October 2025 arXiv papers — page 206

Akshay Anand, Chenxu Guo, Cheol Jun Cho, Jiachen Lian

Current speech production systems predominantly rely on large transformer models that operate as black boxes, providing little interpretability or grounding in the physical mechanisms of human speech. We address this limitation by proposing a new framework: speech generation through explicit articulatory control. This reframes speech as a motor control task

Zeshan Khan

Within the domain of medical image analysis, three distinct methodologies have demonstrated commendable accuracy: Neural Networks, Decision Trees, and Ensemble-Based Learning Algorithms, particularly in the specialized context of genstro institutional track abnormalities detection. These approaches exhibit efficacy in disease detection scenarios where a subs

M. I. Sobirov, K. A. Rogachev, M. A. Bazrov, Zh. Zh. Namsaraev

Expanding the spectrum of 3D magnetic nanostructures requires mastering the interplay between different anisotropy contributions. Here, we fabricate bisegmented jellyfish nanowires with tailored arrangements of Co (strong magnetocrystalline anisotropy) and Ni (dominant shape anisotropy) segments. We uncover a unique magnetic duality: Co segments can be tuned

Ibrahim Salihu Yusuf, Iffanice Houndayi, Rym Oualha, Mohamed Aziz Cherif

Open-access multispectral imagery from missions like Landsat 8-9 and Sentinel-2 has fueled the development of geospatial foundation models (GFMs) for humanitarian and environmental applications. Yet, their deployment remains limited by (i) the absence of automated geospatial data pipelines and (ii) the large size of fine-tuned models. Existing GFMs lack work

Aijaz H. Lone, Meng Tang, Camelia Florica, Bin He

Spintronics offers a promising approach to energy efficient neuromorphic computing by integrating the functionalities of synapses and neurons within a single platform. A major challenge, however, is achieving field-free spin orbit torque SOT control over both synaptic and neuronal devices using an industry-adopted spintronic materials stack. In this study, w

Guangrong Wan, Jun liu, Qiyang Zhou, Tang tang

Tear film break-up (TFBU) analysis is critical for diagnosing dry eye syndrome, but automated TFBU segmentation remains challenging due to the lack of annotated datasets and integrated solutions. This paper introduces the Tear Film Multi-task (TFM) Dataset, the first comprehensive dataset for multi-task tear film analysis, comprising 15 high-resolution video

Meghna Roy Chowdhury, Yi Ding, Shreyas Sen

Electroencephalography (EEG) plays a crucial role in brain-computer interfaces (BCIs) and neurological diagnostics, but its real-world deployment faces challenges due to noise artifacts, missing data, and high annotation costs. We introduce SSL-SE-EEG, a framework that integrates Self-Supervised Learning (SSL) with Squeeze-and-Excitation Networks (SE-Nets) t

Beomjun Choi, Kyeongsu Choi, Dongjun Noh

In this paper, we construct a pancake-like ancient compact solution with flat sides to the Gauss curvature flow, contained in a slab. Also, we construct sausage-like ancient compact solutions to the $\alpha$-Gauss curvature flow with $\alpha >\frac{1}{2}$, asymptotic to a round cylinder.

Surajit Chattopadhyay

In the context of f(G) modified gravity, we address the cosmic application of the most generalized form of holographic dark energy (The European Physical Journal C, 77, (2017): 1-8) in this study, as well as a specific instance of it in the form of Barrow holographic dark energy (Physical Review D, 102(12), p.123525). Holographic dark energy and a well-known

Ziqiao Meng, Qichao Wang, Zhiyang Dou, Zixing Song

Autoregressive point cloud generation has long lagged behind diffusion-based approaches in quality. The performance gap stems from the fact that autoregressive models impose an artificial ordering on inherently unordered point sets, forcing shape generation to proceed as a sequence of local predictions. This sequential bias emphasizes short-range continuity

Utsav Pathak, Amit Mankodi

Accurate query runtime prediction is a critical component of effective query optimization in modern database systems. Traditional cost models, such as those used in PostgreSQL, rely on static heuristics that often fail to reflect actual query performance under complex and evolving workloads. This remains an active area of research, with recent work exploring

Wei-Chieh Huang, Cornelia Caragea

Implicit Attribute Value Extraction (AVE) is essential for accurately representing products in e-commerce, as it infers latent attributes from multimodal data. Despite advances in multimodal large language models (MLLMs), implicit AVE remains challenging due to the complexity of multidimensional data and gaps in vision-text understanding. In this work, we in

Jiaqi Liu, Tao Huang, Chang Xu

Recent advances in autoregressive (AR) models have demonstrated their potential to rival diffusion models in image synthesis. However, for complex spatially-conditioned generation, current AR approaches rely on fine-tuning the pre-trained model, leading to significant training costs. In this paper, we propose the Efficient Control Model (ECM), a plug-and-pla

Junwen Chen, Peilin Xiong, Keiji Yanai

Recent human-object interaction detection (HOID) methods highly require prior knowledge from vision-language models (VLMs) to enhance the interaction recognition capabilities. The training strategies and model architectures for connecting the knowledge from VLMs to the HOI instance representations from the object detector are challenging, and the whole frame

Shuzheng Si, Haozhe Zhao, Kangyang Luo, Gang Chen

Agents based on large language models (LLMs) struggle with brainless trial-and-error and generating hallucinatory actions due to a lack of global planning in long-horizon tasks. In this paper, we introduce a plan-and-execute framework and propose EAGLET, an efficient and effective planner training method to enhance the executor agent's planning abilities wit

Andrew Ly, Pulin Gong

Fundamental limits to predictability are central to our understanding of many physical and computational systems. Here we show that, despite its remarkable capabilities, deep learning exhibits such fundamental limits rooted in the fractal, riddled geometry of its basins of attraction: any initialization that leads to one solution lies arbitrarily close to an

Yasod Ginige, Akila Niroshan, Sajal Jain, Suranga Seneviratne

Penetration testing and vulnerability assessment are essential industry practices for safeguarding computer systems. As cyber threats grow in scale and complexity, the demand for pentesting has surged, surpassing the capacity of human professionals to meet it effectively. With advances in AI, particularly Large Language Models (LLMs), there have been attempt

Rintaro Kanaji, Brittany Reid, Yutaro Kashiwa, Raula Gaikovina Kula

GitHub recommends that projects adopt a security file that outlines vulnerability reporting procedures. However, the effectiveness and operational challenges of such files are not yet fully understood. This study aims to clarify the challenges that security files face in the vulnerability reporting process within open-source communities. Specifically, we cla

Yi-Ping Chen, Tien Hsieh, Da-Shin Lee

The motion of a spinning particle in the exterior of a Kerr-Newman black hole is studied. The dynamics is governed by the Mathisson-Papapetrou equations in the pole-dipole approximation, which includes spin-curvature coupling to the first order of spin. In terms of conserved quantities, the dynamical equations in Mino time can be transformed into the integra

Firuz Rakhmonov

For $n \geq 3$, an asymptotic formula is derived for the number of representations of a sufficiently large natural number $N$ in the form $p_1+p_2+m^n=N$, where $p_1$, $p_2$ $-$ prime numbers, $m$ $-$ natural number satisfying the conditions $$ \left|p_k-\mu_kN\right|\le H, \quad k=1,2,\qquad \left|m^n-\mu_3N\right|\le H,\qquad H \ge N^{1-\frac1{n(n-1)}} {\m

Noboru Ikeya, Jonathan R. Woodward

The importance of spin-correlated radical pairs in biology is increasingly recognized, with roles in biological effects of weak magnetic fields and emerging quantum spin-based biomedical applications. Fluorescence microscopy offers sufficient sensitivity to study magnetic field effects on radical pair reactions in living cells, but conventional techniques ca

Narendranath Layek, Prantik Nandi, Sachindra Naik, Birendra Chhotaray

We present a comprehensive long-term multi-wavelength study of the active galactic nucleus (AGN) NGC 3822, based on 17 years (2008 to 2025) of X-ray, ultraviolet (UV), and optical observations.The dataset includes observations from Swift, XMM-Newton, and NuSTAR, the Very Large Telescope, and the Himalayan Chandra Telescope. Our multiwavelength light curve an

Mingdai Yang, Nurendra Choudhary, Jiangshu Du, Edward W. Huang

Recent agent-based recommendation frameworks aim to simulate user behaviors by incorporating memory mechanisms and prompting strategies, but they struggle with hallucinating non-existent items and full-catalog ranking. Besides, a largely underexplored opportunity lies in leveraging LLMs'commonsense reasoning to capture user intent through substitute and comp

Changyuan Zhao, Ruichen Zhang, Jiacheng Wang, Dusit Niyato

Self-evolving agentic artificial intelligence (AI) offers a new paradigm for future wireless systems by enabling autonomous agents to continually adapt and improve without human intervention. Unlike static AI models, self-evolving agents embed an autonomous evolution cycle that updates models, tools, and workflows in response to environmental dynamics. This

Guangyan Zhu, Yuanyuan Luo, Jixiang Wan

For integers $x$ and $y$, $(x, y)$ and $[x, y]$ stand for the greatest common divisor and the least common multiple of $x$ and $y$ respectively. Denote by $|T|$ the number of elements of a finite set $T$. Let $a,b$ and $n$ be positive integers and let $S=\{x_1, \cdots, x_n\}$ be a set of $n$ distinct positive integers. We denote by $(S^a)$ (resp. $[S^a]$) th

Takao Tomono, Kazuya Tsujimura

The evolution of manufacturing toward smart factories has underscored major challenges in equipment maintenance, particularly the dependence on numerous contact sensors for anomaly detection, leading to increased sensor complexity and computational costs. This study explores the use of quantum kernels to enhance anomaly detection based on noncontact sensors.

Zhiyang Zhang, Ningcong Chen, Xin Zhang, Yanhua Li

The widespread use of GPS devices has driven advances in spatiotemporal data mining, enabling machine learning models to simulate human decision making and generate realistic trajectories, addressing both data collection costs and privacy concerns. Recent studies have shown the promise of diffusion models for high-quality trajectory generation. However, most

Zeqi Gu, Markos Georgopoulos, Xiaoliang Dai, Marjan Ghazvininejad

Autoregressive multimodal large language models have recently gained popularity for image generation, driven by advances in foundation models. To enhance alignment and detail, newer approaches employ chain-of-thought (CoT) reasoning, expanding user inputs into elaborated prompts prior to image synthesis. However, this strategy can introduce unnecessary redun

Aleksandar Badza, Gary Froyland, Roshan J. Samuel, Jörg Schumacher

Horizontally extended plane-layer convection flows are characterized by characteristic patterns of turbulent heat transfer due to the convective fluid motion consisting of a nearly-regular ridge network where hot fluid rises and cold fluid sinks. Here, we analyse this transport behavior by the so-called inflated generator framework, which identifies quasi-st

Jianqiang Li

Given a matrix $A$ of dimension $M \times N$ and a vector $\vec{b}$, the quantum linear system (QLS) problem asks for the preparation of a quantum state $|\vec{y}\rangle$ proportional to the solution of $A\vec{y} = \vec{b}$. Existing QLS algorithms have runtimes that scale linearly with the condition number $\kappa(A)$, the sparsity of $A$, and logarithmical

Bang Chen, Lijun Guo, Houli Fan, Wentao He

Identifying cancer driver genes (CDGs) is essential for understanding cancer mechanisms and developing targeted therapies. Graph neural networks (GNNs) have recently been employed to identify CDGs by capturing patterns in biological interaction networks. However, most GNN-based approaches rely on a single protein-protein interaction (PPI) network, ignoring c

Yiran Wang, Ruican Ma, Haiyun Zhang, Dahai Yan

We analyzed Insight-HXMT data of the black hole X-ray binary MAXI J1348-630 during the type-B QPO phase of its 2019 outburst. Using the Gaussian process method, we applied an additive composite kernel model consisting of an SHO, a DRW, and an additional white noise (AWN) to data from three energy bands: LE (1-10 keV), ME (10-30 keV), and HE (30-150 keV). We

Bin Kang, Bin Chen, Junjie Wang, Yulin Li

Existing Visual Language Models (VLMs) suffer structural limitations where a few low contribution tokens may excessively capture global semantics, dominating the information aggregation process and suppressing the discriminative features in text-driven image retrieval tasks. To address this, we introduce \textbf{CalibCLIP}, a training-free method designed to

Mikhail Anikushin, Andrey Romanov

We apply the iterative nonlinear programming method, previously proposed in our earlier work, to optimize Schur test functions and thereby provide refined upper bounds for the norms of integral operators. As an illustration, we derive such bounds for transfer operators associated with twofold additive compound operators that arise in the study of delay equat

Sourav Pal, Satyajit Seth

We perform a detailed computation of the helicity-dependent next-to-leading power leading logarithms in W+jet production, originating from next-to-soft gluon radiation and soft (anti-)quark emissions. These contributions are systematically captured via helicity-sensitive spinor shifts and soft quark operators. The resulting expressions exhibit full agreement

ElMouatez Billah Karbab, Mourad Debbabi

Malware proliferation is increasing at a tremendous rate, with hundreds of thousands of new samples identified daily. Manual investigation of such a vast amount of malware is an unrealistic, time-consuming, and overwhelming task. To cope with this volume, there is a clear need to develop specialized techniques and efficient tools for preliminary filtering th

Arindam Chowdhury, Massimiliano Lupo Pasini

Graph neural networks (GNNs) are widely used as surrogates for costly experiments and first-principles simulations to study the behavior of compounds at atomistic scale, and their architectural complexity is constantly increasing to enable the modeling of complex physics. While most recent GNNs combine more traditional message passing neural networks (MPNNs)

Jiashu Tao, Reza Shokri

Machine learning models are known to leak sensitive information, as they inevitably memorize (parts of) their training data. More alarmingly, large language models (LLMs) are now trained on nearly all available data, which amplifies the magnitude of information leakage and raises serious privacy risks. Hence, it is more crucial than ever to quantify privacy

Praneeth Vepakomma, Kaustubh Ponkshe

Traditional collaborative learning approaches are based on sharing of model weights between clients and a server. However, there are advantages to resource efficiency through schemes based on sharing of embeddings (activations) created from the data. Several differentially private methods were developed for sharing of weights while such mechanisms do not exi

Chen Li, Zhantao Yang, Han Zhang, Fangyi Chen

Vision-Language-Action (VLA) models show promise in embodied reasoning, yet remain far from true generalists-they often require task-specific fine-tuning, incur high compute costs, and generalize poorly to unseen tasks. We propose MetaVLA, a unified, backbone-agnostic post-training framework for efficient and scalable alignment. MetaVLA introduces Context-Aw

Yanjia Huang, Shuo Liu, Sheng Liu, Qingxiao Xu

Long-horizon robot manipulation tasks remain challenging for Vision-Language-Action (VLA) policies due to drift and exposure bias, often denoise the entire trajectory with fixed hyperparameters, causing small geometric errors to compound across stages and offering no mechanism to allocate extra test-time compute where clearances are tight. To address these c

Thippayawis Cheunchitra, Andrew Melatos, Rachel Webster

The luminosity distance-redshift ($D_{\rm L}$--$z$) relation derived from Type Ia supernovae (SNe Ia) yields evidence for a nonzero cosmological constant. SNe Ia analyses typically fit to the functional form $D_{\rm L}(z)$ derived theoretically from the homogeneous and isotropic Friedmann-Lemaitre-Robertson-Walker (FLRW) metric. Yet, the metric in the epoch

Mao Sheng

In this article, we extend the nonabelian Hodge correspondence in positive characteristic to the nonlinear setting.

Dong Yan, Gaochen Wu, Bowen Zhou

Recent advancements in language agents have led to significant improvements in multi-hop reasoning tasks. However, existing approaches often struggle with handling open-domain problems, which require massive information retrieval due to their reliance on a fixed sequence of actions. To address this, we propose Feedback-Guided Dynamic Interactive Planning (FG

Shakib Daryanoosh

There exist numerous problems in nature inherently described by finite $D$-dimensional states. Formulating these problems for execution on qubit-based quantum hardware requires mapping the qudit Hilbert space to that of multiqubit which may be exponentially larger. To exclude the infeasible subspace, one common approach relies on penalizing the objective fun

Hong-Yi Zhang

Magnification of total image fluxes is typically considered a defining feature of gravitational microlensing. In contrast, I will show that nonminimal couplings to gravity can generate regions of negative gravitational potential curvature, giving rise to the distinctive possibility of demagnification. Such events, appearing as flux troughs in microlensing li

Erick Lee-Guzmán, Egor A. Maximenko, Enrique Abdeel Muñoz-de-la-Colina, Marco Iván Ruiz-Carmona

Let $X$ be a set and $d_1,d_2$ be two distances on $X$. We say that $d_1$ and $d_2$ are locally similar and write $d_1\cong d_2$ if $d_1$ and $d_2$ are topologically equivalent and, for every $a$ in $X$, \[ \lim_{x\to a} \frac{d_2(x,a)}{d_1(x,a)}=1. \] We prove that if $d_1\cong d_2$, then the intrinsic distances induced by $d_1$ and $d_2$ coincide. We also

Hossein Taheri, Avishek Ghosh, Arya Mazumdar

Continual learning, the ability of a model to adapt to an ongoing sequence of tasks without forgetting earlier ones, is a central goal of artificial intelligence. To better understand its underlying mechanisms, we study the limitations of continual learning in a tractable yet representative setting. Specifically, we analyze one-hidden-layer quadratic neural

Xinyu Ma, Chengxin Wang, Meng Wang, Xu Guo

We introduce the Gaussian Ensemble Topology (GET) method, a new explicit and manufacture-ready framework for topology optimization in which design geometries are represented as superpositions of anisotropic Gaussian functions. By combining explicit Gaussian descriptions with a level-set-like Heaviside projection, GET inherently generates smooth, curvature-co

Chengzhi Liu, Yuzhe Yang, Kaiwen Zhou, Zhen Zhang

The promotion of academic papers has become an important means of enhancing research visibility. However, existing automated methods struggle limited storytelling, insufficient aesthetic quality, and constrained self-adjustment, making it difficult to achieve efficient and engaging dissemination. At the heart of those challenges is a simple principle: \emph{

John A. Toth, Xiao Xiao

We prove a quantum ergodic restriction (QER) theorem for real hypersurfaces $\Sigma \subset X,$ where $X$ is the Grauert tube associated with a real-analytic, compact Riemannian manifold. As an application, we obtain $h$ independent upper and lower bounds for the $L^2$ - restrictions of the FBI transform of Laplace eigenfunctions restricted to $\Sigma$ satis

Sheng Xiang, Chenhao Xu, Dawei Cheng, Xiaoyang Wang

Graph simulation has recently received a surge of attention in graph processing and analytics. In real-life applications, e.g. social science, biology, and chemistry, many graphs are composed of a series of evolving graphs (i.e., temporal graphs). While most of the existing graph generators focus on static graphs, the temporal information of the graphs is ig

Nicholas H. Nelsen, Houman Owhadi, Andrew M. Stuart, Xianjin Yang

Methods for solving scientific computing and inference problems, such as kernel- and neural network-based approaches for partial differential equations (PDEs), inverse problems, and supervised learning tasks, depend crucially on the choice of hyperparameters. Specifically, the efficacy of such methods, and in particular their accuracy, stability, and general

Jamie Karthein, Maneesha Pradeep, Krishna Rajagopal, Mikhail Stephanov

We present the first application of the maximum-entropy freeze-out prescription to calculate factorial cumulants of proton multiplicities near the conjectured QCD critical point in thermal equilibrium. We map the Gibbs free energy of the 3D Ising model to a parameterized class of possible EoS near QCD critical point. This equilibrium baseline highlights how

Mohamad Lukman Aidid Mohd Yusoff, Norhasliza Yusof, Hasan Abu Kassim, Jan Steinheimer

We investigate dimuon production in the context of a first-order phase transition in QCD matter using a chiral fluid dynamics model. This approach incorporates non-equilibrium effects such as entropy production and reheating, which emerge during the dynamical evolution through a first-order phase transition. By comparing equilibrium and non-equilibrium scena

Junhyoung Park, Andrea Olivati, Mirko Prato, Min Kim

Co-evaporation of formamidinium tin triiodide (FASnI3) precursors, without any additives or reducing agents, leads to the growth of a highly crystalline thin film which shows a bandgap around 1.31 eV, closely matching the theoretical value predicted from the ideal single crystal structure of FASnI3. The polycrystalline thin film presents a lower tendency of

Cemal Tugrul Yilmaz, Eric Foss, Mamadou Diagne, Miroslav Krstic

This paper presents novel extremum seeking (ES) strategies for maximum power point tracking (MPPT) in photovoltaic (PV) systems that ensure unbiased convergence and prescribed-time performance. Conventional ES methods suffer from steady-state bias due to persistent dither signal. We introduce two novel ES algorithms: the exponential unbiased ES (uES), which

Sheng Xiang, Yidong Jiang, Yunting Chen, Dawei Cheng

Spoofing detection in financial trading is crucial, especially for identifying complex behaviors such as conspiracy spoofing. Traditional machine-learning approaches primarily focus on isolated node features, often overlooking the broader context of interconnected nodes. Graph-based techniques, particularly Graph Neural Networks (GNNs), have advanced the fie

Xuan Zhao, Le-Man Kuang, Jie-Qiao Liao

The dark-state effect, caused by destructive quantum interference, is an important physical effect in atomic physics and quantum optics. It not only deepens the understanding of light-atom interactions, but also has wide applications in quantum physics and quantum information. Therefore, how to efficiently and conveniently determine the number and form of th

Hongchi Xia, Chih-Hao Lin, Hao-Yu Hsu, Quentin Leboutet

Digitizing the physical world into accurate simulation-ready virtual environments offers significant opportunities in a variety of fields such as augmented and virtual reality, gaming, and robotics. However, current 3D reconstruction and scene-understanding methods commonly fall short in one or more critical aspects, such as geometry completeness, object int

Md Rakibul Mowla, Sukhbinder Kumar, Ariane E. Rhone, Brian J. Dlouhy

Statistical significance testing of neural coherence is essential for distinguishing genuine cross-signal coupling from spurious correlations. A widely accepted approach uses surrogate-based inference, where null distributions are generated via time-shift or phase-randomization procedures. While effective, these methods are computationally expensive and yiel

Christopher Hoang, Mengye Ren

Object recognition and motion understanding are key components of perception that complement each other. While self-supervised learning methods have shown promise in their ability to learn from unlabeled data, they have primarily focused on obtaining rich representations for either recognition or motion rather than both in tandem. On the other hand, latent d

Jiakai Xu, Tianle Zhou, Eugene Wu, Kostis Kaffes

Agentic exploration, letting LLM-powered agents branch, backtrack, and search across many execution paths, demands systems support well beyond today's pass-at-k resets. Our benchmark of six snapshot/restore mechanisms shows that generic tools such as CRIU or container commits are not fast enough even in isolated testbeds, and they crumble entirely in real de

Zhongyi Zhang, Julie A. Hides, Enrico De Martino, Abdul Joseph Fofanah

Purpose: To develop and validate No-New SAM2 (nnsam2) for few-shot segmentation of lumbar paraspinal muscles using only a single annotated slice per dataset, and to assess its statistical comparability with expert measurements across multi-sequence MRI and multi-protocol CT. Methods: We retrospectively analyzed 1,219 scans (19,439 slices) from 762 participan

Cade Houston Kennedy, Amr Hilal, Morteza Momeni

With the growth of digital financial systems, robust security and privacy have become a concern for financial institutions. Even though traditional machine learning models have shown to be effective in fraud detections, they often compromise user data by requiring centralized access to sensitive information. In IoT-enabled financial endpoints such as ATMs an

Yan Rui Tan, Wenqi Liu, Wai Lun Leong, John Guan Zhong Tan

Artificial Potential Field (APF) methods are widely used for reactive flocking control, but they often suffer from challenges such as deadlocks and local minima, especially in the presence of obstacles. Existing solutions to address these issues are typically passive, leading to slow and inefficient collective navigation. As a result, many APF approaches hav

Buu Phan, Ashish Khisti

We study channel simulation and distributed matching, two fundamental problems with several applications to machine learning, using a recently introduced generalization of the standard rejection sampling (RS) algorithm known as Ensemble Rejection Sampling (ERS). For channel simulation, we propose a new coding scheme based on ERS that achieves a near-optimal

Onil Boussim

In this paper, I propose a method for correcting sample selection bias when the outcome of interest is categorical, such as occupational choice, health status, or field of study. Classical approaches to sample selection rely on strong parametric distributional assumptions, which may be restrictive in practice. I develop a local representation that decomposes

Sedi Bartz, Heinz H. Bauschke, Yuan Gao

A cornerstone of convex analysis, established by Rockafellar in 1966, asserts that a set has a potential if and only if it is cyclically monotone. This characterization was generalized to hold for any {\color{black} finite-valued} cost function $c$ and lies at the core structure of optimal transport plans. However, this equivalence fails to hold for costs th

Aneesh Jonelagadda, Christina Hahn, Haoze Zheng, Salvatore Penachio

Long-term memory is essential for natural, realistic dialogue. However, current large language model (LLM) memory systems rely on either brute-force context expansion or static retrieval pipelines that fail on edge-constrained devices. We introduce Mnemosyne, an unsupervised, human-inspired long-term memory architecture designed for edge-based LLMs. Our appr

Keller Blackwell, Jeongwan Haah

For fault-tolerant quantum memory defined by periodic Pauli measurements, called Floquet codes, we prove that every correctable, undetectable spacetime error occurring during the steady stage is a product of (i) measurement operators inserted at the time of the measurement and (ii) pairs of identical Pauli operators sandwiching a measurement that commutes wi

Gordon Hung, Salinna Abdullah

Taiwan's high population and heavy dependence on fossil fuels have led to severe air pollution, with the most prevalent greenhouse gas being carbon dioxide (CO2). There-fore, this study presents a reproducible and comprehensive case study comparing 21 of the most commonly employed time series models in forecasting emissions, analyzing both univariate and mul

Eugene Vorobiov, Ammar Jaleel Mahmood, Salim Rezvani, Robin Chhabra

We present ARRC (Advanced Reasoning Robot Control), a practical system that connects natural-language instructions to safe local robotic control by combining Retrieval-Augmented Generation (RAG) with RGB-D perception and guarded execution on an affordable robot arm. The system indexes curated robot knowledge (movement patterns, task templates, and safety heu

Weiguo Chen, Kai Tang

In this paper, we consider general $k$th-mixed curvature $\mathcal{C}^{(k)}_{\alpha,\beta}$ ($\beta\neq0$) for Hermitian manifolds, which is a convex combination of the $k$th Chern Ricci curvature and holomorphic sectional curvature. We prove that any compact Hermitian surface with constant $k$th-mixed curvature is self-dual. Furthermore, we show that if a c

Xinrui Ruan, Xinwei Ma, Yingfei Wang, Waverly Wei

Randomized controlled trials (RCTs) are widely adopted for causal inference, yet cost and sample-size constraints limit power. We introduce CALM (Causal Analysis leveraging Language Models), a statistical framework that integrates insights generated by large language models (LLMs) into the analysis of RCTs using established causal estimators to increase prec

E. Santos, J. L. Costa, R. L. Rodriguez-Suarez, J. B. S. Mendes

We report a comprehensive experimental investigation of orbital-to-charge conversion in metallic and semiconductor materials, emphasizing the fundamental roles of the inverse orbital Hall effect (IOHE) and the inverse orbital Rashba effect. Using spin pumping driven by ferromagnetic resonance (SP-FMR) and the spin Seebeck effect (SSE), we demonstrate efficie

Xilin Jiang, Hannes Gamper, Sebastian Braun

Acoustic scene perception involves describing the type of sounds, their timing, their direction and distance, as well as their loudness and reverberation. While audio language models excel in sound recognition, single-channel input fundamentally limits spatial understanding. This work presents Sci-Phi, a spatial audio large language model with dual spatial a

Qiaoyi Wen, Fanrong Xu

Tensor-current operators, potentially generated by scalar leptoquarks in grand unified theories (GUTs), are among the plausible new physics (NP) candidates suggested by the anomalies observed in $B$-meson decays. As experimental data continue to accumulate, exploring this possibility remains timely and well motivated. In this work, we present a systematic an

Minchul Kam, Hiroshi Nagai, Motoki Kino, Keiichi Asada

We present the results of multi-frequency monitoring of the radio quasar 3C 286, conducted using three instruments: ALMA at 91.5, 103.5, 233.0, and 343.4 GHz, the IRAM 30-m Telescope at 86 and 229 GHz, and SMA at 225 GHz. The IRAM measurements from 2006 to 2024 show that the total flux of 3C 286 is stable within measurement uncertainties, indicating long-ter

N. Emas, A. Porredon, C. Blake, J. DeRose

Combined survey analyses of galaxy clustering and weak gravitational lensing (3x2-pt studies) will allow new and accurate tests of the standard cosmological model. However, careful validation is necessary to ensure that these cosmological constraints are not biased by uncertainties associated with the modelling of astrophysical or systematic effects. In this

Owen Henkel, Bill Roberts, Doug Jaffe, Laurence Holt

Recent advances in multimodal large language models (MLLMs) raise the question of their potential for grading, analyzing, and offering feedback on handwritten student classwork. This capability would be particularly beneficial in elementary and middle-school mathematics education, where most work remains handwritten, because seeing students' full working of

Roni Chatterjee, Smarajit Karmakar, Muhittin Mungan, Damien Vandembroucq

We investigate by atomistic simulations the memory behavior a model glass subjected to random driving protocols. The training consists of a random walk of forward and/or backward shearing sequences bounded by a maximal shear strain of absolute value {\gamma}T . We show that such a stochastic training protocol is able to record the training amplitude. Differe

Mahboubeh Zarei, Robin Chhabra, Farrokh Janabi-Sharifi

Accurate pose and velocity estimation is essential for effective spatial task planning in robotic manipulators. While centralized sensor fusion has traditionally been used to improve pose estimation accuracy, this paper presents a novel decentralized fusion approach to estimate both pose and velocity. We use dual-view measurements from an eye-in-hand and an

Rui Liu, Tao Zhe, Yanjie Fu, Feng Xia

Feature selection eliminates redundancy among features to improve downstream task performance while reducing computational overhead. Existing methods often struggle to capture intricate feature interactions and adapt across diverse application scenarios. Recent advances employ generative intelligence to alleviate these drawbacks. However, these methods remai

Yao Xiao, Jung-jae Kim, Roy Ka-wei Lee, Lidong Bing

Self-play preference optimization has emerged as a prominent paradigm for aligning large language models (LLMs). It typically involves a language model to generate on-policy responses for prompts and a reward model (RM) to guide the selection of chosen and rejected responses, which can be further trained with direct preference optimization (DPO). However, th

Weilong Fu

Large language models are reshaping quantitative investing by turning unstructured financial information into evidence-grounded signals and executable decisions. This survey synthesizes research with a focus on equity return prediction and trading, consolidating insights from domain surveys and more than fifty primary studies. We propose a task-centered taxo

Sam Sartor, Pieter Peers

Large pretrained diffusion models can provide strong priors beneficial for many graphics applications. However, generative applications such as neural rendering and inverse methods such as SVBRDF estimation and intrinsic image decomposition require additional input or output channels. Current solutions for channel expansion are often application specific and

Harshil Vejendla

Test-time adaptation (TTA) aims to adapt a pretrained model to distribution shifts using only unlabeled test data. While promising, existing methods like Tent suffer from instability and can catastrophically forget the source knowledge, especially with small batch sizes or challenging corruptions. We argue that this arises from overly deterministic updates o

Harshil Vejendla

Autoregressive decoding in large language models (LLMs) requires caching a growing list of past key-value (KV) pairs, making long-context inference a memory-bound problem. While recent methods have explored quantizing the cache, evicting tokens, or using binary sketches for keys (e.g., Loki), these approaches often provide an incomplete solution by leaving o

Lawrence Liu, Alexander Liu, Mengdi Wang, Tuo Zhao

Large language models (LLMs) present significant deployment challenges due to their immense computational and memory requirements. While semi-structured pruning, particularly 2:4 sparsity, offers a path to practical hardware acceleration, existing methods often incur substantial performance degradation. To bridge this gap, we introduce ARMOR: (Adaptive Repre

Yuyao Wang, Yu-Hung Cheng, Debarghya Mukherjee, Huimin Cheng

Graphon models provide a flexible nonparametric framework for estimating latent connectivity probabilities in networks, enabling a range of downstream applications such as link prediction and data augmentation. However, accurate graphon estimation typically requires a large graph, whereas in practice, one often only observes a small-sized network. One approa

Tian Xin

Motivated by Heisenberg's observable-only stance, we replace latent "information" (filtrations, hidden diffusions, state variables) with observable transitions between price states. On a discrete price lattice with a Hilbert-space representation, shift operators and the spectral calculus of the price define observable frequency operators and a translation-in

Ziyi Chen, Junyi Li, Peiran Yu, Heng Huang

Reinforcement learning from human feedback (RLHF) and direct preference optimization (DPO) are important techniques to align large language models (LLM) with human preference. However, the quality of RLHF and DPO training is seriously compromised by \textit{\textbf{C}orrupted} preference, reward \textit{\textbf{O}veroptimization}, and bias towards \textit{\t

Lei Wang, Zeren Simon Wang, Haotian Xu

In the type-I two-Higgs-doublet model, the pseudoscalar $A$ can act as a long-lived particle (LLP) for sufficiently large values of $\tan\beta$. At the LHC, the $A$ particles are predominantly produced in pairs through $pp \to W^*/Z^* \to H^\pm/H \, A$, with subsequent decays $H^{\pm}/H \to W^\pm/Z\, A$. For the mass range of our interest, $10~\text{GeV}\les

Kuangshi Ai, Jonathan A. Karr, Meng Jiang, Nitesh V. Chawla

We present Knowledge Extraction on OMIn (KEO), a domain-specific knowledge extraction and reasoning framework with large language models (LLMs) in safety-critical contexts. Using the Operations and Maintenance Intelligence (OMIn) dataset, we construct a QA benchmark spanning global sensemaking and actionable maintenance tasks. KEO builds a structured Knowled

Akatsuki Nishioka

An invex function generalizes a convex function in the sense that every stationary point is a global minimizer. Recently, invex functions and their subclasses have attracted attention in signal processing and machine learning. However, verifying invexity is often difficult because its definition involves an unknown function called a kernel function. This pap

Le Quang Ham, Do Trong Hoang, Le Quang Hung, Doowon Koh

Let $d\ge3$ and $\mathbb{F}_q^{\,d}$ be the $d$-dimensional vector space over a finite field of order $q$, where $q$ is an odd prime power. Let $X_\pi$ be the set of lines through the origin intersecting the slice $\pi\cap S^{d-1}$, where $\pi=\{x_d=\lambda\}$ and $S^{d-1}=\{x:\|x\|=1\}$. For $E\subset\mathbb{F}_q^{\,d}$ and $N\ge1$, we study the exceptional

Guocheng Wang, Qi Su, Long Wang, Joshua B. Plotkin

Evolutionary game theory offers a general framework to study how behaviors evolve by social learning in a population. This body of theory can accommodate a range of social dilemmas, or games, as well as real-world complexities such as spatial structure or behaviors conditioned on reputations. Nonetheless, this approach typically assumes a deterministic payof

Rui Li, Zeyu Zhang, Xiaohe Bo, Zihang Tian

Current Large Language Models (LLMs) are confronted with overwhelming information volume when comprehending long-form documents. This challenge raises the imperative of a cohesive memory module, which can elevate vanilla LLMs into autonomous reading agents. Despite the emergence of some heuristic approaches, a systematic design principle remains absent. To f

Vyoma Raman, Camille Chabot, Betsy Popken

The Universal Declaration of Human Rights and other international agreements outline numerous inalienable rights that apply across geopolitical boundaries. As generative AI becomes increasingly prevalent, it poses risks to human rights such as non-discrimination, health, and security, which are also central concerns for AI researchers focused on fairness and