal and a multi-wavelength pump source via optical Kerr nonlin

Hong Su

Large Language Models (LLMs) excel at extracting common patterns from large-scale corpora, yet they struggle with rare, low-resource, or previously unseen scenarios-such as niche hardware deployment issues or irregular IoT device behaviors-because such cases are sparsely represented in training data. Moreover, LLMs rely primarily on implicit parametric memor

Danai Roumelioti, Stelios Stefas, George Zoupanos

Within the gauge-theoretic approach of gravity, the gauging of an enlarged symmetry of the tangent space in four dimensions allows gravity to be unified with internal interactions. We study the unification of the conformal and noncommutative (fuzzy) gravities with internal interactions based on the SO(10) GUT.

Dein Wong, Songnian Xu, Chi Zhang, Jinxing Zhao

A numerical semigroup $S$ is a cofinite subsemigroup of $ \mathbb{N}$, where $\mathbb{N}$ is the additive monoid of non-negative integers. Denote by $\mathcal{P}_{\rm fin} (S)$ the semigroup consisting of all non-empty finite subsets of $S$ endowed with the operation of setwise addition defined by $$X+Y=\{x+y:x\in X, y\in Y\}, \qquad\text{for all } X, Y \in

Pranav Gupta, Nithin Surendran

Most major retailers today have multiple divisions focused on various aspects, such as marketing, supply chain, online customer experience, store customer experience, employee productivity, and vendor fulfillment. They also regularly collect data corresponding to all these aspects as dashboards and weekly/monthly/quarterly reports. Although several machine l

Tingyan Wen, Haoyu Li, Yihuang Chen, Xing Zhou

Diffusion models achieve remarkable generative quality, but computational overhead scales with step count, model depth, and sequence length. Feature caching is effective since adjacent timesteps yield highly similar features. However, an inherent trade-off remains: aggressive timestep reuse offers large speedups but can easily cross the critical line, hurtin

Guo-Niu Han

We prove and generalize a conjecture of Johann Cigler on the Hankel determinants of convolution powers of Narayana polynomials. Our method follows a "guess-and-prove" strategy, relying on established techniques involving Hankel continued fractions. While the final forms of our theorems are given by simple closed expressions, the proofs require us to formulat

Jingdi Lei, Di Zhang, Soujanya Poria

In this paper, we introduce Exact Flow Linear Attention~(EFLA), an exact-flow formulation of delta-rule linear attention. We show that the delta-rule update can be interpreted as an explicit Euler discretization of an underlying continuous-time system. EFLA replaces this first-order update with the exact closed-form flow. By exploiting the rank-1 structure o

Si Wu, Zhengyan Qin, Tengfei Liu, Zhong-Ping Jiang

This paper investigates the control problem of steering a group of spherical mobile robots to cooperatively transport a spherical object. By controlling the movements of the robots to exert appropriate contact (pushing) forces, it is desired that the object follows a velocity command. To solve the problem, we first treat the robots' positions as virtual cont

Si Wu, Tengfei Liu, Yiguang Hong, Zhong-Ping Jiang

This paper presents a novel feasible-set reshaping technique to optimization-based control with ensured constraint qualification. In our problem setting, the feasible set of admissible control inputs depends on the real-time state of the plant, and the linear independence constraint qualification (LICQ) may not be satisfied in some regions of interest. By fe

Wei Wang, Siqi Wu

Wang and Zhao (Adv. Appl. Math. 173 (2026) 102994) generalized the classic Johnson-Newman theorem on simultaneous similarity of symmetric matrices from a single rank-one perturbation to multiple rank-one perturbations. However, their result applies only to specific rank-one perturbations, and the given condition is quite involved as it relies on multivariate

Miriam Horovicz

LLM agents that use external tools can solve complex tasks, but understanding which tools actually contributed to a response remains a blind spot. No existing XAI methods address tool-level explanations. We introduce AgentSHAP, the first framework for explaining tool importance in LLM agents. AgentSHAP is model-agnostic: it treats the agent as a black box an

Kei Yoshitake, Kento Hosono, Ken Kobayashi, Kazuhide Nakata

In this paper, we propose a method for generating layouts for image-based advertisements by leveraging a Vision-Language Model (VLM). Conventional advertisement layout techniques have predominantly relied on saliency mapping to detect salient regions within a background image, but such approaches often fail to fully account for the image's detailed compositi

Karthikeya KV

This research introduces a transformative framework for integrating Vision-Enhanced Large Language Models (LLMs) with advanced transformer-based architectures to tackle challenges in high-resolution image synthesis and multimodal data interpretation. The proposed model incorporates a rectified flow mechanism that connects noise and data with linear paths, en

Luoxi Meng, Henry Feng, Ilia Shumailov, Earlence Fernandes

Browser-using agents (BUAs) are an emerging class of AI agents that interact with web browsers in human-like ways, including clicking, scrolling, filling forms, and navigating across pages. While these agents help automate repetitive online tasks, they are vulnerable to prompt injection attacks that trick an agent into performing undesired actions, such as l

Tom Lee, Sihoon Lee, Seonghun Kim

Large Language Models (LLMs) challenge the validity of traditional open-ended assessments by blurring the lines of authorship. While recent research has focused on the accuracy of automated scoring (AES), these static approaches fail to capture process evidence or verify genuine student understanding. This paper introduces a novel Human-AI Collaboration fram

Timofei Izhitskii

This paper examines linear binary codes capable of correcting one or more errors. For the single-error-correcting case, it is shown that the Hamming bound is achieved by a constructive method, and an exact expression for the minimal codeword length is derived. For the general case, a simple lower bound for the parameters of linear codes is derived from an an

Indiwara Nanayakkara, Dehan Jayawickrama, Mervyn Parakrama B. Ekanayake

Wire harness inspection process remains a labor-intensive process prone to errors in the modern Electronics Manufacturing Services (EMS) industry. This paper introduces a semiautomated machine vision system capable of verifying correct wire positioning, correctness of the connector polarity and correctness of color sequences for both linear and circular wire

Andre Nies, Philipp Schlicht

Let $\Aut(G)$ denote the group of (bi-)continuous automorphisms %and $\Out(G)$ the outer automorphism group of a non-Archimedean Polish group~$G$. We show that for any such $G$ with an invariant countable basis of open subgroups, the group $\Aut(G)$ carries a unique Polish topology that makes its natural action on $G$ continuous. Furthermore, for any class o

Li-Xuan Zhao, Chen-Yang Xu, Wen-Qiang Li, Bo Wang

In recent years, there has been a notable increase in the use of supervised detection methods of major depressive disorder (MDD) based on electroencephalogram (EEG) signals. However, the process of labeling MDD remains challenging. As a self-supervised learning method, contrastive learning could address the shortcomings of supervised learning methods, which

Jie Guo, Shuyuan Yang, Jinchuan Hou, Xiaofei Qi

Multipartite entanglement underpins quantum technologies but its study is limited by the lack of universal measures, unified frameworks, and the intractability of convex-roof extensions. We establish an axiomatic framework and introduce the first \emph{true} $k$-entanglement measure, $E_w^{(k,n)}$, which satisfies all axioms, establishes $k$-entanglement as

Runqiao Fu, Shun Tang

In this paper, we prove a Lefschetz-Riemann-Roch theorem for singular projective schemes which admit diagonalisable group scheme actions, this result generalizes P. Baum, W. Fulton and G. Quart's Lefschetz-Riemann-Roch theorem for singular varieties (cf. \cite{BFQ}) to general scheme case.

Lixin Chen, Chaomeng Chen, Jiale Zhou, Zhijian Wu

Despite the rapid progress of deep learning in video action recognition (VAR) in recent years, privacy leakage in videos remains a critical concern. Current state-of-the-art privacy-preserving methods often rely on anonymization. These methods suffer from (1) low concealment, where producing visually distorted videos that attract attackers' attention during

Xuechun Liu, Heli Sun, Xuecheng Wu, Ruichen Cao

Multivariate time series anomaly detection (MTSAD) aims to accurately identify and localize complex abnormal patterns in the large-scale industrial control systems. While existing approaches excel in recognizing the distinct patterns under the low-dimensional scenarios, they often fail to robustly capture long-range spatiotemporal dependencies when learning

Zesen Huang, Marco Velli, Yuliang Ding

We present a solitary Alfv\'en wave model that exhibits nontrivial three-dimensional twisting of open magnetic field lines while preserving constant $|B|$. Embedded rotational discontinuities sharply deflect the otherwise uniform field lines, producing localized, large-amplitude field reversals in one-dimensional profiles that closely resemble the ``switchba

Haochen Yuan, Yang Zhang, Xiang He, Quan Z. Sheng

With the rise of cloud-edge collaboration, recommendation services are increasingly trained in distributed environments. Federated Recommendation (FR) enables such multi-end collaborative training while preserving privacy by sharing model parameters instead of raw data. However, the large number of parameters, primarily due to the massive item embeddings, si

Zi-Fan Zhu, Yu Su, Yao Wang, Rui-Xue Xu

In this work, we develop an extended dissipaton theory that generalizes the environmental couplings beyond the conventional linear and quadratic forms, enabling the treatment of arbitrary order of bath couplings. Applying this theoretical framework to the condensed-phase non-Condon spectroscopy, we demonstrate the interplay of anharmonicity, non-Condon and s

Safwan Shaheer, G. M. Refatul Islam, Mohammad Rafid Hamid, Md. Abrar Faiaz Khan

Prompt injection attacks can compromise the security and stability of critical systems, from infrastructure to large web applications. This work curates and augments a prompt injection dataset based on the HackAPrompt Playground Submissions corpus and trains several classifiers, including LSTM, feed forward neural networks, Random Forest, and Naive Bayes, to

Kexin Chen, Jianwei Huang, Yuan Luo

Generative Artificial Intelligence (GenAI) enables digital representatives to make decisions on behalf of team members in collaborative tasks, but faces challenges in accurately representing preferences. While supplying GenAI with detailed personal information improves representation fidelity, feasibility constraints make complete information access impracti