November 2025 arXiv papers — page 78

Weiping Yu, Ye Jiarui, He Mengke, Junfeng Liu

Large language models (LLMs) rely on Key-Value (KV) cache to reduce time-to-first-token (TTFT) latency, but existing disk-based KV cache systems using file-per-object layouts suffer from severe scalability bottlenecks due to file system metadata overhead, I/O inefficiency, and poor spatial locality. This paper presents SGLANG-LSM, a database-inspired system

Ziqiang Li, Jiazhen Yan, Fan Wang, Kai Zeng

The rapid advancement of generative models has made real and synthetic images increasingly indistinguishable. Although extensive efforts have been devoted to detecting AI-generated images, out-of-distribution generalization remains a persistent challenge. We trace this weakness to spurious shortcuts exploited during training and we also observe that small fe

Shujie Yang, Xuzhe Zhao, Yuqi Zhang, Yansong Tang

Metamaterials, known for their ability to manipulate light at subwavelength scales, face significant design challenges due to their complex and sophisticated structures. Consequently, deep learning has emerged as a powerful tool to streamline their design process. Reconfigurable multi-state metamaterials (RMMs) with adjustable parameters can switch their opt

Taejun Kim, Youngbo Aram Shim, Geehyuk Lee

Beyond a simple notification of incoming calls or messages, more complex information such as alphabets and digits can be delivered through spatiotemporal tactile patterns (STPs) on a wrist-worn tactile display (WTD) with multiple tactors. However, owing to the limited skin area and spatial acuity of the wrist, frequent confusions occur between closely locate

Paula Joy B. Martinez, Jose Marie Antonio Miñoza, Sebastian C. Ibañez

Emotion recognition from social media is critical for understanding public sentiment, but accessing training data has become prohibitively expensive due to escalating API costs and platform restrictions. We introduce an interpretability-guided framework where Shapley Additive Explanations (SHAP) provide principled guidance for LLM-based synthetic data genera

Xuan-Quang Phan, Tan-Ha Mai, Thai-Duy Dinh, Minh-Thuan Nguyen

Interacting with relational databases remains challenging for users across different expertise levels, particularly when composing complex analytical queries or performing administrative tasks. Existing systems typically address either natural language querying or narrow aspects of database administration, lacking a unified and intelligent interface for gene

Iolanda Navone, Maria Giovanna Dainotti, Elisa Fazzari, Giovanni Montani

We study a family of cosmological models featuring dynamical dark energy (DE), based on the idea that the creation of its constituents arises from the gravitational field of the expanding universe, whose non-equilibrium physics is described by a non-zero bulk viscosity coefficient. We consider the complete scenario, in which both matter creation and bulk vis

Patrizia Pucci, Jianjun Zhang, Xuexiu Zhong

This paper extends the uniqueness results of Serrin and Tang [\textit{Indiana Univ. Math. J.}, 49 (2000), pp. 897--923] to the low-dimensional case $1\leq N\leq m$ with $m>1$. We consider radial solutions of the overdetermined problem \[ \begin{cases} -\Delta_m u = f(u), \quad u>0 & \text{in } B_R,\\[4pt] u = \partial_\nu u = 0 & \text{on } \partial B_R, \te

Zhongxian Cao

Let $(M^5,g)$ be a five-dimensional non-trivial simply-connected compact quasi-Einstein manifold with boundary. If $M$ has constant scalar $R$, Johnatan Costa, Ernani Ribeiro Jr, and Detang Zhou show that $R$ = $((m-5)k+20)/(m-k+4)\lambda$ for some $k\in\{0,2,3,4\}$. Both cases of $k=0$ and $k=4$ are already classified. In this paper we will prove that the c

Livia Betz

We derive necessary conditions for locally optimal shapes of a design problem governed by a non-smooth PDE. The main particularity of the state system is the lack of differentiability of the nonlinearity. We work in the framework of the functional variational approach (FVA), which has the capacity to transfer geometric optimization problems into optimal cont

Ryo Aihara, Yoshiki Masuyama, Francesco Paissan, François G. Germain

Neural audio codecs (NACs) provide compact representations that can be leveraged in many downstream applications, in particular large language models. Yet most NACs encode mixtures of multiple sources in an entangled manner, which may impede efficient downstream processing in applications that need access to only a subset of the sources (e.g., analysis of a

D. V. Vinnikov, O. M. Ozerov, V. V. Katrechko, V. I. Tkachov

An experimental study of the parameters of electromagnetic and X-ray radiation was carried out on a small-sized direct-acting electron accelerator with inductive storage and a plasma opening switch. Frequency spectra have been determined, and the diagnostics of the power of the microwave component of the spectrum has been tested. An analytical comparison of

Liyuan Deng, Yunpeng Bai, Yongkang Dai, Xiaoshui Huang

Parametric Computer-Aided Design (CAD) is crucial in industrial applications, yet existing approaches often struggle to generate long sequence parametric commands due to complex CAD models' geometric and topological constraints. To address this challenge, we propose MamTiff-CAD, a novel CAD parametric command sequences generation framework that leverages a T

Chenyang Wu, Jiayi Fu, Chun-Le Guo, Shuhao Han

Due to large pixel movement and high computational cost, estimating the motion of high-resolution frames is challenging. Thus, most flow-based Video Frame Interpolation (VFI) methods first predict bidirectional flows at low resolution and then use high-magnification upsampling (e.g., bilinear) to obtain the high-resolution ones. However, this kind of upsampl

Linyi Han, Shidong Pan, Zhenchang Xing, Sofonias Yitagesu

Textual Vulnerability Descriptions (TVDs) are crucial for security analysts to understand and address software vulnerabilities. However, the key aspect inconsistencies in TVDs from different repositories pose challenges for achieving a comprehensive understanding of vulnerabilities. Existing approaches aim to mitigate inconsistencies by aligning TVDs with ex

Qing Zhang, Bing Xu, Xudong Zhang, Yifan Shi

The remarkable performance of Large Language Models (LLMs) highly relies on crafted prompts. However, manual prompt engineering is a laborious process, creating a core bottleneck for practical application of LLMs. This phenomenon has led to the emergence of a new research area known as Automatic Prompt Optimization (APO), which develops rapidly in recent yea

Haruo Minami

Consider the quotient $G/B$ of a simple matrix Lie group $G$ by a subgroup $B$ isomorphic to a direct product of some of $S^1$s and $S^3$s such that its adjoint representation can be extended over $G$. Then it naturally inherits a stable framing from a twisted left invariant framing $\mathscr{L}^\alpha$ of $G$ where $\alpha$ is the realization of a complex r

Shuo Zhai, Jian-Min Wang, Yan-Rong Li, Wei-Jian Guo

JWST/NIRCam observations reveal dense stellar cores in high-redshift galactic bulges, indicative of sustained star formation and potential stellar accretion. We introduce accretion-modified star (AMS) as a new component in the chemical evolution of high-redshift bulges hosting active galactic nuclei (AGNs). The gas-phase chemical evolution of bulge environme

Susmita Das, Anupam Bhardwaj, Marcella Marconi

Type II Cepheids are a class of pulsating variable stars that play a critical role in our understanding of stellar evolution, distance measurement and tracing the structure and kinematics of old stars in nearby galaxies. This review provides a comprehensive summary of the current state of research on Type II Cepheids, including their observed properties, pul

Tianxiong Zhong, Xingye Tian, Xuebo Wang, Boyuan Jiang

Existing latent diffusion models typically couple scale with content complexity, using more latent tokens to represent higher-resolution images or higher-frame rate videos. However, the latent capacity required to represent visual data primarily depends on content complexity, with scale serving only as an upper bound. Motivated by this observation, we propos

A. Chilingarian, L. Hovhannisyan, B. Sargsyan, M. Zazyan

This work presents a unified conceptual and observational framework that reinterprets these radiation bursts as manifestations of the same runaway processes happening at different atmospheric depths (Dual-stage model, DSM). We review recent results from satellite (ASIM), aircraft (ALOFT), balloon (HELEN), and ground-based (SEVAN and KANAZAWA) experiments to

Tan-Dat Khuu, Trung-Hieu Huynh, Hoang-Hung Vo

In this paper we study Liouville-type properties for a class of degenerate elliptic equations driven by the fractional infinity Laplacian with nonlinear lower-order terms, \[ \Delta_\infty^{\beta}u - c\,H(u,\nabla u) - \lambda\, f(|x|,u)=0 \qquad \text{in }\mathbb{R}^n, \] where $\beta\in[0,2]$, $\Delta_\infty^\beta$ denotes the fractional infinity Laplace o

Zhongwei Jin, Keyi Chen, Qiuyu Ren, Zhigang Dai

Breaking the diffraction limit in optical imaging is crucial for resolving subwavelength details in a wide range of applications, where superoscillatory imaging and subtraction imaging are two common strategies for surpassing conventional resolution limits. We propose an end-to-end deep learning framework that integrates superoscillatory focusing and subtrac

Rui Sang, Yuxuan Liu

Voice cloning technology poses significant privacy threats by enabling unauthorized speech synthesis from limited audio samples. Existing defenses based on imperceptible adversarial perturbations are vulnerable to common audio preprocessing such as denoising and compression. We propose SceneGuard, a training-time voice protection method that applies scene-co

Hongfu Lou

Predicting the binding affinity of protein protein complexes directly from sequence remains a challenging problem, particularly in the absence of reliable structural information. Here I present ProtT Affinity, a sequence only model that combines ProtT5 embeddings with a lightweight Transformer architecture. The model is trained and evaluated on homology filt

Taeho Kang, Jaeyeon Park, Kyungjin Lee, Youngki Lee

Existing 4D Gaussian Splatting (4DGS) methods struggle to accurately reconstruct dynamic scenes, often failing to resolve ambiguous pixel correspondences and inadequate densification in dynamic regions. We address these issues by introducing a novel method composed of two key components: (1) Elliptical Error Clustering and Error Correcting Splat Addition tha

Yijun Yang, Lichao Wang, Jianping Zhang, Chi Harold Liu

The growing misuse of Vision-Language Models (VLMs) has led providers to deploy multiple safeguards, including alignment tuning, system prompts, and content moderation. However, the real-world robustness of these defenses against adversarial attacks remains underexplored. We introduce Multi-Faceted Attack (MFA), a framework that systematically exposes genera

Tony J. Puthenpurakal

Let $(A,\mathfrak{m})$ be a Cohen-Macaulay local ring with residue field $k$. If $M$ is a finitely generated $A$-module then set $\text{curv}(M) = \limsup_n\sqrt[n]{\beta_n^A(M)}$. We show that under mild hypotheses the existence of a single module $M$ with $1 \leq \text{curv}(M) < \text{curv}(k)$ imposes obstructions to both $\text{curv}(k)$ and $\text{curv

Shiyi Cao, Dacheng Li, Fangzhou Zhao, Shuo Yuan

We introduce SkyRL-Agent, a framework for efficient, multi-turn, long-horizon agent training and evaluation. It provides efficient asynchronous dispatching, lightweight tool integration, and flexible backend interoperability, enabling seamless use with existing RL frameworks such as SkyRL-train, VeRL, and Tinker. Using SkyRL-Agent, we train SA-SWE-32B, a sof

Archish S, Ankit Garg, Kirankumar Shiragur, Neeraj Kayal

ColBERT introduced a late interaction mechanism that independently encodes queries and documents using BERT, and computes similarity via fine-grained interactions over token-level vector representations. This design enables expressive matching while allowing efficient computation of scores, as the multi-vector document representations could be pre-computed o

Yuanbo Tang, Yan Tang, Zixuan Zhang, Zihui Zhao

Trajectory generation has recently drawn growing interest in privacy-preserving urban mobility studies and location-based service applications. Although many studies have used deep learning or generative AI methods to model trajectories and have achieved promising results, the robustness and interpretability of such models are largely unexplored. This limits

Vladimir Danilov

A longer and more correct title is `a short and direct path to the theory of stable contract systems in a bipartite market'. There is no new meaningful results in the article. It is dedicated to the presentation of a short method for obtaining the main body of stability theory: existence, polarization, and latticing. The brevity and uniformity are achieved t

Avijit Sen Majumder, Sourav Bhattacharya

In this work, we compute the graviton mediated scattering amplitude of a massive spin-2 Fierz-Pauli field with various other massive spin fields, and in the non-relativistic limit, find out the corresponding two-body gravitational potentials. The massive spin-2 field does not represent gravity here. The theory of gravity is taken to be the usual massless gen

Bankitdor M Nongrum, Adarsha Kumar Jena

Measures of relative variability, such as the Pearson's coefficient of variation (CV$_p$), give much insight into the spread of lifetime distributions, like the Weibull distribution. The estimation of the Weibull CV$_p$ in modern statistics has traditionally been prioritized only when complete data is available. In this article, we estimate the Weibull CV$_p

Huseyin Goksu

Spectral Graph Neural Networks offer a principled approach to graph filtering but face a fundamental "Stability-vs-Adaptivity" trade-off. This trade-off is dictated by the choice of spectral domain. Filters in the finite [-1, 1] domain (e.g., ChebyNet) are numerically stable at high polynomial degrees (K) but are static and low-pass, causing them to fail on

Quan-Hoang Vuong, Viet-Phuong La, Minh-Hoang Nguyen

This study explores Bitcoin's value formation through the Granular Interaction Thinking Theory-Value Theory (GITT-VT). Rather than stemming from material utility or cash flows, Bitcoin's value arises from informational attributes and interactions of multiple factors, including cryptographic order, decentralization-enabled autonomy, trust embedded in the cons

Ken-ichi Kawarabayashi, Hirotaka Yoneda, Masataka Yoneda

We study the problem of online graph coloring for $k$-colorable graphs. The best previously known deterministic algorithm uses $\widetilde{O}(n^{1-\frac{1}{k!}})$ colors for general $k$ and $\widetilde{O}(n^{5/6})$ colors for $k = 4$, both given by Kierstead in 1998. In this paper, we finally break this barrier, achieving the first major improvement in nearl

Kun Cheng, Yurui Tang

The bipartite-hole-number of a graph $G$, denoted by $\widetilde{\alpha}(G)$, is the minimum number $k$ such that there exist positive integers $s$ and $t$ with $s+t=k+1$ with the property that for any two disjoint sets $A,B\subseteq V(G)$ with $|A|=s$ and $|B|=t$, there is an edge between $A$ and $B$. In this paper, we first prove that any $2$-connected gra

Jing-Cheng Chang, Yang He, Yu-Xiao Liu, Yuan Sun

In de Sitter (dS) holography, both the dS/CFT correspondence and the dS static patch holography have been extensively studied. In these two holographic frameworks, the dual field theories are defined on spacelike and timelike boundaries, respectively, where the inward motion of the holographic boundary into the bulk corresponds to the $T\bar{T}$ and $T\bar{T

Jun-Hui Yan, Yan Gong, Qi Xiong, Xuelei Chen

We forecast the constraint on the Hu-Sawicki $f(R)$ model from the photometric survey operated by the Chinese Space Station Survey Telescope (CSST). The simulated $3\times2$pt data of galaxy clustering, weak lensing, and galaxy-galaxy lensing measurements within 100 deg$^{2}$ are used in the analysis. The mock observational maps are constructed from a light

Anh D. Pham, Robert Moss, Wirichada Pan-ngum, Rose McGready

Introduction: In Thailand, Hepatitis B is still endemic despite a strong program to eliminate the disease. A higher prevalence is reported in the border region and among migrants due to physical, financial and cultural barriers. Policies and programs targeting the border region and migrant communities have been suggested. Models can be used to understand and

Manfred Einsiedler, Dmitry Kleinbock, Anurag Rao

Let $X = G/\Gamma$ be a quotient of a real Lie group by a non-uniform lattice. Consider a one-parameter subgroup $F$ of $G$ that is $\operatorname{Ad}$-diagonalizable over $\mathbb{C}$ and whose action on $(X,m_X)$ is mixing. In this dynamical system we study the set of points $x \in X$ with a precompact orbit, written as $E(F,\infty)$, which is known to be

Michael Dinitz, Jeremy T. Fineman, Seeun William Umboh

This paper considers using predictions in the context of the online Joint Replenishment Problem with Deadlines (JRP-D). Prior work includes asymptotically optimal competitive ratios of $O(1)$ for the clairvoyant setting and $O(\sqrt{n})$ of the nonclairvoyant setting, where $n$ is the number of items. The goal of this paper is to significantly reduce the com

Xinyuan Liao, Shaowei Chen, Shuai Zhao

Accurate and efficient thermal dynamics models of permanent magnet synchronous motors are vital to efficient thermal management strategies. Physics-informed methods combine model-based and data-driven methods, offering greater flexibility than model-based methods and superior explainability compared to data-driven methods. Nonetheless, there are still challe

Xing Hu, Raula Gaikovina Kula, Christoph Treude

Generative Artificial Intelligence (GenAI) models are achieving remarkable performance in various tasks, including code generation, testing, code review, and program repair. The ability to increase the level of abstraction away from writing code has the potential to change the Human-AI interaction within the integrated development environment (IDE). To explo

Renxiang Xiao, Wei Liu, Yuanfan Zhang, Yushuai Chen

We present Rad-GS, a 4D radar-camera SLAM system designed for kilometer-scale outdoor environments, utilizing 3D Gaussian as a differentiable spatial representation. Rad-GS combines the advantages of raw radar point cloud with Doppler information and geometrically enhanced point cloud to guide dynamic object masking in synchronized images, thereby alleviatin

Haohui Chen, Zhiyong Chen, Aoxiang Liu, Wentuo Fang

Deterministic policy gradient algorithms for continuous control suffer from value estimation biases that degrade performance. While double critics reduce such biases, the exploration potential of double actors remains underexplored. Building on temporal-difference error-driven regularization (TDDR), a double actor-critic framework, this work introduces enhan

Cangtao Yin, Meenu Upadhyay, Markus Meuwly

The dynamics and spectroscopy of the small (H$_2$COO) and large (CH$_3$CHOO) Criegee intermediates (CIs) in the gas phase, inside/on water droplets, on amorphous solid water (ASW) and in bulk water are investigated using validated energy functions. For both species, facile diffusion between surface and inside positions for water droplets are found whereas on

Vu Van Than

Traditional threat modeling remains reactive-focused on known TTPs and past incident data, while threat prediction and forecasting frameworks are often disconnected from operational or architectural artifacts. This creates a fundamental weakness: the most serious cyber threats often do not arise from what is known, but from what is assumed, overlooked, or no

Vincent Fan, Regina Barzilay

The performance of machine learning models in drug discovery is highly dependent on the quality and consistency of the underlying training data. Due to limitations in dataset sizes, many models are trained by aggregating bioactivity data from diverse sources, including public databases such as ChEMBL. However, this approach often introduces significant noise

Shoichi Murakami, Shunsuke Hiraoka, Toshiki Kobayashi, Takashi Yamamoto

We demonstrate channel-selective frequency up-conversion from telecom wavelengths around 1540 nm for optical fiber communication to visible wavelengths around 780 nm, based on second-order optical nonlinearity in a cavity of the converted modes. In our experiment, we selectively convert a light from any frequency mode within frequency-multiplexed telecom sig

Colin J. Burke, Zachary Stone, Yue Shen, Yan-Fei Jiang

Several local ($z\lesssim 0.2$) metal-poor dwarf AGNs have remarkably similar properties to those of high-redshift Little Red Dots (LRDs), and are recently proposed to be local analogs of LRDs. We use long-term photometric and spectroscopic observations of three local LRDs spanning $\sim 20$ years to measure variability in their rest-frame optical continuum

Pratik Mishra, Caner Gözübüyük, Seema Nagar, Prateeti Mohapatra

Manual creation of IT monitoring dashboard widgets is slow, error-prone, and a barrier for both novice and expert users. We present NOVAID, an interactive chatbot that leverages Large Language Models (LLMs) to generate IT monitoring widgets directly from natural language queries. Unlike general natural language-to-visualization tools, NOVAID addresses IT ope

Huseyin Goksu

Spectral GNNs, like ChebyNet, are limited by heterophily and over-smoothing due to their static, low-pass filter design. This work investigates the "Adaptive Orthogonal Polynomial Filter" (AOPF) class as a solution. We introduce two models operating in the [-1, 1] domain: 1) `L-JacobiNet`, the adaptive generalization of `ChebyNet` with learnable alpha, beta

Sungbin Moon, Jiho Park, Suyoung Hwang, Donghyun Koh

Modern data processing workflows frequently encounter ragged data: collections with variable-length elements that arise naturally in domains like natural language processing, scientific measurements, and autonomous AI agents. Existing workflow engines lack native support for tracking the shapes and dependencies inherent to ragged data, forcing users to manag

Mitsuo Higaki

We establish the first quantitative Runge approximation theorem, with explicit $L^2$-estimates, for the 3d nonstationary Stokes system on a bounded spatial domain. This result addresses the two primary limitations of the qualitative result [H.-Sueur, 2025] obtained in collaboration with Franck Sueur: first, it bypasses the non-constructive Hahn-Banach theore

Shuo Huang, Ryohei Kawabe, Hideki Umehata, Kotaro Kohno

Bar structures are present in about half of local disk galaxies and play pivotal roles in secular galaxy evolution. Bars impose a non-axisymmetric perturbation to the rotating disk and transport gas inward to feed central starburst and, possibly, the activity of the nuclear supermassive black hole. They are believed to be long-lived structures and are now id

Zishan Xu, Yifu Guo, Yuquan Lu, Fengyu Yang

Traditional video reasoning segmentation methods rely on supervised fine-tuning, which limits generalization to out-of-distribution scenarios and lacks explicit reasoning. To address this, we propose \textbf{VideoSeg-R1}, the first framework to introduce reinforcement learning into video reasoning segmentation. It adopts a decoupled architecture that formula

Lingyi Kong, Jiangpei Dou, Wei Guo, Mingming Xu

This paper optimizes the design and analysis of the mechanical test platform for the scientific probe module of the Cool Planet Imaging Coronagraph, which is the fifth part of the China Space Station survey Telescope. First, according to the module layout and economic requirements, the preliminary structural design of the module mechanical test platform is c

Hrikshesh Kumar, Anika Garg, Anshul Gupta, Yashika Agarwal

Old cloud edge workload resource management is too reactive. The problem with relying on static thresholds is that we are either overspending for more resources than needed or have reduced performance because of their lack. This is why we work on proactive solutions. A framework developed for it stops reacting to the problems but starts expecting them. We de

Masayuki Hayakawa, Biplab Bhattacherjee, Hidekazu Kuwayama, Tatsuo Shibata

Chirality organize living and active matter systems into striking collective states, yet the principles that govern chiral ordering in open systems, where elements are continuously added or removed, remain unclear. A mutant strain of Dictyostelium discoideum deficient in chemotaxis (KI cells) forms centimeter-scale, clockwise multi-armed spirals. Each arm is

Shreyansh Jain, Madhav Singhvi, Shreya Rahul Jain, Pranav S

Most of the traditional Applicant Tracking Systems (ATS) depend on strict matching using keywords, where candidates that are highly qualified are many times disqualified because of minor semantic differences. In this article, the two-stage process of developing a more comprehensive resume assessment system based on a small language model that is trained with

Sandro Andric

Mechanistic interpretability aspires to reverse-engineer neural networks into explicit algorithms, while model editing seeks to modify specific behaviours without retraining. Both areas are typically evaluated with informal evidence and ad-hoc experiments, with few explicit guarantees about how far an extracted or edited model can drift from the original on

Sébastien Bubeck, Christian Coester, Ronen Eldan, Timothy Gowers

AI models like GPT-5 are an increasingly valuable tool for scientists, but many remain unaware of the capabilities of frontier AI. We present a collection of short case studies in which GPT-5 produced new, concrete steps in ongoing research across mathematics, physics, astronomy, computer science, biology, and materials science. In these examples, the author

Rashid Rafeek V Valappil, Sayan Maity, Varadharajan Srinivasan

High-pressure phases of the hybrid perovskite MAPbBr3 have been investigated in detail using a novel machine learning force field (MLFF). MLFF simulations successfully reproduce the sequence of pressure-induced phase transitions from the $\alpha$ ($Pm\bar{3}m$) to the $\beta$ ($Im\bar{3}$) and finally the $\gamma$ ($Pnma$/$Pmn2_1$) phase. In the $\alpha$ pha

Fang Feng, Yuanyi Sun, Yue Yu

In recent studies \cite{ZZ24, FY24}, the Interior Penalty Virtual Element Method (IPVEM) has been developed for solving a fourth-order singular perturbation problem, with uniform convergence established in the lowest-order case concerning the perturbation parameter. However, the resulting uniform convergence rate is only of half-order, which is suboptimal. I

Junchao Zhou, Junkang Liu, Fanhua Shang

Federated Learning with Low-Rank Adaptation (LoRA) faces three critical challenges under client heterogeneity: (1) Initialization-Induced Instability due to random initialization misaligning client subspaces; (2) Rank Incompatibility and Aggregation Error when averaging LoRA parameters of different ranks, which biases the global model; and (3) exacerbated Cl

Jilong Shi, Qiangpeng Fang, Xiaobin Rui, Jian Zhang

Adversarial Influence Blocking Maximization (AIBM) aims to select a set of positive seed nodes that propagate synchronously with the known negative seed nodes to counteract their negative influence. Time factor plays a particularly vital role for many AIBM application scenarios. However, the AIBM problem with time constraint remains unexplored. More importan

Huan Lin, Dakai Liu, Lianghui Ding, Lin Wang

Unmanned aerial vehicle (UAV) swarms encounter the challenge of high overhead due to both network management and formation control requirements. In this paper, we propose a Bio-inspired Integrated Networking and Control (BINC) scheme, enabling efficient formation management for swarms comprising thousands of UAVs. The scheme forms a two-layer hierarchical st

Shreyan Banerjee, Aasifa Rounak, Vikram Pakrashi

Application of neuromorphic edge devices for control is limited by the constraints on gradient-free online learning and scalability of the hardware across control problems. This paper introduces a synaptic Q-learning algorithm for the control of the classical Cartpole, where the Bellman equations are incorporated at the synaptic level. This formulation enabl

Cheol-Yeon Cheon, Volodymyr Multian, Kenji Watanabe, Takashi Taniguchi

Two-dimensional antiferromagnetism has long attracted significant interest in many areas of condensed matter physics, but only recently has experimental exploration become feasible due to the isolation of van der Waals antiferromagnetic monolayers. Probing the magnetic phase diagram of these monolayers remains however challenging because established experime

John Lott

We look at smooth manifolds equipped with a possibly singular Riemannian metric. We give sufficient conditions for the existence of scalar curvature measures and Dirac operators.

Jason Gerard, Juan A. Fraire, Sandra Céspedes

Free-space optical inter-satellite links (OISLs) enable high-capacity space communications but require precise Pointing, Acquisition, and Tracking (PAT) between links. Current scheduling approaches often overlook or oversimplify PAT delays, leading to inefficient contact planning and overestimated network capacities. We present a validated model for quantify

Andrew J. Groszek, Charles W. Woffinden, Michael D. Harvey, Andrew G. White

All modern wireless communication technologies are based on electromagnetism. However, electromagnetic signals are susceptible to screening and blocking, so their availability cannot be guaranteed in adverse environments. This raises a fundamental question: Can information be transmitted through a truly unblockable channel? Here we show that gravity, unlike

Alex Ning, Vainateya Rangaraju

Structured pruning removes entire neurons or channels, but its effectiveness depends on how importance is distributed across the representation space. Change-of-basis (CoB) pruning addresses this challenge by applying orthogonal linear transformations that concentrate importance within certain dimensions. However, many standard deep learning architectures ar

Nicholas J. Pritchard, Andreas Wicenec, Richard Dodson, Mohammed Bennamoun

Imminent radio telescope observatories provide massive data rates making deep learning based processing appealing while simultaneously demanding real-time performance at low-energy; prohibiting the use of many artificial neural network based approaches. We begin tackling the scientifically existential challenge of Radio Frequency Interference (RFI) detection

Qi-Jun Hong, Qing Chen, Ligen Wang, Dallin Fisher

We present an extension of the SLUSCHI package (Solid and Liquid in Ultra Small Coexistence with Hovering Interfaces) to enable automated diffusion calculations from first-principles molecular dynamics. While the original SLUSCHI workflow was designed for melting temperature estimation via solid-liquid coexistence, we adapt its input and output handling to i

Yuxin Dong, Hezi Lin, Weihao Zheng

In this paper, we study $(p,V)$-harmonic functions on complete Riemannian manifolds using the Moser iteration method. A volume comparison theorem and a Sobolev embedding theorem are established under the Bakry-$\acute{E}$mery curvature condition. Moreover, we obtain an explicit global gradient estimate for positive entire $(p,V)$-harmonic functions.

Stephen A. Giles, Hugh Shortt, Peter K. Liaw, Debasis Sengupta

Refractory high-entropy alloys (RHEAs) are compositionally complex materials which have been demonstrated to have the potential for exceptional strength at high operating temperatures. However, their composition space is vast, and other property requirements, such as acceptable plasticity at room-temperature, must be met. Here, we leverage recently published

Yukun Huang, Brett Gladman, Eiichiro Kokubo

Gravitational scattering of small bodies (planetesimals) by a planet remains a fundamental problem in celestial mechanics. It is traditionally modeled within the circular restricted three-body problem (CR3BP), where individual particle trajectories are obtained via numerical integrations. Here, we use {\"O}pik's close-encounter framework to study the random

Theophanis C. Stratopoulos, Victor Xiaoqi Wang

Recent advances in artificial intelligence, particularly generative AI (GenAI) and large language models (LLMs), are fundamentally transforming accounting research, creating both opportunities and competitive threats for scholars. This paper proposes a framework that classifies AI-accounting research along two dimensions: research focus (accounting-centric v

Gwen Yidou-Weng, Ian Li, Anji Liu, Oliver Broadrick

Controlled generation imposes sequence-level constraints (syntax, style, safety) that depend on future tokens, making exact conditioning of an autoregressive LM intractable. Tractable surrogates such as HMMs can approximate continuation distributions and steer decoding, but standard surrogates are often weakly context-aware. We propose Learning to Look Ahead

Yifei Huang, Pascal Jahan Elahi, Ugo Varetto, Kan He

Noisy intermediate-scale quantum (NISQ) devices impose dual challenges on quantum circuit execution: limited qubit connectivity requires extensive SWAP-gate routing, while time-dependent decoherence progressively degrades quantum information. Existing qubit mapping algorithms optimize for hardware topology and static calibration metrics but systematically ne

Takeru Tsunoori, Masato Kobayashi, Yuki Uranishi

Underwater robotic manipulation remains challenging because lighting variation, color attenuation, scattering, and reduced visibility can severely degrade visuomotor policies. We present Bi-AQUA, the first underwater bilateral control-based imitation learning framework for robot arms that explicitly models lighting within the policy. Bi-AQUA integrates trans

Pei Liu, Songtao Wang, Lang Zhang, Xingyue Peng

Synthesizing high-fidelity and controllable 4D LiDAR data is crucial for creating scalable simulation environments for autonomous driving. This task is inherently challenging due to the sensor's unique spherical geometry, the temporal sparsity of point clouds, and the complexity of dynamic scenes. To address these challenges, we present LiSTAR, a novel gener

Qing Zhang, Jing Huang, Mingyang Xu, Jun Rekimoto

While mainstream robotics pursues metric precision and flawless performance, this paper explores the creative potential of a deliberately "lo-fi" approach. We present the "Semantic Glitch," a soft flying robotic art installation whose physical form, a 3D pixel style cloud, is a "physical glitch" derived from digital archaeology. We detail a novel autonomous

Boxun Xu, Yu Wang, Zihu Wang, Peng Li

Visual autoregressive modeling (VAR) via next-scale prediction has emerged as a scalable image generation paradigm. While Key and Value (KV) caching in large language models (LLMs) has been extensively studied, next-scale prediction presents unique challenges, and KV caching design for next-scale based VAR transformers remains largely unexplored. A major bot

Mohan Shi, Xiong Xiao, Ruchao Fan, Shaoshi Ling

Joint automatic speech recognition (ASR) and speaker diarization aim to answer the question "who spoke what" in multi-speaker scenarios. In this paper, we present an end-to-end speech large language model (Speech-LLM) for Joint strEamable DIarization and aSr (JEDIS-LLM). The model is trained only on short audio under 20s but is capable of streamable inferenc

Jorge A. Huertas, Pascal Van Hentenryck

In serial batch (s-batch) scheduling, jobs from similar families are grouped into batches and processed sequentially to avoid repetitive setups that are required when processing consecutive jobs of different families. Despite its large success in scheduling, only three Constraint Programming (CP) models have been proposed for this problem considering minimum

Yiding Feng, Rad Niazadeh, Amin Saberi

In classic adversarial online resource allocation problems such as AdWords, customers arrive online while products are given offline with a fixed initial inventory. To ensure revenue guarantees under uncertainty, the decision maker must balance consumption across products. Based on this, the prevalent policy "inventory balancing (IB)" has proved to be optima

Peng Xia, Kaide Zeng, Jiaqi Liu, Can Qin

Large Language Model (LLM) Agents, often trained with Reinforcement Learning (RL), are constrained by a dependency on human-curated data, limiting scalability and tethering AI to human knowledge. Existing self-evolution frameworks offer an alternative but are typically restricted by the model's inherent capabilities and single-round interactions, hindering t

S. Koonkor, C. M. Baugh, G. Manzoni, D. Navarro-Gironés

We present a measurement of the $i$-band galaxy luminosity function from the present-day to $z = 2$, using over 1.1 million galaxies from the Physics of the Accelerating Universe Survey (PAUS). PAUS combines broad-band imaging from the Canada-France-Hawaii Telescope Lensing Survey with narrow-band photometry from PAUCam, enabling high-precision photometric r

Chengyue Wang, Wesley Pang, Xinrui Wu, Gregory Jun

General matrix multiplication (GEMM) is the computational backbone of modern AI workloads, and its efficiency is critically dependent on effective tiling strategies. Conventional approaches employ symmetric tile buffering, where the buffered tile size of the input $A$ along the dimension $M$ matches the output tile size of $C$. In this paper, we introduce as

Garritt L. Page, Andrés F. Barrientos, David B. Dahl, David B. Dunson

Bayesian clustering methods have the widely touted advantage of providing a probabilistic characterization of uncertainty in clustering through the posterior distribution. An amazing variety of priors and likelihoods have been proposed for clustering in a broad array of settings. There is also a rich literature on Markov chain Monte Carlo (MCMC) algorithms f

Eddie O'Sullivan, Henry Stone, Swati, Xiaolan Jin

In 1973, Swinnerton-Dyer completely classified all congruences for coefficients of normalized eigenforms in weights $k \in \{12, 16, 18, 20, 22, 26\}$ on $\Gamma_{0}(1) = \operatorname{SL}_{2}(\mathbb{Z})$ using the theory of modular Galois representations. In this paper, we classify congruences of Type I and Type II considered by Swinnerton-Dyer for the coe

Qing Zhang, Jing Huang, Yifei Huang, Jun Rekimoto

Current text-to-image models struggle to render the nuanced facial expressions required for compelling manga narratives, largely due to the ambiguity of language itself. To bridge this gap, we introduce an interactive system built on a novel, dual-hybrid pipeline. The first stage combines landmark-based auto-detection with a manual framing tool for robust, a

Qing Wang, Chong-Wah Ngo, Ee-Peng Lim, Qianru Sun

Training a model for food recognition is challenging because the training samples, which are typically crawled from the Internet, are visually different from the pictures captured by users in the free-living environment. In addition to this domain-shift problem, the real-world food datasets tend to be long-tailed distributed and some dishes of different cate

Isaac Pérez Castillo, Simon Leturcq, Sylvain Domitin, Ashley L. Nord

Fast and accurate 3D position detection in optical tweezers (OT) is essential for quantitatively monitoring subtle variations in the mechanical properties of microscopic systems ranging from biomolecules to cells and colloids. Because standard OT configurations do not provide direct access to the axial position, axial detection typically relies on temporal f

Kieron Kretschmar, Walter Laurito, Sharan Maiya, Samuel Marks

Prior work has introduced techniques for detecting when large language models (LLMs) lie, that is, generate statements they believe are false. However, these techniques are typically validated in narrow settings that do not capture the diverse lies LLMs can generate. We introduce LIARS' BENCH, a testbed consisting of 72,863 examples of lies and honest respon

Ashwin Poudel, Utsav Poudel, Dikshyanta Aryal, Anuj Nepal

The rapid growth of quantum computing poses a threat to the cryptographic foundations of digital systems, requiring the development of secure and scalable electronic voting (evoting) frameworks. We introduce a post-quantum-secure evoting architecture that integrates Falcon lattice-based digital signatures, biometric authentication via MobileNetV3 and AdaFace