November 2025 arXiv papers — page 134

Uday Bhaskar, Rishabh Bhattacharya, Avinash Patel, Sarthak Khoche

Foundation models, especially vision-language models (VLMs), offer compelling zero-shot object detection for applications like autonomous driving, a domain where manual labelling is prohibitively expensive. However, their detection latency and tendency to hallucinate predictions render them unsuitable for direct deployment. This work introduces a novel pipel

O. K. Khattab, M. D. Filipovic', Z. J. Smeaton, R. Z. E. Alsaberi

We present two new radio continuum images from the ASKAP POSSUM survey in the direction of the Small Magellanic Cloud. The two new source lists contain 36,571 radio continuum sources detected at 944 MHz and 15,227 sources at 1367 MHz, with beam sizes of approximately 14.5 by 12.2 arcsec and 8.7 by 8.2 arcsec, respectively. We used the Aegean software package

Pengqian Lu, Jie Lu, Anjin Liu, En Yu

Existing drift detection methods focus on designing sensitive test statistics. They treat the detection threshold as a fixed hyperparameter, set once to balance false alarms and late detections, and applied uniformly across all datasets and over time. However, maintaining model performance is the key objective from the perspective of machine learning, and we

Jasleen Birdi, Tamal Majumder, Debanjan Halder, Muskan Kularia

Inverse problems in imaging are typically ill-posed and are usually solved by employing regularized optimization techniques. The usage of appropriate constraints can restrict the solution space, thus making it feasible for a reconstruction algorithm to find a meaningful solution. In recent years, deep network based ideas aimed at learning the end-to-end mapp

Remmy Zen, Maximilian Nägele, Florian Marquardt

Quantum computing has the potential to solve problems that are intractable for classical computers, with possible applications in areas such as drug discovery and high-energy physics. However, the practical implementation of quantum computation is hindered by the complexity of executing quantum circuits on hardware. In particular, minimizing the number of T-

Monu Sharma

The incorporation of Artificial Intelligence (AI) into Enterprise Resource Planning (ERP) is a dramatic transition from static, on-premises systems to systems that can adapt and operate in cloud-native architectures. Cloud ERP solutions like Workday illustrate this evolution by incorporating machine learning, deep learning, and natural language processing in

Frida Brogren, Christoffer Olofsson, Joel Magnusson, Arkady Gonoskov

Particle-in-cell (PIC) codes are indispensable tools for studying plasma-based interactions across a wide range of physical regimes. The predictive capabilities of these codes have dramatically increased by continuous advances in algorithms and computing hardware. Nevertheless, greater computational performance and broader framework capabilities often lead t

Peilin He, Tananun Songdechakraiwut

Functional brain connectivity changes dynamically over time, making its representation challenging for learning on non-Euclidean data. We present a framework that encodes dynamic functional connectivity as an image representation of evolving network topology. Persistent graph homology summarizes global organization across scales, yielding Wasserstein distanc

Zhicheng Liao, Dongxu Wu, Zhenshan Shi, Sijie Mai

Recent efforts have repurposed the Contrastive Language-Image Pre-training (CLIP) model for No-Reference Image Quality Assessment (NR-IQA) by measuring the cosine similarity between the image embedding and textual prompts such as "a good photo" or "a bad photo." However, this semantic similarity overlooks a critical yet underexplored cue: the magnitude of th

Sha Zhao, Mingyi Peng, Haiteng Jiang, Tao Li

Scalable and generalizable analysis of brain activity is essential for advancing both clinical diagnostics and cognitive research. Electroencephalography (EEG), a non-invasive modality with high temporal resolution, has been widely used for brain states analysis. However, most existing EEG models are usually tailored for individual specific tasks, limiting t

Susan Eldhose, Bhargava Rama Chilukuri, Chandrasekharan Rajendran

Road traffic in developing countries, such as India, features a heterogeneous mix of vehicles operating under weak lane discipline (HWLD), encompassing both motorised and non-motorised modes with diverse sizes and manoeuvrability. These conditions lead to complex driver interactions, complicating the reliable identification of vehicle-following (VF) behaviou

Amol Upadhye, Yin Li

Recent cosmological bounds on the sum of neutrino masses, M_nu = sum m_nu, are in tension with laboratory oscillation experiments, making cosmological tests of neutrino free-streaming imperative. In order to study the scale-dependent clustering of massive neutrinos, we develop a fast linear response method, FAST-nu f, applicable to neutrinos and other non-re

Zhiyuan Xu, Nan Min, Yuhang Guo, Tong Wei

3D Gaussian Splatting offers a strong speed-quality trade-off but struggles to reconstruct semi-transparent surfaces because most methods assume a single depth per pixel, which fails when multiple surfaces are visible. We propose TSPE-GS (Transparent Surface Probabilistic Extraction for Gaussian Splatting), which uniformly samples transmittance to model a pi

Bimal Gaudel, Robert G. Adam, Ajay Melekamburath, Conner Masteran

SeQuant is an open-source library for symbolic algebra of tensors over commutative (scalar) and non-commutative (operator) rings. The key innovation supporting most of its functionality is a graph-theoretic tensor network (TN) canonicalizer that can handle tensor networks with symmetries faster than their standard group-theoretic counterparts. The TN canonic

Mustafa Munir, Md Mostafijur Rahman, Radu Marculescu

Vision Graph Neural Networks (ViGs) offer a new direction for advancements in vision architectures. While powerful, ViGs often face substantial computational challenges stemming from their graph construction phase, which can hinder their efficiency. To address this issue we propose AdaptViG, an efficient and powerful hybrid Vision GNN that introduces a novel

Min Jeong Cho, Hyeong Seok Shim, Sungyu Kim, Jae Sung Lee

Accurate attenuation and scatter corrections are crucial in positron emission tomography (PET) imaging for accurate visual interpretation and quantitative analysis. Traditional methods relying on computed tomography (CT) or magnetic resonance imaging (MRI) have limitations in accuracy, radiation exposure, and applicability. Deep neural networks provide poten

Ruyu Liu, Shaohua Pan

This paper concerns a class of constrained optimization problems in which, the objective and constraint functions are both upper-$\mathcal{C}^2$. For such nonconvex and nonsmooth optimization problems, we develop an inexact moving balls approximation (MBA) method by a workable inexactness criterion for the solving of subproblems. By leveraging a global error

Zekai Shi, Zhixi Cai, Kalin Stefanov

Typically, children start to learn their first words between 6 and 9 months, linking spoken utterances to their visual referents. Without prior knowledge, a word encountered for the first time can be interpreted in countless ways; it might refer to any of the objects in the environment, their components, or attributes. Using longitudinal, egocentric, and eco

Yu Gan, Hirad Alipanah, Jinglei Cheng, Zeguan Wu

Quantum computers have long been expected to efficiently solve complex classical differential equations. Most digital, fault-tolerant approaches use Carleman linearization to map nonlinear systems to linear ones and then apply quantum linear-system solvers. However, provable speedups typically require digital truncation and full fault tolerance, rendering su

Hengkai Wu, Taishi Nishihara, Mioko Hizukuri, Takeshi Tanaka

Spectrally selective absorbers that maximize solar absorption and minimize thermal radiation loss are crucial for efficient solar thermal energy harvesting. However, limitations imposed by the intrinsic properties of conventional materials hinder the fabrication of interference-based absorbers with desired optical properties. Herein, a high-performance solar

Jiayu Wan

Let {\Gamma} be a finitely generated group and consider the set of all characters of representations of {\Gamma} into SL2(C). This set, denoted by X({\Gamma}), admits an algebraic structure and is called the character variety of {\Gamma}. When {\Gamma} is the fundamental group of a hyperbolic 3-manifold M, X({\Gamma}) turns out to be a powerful tool in the s

Hao Zheng, Longxiang Wang, Yun Xu, Qiang Wang

The growth of cloud computing drives data centers toward higher density and efficiency. Data processing units (DPUs) enhance server network and storage performance but face challenges such as long hardware upgrade cycles and limited resources. To address these, we propose Taiji, a resource-elasticity architecture for DPUs. Combining hybrid virtualization wit

Canwen Wang, Jionghao Lin, Kenneth R. Koedinger

Knowledge Components (KCs) are foundational to adaptive learning systems, but their manual identification by domain experts is a significant bottleneck. While Large Language Models (LLMs) offer a promising avenue for automating this process, prior research has been limited to small datasets and has been shown to produce superfluous, redundant KC labels. This

David X. Lin, Giannis Fikioris, Siddhartha Banerjee, Éva Tardos

A canonical setting for non-monetary online resource allocation is one where agents compete over multiple rounds for a single item per round, with i.i.d. valuations and additive utilities across rounds. With $n$ symmetric agents, a natural benchmark for each agent is the utility realized by her favorite $1/n$-fraction of rounds; a line of work has demonstrat

Yuhang Zhou, Yanxiang Zhao, Zhongyun Hua, Zhipu Liu

Person re-identification (ReID) is a fundamental task in many real-world applications such as pedestrian trajectory tracking. However, advanced deep learning-based ReID models are highly susceptible to adversarial attacks, where imperceptible perturbations to pedestrian images can cause entirely incorrect predictions, posing significant security threats. Alt

Martín A. Díaz-Viera, Arturo Ortiz-Tapia, Joaquín R. Hernández-Pérez, Gladys Castorena-Cortés

A general 3D flow-and-transport model in porous media is derived using an axiomatic continuum-mechanics approach and implemented with the finite element method to simulate microbial enhanced oil recovery (MEOR) at core scale under laboratory conditions. The development pipeline (conceptual -> mathematical -> numerical -> computational) is detailed. The model

Mohammed Abboodi

The computer service industry has expanded rapidly over the past two decades, driven by the proliferation of computing technologies, the entry of large firms, and the availability of online diagnostic and troubleshooting tools. In this increasingly competitive environment, many small and medium sized enterprises struggle to maintain customer satisfaction as

Hanwen Wang

The generalization ability of visuomotor policy is crucial, as a good policy should be deployable across diverse scenarios. Some methods can collect large amounts of trajectory augmentation data to train more generalizable imitation learning policies, aimed at handling the random placement of objects on the scene's horizontal plane. However, the data generat

Hongda Qiu

Let $(M,g)$ be a $C^\infty$-smooth, $n$-dimensional Riemannian manifold which is diffeomorphic to $\RR^n$ and admit an action of a properly discontinuous and cocompact group. This work proves the existence of a $C^\infty$ equivariant isometric embedding of $M$ in some Euclidean space $\RR^q$ where $q = \max\{s_n+2n, s_n+n+5\}$ is the same as the dimension of

Masanori Hino

We establish a general analytic framework for determining the AF-martingale dimension of diffusion processes associated with strongly local regular Dirichlet forms on metric measure spaces. While previous approaches typically relied on self-similarity, our argument is based instead on purely analytic balance conditions between energy measures and relative ca

Zhentian Zhang, David Morales-Jimenez, Hao Jiang, Christos Masouros

Fluid antenna systems (FASs) offer genuine simplicity for communication network design by eliminating expensive hardware overhead and reducing the complexity of access protocol architectures. Through the discovery of significant spatial diversity within a compact antenna space, FASs enable the implementation of reconfigurable-antenna-based architectures. How

Xinke Lyu, Alex Mayer, Grace McLaughlin, Amy Gladfelter

mRNA-protein assemblies play a fundamental role in forming membraneless compartments within cells, whose functions may include activating, inhibiting, and localizing reactions. Recruitment of proteins into droplets can diminish cell to cell variability in protein abundance. However, the extent to which mRNA-protein assemblies may also buffer noise arising fr

Yuan Zhao, Hualei Zhu, Tingyu Jiang, Shen Li

Graphical User Interface (GUI) task automation constitutes a critical frontier in artificial intelligence research. While effective GUI agents synergistically integrate planning and grounding capabilities, current methodologies exhibit two fundamental limitations: (1) insufficient exploitation of cross-model synergies, and (2) over-reliance on synthetic data

Xuan Rao, Simian Xu, Zheng Li, Bo Zhao

Recent advances have shown that sequential fine-tuning (SeqFT) of pre-trained vision transformers (ViTs), followed by classifier refinement using approximate distributions of class features, can be an effective strategy for class-incremental learning (CIL). However, this approach is susceptible to distribution drift, caused by the sequential optimization of

Minrui Luo, Weihang Xu, Xiang Gao, Maryam Fazel

Gradient descent dynamics on the deep matrix factorization problem is extensively studied as a simplified theoretical model for deep neural networks. Although the convergence theory for two-layer matrix factorization is well-established, no global convergence guarantee for general deep matrix factorization under random initialization has been established to

Hu Zhang, Zhien Dai, Zhaohui Tang, Yongfang Xie

Time series forecasting is essential across diverse domains. While MLP-based methods have gained attention for achieving Transformer-comparable performance with fewer parameters and better robustness, they face critical limitations including loss of weak seasonal signals, capacity constraints in weight-sharing MLPs, and insufficient channel fusion in channel

Ethan Hirschowitz, Fabio Ramos

Improving competent robot policies with on-policy RL is often hampered by noisy, low-signal gradients. We revisit Evolution Strategies (ES) as a policy-gradient proxy and localize exploration with bounded, antithetic triangular perturbations, suitable for policy refinement. We propose Triangular-Distribution ES (TD-ES) which pairs bounded triangular noise wi

Leping Si, Meimei Yang, Hui Xue, Shipeng Zhu

Hierarchical data pervades diverse machine learning applications, including natural language processing, computer vision, and social network analysis. Hyperbolic space, characterized by its negative curvature, has demonstrated strong potential in such tasks due to its capacity to embed hierarchical structures with minimal distortion. Previous evidence indica

Yaoyuan Zhang, Aishan Liu, Zonghao Ying, Xianglong Liu

Large language models (LLMs) face growing trustworthiness concerns (\eg, deception), which hinder their safe deployment in high-stakes decision-making scenarios. In this paper, we present the first systematic investigation of strategic egoism (SE), a form of rule-bounded self-interest in which models pursue short-term or self-serving gains while disregarding

Ketong Chen, Yuhao Chen, Yang Xue

Despite the rapid progress of Vision-Language Models (VLMs), their capabilities are inadequately assessed by existing benchmarks, which are predominantly English-centric, feature simplistic layouts, and support limited tasks. Consequently, they fail to evaluate model performance for Visually Rich Document Understanding (VRDU), a critical challenge involving

Pritish Sahu, Anirudh Som, Dimitra Vergyri, Ajay Divakaran

Social norms are implicit, culturally grounded expectations that guide interpersonal communication. Unlike factual commonsense, norm reasoning is subjective, context-dependent, and varies across cultures, posing challenges for computational models. Prior works provide valuable normative annotations but mostly target isolated utterances or synthetic dialogues

Jiazhen Chen, Xiuqin Liang, Sichao Fu, Zheng Ma

Unsupervised graph anomaly detection (GAD) has received increasing attention in recent years, which aims to identify data anomalous patterns utilizing only unlabeled node information from graph-structured data. However, prevailing unsupervised GAD methods typically presuppose complete node attributes and structure information, a condition hardly satisfied in

Kunjing Yang, Libin Zheng, Minru Bai

Recently, triple decomposition has attracted increasing attention for decomposing third-order tensors into three factor tensors. However, this approach is limited to third-order tensors and enforces uniformity in the lower dimensions across all factor tensors, which restricts its flexibility and applicability. To address these issues, we propose the Multiple

Zhiming Ma, Shiyu Gan, Junhao Zhao, Xianming Li

Hearing-impaired individuals often face significant barriers in daily communication due to the inherent challenges of producing clear speech. To address this, we introduce the Omni-Model paradigm into assistive technology and present HI-TransPA, an instruction-driven audio-visual personal assistant. The model fuses indistinct speech with lip dynamics, enabli

Xuan Shen, Brian Wingenroth, Zichao Wang, Jason Kuen

The opioid crisis represents a significant moment in public health that reveals systemic shortcomings across regulatory systems, healthcare practices, corporate governance, and public policy. Analyzing how these interconnected systems simultaneously failed to protect public health requires innovative analytic approaches for exploring the vast amounts of data

Sen Hu

John Mather is a great scholar who was dedicated to mathematics in his whole life. His works in mathematics can be characterized as original and foundational. He laid out the foundation of singularity theory while he was a graduate student. He also laid out the foundation of modern Hamiltonian dynamical systems. Those fields became main stream in mathematics

Sara C. Beck, Jean L. Turner, Elm Zweig, John H. Black

The nearby dwarf starburst NGC 5253 is dominated by a compact radio-infrared supernebula powered by a very young and bright embedded Super Star Cluster (SSC) of $\sim 10^9 L_\odot$. We observed this source and its surroundings over the 5-25$\mu$m range with MIRI/MRS on JWST and in Paper I presented the JWST view of the region and its continuum features. We n

Jean L. Turner, Sara C. Beck, Elm A. Zweig, L. Barcos-Muñoz

We present imaging spectroscopy of the "supernebula" in the nearby dwarf galaxy NGC 5253 with the MIRI-MRS integral field spectrometer of the JWST. NGC 5253 is host to an luminous ($L\sim 10^9~\rm L_\odot$) HII region, powered by a giant young star cluster, a possible local analogue to super star cluster formation at Cosmic Dawn and Noon. In this paper, the

Zihao Zhang, Yang Li, Aming Wu, Yahong Han

In this paper, we focus on Single-Domain Generalized Object Detection (Single-DGOD), aiming to transfer a detector trained on one source domain to multiple unknown domains. Existing methods for Single-DGOD typically rely on discrete data augmentation or static perturbation methods to expand data diversity, thereby mitigating the lack of access to target doma

Elizabeth J. Iles, Joss Bland-Hawthorn, Courtney Crawford, Scott Croom

Bars are ubiquitous morphological features in the observed distribution of galaxies. There are similarly many methods for classifying these features and, without a strict theoretical definition or common standard practice, this is often left to circumstance. So, we were concerned whether astronomers even agree on the bar which they perceive in a given galaxy

Yongxian Wei, Yilin Zhao, Zixuan Hu, Li Shen

Data synthesis for training large reasoning models offers a scalable alternative to limited, human-curated datasets, enabling the creation of high-quality data. However, existing approaches face several challenges: (i) indiscriminate generation that ignores the solver's ability and yields low-value problems, or reliance on complex data pipelines to balance p

Hao Xu, Yuntian Chen, Dongxiao Zhang

Constitutive models are fundamental to solid mechanics and materials science, underpinning the quantitative description and prediction of material responses under diverse loading conditions. Traditional phenomenological models, which are derived through empirical fitting, often lack generalizability and rely heavily on expert intuition and predefined functio

Brian B. Moser, Shalini Sarode, Federico Raue, Stanislav Frolov

Dataset distillation (DD) promises compact yet faithful synthetic data, but existing approaches often inherit the inductive bias of a single teacher model. As dataset size increases, this bias drives generation toward overly smooth, homogeneous samples, reducing intra-class diversity and limiting generalization. We present PRISM (PRIors from diverse Source M

Francis Rhys Ward, Teun van der Weij, Hanna Gábor, Sam Martin

AI systems are increasingly able to autonomously conduct realistic software engineering tasks, and may soon be deployed to automate machine learning (ML) R&D itself. Frontier AI systems may be deployed in safety-critical settings, including to help ensure the safety of future systems. Unfortunately, frontier and future systems may not be sufficiently trustwo

Zahra Aslani, Fabio Taddei, Fabrizio Dolcini, Alessandro Braggio

Semiconductor nanowires (NWs) with strong Rashba spin-orbit coupling (RSOC), when exposed to a suitably applied Zeeman field, exhibit one-dimensional helical channels with a spin orientation locked to the propagation direction within the magnetic energy gap. Here, by adopting a scattering-matrix approach applied to a tight-binding model of the NW, we demonst

Hossein Rouhvarzi, Anastasis Kratsios

Incremental flow-based denoising models have reshaped generative modelling, but their empirical advantage still lacks a rigorous approximation-theoretic foundation. We show that incremental generation is necessary and sufficient for universal flow-based generation on the largest natural class of self-maps of $[0,1]^d$ compatible with denoising pipelines, nam

Weilin Wan, Fan Yi, Weizhong Zhang, Quan Zhou

Modern deep neural networks rely heavily on massive model weights and training samples, incurring substantial computational costs. Weight pruning and coreset selection are two emerging paradigms proposed to improve computational efficiency. In this paper, we first explore the interplay between redundant weights and training samples through a transparent anal

Yaodong Yang, Yang Wang, Jinpeng Li, Pei Guo

Protein evolution through amino acid mutations is a cornerstone of life sciences. Recent advances in protein language models have shown rich evolutionary patterns, offering unprecedented potential for in-silicon directed evolution. However, existing directed evolution methods largely rely on heuristic evolution strategies and have yet to efficiently integrat

Chen Li, Ryan Requist, E. K. U. Gross

We formulate a time-dependent density functional theory for the coupled dynamics of electrons and nuclei that goes beyond the Born-Oppenheimer (BO) approximation. We prove that the time-dependent marginal nuclear probability density $|\chi({\bdu R},t)|^2$, the conditional electronic density $n_{\bdu R}(\br,t)$, and the current density $\bm J_{\bdu R}(\br,t)$

Yitao Lin, Dahai Dai, Shilong Sun, Yuchen Wu

A novel electromagnetic quantitative inversion scheme for translationally moving targets via phase correlation registration of back-projection (BP) images is proposed. Based on a time division multiplexing multiple-input multiple-output (TDM-MIMO) radar architecture, the scheme first achieves high-precision relative positioning of the target, then applies re

Samih Fadli

We propose that unconstrained artificial intelligence obeys a Second Law analogous to thermodynamics, where ethical entropy, defined as a measure of divergence from intended goals, increases spontaneously without continuous alignment work. For gradient-based optimizers, we define this entropy over a finite set of goals {g_i} as S = -{\Sigma} p(g_i; theta) ln

Shunan Sheng, Bohan Wu, Bennett Zhu, Sinho Chewi

Structured variational inference constitutes a core methodology in modern statistical applications. Unlike mean-field variational inference, the approximate posterior is assumed to have interdependent structure. We consider the natural setting of star-structured variational inference, where a root variable impacts all the other ones. We prove the first resul

Marco Bernardi

The quantum period-finding (QPF) algorithm can compute the period of a function exponentially faster than the best-known classical algorithm. In standard QPF, the output state has a primary contribution from $r$ high-probability bit strings, where $r$ is the period. Measurement of this state, combined with continued fraction analysis, reveals the unknown per

Xiaoda Wang, Kaiqiao Han, Yuhao Xu, Xiao Luo

Cardiovascular disease (CVD) is a leading cause of mortality worldwide. Electrocardiograms (ECGs) are the most widely used non-invasive tool for cardiac assessment, yet large, well-annotated ECG corpora are scarce due to cost, privacy, and workflow constraints. Generating ECGs can be beneficial for the mechanistic understanding of cardiac electrical activity

Keshara Weerasinghe, Xueren Ge, Tessa Heick, Lahiru Nuwan Wijayasingha

Emergency Medical Services (EMS) are critical to patient survival in emergencies, but first responders often face intense cognitive demands in high-stakes situations. AI cognitive assistants, acting as virtual partners, have the potential to ease this burden by supporting real-time data collection and decision making. In pursuit of this vision, we introduce

Zhiyuan Yao, Anita Schöbel, Lei Nie, Sven Jäger

Periodic timetables are widely adopted in passenger railway operations due to their regular service patterns and well-coordinated train connections. However, fluctuations in passenger demand require varying train services across different periods, necessitating adjustments to the periodic timetable. This study addresses a hybrid periodic train timetabling pr

Jinfu Li, Yuqi Huang, Hong Song, Ting Wang

Recently, despite the remarkable advancements in object detection, modern detectors still struggle to detect tiny objects in aerial images. One key reason is that tiny objects carry limited features that are inevitably degraded or lost during long-distance network propagation. Another is that smaller objects receive disproportionately greater regression pena

Shengfei Wei, Suyuan Liu, Jun Wang, Ke Liang

Fair clustering is crucial for mitigating bias in unsupervised learning, yet existing algorithms often suffer from quadratic or super-quadratic computational complexity, rendering them impractical for large-scale datasets. To bridge this gap, we introduce the Anchor-based Fair Clustering Framework (AFCF), a novel, general, and plug-and-play framework that em

Koji Fujiwara, Takuya Kawada, Natsumi Nikaido, Jihoon Park

Recent development in nanotechnology has enabled us to investigate the dynamic properties of van der Waals materials on a piezoelectric substrate. Here we report on the dynamics of charge density wave (CDW) in NbSe$_{3}$ nanowires induced by surface acoustic waves (SAWs). Clear peaks in the differential resistance were observed at the resonant frequency of t

Xingyu Cheng

We consider the problem of existence of semistable systems of Hodge bundles with parabolic structure over a finite set $S \subset \mathbb P^1$ of type $(1,n)$. That is, we consider parabolic Higgs bundles $(\mathcal E, \theta)$, where $\mathcal E = \mathcal L \oplus \mathcal V$ and $\theta (\mathcal L) \subset \mathcal V \otimes \Omega_{\mathbb P^1}^1 (\log

Sidney Shapiro

Hybrid quantum-classical machine learning represents a frontier in computational research, combining the potential advantages of quantum computing with established classical optimization techniques. PennyLane provides a Python framework that seamlessly bridges quantum circuits and classical machine learning, enabling researchers to build, optimize, and deplo

Feifei Chen, Kaiming Zhang, Yanni Zhang, Hua Liang

Assessing model adequacy is a crucial step in regression analysis, ensuring the validity of statistical inferences. For Generalized Functional Linear Models (GFLMs), which are widely used for modeling relationships between scalar responses and functional predictors, there is a recognized need for formal goodness-of-fit testing procedures. Current literature

Shashwat Singh, Zilin Si, Zeynep Temel

Amphibians adapt their morphologies and motions to accommodate movement in both terrestrial and aquatic environments. Inspired by these biological features, we present PuffyBot, an untethered shape morphing robot capable of changing its body morphology to navigate multiple environments. Our robot design leverages a scissor-lift mechanism driven by a linear a

Liheng Zhang, Jin Wang, Hui Li, Bingfeng Zhang

3D understanding has drawn significant attention recently, leveraging Vision-Language Models (VLMs) to enable multi-modal reasoning between point cloud and text data. Current 3D-VLMs directly embed the 3D point clouds into 3D tokens, following large 2D-VLMs with powerful reasoning capabilities. However, this framework has a great computational cost limiting

Shen Zhang, Lei Wang, Huanyuan Shan, Ran Li

We propose fiDrizzleMU, an algorithm for co-adding exposures via iterative multiplicative updates, replacing the additive correction framework. This method achieves superior anti-aliasing and noise reduction in stacked images. When applied to James Webb Space Telescope data, the fiDrizzleMU algorithm reconstructs a gravitational lensing candidate that was si

Jialin Wu, Kecen Li, Zhicong Huang, Xinfeng Li

Many machine learning models are fine-tuned from large language models (LLMs) to achieve high performance in specialized domains like code generation, biomedical analysis, and mathematical problem solving. However, this fine-tuning process often introduces a critical vulnerability: the systematic degradation of safety alignment, undermining ethical guideline

Catherine Xia, Manar H. Alalfi

AI programming assistants have demonstrated a tendency to generate code containing basic security vulnerabilities. While developers are ultimately responsible for validating and reviewing such outputs, improving the inherent quality of these generated code snippets remains essential. A key contributing factor to insecure outputs is the presence of vulnerabil

Islem Sahraoui

This thesis explores a multimodal AI framework for enhancing construction safety through the combined analysis of textual and visual data. In safety-critical environments such as construction sites, accident data often exists in multiple formats, such as written reports, inspection records, and site imagery, making it challenging to synthesize hazards using

Wenzhe He, Xiaojun Chen, Wentang Chen, Hongyu Wang

Semantic Scene Completion (SSC) aims to generate a complete semantic scene from an incomplete input. Existing approaches often employ dense network architectures with a high parameter count, leading to increased model complexity and resource demands. To address these limitations, we propose RWKV-PCSSC, a lightweight point cloud semantic scene completion netw

Fangyan Wang, Sai Liang, Zaiyan Wei

Digital platforms increasingly face a common challenge in the age of artificial intelligence (AI): how to elicit richer and more useful user-generated content (UGC) without fully automating content production. We study this question in the context of online reviews by examining Yelp's introduction of an AI-enabled topic nudging tool in 2023, which provides r

Shuo Shi, Jinghuai Zhang, Shijie Jiang, Chunyi Zhou

Dataset distillation (DD) compresses large datasets into smaller ones while preserving the performance of models trained on them. Although DD is often assumed to enhance data privacy by aggregating over individual examples, recent studies reveal that standard DD can still leak sensitive information from the original dataset due to the lack of formal privacy

Yingchun Zhang, Zijun Zhou

We apply the abelianization technique to obtain an explicit ring presentation for the quasimap quantum cohomology of GIT quotients. As an application, for quiver varieties associated with oriented-acyclic quivers, we establish a cluster algebra structure on their equivariant quasimap quantum cohomology rings.

Lia Feital, Naamã Galdino, Renato Vidal Martins, Átila Felipe de Souza

We propose a version of the Enriques-Babagge Theorem for a singular curve $C$, involving its canonical model $C'$. We provide a partial proof for an arbitrary curve $C$ and complete the proof for unicuspidal monomial curves by describing the generators of the ideal of $C'\subset\mathbb{P}^{g-1}$.

Soumyendu Sarkar, Antonio Guillen-Perez, Zachariah J Carmichael, Avisek Naug

Reducing energy consumption and carbon emissions in data centers by enabling real-time temperature prediction is critical for sustainability and operational efficiency. Achieving this requires accurate modeling of the 3D temperature field to capture airflow dynamics and thermal interactions under varying operating conditions. Traditional thermal CFD solvers,

Nikunj Gupta, Bill Guo, Rajgopal Kannan, Viktor K. Prasanna

Large Language Models (LLMs) deliver state-of-the-art performance across many tasks but impose high computational and memory costs, limiting their deployment in resource-constrained or real-time settings. To address this, we propose HierRouter, a hierarchical routing approach that dynamically assembles inference pipelines from a pool of specialized, lightwei

Dong-Yue Xie, Xi Yang

To conduct a more in-depth investigation of randomized solvers for solving linear systems, we adopt a unified randomized batch-sampling Kaczmarz framework with per-iteration costs as low as cyclic block methods, and develop a general analysis technique to establish its convergence guarantee. With concentration inequalities, we derive new expected linear conv

Arham Rajendra Lodha

The Discrete Schwarz-Pick Lemma is a discrete analogue of the classical result from complex analysis, arising from the connection between circle packings and conformal maps established by Thurston. Previous works by Beardon-Stephanson and Van Eeuwen proved this lemma for circle packings where circles are tangent or intersect at non-obtuse angles, correspondi

Hyung-Jun Moon, Sung-Bae Cho

Continual learning methods used to force neural networks to process sequential tasks in isolation, preventing them from leveraging useful inter-task relationships and causing them to repeatedly relearn similar features or overly differentiate them. To address this problem, we propose a fully differentiable, exemplar-free expandable method composed of two com

Jia Lin, Xiaofei Zhou, Jiyuan Liu, Runmin Cong

Recently segment anything model (SAM) has attracted widespread concerns, and it is often treated as a vision foundation model for universal segmentation. Some researchers have attempted to directly apply the foundation model to the RGB-D video salient object detection (RGB-D VSOD) task, which often encounters three challenges, including the dependence on man

Xue-Chun Jiang, Jia-Lan Chen, Wei-Xue Li, Jin-Xun Liu

Electrocatalyst surfaces continuously reorganize on the timescale of catalytic turnover, obscuring the identification of active sites under operando conditions and hindering rational catalyst design. Here, we resolve the operando Cu(111) electrolyte interface for nitrate-to-ammonia electroreduction (NO3RR) via a multiscale modeling framework accelerated by a

Peng Gao, Yujian Lee, Xiaofeng Zhang, Zailong Chen

Large Vision-Language Models (LVLMs) have achieved impressive performance across a wide range of multimodal tasks. However, they still face critical challenges in modeling long-range dependencies under the usage of Rotary Positional Encoding (ROPE). Although it can facilitate precise modeling of token positions, it induces progressive attention decay as toke

Shogo Sato, Takuhiro Kaneko, Shoichiro Takeda, Tomoyasu Shimada

Point clouds are widely used in various fields, including augmented reality (AR) and robotics, where relighting and texture editing are crucial for realistic visualization. Achieving these tasks requires accurately separating albedo from shade. However, performing this separation on point clouds presents two key challenges: (1) the non-grid structure of poin

Mingye Zhu, Yi Liu, Zheren Fu, Quan Wang

Training Large Language Models (LLMs) for chain-of-thought reasoning presents a significant challenge: supervised fine-tuning on a single "golden" rationale hurts generalization as it penalizes equally valid alternatives, whereas reinforcement learning with verifiable rewards struggles with credit assignment and prohibitive computational cost. To tackle thes

Manh Nguyen, Dung Nguyen, Dai Do, Svetha Venkatesh

Reinforcement learning (RL) finetuning is crucial to aligning large language models (LLMs), but the process is notoriously unstable and exhibits high variance across model checkpoints. In practice, selecting the best checkpoint is challenging: evaluating checkpoints on the validation set during training is computationally expensive and requires a good valida

Li-rong Xie, Shi-Qiang Zhou, Di-Fan Yi, Huan-Bo Feng

POLAR-2 is a next-generation space astronomy platform led by China, with its core scientific objective focused on high-precision polarization measurements of gamma-ray bursts. As one of its key payloads, the Low-energy Polarization Detector (LPD) is designed to perform wide-field surveys to capture X-ray polarization information from gamma-ray bursts in the

Yi-Ming Zhu, Gang Zhao, Jiang-Pei Dou, Zhong-Hua Lv

We introduce CPISM, a simulation program developed for the Cool Planet Imaging Coronagraph (CPI-C) on the China Space Station Telescope (CSST). CPISM supports high-contrast exoplanet imaging by simulating observational conditions and instrumental effects to optimize target selection and observation strategies. The modular design includes target modeling, ima

Zhongchen Xu, Xinyang Liu, Cuiwei Zhang, Shuai Zhang

Intense quantum fluctuations arising from geometrical frustrations in kagome-lattice magnets provide a feasible approach to exotic quantum states. Here, we document an unexpected isosymmetric first-order magnetic transition in the recently synthesized frustrated kagome-lattice antiferromagnet Nd3ScBi5, which is characterized by significant latent heat and a

Shreya Shukla, Abhijith Jayakumar, Andrey Y. Lokhov

Recovering microscopic couplings directly from data provides a route to solving the inverse problem in statistical field theories, one that complements the traditional-often computationally intractable-forward approach of predicting observables from an action or Hamiltonian. Here, we propose an approach for the inverse problem that uniformly accommodates sys

Shufei Liu, Hu Zou, Jinfu Gou, Weijian Guo

This study investigates the evolution of the star-forming fraction ($F_{\mathrm{sf}}$) of Brightest Cluster Galaxies (BCGs) at $z<0.8$, using the galaxy clusters identified from the Legacy Imaging Surveys from the Dark Energy Spectroscopic Instrument (DESI). Star-forming galaxies are identified using the $g-z$ color, and $F_{\mathrm{sf}}$ is measured as a fu

Xiaokang Wang

In this paper, we prove several structure theorems for locally conformally flat, positive Yamabe orbifolds and nonnegative scalar curvature, ALE manifolds. These two kinds of spaces can be related by conformal blow-up and conformal compactification. For the orbifolds, we prove that such orbifolds admit a manifold cover. For the ALE manifolds, the homomorphis