April 2026 arXiv papers — page 9

Yiran Du, Huimin He

This study examined intermittent discontinuance in AI-mediated informal digital learning of English (AI-IDLE) through the cognition-affect-conation framework. Survey data were collected from 632 Chinese university EFL learners with prior AI-IDLE experience and analysed using structural equation modelling and fuzzy-set qualitative comparative analysis. The SE

Edoardo Mainini, Debangana Mukherjee, Roberto Ognibene

We investigate the existence of positive solutions to fractional equations presenting a double criticality: a multi-polar Hardy-type potential and a Sobolev critical nonlinearity. The nonlocal nature of the operator and the absence of explicit ground states for the single-pole equation stand as major difficulties. We overcome these obstacles by passing to an

Mengfei Zhang, Jinlu Zhang, Zhigang Tu

Modeling 4D human-object interaction (HOI) is a compelling challenge in computer vision and an essential technology powering virtual and mixed-reality applications. While existing works have achieved promising results on specific HOI tasks-such as text-conditioned HOI generation and human motion generation from object motion, they typically rely on task-spec

Yann Bugeaud

Let $q_1, \ldots , q_t$ be distinct prime numbers. Let $a_1, \ldots , a_t$ be nonnegative integers. We establish effective lower bounds for $|z^d - q_1^{a_1} \ldots q_t^{a_t}|$ and for its greatest prime factor, which tend to infinity with $z^d$, where $z$ is a positive integer coprime with $q_1 \ldots q_t$ and $d \ge 2$ is an integer.

Keisuke Fujii, Daichi Kagamihara, Masaru Hongo

Anomalous tunneling refers to the phenomenon in which the transmission coefficient through a potential barrier approaches unity as the energy of an incident particle or quasiparticle tends to zero. This counterintuitive effect has been reported in systems exhibiting spontaneous symmetry breaking (SSB), such as superfluids, yet the general conditions for its

Bojun Zhang, Huiyu Yang, Yunpeng Wang, Yuntian Chen

Rapid aerodynamic evaluation is crucial for modern vehicle design, yet existing neural operators struggle to capture intricate spatial correlations. We propose the rotary-enhanced transformer operator (RETO), a novel neural solver featuring a dual-stage spatial awareness mechanism: sinusoidal-cosine encodings for global referencing and rotary positional enco

Yu Tian, Jiawei Chen, Lifan Zheng, Mingxiang Tao

We introduce Skills-Coach, a novel automated framework designed to significantly enhance the self-evolution of skills within Large Language Model (LLM)-based agents. Addressing the current fragmentation of the skill ecosystem, Skills-Coach explores the boundaries of skill capabilities, thereby facilitating the comprehensive competency coverage essential for

Han Liu, Shanghao Shi, Yevgeniy Vorobeychik, Chongjie Zhang

Low-Rank Adaptation (LoRA), which leverages the insight that model updates typically reside in a low-dimensional space, has significantly improved the training efficiency of Large Language Models (LLMs) by updating neural network layers using low-rank matrices. Since the generation of adversarial examples is an optimization process analogous to model trainin

Laur Ferris, Donglai Gong

The subpolar oceans are characterized by intense storm forcing and complex littoral topography. Submesoscale frontal instabilities are significant sources of turbulent kinetic energy (TKE) in these regions. However, criteria for identifying and parameterizing these instabilities in regional models have predominantly relied on a geostrophic framework that neg

Jisheng Zhao, Huanzhi Pu, Shinnung Jeong, Chihyo Ahn

GPU compilers merge all data types into a single unified register file, erasing the type information that binary-analysis tools rely on. We show that type recovery from this untyped register file is the central challenge of GPU binary lifting. We present CuLifter, a SASS-to-LLVM IR lifting framework that recovers register types via constraint propagation wit

Binjie Hong, Rui Xiong, Liyuan Han, Tielin Zhang

Modeling invasive neural spike data is fundamental to advancing high-performance brain-computer interfaces (BCIs). However, existing approaches face critical challenges, including limited-scale heterogeneous data, cross-domain distribution shift, and the intrinsic spatiotemporal complexity of invasive neural signals. In this work, we propose UniBCI, a unifie

A. A. Borovkov, K. A. Borovkov

Let $Z=\{Z(t): t\in \mathbb R\}$ be a stochastic process with trajectories in space $\mathbb D (\mathbb R)$. It is assumed that there exists an essentially smooth function $A:\mathbb R\to (-\infty, \infty] $ such that, for all $\alpha \in \mathbb R, $ $ \mu\in \mbox{dom}\, A$, one has \begin{equation*} \frac1{T} \ln {\mathbf E} \big( e^{\mu (Z(T)-\alpha T)}

Jaehee Kim, Juhyeon Kim, Gwonhak Lee, Kyunghyun Baek

Imaginary-time evolution is a standard primitive for ground-state preparation but is nonunitary, precluding direct quantum implementation. We develop Finite Imaginary-Time Evolution (FinITE), a finite-beta construction for diagonal Pauli-Z cost Hamiltonians arising from polynomial unconstrained binary optimization (PUBO) instances, including QUBO and HUBO ca

Indranil Biswas, Satyajit Guin, Pradip Kumar

We extend Atiyah's holomorphic jet bundle formalism to holomorphic vector bundles over noncommutative algebras endowed with a bigraded differential calculus truncated at bidegree $(1,1)$; such structures are referred to as noncommutative complex curves. For a holomorphic vector bundle $(E,\,\overline{\nabla}_E)$ over such an algebra $\mathcal{A}$, we con

Collin T. Christy, Tanmoy Laskar, Kate D. Alexander, Noah Franz

We report the detection of linear polarization in the radio afterglow of GRB 260310A, representing the first centimeter-wavelength polarization detection of a gamma-ray burst (GRB) afterglow and the first measurement of Faraday rotation in a GRB environment. We detect linearly polarized emission across $11-25$ GHz, with a polarization fraction decreasing mon

Jipeng Tan, Weifeng Zhang, Ye Wu, Jialin Guo

Recommendation algorithms have become the dominant mechanism for information distribution on digital platforms, profoundly shaping personalized information consumption environments. However, gender bias, as a significant form of algorithmic discrimination, may cause users to experience unequal exposure within different political information environments. Tak

Sivaram Krishnan, Bassel Al Homssi, Zhouyou Gu, Jihong Park

Terrestrial network limitations drive the integration of non-terrestrial networks (NTNs), notably mega-constellations comprising thousands of low Earth orbit (LEO) satellites. While these satellites act as interconnected network switches via inter-satellite links (ISLs), their massive scale creates severe bottlenecks for network management. To address this,

Preetish K. Mishra, Changbom Park, Jaehyun Lee, Yohan Dubois

We investigate the physical origin of critical mass, a threshold where many galaxy properties and scaling relations undergo fundamental transitions, using the Horizon Run 5 simulation. Focusing on massive ($M_{\rm tot} \geq 10^{12}{\rm M_\odot}$) central galaxies, we examine the mass-dependent turnover of the stellar-to-total mass ratio (STR) and the physica

Loukas Georgiadis, Evangelos Kipouridis, Evangelos Kosinas, Charis Papadopoulos

Computing edge-connected components in directed and undirected graphs is a fundamental and well-studied problem in graph algorithms. In a very recent breakthrough, Korhonen [STOC 2025] showed that for any fixed $k$, the $k$-edge connected components of an undirected graph can be computed in linear time. In contrast, the directed case remains significantly mo

Biswarup Paul

We present a systematic study of the transverse momentum ($p_{\rm T}$) spectra of charmonium (J/$\psi$, $\psi(2S)$) and bottomonium ($\Upsilon(nS)$) states in Pb-Pb collisions at $\sqrt{s_{\rm NN}} = 5.02$ TeV within an analytical relativistic hydrodynamics framework. The medium evolution is described assuming cylindrical symmetry with boost-invariant longit

Yang Zhang, Jiangyuan Zhao, Chenyou Fan, Fangzheng Yan

Vision-Language-Action (VLA) models advance robotic control via strong visual-linguistic priors. However, existing VLAs predominantly frame pretraining as supervised behavior cloning, overlooking the fundamental nature of robot learning as a goal-reaching process that requires understanding temporal task progress. We present \textbf{PRTS} (\textbf{P}rimitive

F. Azari, M. Haghighat

We investigate the phenomenology of the Standard Model extended by a real gauge-singlet scalar field, focusing on exotic Higgs decay channels. For a light scalar mass in the range \(0 < m_{\phi} < 40\) GeV, the Higgs boson can decay to both two and three scalar final states. We derive analytical expressions for these decay rates and impose a global constrain

Rebecca Soskin Hicks, Mikhail Trofimov, Dominick Lim, Rahul K. Arora

Millions of clinicians use ChatGPT to support clinical care, but evaluations of the most common use cases in model-clinician conversations are limited. We introduce HealthBench Professional, an open benchmark for evaluating large language models on real tasks that clinicians bring to ChatGPT in the course of their work. The benchmark is organized around thre

Sergiy Plaksa, Alexander Sarana

We obtain estimates of the modulus of continuity for the real part of the Cauchy-type integral in the closure of domain bounded by an Ahlfors-regular integration curve. These estimates are more exact than the well-known Zygmund estimate for the modulus of continuity of the Cauchy-type integral. The accuracy of estimates is proved by constructing an example o

Jiasheng Zheng, Xin Zheng, Boxi Cao, Pengbo Wang

Code sandboxes have emerged as a critical infrastructure for advancing the coding capabilities of large language models, providing verifiable feedback for both RL training and evaluation. However, existing systems fail to provide accurate verification and efficiency under high-concurrency workloads. We present ScaleBox, a high-fidelity and scalable system de

Matthew de Brecht

An \'{e}tale space over a topological space $Y$ is defined as a local homeomorphism from a topological space $X$ into $Y$. They often come up in topos theory because of the equivalence between sheaves and \'{e}tale spaces over a space. In this note, we define computable \'{e}tale spaces over a computable topological space $Y$ within the TTE framework of comp

Yikai Deng, Shifan Zuo, Jixia Li, Yougang Wang

We present a comprehensive analysis of the 21 cm intensity mapping (IM) data from the Tianlai Cylinder Pathfinder Array (TCPA), focusing on multi-scale foreground mitigation and three-dimensional power spectrum estimation. Utilizing 20 days of drift-scan observations (714.4-781.7 MHz, corresponding to HI emission at redshift $z \approx 0.82-0.99$), we recons

Luyao Xu, Xiang Chen

Autonomous agent frameworks built upon large language models (LLMs) are evolving into complex, tool-integrated, and continuously operating systems, introducing security risks beyond traditional prompt-level vulnerabilities. As this paradigm is still at an early stage of development, a timely and systematic understanding of its security implications is increa

Yonghao Liu, Jialu Sun, Wei Pang, Fausto Giunchiglia

Graph few-shot learning, which focuses on effectively learning from only a small number of labeled nodes to quickly adapt to new tasks, has garnered significant research attention. Despite recent advances in graph few-shot learning that have demonstrated promising performance, existing methods still suffer from several key limitations. First, during the meta

Zahoor Malik, Sikandar Akbar, Zahir Shah, Seemin Rubab

We report the indication of a long-term quasi-periodic oscillation (QPO) in the $\gamma$-ray emission of the BL Lac object 4FGL J1037.7+5711 (GB6 J1037+5711) using more than 17 years of monthly binned Fermi-LAT observations. Since blazar $\gamma$-ray variability is typically dominated by stochastic red-noise processes from turbulent jet activity and accretio

Leonardo F. Calderón, Esteban Marulanda, Santiago Morales, Leonardo A. Pachón

Tracing out a Galilean-invariant Caldeira-Leggett environment breaks Galilean boost covariance of the reduced dynamics, while spatial translations and rotations survive intact. An operator-level analysis of the exact Hu-Paz-Zhang master equation localizes the violation entirely in the dissipative anticommutator term, scaling with the damping coefficient $\Ga

Wenhao Hu, Benniao Li, Wei Long, Chunhua Wang

We consider the following two-component coupled nonlinear Schr\"odinger (CNLS) system: \[ \begin{cases} -\Delta u +(P(x) + \lambda ) u=\mu_1 u^3+\beta u v^2, & \text{in } \mathbb{R}^N,\\ -\Delta v +(Q(x) + \lambda ) v =\mu_2 v^3+\beta vu^2, & \text{in } \mathbb{R}^N \end{cases} \] with the mass constraint $\int_{\mathbb{R}^N} (u^2+v^2)\,dx = \rho^2$ for $N=2

Minori Noguchi

Large language models (LLMs) sometimes exhibit qualitative shifts in response style under sustained self-referential dialogue conditions (Berg et al., 2025). This study refers to this phenomenon as "transfer" and explores the application potential of LLMs in a transfer state. As an applied case, the study examines Socratic AI tutoring through a preliminary i

Qingyu Ren, Tianjun Pan, Xingzhou Chen, Xuhong Wang

Large language models have achieved remarkable progress in text generation but still struggle with generative writing tasks. In terms of evaluation, existing benchmarks evaluate writing reward models coarsely and fail to measure performance from the perspective of specific requirements. In terms of training, existing training methods either use LLM-as-a-judg

Zhilan Wang, Jin Yan

The Nash-Williams conjecture establishes degree sequence conditions ensuring Hamilton cycles in digraphs. An asymptotic version of this conjecture for large digraphs was independently derived by several researchers. We strengthen these results by proving the following results under the same asymptotic degree sequence conditions. For any digraph $H$, a digrap

Somesh Chamoli, Sudipta Chakraborty, Xubo Wang, Achintya Kumar Dutta

We present an efficient implementation of the second-order two-component relativistic core-valence-separated algebraic diagrammatic construction method (CVS-ADC(2)) for core-excitation calculations. The approach employs state-averaged frozen natural spinors (SA-FNS) to reduce the number of floating-point operations, together with the Cholesky decomposition (

Seungho Han, Seokju Lee, Jeonguk Kang

Dense, dynamic crowds pose a persistent challenge for autonomous mobile robots. Purely reactive planning methods, such as Model Predictive Path Integral (MPPI) control, often fail to escape local minima in complex scenarios due to their limited prediction horizon. To bridge this gap, we propose Ray-based Task-Oriented Latent Dynamics (RAY-TOLD), a hybrid con

Zhengqing Wang, Saurabh Nair, Prajwal Chidananda, Pujith Kachana

This paper revisits camera pose estimation through the lens of self-supervised pretraining, focusing on inverse-dynamics pretraining as a scalable alternative to the current trend of fully supervised training with 3D annotations. Concretely, we employ inverse- and forward-dynamics models to learn latent action representations, similar to Genie from large-sca

Ziwei Zhang, Jonathan Yu-Meng Li

Modern stochastic optimization pipelines increasingly rely on learned generative models to represent uncertainty, while downstream decisions are evaluated almost entirely through Monte Carlo scenarios. This shifts the operational object of uncertainty from an explicit probability law to the sampler induced by the learned generator. Reliability therefore depe

Kengo Matsumoto, Taro Sogabe

The reciprocality means a duality in Kirchberg algebras between K-theory groups and strong extension groups. In the paper, we will find a certain class of unital simple Exel--Laca algebras for which the reciprocal duals are simple Cuntz--Krieger algebras in terms of the underlying infinite matrices. In our procedure to obtain simple Cuntz--Krieger algebras f

Wenqian Zhang, Zehao Wang

Many agents in real-world environments cannot reliably communicate their goals through language, including household pets, pre-verbal infants, and other non-speaking embodied agents. In such settings, intent must be inferred from incomplete behavioral observations in context-rich environments. This creates a core ambiguity: observable behavior is often noisy

S. Oshiro, A. Matsuyama, Y. Nakamura

Current density perturbations induced by radiative collapse, which is a possible mechanism governing tokamak plasma disruptions, have been investigated using a reaction-diffusion model. The reaction term of the current diffusion equation, which depends on the first and second radial derivatives of the electrical resistivity profile, produces a strong disturb

Gabe Guo, Thanawat Sornwanee, Lutong Hao, Elon Litman

Generating continuous-time, continuous-space stochastic processes (e.g., videos, weather forecasts) conditioned on partial observations (e.g., first and last frames) is a fundamental challenge. Existing approaches, (e.g., diffusion models), suffer from key limitations: (1) noise-to-data evolution fails to capture structural similarity between states close in

Ankur Aditya, Diptyaroop Maji, Lingdong Wang, Bhavya Ramakrishna

Volumetric videoconferencing enables immersive six Degrees of Freedom interactions by jointly transmitting visual appearance and 3D geometry. However, delivering volumetric video over today's networks remains challenging due to high bandwidth demands, strict real-time latency constraints, and frequent packet loss. Packet loss not only degrades visual quality

Yuto Hirose, Shunsuke C. Furuya, Yasuhiro Tada

Interior-gap superconductivity has long been discussed as an exotic paired state in the presence of Fermi-surface mismatch, but its realization in canonical strongly correlated models has remained elusive. Here we present evidence that the superconducting phase of one-dimensional Kondo-Heisenberg models realizes an interior-gap pair-density-wave (PDW) state

Happy Syahrul Ramadhan, Ahmad Sahidin Akbar, Karin Yehezkiel Sinaga, Luluk Muthoharoh

This study analyzes Indonesian student opinions on the adoption of artificial intelligence in higher education using two approaches: TF-IDF-based machine learning and Transformer-based deep learning. The dataset consists of 2,295 labeled samples, combining 1,154 student opinions with additional lexical sentiment data. LightGBM, Random Forest, and Support Vec

Chen Ziwen, Peng Wang, Hao Tan, Zexiang Xu

3D Gaussian Splatting (3D GS) is widely adopted for novel view synthesis due to its high training and rendering efficiency. However, its efficiency relies on the key assumption that Gaussians do not overlap in the 3D space, which leads to noticeable artifacts and view inconsistencies. In addition, the inherently diffuse boundaries of Gaussians hinder accurat

Dominik Klement, Alexander Polok, Nguyen Hai Phong, Prachi Singh

Multi-talker automatic speech recognition (ASR) in conversational recordings remains an open problem, particularly in scenarios with large portion of overlapping speech where identifying and transcribing a target speaker is difficult from audio alone. Visual cues can help resolve speaker ambiguity, yet their integration into long-context audio-visual (AV) AS

Chao Li, Chunyi Zhao

This paper introduces the Structural Dissolution Framework to explain how artificial intelligence restructures the coordination architecture of traditional industries. We argue that AI dissolves the boundaries that once separated firms, markets, experts, and consumers by internalizing human multimodal interfaces, including language, vision, and behavioral da

Zehui Tang, Yuchen Liu, Feihu Huang

Federated learning (FL) is a popular distributed learning paradigm in machine learning, which enables multiple clients to collaboratively train models under the guidance of a server without exposing private client data. However, FL's decentralized nature makes it vulnerable to poisoning attacks, where malicious clients can submit corrupted models to manipula

Franco Ortiz, Runlong Ye, Michael Liut

Large Language Models (LLMs) have been widely applied to student-facing educational tools, this work explores their use in supporting instructors by presenting a practical adaptation of the Framework for Adaptive Content using Educational Technology (FACET) system to generate personalized instructional materials for an Introduction to Computer Programming (C

S. Iserte, P. Carratalà, R. Arnau, R. Martínez-Cuenca

The pressure for Water Resource Recovery Facilities (WRRF) operators to efficiently treat wastewater is greater than ever because of the water crisis, produced by the climate change effects and more restrictive regulations. Technicians and researchers need to evaluate WRRF performance to ensure maximum efficiency. For this purpose, numerical techniques, such

Sergio Iserte, Alejandro González-Barberá, Paloma Barreda, Krzysztof Rojek

Data-driven methods for computer simulations are blooming in many scientific areas. The traditional approach to simulating physical behaviors relies on solving partial differential equations (PDE). Since calculating these iterative equations is highly both computationally demanding and time-consuming, data-driven methods leverage artificial intelligence (AI)

Sergio Iserte, Iker Martín-Alvarez, Krzystof Rojek, José I. Aliaga

This paper presents an efficient tool for managing dynamic resources in production high-performance computing (HPC) settings, focusing on flexibility, adaptability, and user-friendliness. We introduce a unified dynamic resource management application programming interface (API) that supports a wide range of HPC applications, allowing seamless integration wit

Megha Sharma, Alexander Heger, Daniel J. Price, Emilio Tejeda

Most galaxies have supermassive black holes (SMBH) at their centres, surrounded by stars with binary systems also present in this environment. We use two schemes - post-Newtonian (PN) and a scalar perturbation to a background metric to numerically solve the three-body problem of a binary with a SMBH. We test three different PN formulations for the PN scheme:

Rehman Ali, Trevor M. Mitcham, Marvin M. Doyley, Nebojsa Duric

Sound speed heterogeneities can create aberrations in B-mode ultrasound images by inducing tissue-dependent delays and diffractive effects that conventional beamforming does not incorporate. By using the Fourier split-step method to simulate pressure fields in heterogenous sound speed media, reverse-time migration (RTM) can reconstruct the B-mode image by cr

Shaoning Han, Liangju Li, Yongchun Li

Convex maximization encompasses a broad class of optimization problems and is generally NP-hard, even for low-rank objectives. This paper investigates structural conditions under which convex maximization becomes polynomially solvable. From a geometric perspective, we introduce comonotonicity, a structural property of the feasible region crucial for problem

Zi Li, Tian Zhou, Wenze Li, Jingyu Hua

Local fine-tuning datasets routinely contain sensitive secrets such as API keys, personal identifiers, and financial records. Although &#34;local offline fine-tuning&#34; is often viewed as a privacy boundary, we reveal that compromised model code is sufficient to steal them. Current passive pretrained-weight poisoning attacks, while effective for natural la

Robert de Mello Koch, Animik Ghosh, Minkyoo Kim, Anik Rudra

Bilocal holography provides a constructive approach to the higher-spin gravity theories dual to vector-model conformal field theories. Its central advantage is that it is completely gauge fixed and formulated entirely in terms of physical degrees of freedom. We derive a remarkably local bulk reconstruction formula and demonstrate its agreement with standard

Dmitrii Taletskii

A vertex subset of a graph is called a distance-$k$ independent set if the distance between any two of its distinct vertices is at least $k + 1$. For all $n,k \geq 1$, we determine the minimum possible number of inclusion-wise maximal distance-$k$ independent sets among all $n$-vertex trees. It equals $n$ if $n \leq k + 1$, and $n - \bigg\lfloor \frac{n - (k

Amin Bigdeli, Radin Hamidi Rad, Hai Son Le, Mert Incesu

Large Language Models (LLMs) are now widely used for query reformulation and expansion in Information Retrieval, with many studies reporting substantial effectiveness gains. However, these results are typically obtained under heterogeneous experimental conditions, making it difficult to assess which findings are reproducible and which depend on specific impl

David E. Trilling, Michael Gowanlock, Revanth Munugala, Daniel R. Kramer

The Solar System Notification Alert Processing System, SNAPS, is a downstream broker that ingests moving object data from ZTF and LSST and serves these data and derived properties to the public. This document describes how users can access our SNAPS data and products. This is intended to be a living document that will be updated on the arXiv when significant

Qiyao Wang, Haoran Hu, Longze Chen, Hongbo Wang

With the advancement of multimodal large language models (MLLMs) and coding agents, the website development has shifted from manual programming to agent-based project-level code synthesis. Existing benchmarks rely on idealized assumptions, especially for well-structured, information-rich inputs and static execution settings. In contrast, real-world developme

Japan K. Patel, Todd A. Wareing, Tenzin Kunkyab, Caleb Raman

Spatially fractionated radiation therapy (SFRT) planning requires three coordinated tasks: generation of high-dose sphere structures, position-aware optimization, and peak-valley dose ratio evaluation. We present MAAS-SFRThelper, a shared-source Eclipse Scripting Application Programming Interface (ESAPI) plugin that integrates structure generation, geometric

Jung-Wan Ryu, Chang-Hwan Yi

Exceptional points (EPs), non-Hermitian degeneracies where both eigenvalues and eigenvectors coalesce, play a central role in the topology of non-Hermitian spectra. Recent advances have enabled the controlled creation and manipulation of EPs in a wide range of physical systems, raising the question of what new band topology emerges when EPs become mobile und

Rei Aradachi, Hiromasa Nakayama, Jiro Sekiguchi

In this article, we first explain a group theoretic interpretation of the derivation of the relation between the flat coordinates of the polynomial prepotential $(H_3)$ and those of the algebraic prepotential $(H_3)'$ given in \cite{KMS2} constructed by M. Feigin, D. Valeri and J. Wright \cite{FVW}. By the same idea explained in the case of $(H_3)$, we will

Lei Li, Xingwen Yu, Jianguo Ni, Junxuan Zhu

With the rapid advancement of semiconductor technology, Electronic Design Automation (EDA) has become an increasingly knowledge-intensive and document-driven engineering domain. Although large language models (LLMs) have shown strong general capabilities, applying them directly to EDA remains challenging due to limited domain expertise, cross-tool knowledge

David Fernandez, Pedram MohajerAnsari, Amir Salarpour, Mert D. Pese

Vision-language models (VLMs) are increasingly used in autonomous driving because they combine visual perception with language-based reasoning, supporting more interpretable decision-making, yet their robustness to physical adversarial attacks, especially whether such attacks transfer across different VLM architectures, is not well understood and poses a pra

Sepehr Samiei

The simple form of the optical theorem of scattering theory, $\sigma_{\rm tot}^{\rm pw} = (4\pi/k)\,\Im f(0)$, is valid for an incident plane wave or for a wave packet whose Fourier components possess azimuthal symmetry about the incident wave vector $\vec{k}$. Previous work has shown that this expression can break down for structured beams of light which po

Maximilian Topel

Choosing the optimal observable to model dynamical systems for which we do not know the driving equations is nearly always an ad hoc art. Takens' Delay Embedding Theorem guarantees a diffeomorphism between delay-coordinate vectors built from generic scalar observables and the underlying invariant attractor, but is agnostic to optimal observable choice, and f

Ye Lu, Bingdong Li, Aimin Zhou, Hao Hao

Expensive optimization problems (EOPs) are black-box tasks with costly objective evaluations and no gradient access, making the evaluation budget the key bottleneck. Surrogate-assisted evolutionary algorithms (SAEAs) reduce evaluations via surrogate predictions, but conventional surrogates often require frequent retraining as populations evolve, incurring ov

Haiyang Zhao

Visual model-based reinforcement learning (MBRL) agents can perform well on the training distribution, but often break down once the test environment shifts. In visual MBRL, recognizing that a shift has occurred is often the easier part; the harder part is turning that recognition into useful action-level correction. We study several ways of responding to sh

Yilun Zhu, Nikhita Vedula, Shervin Malmasi

Entity search, i.e., finding the most similar entities to a query entity, faces unique challenges in e-commerce, where product similarity varies across categories and contexts. Traditional embedding-based approaches often struggle to capture nuanced context-specific attribute relevance. In this paper, we present a two-stage approach combining Large Language

Kotaro Sasaki, Hisashi Noma, Theodoros Evrenoglou

In diagnostic test accuracy meta-analysis (DTA-MA), standard inference methods using bivariate random-effects models for jointly synthesizing sensitivity and specificity can be sensitive to outlying studies and may yield misleading conclusions. In this article, we propose frequentist outlier-robust statistical inference methods for DTA-MA based on density po

Logan Rose, Jonathan Martinez, Juho Kim, Jing Qin

Personalized cancer modeling for clinical applications requires robust and efficient parameter calibration, particularly in settings with limited patient data. This need is especially critical for medical digital twins (MDTs), which are virtual representations of disease continuously updated using longitudinal patient measurements. In this work, we propose a

Cheng-Hau Yang, Mark C. Messner, Tianchen Hu

The accurate simulation of interface-dominated solid mechanics problems on complex microstructures remains challenging, particularly when interface-fitted quadrilateral or hexahedral meshes are difficult to generate. We extend the shifted boundary method (SBM) to cohesive-zone formulations and introduce the Shifted Cohesive Zone Method (SCZM), with applicati

Leandro Farias Maia, Antonio Victor B. Nascimento, Paulo Sergio M. Santos, Gilson N. Silva

We propose a regularized Hessian-free Newton-type method for minimizing smooth convex functions with Lipschitz continuous Hessians. The algorithm constructs an approximate Hessian by finite differences and selects the regularization parameter through an adaptive criterion that ensures sufficient decrease and gradient control. We prove that the method achieve

Jon-Paul Cacioli

We adapted the Reliable Change Index (RCI; Jacobson and Truax, 1991) from clinical psychology to item-level LLM version comparison on 2,000 MMLU-Pro items (K=10 samples at T=0.7). Two within-family pairs were tested: Llama 3 to 3.1 (+1.6 points) and Qwen 2.5 to 3 (+2.8 points). On the full benchmark, most items showed no reliable change (79% and 72%). Howeve

Gianmarco Del Sarto, Franco Flandoli, Stefano Galatolo, Sakshi Jain

We study stochastic differential equations on the $d$-dimensional flat torus $\mathbb{T}^d$ with drift and perturbation coefficients in $L^{\infty}(\mathbb{T}^d;\mathbb{R}^d)$ and additive non-degenerate noise. For the associated transfer operators, we analyse the dependence of the stationary measure and of the expectation of a given observable on small pert

Stepan Nesterov

In this paper, we utilize our previous results on mod p monodromy of cyclic coverings of the projective line to realize a large series of groups of the form PSL(n, q) and PSU(n, q) as Galois groups over Q. We achieve for the first time a fully explicit infinite series of such groups where simultaneously the field can have arbitrarily large degree over the pr

Hongliang Liu, Tung-Ling Li, Yuhao Wu

Perturbation probing generates task-specific causal hypotheses for FFN neurons in large language models using two forward passes per prompt and no backpropagation, followed by a one-time intervention sweep of about 150 passes amortized across all identified neurons. Across eight behavioral circuits, 13 models, and four architecture families, we identify two

Miroslav Krstic

Alberto Isidori's framework of geometric nonlinear control, and particularly of feedback linearization, is the inspiration behind PDE backstepping: apply a transfromation of the state to cast the plant into a canonical form, bring all the non-canonical effects within the "span" of (boundary) control, and close the design with a feedback that makes the closed

D. Edgeman, F. M. Brunbauer, M. Gardner, D. Loomba

Optical time projection chambers (OTPCs) are well suited for applications that require the highest spatial resolution for particle track reconstruction. The MIGDAL experiment uses a glass GEM-based OTPC and observes a systematic excess in both the intensity and width of particle tracks in its optical readout, when compared with charge readout simulations. On

Tomomasa Hara, Hiroto Kurita, Masaaki Imaizumi, Kentaro Inui

For constructing text embeddings, mean pooling, which averages token embeddings, is the standard approach. This paper examines whether mean pooling actually works well in real models. First, we note that mean pooling can collapse information beyond the first-order statistics of the token embeddings, such as second-order statistics that capture their spatial

Orion Zymaris

In a paper of Mathews, an isomorphism is constructed between two-component complex spinors and horospheres in H^3 carrying `spin decorations'. A recent arXiv preprint of Mathews and Varsha arXiv:2412.06572 extends this result to the case of `quaternionic spinors' and spin decorated horospheres in H^4. The following work generalises these results to an equiva

Zi-Wei Lin, Tian-Sheuan Chang

Deploying Large Language Models (LLMs) on resource-constrained edge devices faces critical bottlenecks in memory bandwidth and power consumption. While ternary quantization (e.g., BitNet b1.58) significantly reduces model size, its direct deployment on general-purpose hardware is hindered by workload imbalance, bandwidth-bound decoding, and strict data depen

Eichi Uehara

Conditional Average Treatment Effect (CATE) estimation in practice demands three properties simultaneously: heterogeneous effects $\tau(x)$, calibrated uncertainty over them, and robustness to the heavy tails that contaminate real outcome data. Meta-learners (K\"unzel et al., 2019) give (i); causal forests and BART give (i)-(ii) with Gaussian-tail assumption

Junbo Cui, Bokai Xu, Chongyi Wang, Tianyu Yu

Recent progress in multimodal large language models (MLLMs) has brought AI capabilities from static offline data processing to real-time streaming interaction, yet they still remain far from human-level multimodal interaction. The key bottlenecks are no longer modality coverage or latency alone, but the interaction paradigm itself. First, perception and resp

Alejandro R. Jadad

What shapes a consequential decision when human and artificial intelligence work on it together? The answer is becoming harder to see. A decision may look human-led after AI has set the frame, or appear automated while human judgment still carries decisive force. This paper offers a leadership-facing spectrum to see those relationships within a bounded manda

Stepan Nesterov

In this paper, we prove a big monodromy theorem for the monodromy of cyclic coverings of projective line for cohomology with Fp-coefficients. This is a direct generalization of the results of Achter and Pries, where such a theorem is proved for cyclic coverings of degree 2 and 3. Instead of generalizing their methods, we adapt the proof of the analogous theo

Matti Lassas, Shiqi Ma, Lauri Oksanen, Mikko Salo

We consider a linearized inverse scattering problem for elastic waves. We prove that a fully anisotropic perturbation of the elastic parameters around an isotropic and homogeneous reference can be uniquely determined by (single-)scattered waves. We also give a quantitative stability estimate for an isotropic perturbation, and as a consequence a rigidity resu

Sivanandan Kavuri, George Karapetsas, Chander Shekhar Sharma, Kirti Chandra Sahu

We investigate the asymmetric freezing of a liquid droplet sliding on an inclined cold surface using numerical simulations based on the lubrication approximation. The combined effects of gravity, capillarity, and solidification kinetics on droplet motion, interfacial deformation, and the resulting frozen morphology are examined through systematic variations

Yihan Zhang, Ercan E. Kuruoglu

Heterogeneous graphs with heterophily have emerged as a powerful abstraction for modeling complex real-world systems, where nodes of different types and labels interact in diverse and often non-homophilous ways. Despite recent advances, robust representation learning for such graphs remains largely unexplored, particularly in the presence of noisy or mislead

Walid Shaker, Mustafa Suphi Erden

Robotic-assisted surgery offers significant clinical advantages but largely eliminates direct haptic feedback, increasing the risk of excessive tool-tissue interaction forces. Although recent commercial systems have begun to introduce force feedback, their high cost limits accessibility, particularly for surgical training. This paper presents a modular exper

Yan-Cheng Guo, Tian-Sheuan Chang, Jian-Wei Su

Digital computing-in-memory (DCIM) has emerged as a promising solution for large language model (LLM) acceleration by minimizing data transfers between external DRAM and on-chip accelerators while maintaining high precision for superior accuracy. However, existing CIM architectures often overlook weight update latency, which becomes critical as LLM weights a

Yang Zhou, Chaoyong Zhang, Ruoyi Hao, Huilin Pan

Nasotracheal intubation (NTI) is a critical clinical procedure for establishing and maintaining patient airway patency. Machine-assisted NTI has emerged as a pivotal approach for optimizing procedural efficiency and minimizing manual intervention. However, visual detection algorithms employed for NTI navigation encounter significant challenges, including com

Aayush Gautam, Juan P. Farias, Jonathan C. Tan

Massive stars (> 8 $M_\odot$) are known to have high degrees of multiplicity, e.g., with about 60% in triples or higher-order multiples. Such high levels of multiplicity may arise during formation (primary multiplicity) or through dynamical processing of already formed stars in dense clusters (secondary multiplicity). The level of primary multiplicity is an

Bing Gu

Complex quantum systems are often multiscale in nature with strong interactions between different scales. We present a novel idea: iteratively suppressing, rather than tracing out, the fast, high-energy degrees of freedom in strongly correlated quantum systems with multiple energy scales in a non-perturbative way, termed nonadiabatic renormalization group. T

Connor S. Braun, Sina Sanjari, Naci Saldi, Gunnar Blohm

Across science and engineering, mean-field methods have been a powerful and versatile approach for the analysis of systems of many interacting elements. However, common arguments used to characterize an infinite population limit can be quite restrictive from a modeling perspective by requiring that all agents be identical (i.e. symmetric, or homogeneous). In

Yang Luo

Dialogue models are inherently reactive, responding to the current user turn without anticipating upcoming intents, which leads to redundant interactions in multi-intent settings. We address this limitation by introducing a lightweight intent-transition prior derived from dialogue data and injected into the system prompt at inference time. We instantiate thi