October 2025 arXiv papers — page 219

Kai Yang, Yuqi Huang, Junheng Tao, Wanyu Wang

Modeling 3D dynamics is a fundamental problem in multi-body systems across scientific and engineering domains and has important practical implications in object trajectory prediction and simulation. While recent GNN-based approaches have achieved strong performance by enforcing geometric symmetries, encoding high-order features or incorporating neural-ODE me

Amin Ahmadi Kasani, Hedieh Sajedi

Convolutional neural networks are continually evolving, with some efforts aimed at improving accuracy, others at increasing speed, and some at enhancing accessibility. Improving accessibility broadens the application of neural networks across a wider range of tasks, including the detection of eye diseases. Early diagnosis of eye diseases and consulting an op

Stefan Dirnstorfer

The problem of image registration is finding a transformation that aligns two images, such that the corresponding points are in the same location. This paper introduces a simple, end-to-end trainable algorithm that is implementable in a few lines of Python code. The approach is shown to work with very little training data and training time, while achieving a

Guijin Son, Donghun Yang, Hitesh Laxmichand Patel, Amit Agarwal

Recent frontier models employ long chain-of-thought reasoning to explore solution spaces in context and achieve stonger performance. While many works study distillation to build smaller yet capable models, most focus on English and little is known about language-specific reasoning. To bridge this gap, we first introduct **Language-Mixed CoT**, a reasoning sc

Rikuo Sasaki, Michimasa Inaba

Recent advancements in AI have highlighted its application in captology, the field of using computers as persuasive technologies. We hypothesized that the "conformity effect," where individuals align with others' actions, also occurs with AI agents. This study verifies this hypothesis by introducing a "Persuadee Agent" that is persuaded alongside a human par

Xiaoshu Zhao, Xu Yuan, Hongxia Zheng, Huajin Chen

We propose a framework for designing coherent optical environments that enable versatile and dynamic optical manipulation. In contrast to conventional material-based near-field platforms, our approach employs a structured coherent light field -- optimized via a back-propagation-based inverse design algorithm -- as the manipulation environment. This light-bas

Shengjiao Ji, Yujin Zhang, Zihan Zou, Bin Jiang

Accurate and efficient simulation of infrared (IR) and Raman spectra is essential for molecular identification and structural analysis. Traditional quantum chemistry methods based on the harmonic approximation neglect anharmonicity and nuclear quantum effects, while ab initio molecular dynamics (AIMD) remains computationally expensive. Here, we integrate our

Dustin Wright, Sarah Masud, Jared Moore, Srishti Yadav

Large language models (LLMs) are increasingly used as primary knowledge sources, yet their epistemic diversity - defined as the diversity of real-world claims in their outputs - has never been measured. Low epistemic diversity would pose a risk of knowledge collapse as homogeneous LLMs mediate a shrinking in the range of accessible information over time. The

Yikun Ji, Yan Hong, Bowen Deng, Jun Lan

The rapid growth of AI-generated imagery has blurred the boundary between real and synthetic content, raising practical concerns for digital integrity. Vision-language models (VLMs) can provide natural language explanations, but standard one-pass classifiers often miss subtle artifacts in high-quality synthetic images and offer limited grounding in the pixel

Zeyu Chen, Yu-Bo Liu, Fan Yang

The discovery of high-temperature superconductivity (HTSC) in La$_3$Ni$_2$O$_7$ has aroused significant interest in exploring the pairing mechanism. Previous studies have proposed an effective d$_{x^2-y^2}$-orbital bilayer $t-J_{\parallel}-J_{\perp}$ model, in which the electrons of the d$_{x^2-y^2}$ orbital are charge carriers, which are subject to the intr

Yan Peng, Guohua Liu

This paper investigates the bounds on the minimum orbital period for test objects around d-dimensional charged black holes in asymptotically flat spacetimes. We find numerically that the minimum orbital period decreases as the charge of the black hole increases. Thus, the upper limit is reached for an uncharged black hole, while the lower limit is attained f

Valeriia Liakh, Jack Jenkins

This review surveys recent advances in the numerical modeling of solar prominences and coronal rain achieved with the fully open-source adaptive-grid, parallelized Adaptive Mesh Refinement Versatile Advection Code (MPI-AMRVAC). We examine how these models have contributed to our understanding of the formation and evolution of cool plasma structures in the so

Vasiliy Volkov, Vladimir Stukopin

For affine special linear superalgebra $\widehat{sl}(m|n, \Pi)$ defined by an arbitrary system of simple roots $\Pi$ we define the affine super Yangian $Y_{\hbar}(\widehat{sl}(m|n, \Pi))$ as Hopf superalgebra which is a quantization of superbialgebra $\widehat{sl}(m|n, \Pi)[t]$ and describe super Yangian in terms of minimalistic system of generators. We cons

Lixuan He, Shikang Zheng, Linfeng Zhang

Autoregressive (AR) models have shown great promise in image generation, yet they face a fundamental inefficiency stemming from their core component: a vast, unstructured vocabulary of visual tokens. This conventional approach treats tokens as a flat vocabulary, disregarding the intrinsic structure of the token embedding space where proximity often correlate

Ritesh Chandra, Sonali Agarwal, Navjot Singh

Timely detection of critical health conditions remains a major challenge in public health analytics, especially in Big Data environments characterized by high volume, rapid velocity, and diverse variety of clinical data. This study presents an ontology-enabled real-time analytics framework that integrates Complex Event Processing (CEP) and Large Language Mod

Zhengjun Yue, Devendra Kayande, Zoran Cvetkovic, Erfan Loweimi

Large-scale end-to-end models such as Whisper have shown strong performance on diverse speech tasks, but their internal behavior on pathological speech remains poorly understood. Understanding how dysarthric speech is represented across layers is critical for building reliable and explainable clinical assessment tools. This study probes the Whisper-Medium mo

Jonathan K. Doyon, Sujin Kim, Alex D. Hwang, Jae-Hyun Jung

Homonymous hemianopia (HH) patients report difficulties in avoiding collisions with other pedestrians. We evaluated pedestrian collision detection and avoidance behaviors in HH patients and healthy controls using a novel virtual reality (VR) walking with pedestrians, which enables natural walking behavior in an empty real-world corridor while viewing an imme

Chenlu Ding, Jiancan Wu, Leheng Sheng, Fan Zhang

Multimodal large language models (MLLMs) have demonstrated remarkable capabilities across vision-language tasks, yet their large-scale deployment raises pressing concerns about memorized private data, outdated knowledge, and harmful content. Existing unlearning approaches for MLLMs typically adapt training-based strategies such as gradient ascent or preferen

Jihang Li, Bing Xu, Zulong Chen, Chuanfei Xu

Talent search is a cornerstone of modern recruitment systems, yet existing approaches often struggle to capture nuanced job-specific preferences, model recruiter behavior at a fine-grained level, and mitigate noise from subjective human judgments. We present a novel framework that enhances talent search effectiveness and delivers substantial business value t

Robert Barish, Hoi Ping Luk, Min Yan

Congruent polygons are congruent in angles as well as in edge lengths. We concentrate on the angle aspect, and investigate how tilings of the sphere by congruent pentagons can be determined by the angle information only. We also investigate how the features of tilings are changed under reductions, i.e., by ignoring the difference among the angles.

Jing Jiang, Siying Wu

Let g be a exceptional complex simple Lie algebra and q be a parabolic subalgebra. A generalized Verma module M is called a scalar generalized Verma module if it is induced from a one-dimensional representation of q. In this paper, we will determine the first diagonal-reducible point of scalar generalized Verma modules associated to minimal parabolic subalge

Xia Zeng, Yihan Chen, Luhui Liu, Chao Luo

We deploy large language models (LLMs) as business development (BD) agents for persuasive price negotiation in online travel agencies (OTAs). The agent must follow a multi-stage Standard Operating Procedure (SOP) and strict guardrails (no over-promising and no hallucinations), while remaining human-like and effective over long, multi-turn dialogues. We propo

Ze Li, Ming Cheng, Ming Li

Large-scale self-supervised Pre-Trained Models (PTMs) have shown significant improvements in the speaker verification (SV) task by providing rich feature representations. In this paper, we utilize w2v-BERT 2.0, a model with approximately 600 million parameters trained on 4.5 million hours of unlabeled data across 143 languages, for the SV task. The MFA struc

Yuan Tan, Youjun Lu, Kunyu Song

Recently, the \textit{EHT} collaboration unveiled the shadow images of the supermassive black hole (SMBH) M87* and Sgr A*, with angular radii of $42\pm3$\,$\mu$as and $48.7\pm7.0$\,$\mu$as, respectively. These observations are consistent with the shadow of a Kerr black hole in general relativity (GR). Observations of the shadow of SMBHs can be used to test m

Shyamalee Bora, Dhruba Jyoti Gogoi, Pralay Kumar Karmakar

We explore the modified thermodynamics of a Hayward-Anti de Sitter (H-AdS) black hole in atypical conditions, incorporating a string fluid, Hayward regularisation, and quantum entropy corrections. Our analysis reveals a first-order phase transition between small and large black hole phases, characterised by a swallowtail behaviour in the Gibbs free energy pr

João Barata

We discuss recent advances in applying Quantum Information Science to problems in high-energy nuclear physics. After outlining key developments, open challenges, and emerging connections between these disciplines, we highlight recent results on the study of matter states, hard probes, and spin correlations using novel quantum technologies. This work summariz

Hanchen Zhang, Xiao Liu, Bowen Lv, Xueqiao Sun

Recent advances in large language models (LLMs) have sparked growing interest in building generalist agents that can learn through online interactions. However, applying reinforcement learning (RL) to train LLM agents in multi-turn, multi-task settings remains challenging due to lack of scalable infrastructure and stable training algorithms. In this work, we

Di Zhang

Kolmogorov-Arnold Networks (KANs) have emerged as a promising alternative to traditional Multi-Layer Perceptrons (MLPs), offering enhanced interpretability and a solid mathematical foundation. However, their parameter efficiency remains a significant challenge for practical deployment. This paper introduces PolyKAN, a novel theoretical framework for KAN comp

Zhengyang Tang, Zihan Ye, Chenyu Huang, Xuhan Huang

Large Reasoning Models (LRMs) have demonstrated strong capabilities in complex multi-step reasoning, opening new opportunities for automating optimization modeling. However, existing domain adaptation methods, originally designed for earlier instruction-tuned models, often fail to exploit the advanced reasoning patterns of modern LRMs -- In particular, we sh

Aayushya Agarwal, Larry Pileggi, Gauri Joshi

Hyperparameter selection is critical for stable and efficient convergence of heterogeneous federated learning, where clients differ in computational capabilities, and data distributions are non-IID. Tuning hyperparameters is a manual and computationally expensive process as the hyperparameter space grows combinatorially with the number of clients. To address

Haiquan Qiu, You Wu, Yingjie Tan, Yaqing Wang

Loss explosions in training deep neural networks can nullify multi-million dollar training runs. Conventional monitoring metrics like weight and gradient norms are often lagging and ambiguous predictors, as their values vary dramatically across different models and even between layers of the same model, making it difficult to establish a unified standard for

Moo Hyun Son, Jintaek Oh, Sun Bin Mun, Jaechul Roh

While text-to-image (T2I) models can synthesize high-quality images, their performance degrades significantly when prompted with novel or out-of-distribution (OOD) entities due to inherent knowledge cutoffs. We introduce World-To-Image, a novel framework that bridges this gap by empowering T2I generation with agent-driven world knowledge. We design an agent

Kamila Kowalska, Enrico Maria Sessolo

In the context of entanglement in relativistic $2\to 2$ scattering described by a perturbative $S$-matrix, we derive analytically the concurrence for a mixed final state of two qubits corresponding to a discrete quantum number of the scattered particles. The qubit density matrix is obtained by tracing the momentum degrees of freedom out of the full density m

Danielle Witt

Given the norms of powers $(\lVert x^n\rVert)_{n\geq 0}$ of a Banach algebra element $x$, the largest possible value of the minimum modulus on the spectrum of $x$ is determined. It is also shown that, given a Banach algebra element $x$ and a compact set $K\subset\mathbb{C}$ with maximum modulus no more than the spectral radius of $x$, there exists a Banach a

Palwinder Singh, Jasleen Kaur Jagde, Megha Jain, Edith Yeung

Localized quantum emitters in transition-metal dichalcogenides (TMDs) have recently emerged as solid-state candidates for on-demand sources of single photons. Due to the role of strain in the site-selective creation of TMD emitters, their hybrid integration into photonic structures such as cavities and waveguides is possible using pick-and-place methods. Her

R. M. Strässle, S. A. Hosseini, I. V. Karlin

A consistent kinetic modeling and discretization strategy for compressible flows across all Prandtl numbers and specific heat ratios is developed using the quasi-equilibrium approach within two of the most widely used double-distribution frameworks. The methodology ensures accurate recovery of the Navier-Stokes-Fourier equations, including all macroscopic mo

Yizhuo Ding, Mingkang Chen, Qiuhua Liu, Fenghua Weng

Large Multimodal Reasoning Models (LMRMs) are moving into real applications, where they must be both useful and safe. Safety is especially challenging in multimodal settings: images and text can be combined to bypass guardrails, and single objective training can cause policy drift that yields over-refusal on benign inputs or unsafe compliance on risky ones.

Lekshmi Thulasidharan, Elena D'Onghia, Robert Benjamin

We investigate the origin of warps in stellar disks using high-resolution Milky Way analogs from the IllustrisTNG50 simulation. Focusing on galaxies that experienced a major merger, we identify a characteristic azimuthal misalignment between the warp structures of stellar populations formed before and after the merger. This misalignment persists even after c

Simon F. Martina-Perez

High-resolution imaging provides dense trajectories of migrating cells, flocking animals, and synthetic active particles, from which interaction laws can be determined with a wide variety of methods. Yet, distinguishing whether front-back or lateral biases seen in such data reflect intrinsic anisotropy in the interaction kernel or emergent correlations that

Rabiya Khalid, Evangelos Pournaras

The growing electricity demand and use of smart appliances are placing pressure on power grids, making efficient energy management more important than ever. The existing energy management systems often prioritize system efficiency (balanced energy demand and supply) at the expense of consumer comfort. This paper addresses this gap by proposing a novel decent

Mikel Artola, Ismael Ayuso, Ruth Lazkoz, Vincenzo Salzano

Recent observations from the Dark Energy Spectroscopic Instrument (DESI) raise doubts about the standard cosmological model, $\Lambda$CDM, suggesting a preference for an inherently dynamical dark energy component. The Chevallier-Polarski-Linder (CPL) parametrization -- a widely used two-parameter model for the dark energy equation of state -- displays marked

Jian-jie Zheng, Chih-kai Yang, Po-han Chen, Lyn Chao-ling Chen

In the study, the social robot act as a patrol to recognize and notify illegal parking in real-time. Dual-model pipeline method and large multimodal model were compared, and the GPT-4o multimodal model was adopted in license plate recognition without preprocessing. For moving smoothly on a flat ground, the robot navigated in a simulated parking lot in the ex

Prashansa Panda, Shalabh Bhatnagar

Recent studies have increasingly focused on non-asymptotic convergence analyses for actor-critic (AC) algorithms. One such effort introduced a two-timescale critic-actor algorithm for the discounted cost setting using a tabular representation, where the usual roles of the actor and critic are reversed. However, only asymptotic convergence was established the

Shikang Zheng, Guantao Chen, Qinming Zhou, Yuqi Lin

Diffusion Transformers offer state-of-the-art fidelity in image and video synthesis, but their iterative sampling process remains a major bottleneck due to the high cost of transformer forward passes at each timestep. To mitigate this, feature caching has emerged as a training-free acceleration technique that reuses or forecasts hidden representations. Howev

Hagen Holthusen, Ellen Kuhl

We propose a complement to constitutive modeling that augments neural networks with material principles to capture anisotropy and inelasticity at finite strains. The key element is a dual potential that governs dissipation, consistently incorporates anisotropy, and-unlike conventional convex formulations-satisfies the dissipation inequality without requiring

Zhijun Liu, Jiang Hu, Zhidong Bai, Zhihui Lv

In this paper, we establish the Central Limit Theorem (CLT) for linear spectral statistics (LSSs) of large-dimensional generalized spiked sample covariance matrices, where the spiked eigenvalues may be either bounded or diverge to infinity. Building upon this theorem, we derive the asymptotic distributions of linear hypothesis test statistics under the gener

Meng Zhu, Jianting Dong, Xinlu Li, Jiahao Shentu

A vital goal in spintronics is the efficient electrical generation of spin currents, a pursuit that has recently focused on using antiferromagnets (AFMs) as spin current sources. It has been demonstrated that antiferromagnets with broken PT symmetry (parity + time reversal) can efficiently generate longitudinal and transverse spin currents. At the same time,

Lucas Pacheco, Torsten Braun, Kaushik Chowdhury, Denis Rosário

Beamforming techniques use massive antenna arrays to formulate narrow Line-of-Sight signal sectors to address the increased signal attenuation in millimeter Wave (mmWave). However, traditional sector selection schemes involve extensive searches for the highest signal-strength sector, introducing extra latency and communication overhead. This paper introduces

Wengao Ye, Yan Liang, Lianlei Shan

Recent advancements in Large Language Models (LLMs) have shifted from explicit Chain-of-Thought (CoT) reasoning to more efficient latent reasoning, where intermediate thoughts are represented as vectors rather than text. However, latent reasoning can be brittle on challenging, out-of-distribution tasks where robust reasoning is most critical. To overcome the

B. Sartayev, A. Ydyrys

In this paper, we study free algebras in subvarieties of the variety of associative algebras singled out by Mal'cev's classification. For each subvariety, we construct the bases for the corresponding free algebras and describe the space of symmetric polynomials they contain.

Léa Pistorius, Namrata U. Nayar, Phillip Tran, Sammy Elmariah

Transcatheter valve repair presents significant challenges due to the mechanical limitations and steep learning curve associated with manual catheter systems. This paper investigates the use of robotics to facilitate transcatheter procedures in the context of mitral valve edge-to-edge repair. The complex handle-based control of a clinical repair device is re

Thaís Maria Dalbelo, Danilo da Nóbrega Santos

In this paper, we establish conditions for a family $\{f_t\}$ of functions, with not necessarily isolated singularities, defined on a toric variety so that the associated family of hypersurfaces $\{f_t^{-1}(0)\}$ is Whitney equisingular. We work in the setting of toric varieties with arbitrary singular sets. This extends previous results by Eyral and Oka con

Yan Li, Teng Hui, Xinhui Zhang, Zihan Cao

Due to the ever-rising global incidence rate of inflammatory bowel disease (IBD) and the lack of effective clinical treatment drugs, elucidating the detailed pathogenesis, seeking novel targets, and developing promising drugs are the top priority for IBD treatment. Here, we demonstrate that the levels of microRNA (miR)-103a were significantly downregulated i

Jakub Frac, Alexander Schmatz, Qiang Li, Guido Van Wingen

Functional magnetic resonance imaging (fMRI) analysis faces significant challenges due to limited dataset sizes and domain variability between studies. Traditional self-supervised learning methods inspired by computer vision often rely on positive and negative sample pairs, which can be problematic for neuroimaging data where defining appropriate contrasts i

Zbigniew Kozioł

A Chain of Springs and Masses (CSM) model is used in the interpretation of molecular dynamics (MD) simulations of movement of atoms in orientated FCC crystals. A force of dynamic origin is found that is perpendicular to the direction of the external shear pressure. It is proportional to the square of the applied pressure; It causes breaking of axial symmetry

Piyush Arora, Navlika Singh, Vasubhya Diwan, Pratik Mazumder

Neural networks have revolutionized numerous fields, yet they remain vulnerable to a critical flaw: the tendency to learn implicit biases, spurious correlations between certain attributes and target labels in training data. These biases are often more prevalent and easier to learn, causing models to rely on superficial patterns rather than task-relevant feat

Soufiane Amini, Yassine Benajiba, Cesare Bernardis, Paul Cayet

The proliferation of agent frameworks has led to fragmentation in how agents are defined, executed, and evaluated. Existing systems differ in their abstractions, data flow semantics, and tool integrations, making it difficult to share or reproduce workflows. We introduce Open Agent Specification (Agent Spec), a declarative language that defines AI agents and

Lakshadeep Naik, Adam Fischer, Daniel Duberg, Danica Kragic

In Mobile Manipulation, selecting an optimal mobile base pose is essential for successful object grasping. Previous works have addressed this problem either through classical planning methods or by learning state-based policies. They assume access to reliable state information, such as the precise object poses and environment models. In this work, we study b

Longze Tan

The random feature method (RFM), a mesh-free machine learning-based framework, has emerged as a promising alternative for solving PDEs on complex domains. However, for large three-dimensional nonlinear problems, attaining high accuracy typically requires domain partitioning with many collocation points and random features per subdomain, which leads to extrem

Ji Zhang, Yiwei Li, Shaoxiong Feng, Peiwen Yuan

KV cache in autoregressive LLMs eliminates redundant recomputation but has emerged as the dominant memory and bandwidth bottleneck during inference, notably with long contexts and test-time scaling. KV quantization is a key lever for reducing cache cost, but accuracy drops sharply as the native KV distribution lacks flatness and thus maintains a wide quantiz

Shao-Qin Zhang

Due to the existence of multiple stationary distributions, we study the stability and instability of a stationary distribution for distribution dependent stochastic differential equations. This note is devoted to the instability of a stationary distribution, and links the instability to a spectral property of the generator of the corresponding linearized sem

Amirmasoud Molaei, Mohammad Heravi, Reza Ghabcheloo

Rock capturing with standard excavator buckets is a challenging task typically requiring the expertise of skilled operators. Unlike soil digging, it involves manipulating large, irregular rocks in unstructured environments where complex contact interactions with granular material make model-based control impractical. Existing autonomous excavation methods fo

Dmitry Rybin, Yushun Zhang, Ding Tian, Zhihang Lin

We present Exact Causal Attention (ECA), a Strassen-style algorithm that computes exact Causal Attention using 10\% fewer operations. ECA improves a special class of matrix multiplications where either one operand or the output matrix is upper- or lower-triangular. This includes all matrix multiplication operations in the forward and backward pass of Causal

Marco Edoardo Palma, Pooja Rani, Harald C. Gall

Syntax highlighting is a critical feature in modern software development environments, enhancing code readability and developer productivity. However, delivering accurate highlighting in real time remains challenging for online and web-based development tools due to strict time and memory constraints on backend services. These systems must serve highlights r

Leonardo Modesto, Edoardo Rattu

We carefully investigate, extend, and shed new light on the McVittie exact solution of Einstein's gravity (EG) with the focus on the implications in the Universe we live in. It turns out that the only known exact solution of EG, which interpolates between an asymptotic homogeneous and isotropic Universe and a Schwarzschild black hole, is actually singular in

Yu-Chuan Yu, Chi Ho Yuen

White's conjecture asserts that any two tuples of matroid bases that have the same multi-set union can be transformed from one to another by symmetric exchanges; it also implies that the toric ideals of matroids are generated by the binomials encoding these exchanges. We prove White's conjecture for the class of paving matroids. Our strategy is to generalize

Aviv Navon, Aviv Shamsian, Neta Glazer, Yael Segal-Feldman

Diffusion and flow-based non-autoregressive (NAR) models have shown strong promise in large language modeling, however, their potential for automatic speech recognition (ASR) remains largely unexplored. We propose Drax, a discrete flow matching framework for ASR that enables efficient parallel decoding. To better align training with inference, we construct a

Longrui Yang, Yiyu Wang, Jingfan Tang, Yunpeng Lv

This paper considers the path planning problem for autonomous exploration of an unknown environment using multiple heterogeneous robots such as drones, wheeled, and legged robots, which have different capabilities to traverse complex terrains. A key challenge there is to intelligently allocate the robots to the unknown areas to be explored and determine the

Mohammad Kazzazi, Mohammad Morsali, Rouhollah Amiri

This paper presents CLEAR -- a closed-form localization estimator with a reduced sensor network. The proposed method is a computationally efficient, two-stage estimator that fuses time-difference-of-arrival (TDOA) and frequency-difference-of-arrival (FDOA) measurements with a minimal number of sensors. CLEAR localizes a moving source in N-dimensional space u

Minki Hhan, Tomoyuki Morimae, Yasuaki Okinaka, Takashi Yamakawa

With the rapid advances in quantum computer architectures and the emerging prospect of large-scale quantum memory, it is becoming essential to classically verify that remote devices genuinely allocate the promised quantum memory with specified number of qubits and coherence time. In this paper, we introduce a new concept, proofs of quantum memory (PoQM). A P

Emad Jacob Maroun

Instruction density and encoding efficiency are some of the few things directly affected by an instruction set architecture's design. In contrast, a processor's implementation often significantly influences performance, power efficiency, and area usage. Therefore, a major goal of instruction set design should be maximizing instruction density and encoding ef

Frank Calegari, Vesselin Dimitrov, Yunqing Tang

In this paper, we explore several threads arising from our recent joint work on arithmetic holonomy bounds, which were originally devised to prove new irrationality results based on the method of Ap\'ery limits. We propose a new method to address effective Diophantine approximation on the projective line and the multiplicative group. This method, and all our

Sourav Bhattacharya

We obtain a complete characterization of \emph{topologically exact patterns} on \emph{triods}. Based on their \emph{rotation number} $\rho$, these \emph{exact patterns} are grouped into three classes: \emph{slow} ($\rho < \frac{1}{3}$), \emph{fast} ($\rho > \frac{1}{3}$) and \emph{ternary} ($\rho = \frac{1}{3}$). For each category, we derive a \emph{linear o

Nicolas Kossowski, Christina Kyrou, Rémi Colom, Pierre-Marie Coulon

Metasurfaces provide a compact, flexible, and reliable solution for controlling the wavefront of light. In imaging systems, micro-lens arrays are integrated with pixel matrices to reduce optical crosstalk, enhance photon collection efficiency, and improve spatial resolution. However, as the aperture size of the photonic devices decreases, fundamental limitat

Haoqi Wu, Wei Dai, Ming Xu, Li Wang

Diffusion Models have gained significant popularity due to their remarkable capabilities in image generation, albeit at the cost of intensive computation requirement. Meanwhile, despite their widespread deployment in inference services such as Midjourney, concerns about the potential leakage of sensitive information in uploaded user prompts have arisen. Exis

Tomasz Cieślak, Sebastian Owczarek, Karolina Wielgos

In this article, we study a thermodynamically consistent thermo-visco-elastic model describing the balance of internal energy in a heat-conducting inelastic body. In the considered problem, the temperature dependence appears in both the elastic and inelastic constitutive relations. For such a system, we introduce the concept of a weak entropy-energy solution

Haonan Qu, Qiulan Zhao

We study the local bihamiltonian structures of the asymmetric rational reductions of the 2D-Toda hierarchy (RR2T) of types $(2,1)$ and $(1,2)$ at the full-dispersive level, and construct a three-dimensional generalized Frobenius manifold with non-flat unity associated with the $(2,1)$-type. Furthermore, we explicitly relate the $(2,1)$-type RR2T to the bi-gr

Liam Hanany, Alexander Lubotzky

Let $s_n^\mathrm{ch}(\Gamma)$ denote the number of characteristic subgroups of index at most $n$ in a finitely generated group $\Gamma$. In response to a question of I. Rivin we show that if $\Gamma = F_r$ is the free group on $r \geq 2$ generators then the growth type of $s_n^{\mathrm{ch}}(F_r)$ is $n^{\mathrm{log}(n)}$. This is in contrast with the expecta

Pritam Bhattacharyya, Nikolay A. Bogdanov, Liviu Hozoi

A magnet is a collection of magnetic moments. How those interact is determined by what lies in between. In transition-metal and rare-earth magnetic compounds, the configuration of the ligands around each magnetic center and the connectivity of the ligand cages are therefore pivotal -- for example, the mutual interaction of magnetic species connected through

Xuezhu Liu, Ming Lu, Haiwen Liu

We theoretically explore a non-Hermitian superfluid model with complex-valued interaction, inspired by two-body loss stemming from inelastic scattering observed in ultracold atomic experiments. Utilizing both the right-eigenstate-based mean-field theory and its biorthogonal counterpart, we study the properties of the system. Notably, the right-eigenstate-bas

Yifeng Gao, Ziang Ji, Yuxuan Wang, Biqing Qi

Diffusion-based Large Language Models (dLLMs) have emerged as a competitive alternative to autoregressive models, offering unique advantages through bidirectional attention and parallel generation paradigms. However, the generation results of current parallel decoding methods deviate from stepwise decoding, introducing potential performance degradation, whic

Minseo Kim, Coleman Hooper, Aditya Tomar, Chenfeng Xu

Large Language Models (LLMs) have achieved state-of-the-art performance on a broad range of Natural Language Processing (NLP) tasks, including document processing and code generation. Autoregressive Language Models (ARMs), which generate tokens sequentially conditioned on all previous tokens, have been the predominant paradigm for LLMs. While these models ha

Chenxin Wang, Elyas Asadi Shamsabadi, Zhaohui Chen, Luming Shen

Conventional construction safety inspection methods are often inefficient as they require navigating through large volume of information. Recent advances in large vision-language models (LVLMs) provide opportunities to automate safety inspections through enhanced visual and linguistic understanding. However, existing applications face limitations including i

Konstantinos Kitsios, Francesco Sovrano, Earl T. Barr, Alberto Bacchelli

Semantic code clone detection is the task of detecting whether two snippets of code implement the same functionality (e.g., Sort Array). Recently, many neural models achieved near-perfect performance on this task. These models seek to make inferences based on their training data. Consequently, they better detect clones similar to those they have seen during

Xiaoyu Yang, En Yu, Wei Duan, Jie Lu

This paper identifies a critical yet underexplored challenge in reasoning alignment from multiple multi-modal large language models (MLLMs): In non-stationary environments, the diverse reasoning distributions of source models often evolve unpredictably, transmitting systematic biases and drift to the target model. To address this, we formulate multi-source r

Mayank Ravishankara, Varindra V. Persad Maharaj

This survey paper chronicles the evolution of evaluation in multimodal artificial intelligence (AI), framing it as a progression of increasingly sophisticated "cognitive examinations." We argue that the field is undergoing a paradigm shift, moving from simple recognition tasks that test "what" a model sees, to complex reasoning benchmarks that probe "why" an

Tim Bakkenes, Daniel Wang, Anton Johansson

The rise of Large Language Models has not been inclusive of all cultures. The models are mostly trained on English texts and culture which makes them underperform in other languages and cultural contexts. By developing a generalizable method for preparing culturally relevant datasets and post-training the Gemma 2 model, this project aimed to increase the per

Muhao Guo, Haoran Li, Yang Weng

Neural ordinary differential equations (NODE) have garnered significant attention for their design of continuous-depth neural networks and the ability to learn data/feature dynamics. However, for high-dimensional systems, estimating dynamics requires extensive calculations and suffers from high truncation errors for the ODE solvers. To address the issue, one

Matteo Gallone, Beatrice Langella

We study the dynamics of a quantum many-body lattice system with a local Hamiltonian subjected to a quasi-periodic driving with finite regularity. For sufficiently large driving frequencies, we prove that the system remains in a prethermal state for times growing polynomially with the frequency, and we show the optimality of this bound by constructing an exp

Umberto Cappellazzo, Minsu Kim, Pingchuan Ma, Honglie Chen

Large language models (LLMs) have recently shown strong potential in audio-visual speech recognition (AVSR), but their high computational demands and sensitivity to token granularity limit their practicality in resource-constrained settings. Token compression methods can reduce inference cost, but they require fixing a compression rate in advance and produce

Jingzhi Gong, Yixin Bian, Luis de la Cal, Giovanni Pinna

Coding agents powered by LLMs face critical sustainability and scalability challenges in industrial deployment, with single runs consuming over 100k tokens and incurring environmental costs that may exceed optimization benefits. This paper introduces GA4GC, the first framework to systematically optimize coding agent runtime (greener agent) and code performan

Yiming Niu, Jinliang Deng, Yongxin Tong

Periodicity is a fundamental characteristic of time series data and has long played a central role in forecasting. Recent deep learning methods strengthen the exploitation of periodicity by treating patches as basic tokens, thereby improving predictive effectiveness. However, their efficiency remains a bottleneck due to large parameter counts and heavy compu

Muhao Guo, Yang Weng

Neural ODEs (NODEs) have emerged as powerful tools for modeling time series data, offering the flexibility to adapt to varying input scales and capture complex dynamics. However, they face significant challenges: first, their reliance on time-domain representations often limits their ability to capture long-term dependencies and periodic structures; second,

Peigui Qi, Kunsheng Tang, Wenbo Zhou, Weiming Zhang

Text-to-image models have shown remarkable capabilities in generating high-quality images from natural language descriptions. However, these models are highly vulnerable to adversarial prompts, which can bypass safety measures and produce harmful content. Despite various defensive strategies, achieving robustness against attacks while maintaining practical u

Nicolau Sobrosa, Miguel Gonçalves, Bruno Amorim, Eduardo V. Castro

We demonstrate that a mean field approximation can be confidently employed in quasiperiodic moir\'e systems to treat interactions and quasiperiodicity on equal footing. We obtain the mean field phase diagram for an illustrative one-dimensional moir\'e system that exhibits narrow bands and a regime with non-interacting multifractal critical states. By systema

Keren Wang, Kaili Sun, Jing Du, Peijuan Dai

Nonlocal metasurfaces provide a compact route to generating momentum-space optical vortices but are limited by steep dispersion typically associated with high-quality (Q) factor resonances, resulting in narrowband and inefficient operation. Here, we introduce a reflection-type nonlocal metasurface that hybrid-couples a bound state in the continuum (BIC) with

Mahdi Cherakhloo, Arash Abbasi, Mohammad Saeid Sarafraz, Bijan Vosoughi Vahdat

Large Language Models (LLMs) have demonstrated remarkable capabilities across numerous languages; however, their effectiveness in low-resource languages like Persian requires thorough investigation. This paper presents a comprehensive benchmark of several open-source LLMs for Persian Natural Language Processing (NLP) tasks, utilizing both zero-shot and few-s

Yang Chen, Yitao Liang, Zhouchen Lin

In Transformers, Position Embeddings (PEs) significantly influence Length Generalization (LG) performance, yet their fundamental role remains unclear. In this work, we investigate the limitations and capabilities of PEs in achieving LG. We theoretically analyze PEs in Position-Only Linear Attentions (POLAs), introducing Linear Representation Complexity (LRC)

Charles-Edouard Bréhier, Ibrahima Faye

We prove the averaging principle for a class of stochastic systems. The slow component is solution to a fractional differential equation, which is coupled with a fast component considered as solution to an ergodic stochastic differential equation driven by a standard Brownian motion. We establish the convergence of the slow component when the time-scale sepa

Dmitrii Troitskii, Koyena Pal, Chris Wendler, Callum Stuart McDougall

Prior work has shown that a significant driver of performance in reasoning models is their ability to reason and self-correct. A distinctive marker in these reasoning traces is the token wait, which often signals reasoning behavior such as backtracking. Despite being such a complex behavior, little is understood of exactly why models do or do not decide to r