December 2025 arXiv papers — page 100

Xiang-Yang Liu, Xiang-Feng Pan, Yong-Yi Jin, Li-Cheng Li

The notion of resistance distance, introduced by Klein and Randi\'c, has become a fundamental concept in spectral graph theory and network analysis, as it captures both the structural and electrical properties of a graph. The associated resistance spectrum serves as a graph invariant and plays an important role in problems related to graph isomorphism. For a

Yasuhiro Matsumoto, Kei Matsushima

This paper proposes a direct-indirect mixed Burton-Miller boundary integral equation for solving Helmholtz scattering problems with transmissive scatterers. The proposed formulation has three unknowns, one more than the number of unknowns for the ordinary formulation. However, we can construct efficient numerical solvers based on LU factorization by exploiti

Alexandr Malijevský

We present a macroscopic theory of capillary condensation in slits formed by parallel walls of unequal length. Using the concept of an edge contact angle, we identify four distinct condensation states and derive Kelvin-like relations for their onset. The resulting phase diagrams, expressed in terms of wall geometry and contact angle, reveal two central organ

Caolin Ua Tuiscint, Jonathan N Coleman

Lithium-ion batteries dominate battery research and industry due to their long research history and high energy density. However, increasing demand and limited lithium resources have raised lithium prices and battery costs, motivating interest in alternative chemistries. Potassium-ion batteries have recently attracted attention because potassium is more abun

Stefano Goria, Levent A. Mengütürk, Murat C. Mengütürk, Berkan Sesen

This paper motivates the use of random-bridges -- stochastic processes conditioned to take target distributions at fixed timepoints -- in the realm of generative modelling. Herein, random-bridges can act as stochastic transports between two probability distributions when appropriately initialized, and can display either Markovian or non-Markovian, and either

Johannes A. Gaus, Daniel Häufle, Woo-Jeong Baek

While many visual odometry (VO), visual-inertial odometry (VIO), and SLAM systems achieve high accuracy, the majority of existing methods miss to assess risks at runtime. This paper presents SUPER (Sensitivity-based Uncertainty-aware PErformance and Risk assessment) that is a generic and explainable framework that propagates uncertainties via sensitivities f

Wei Tao, Sheng Long, Xin Liu, Wei Li

Generating adversarial examples (AEs) can be formulated as an optimization problem. Among various optimization-based attacks, the gradient-based PGD and the momentum-based MI-FGSM have garnered considerable interest. However, all these attacks use the sign function to scale their perturbations, which raises several theoretical concerns from the point of view

Xiaoning Lei, Jianwei Sun, Wenhao Cai, Xichen Xu

Task-based measures of image quality (IQ) are critical for evaluating medical imaging systems, which must account for randomness including anatomical variability. Stochastic object models (SOMs) provide a statistical description of such variability, but conventional mathematical SOMs fail to capture realistic anatomy, while data-driven approaches typically r

Roberto Emparan, Jordi Rafecas-Ventosa, Benson Way

We derive the general form of the effective equations governing black hole dynamics in the limit of a large number of dimensions $D$. These split into a universal \emph{soap-bubble} embedding condition for stationary configurations and a set of nonlinear dynamical evolution equations describing near-horizon fluctuations of $O(1/D)$ amplitude over horizon sca

Alberto Rebassa-Mansergas, Roberto Raddi, Anna F. Pala, Alejandro Santos-García

Deriving precise stellar ages is a challenging task. Consequently, age-dependent relations - such as the age-metallicity and age-velocity dispersion relations of the Milky Way, or the age-rotation-activity relation of low-mass stars - are subject to potentially large uncertainties, despite the well-defined trends observed at the population level. White dwarf

Emanuele Artioli, Farzad Tashtarian, Christian Timmerer

The primary challenge of video streaming is to balance high video quality with smooth playback. Traditional codecs are well tuned for this trade-off, yet their inability to use context means they must encode the entire video data and transmit it to the client. This paper introduces ELVIS (End-to-end Learning-based VIdeo Streaming Enhancement Pipeline), an en

Gill Barequet, Sariel Har-Peled

The 3SUM problem represents a class of problems conjectured to require $\Omega (n^2)$ time to solve, where $n$ is the size of the input. Given two polygons $P$ and $Q$ in the plane, we show that some variants of the decision problem, whether there exists a transformation of $P$ that makes it contained in $Q$, are 3SUM-Hard. In the first variant $P$ and $Q$ a

Tsuyoshi Kato, Daisuke Kishimoto, Nobuhiro Nakamura, Kouichi Yasui

By extending the notion of simple type for the Seiberg--Witten invariant of a 4-manifold, we introduce notions of BF blowup simple type and BF homogeneous type for the Bauer--Furuta invariant and study their applications. Specifically, we show that the existence of an immersed 2-sphere with a certain condition guarantees BF blowup simple type. As an applicat

Ling-Zhi Tang, Xiao Li, Z. D. Wang, Dan-Wei Zhang

Discrete time crystals (DTCs) are non-equilibrium phases of matter that break the discrete time-translation symmetry and is characterized by a robust subharmonic response in periodically driven quantum systems. Here, we explore the DTC in a disorder-free, periodically kicked XXZ spin chain, which is stabilized by the Floquet strong Hilbert space fragmentatio

Karri Koljonen, Claudio Ricci, Thomas Stanke, Doug Johnstone

The millimeter/sub-millimeter (mm/sub-mm) sky remains a rich but under-explored frontier for transient and variable phenomena. A wide-field, high-sensitivity instrument with a large aperture and degree-scale field of view would open this regime, enabling both systematic survey monitoring and rapid-response follow-up. Key science opportunities include Galacti

Shaolun Ruan, Feng Liang, Rohan Ramakrishna, Chao Ren

Quantum Neural Networks (QNNs) represent a promising fusion of quantum computing and neural network architectures, offering speed-ups and efficient processing of high-dimensional, entangled data. A crucial component of QNNs is the encoder, which maps classical input data into quantum states. However, choosing suitable encoders remains a significant challenge

K. M. Jubair Sami, Dipto Sumit, Ariyan Hossain, Farig Sadeque

Translating from a standard language to its regional dialects is a significant NLP challenge due to scarce data and linguistic variation, a problem prominent in the Bengali language. This paper proposes and compares two novel RAG pipelines for standard-to-dialectal Bengali translation. The first, a Transcript-Based Pipeline, uses large dialect sentence conte

Suratno Basu, Krishanu Dan, Aanjaneya Rath

In this article we study the stability of Kernel sheaf obtained from a generating subspace of rank one torsion-free sheaf on an integral nodal curve.

N. Ponthieu, F. -X. Désert, A. Beelen, R. Adam

The NIKA2 Cosmological Legacy Survey (N2CLS) is a large programme using the NIKA2 dual-band camera on the IRAM 30\,m telescope. Its goal is to improve our understanding of the physics of distant Dusty Star Forming Galaxies (DSFGs) by carrying out deep surveys of two fields, GOODS-North and COSMOS. This work is focussed on GOODS-North, which was observed for

Qijun Zhang, Shang Liu, Yao Lu, Mengming Li

Power is a primary objective in modern processor design, requiring accurate yet efficient power modeling techniques. Architecture-level power models are necessary for early power optimization and design space exploration. However, classical analytical architecture-level power models (e.g., McPAT) suffer from significant inaccuracies. Emerging machine learnin

Jonathan Spiegelman, Guy Amir, Guy Katz

Deep Active Learning (DAL) aims to reduce labeling costs in neural-network training by prioritizing the most informative unlabeled samples for annotation. Beyond selecting which samples to label, several DAL approaches further enhance data efficiency by augmenting the training set with synthetic inputs that do not require additional manual labeling. In this

Unai Laskurain, Aitor Aguirre-Ortuzar, Urko Zurutuza

Federated Learning (FL) is an emerging machine learning paradigm that enables multiple parties to collaboratively train models without sharing raw data, ensuring data privacy. In Vertical FL (VFL), where each party holds different features for the same users, a key challenge is to evaluate the feature contribution of each party before any model is trained, p

Parsa Kafashi, Mozhgan Orujlu

We present a quantum information-inspired framework for analyzing complex systems through multivariate time series. In this approach the system's state is encoded into a density matrix, providing a compact representation of higher-order correlations and dependencies. This formulation enables precise quantification of the relative influence among time series,

Juan Diego Mazo-Vasquez, Julius T. Gohsrich, Flore K. Kunst, Lewis Hill

Nonlinear phenomena such as optical bistability and spontaneous symmetry breaking play a central role in Kerr resonators, and are increasingly exploited in photonic integrated circuits for all-optical information processing. In this work, we present an analytical framework allowing to find the stationary states and their bifurcations for the propagating fiel

Jakub Kušnír, Tomáš Grabec, Petr Sedlák, Pavla Stoklasová

Transient grating spectroscopy (TGS) is a material characterization technique based on laser-induced thermoelastic excitation of thermal and acoustic gratings. On opaque samples, these gratings are dynamic surface displacements that reflect the sample's elastic and thermal properties, enabling both types of parameters to be determined from a single experimen

Yunhao Yao, Zhiqiang Wang, Haoran Cheng, Yihang Cheng

The evolution of Large Language Models (LLMs) into Agentic AI has established the Model Context Protocol (MCP) as the standard for connecting reasoning engines with external tools. Although this decoupled architecture fosters modularity, it simultaneously shatters the traditional trust boundary. We uncover a novel privacy vector inherent to this paradigm: th

Wenzhuo Zou, Ming-Min Zhao, An Liu, Min-Jian Zhao

Robust beamforming design under imperfect channel state information (CSI) is a fundamental challenge in multiuser multiple-input multiple-output (MU-MIMO) systems, particularly when the channel estimation error statistics are unknown. Conventional model-driven methods usually rely on prior knowledge of the error covariance matrix and data-driven deep learnin

Evgeny A. Mityushov

We develop a geometric atlas of dynamical regimes on the rotation group SU(2), combining geodesic flows, heavy rigid body dynamics, and a curvature-based decomposition of the Euler-Poisson equations. We represent the equations of motion in a curvature form that interprets rigid-body motion as the interaction of inertial and external curvature fields. This un

Ole Løseth Elvetun, Bjørn Fredrik Nielsen, Niranjana Sudheer

We investigate the weighted Group Lasso formulation for the static inverse electroencephalography (EEG) problem, aiming at reconstructing the unknown underlying neuronal sources from voltage measurements on the scalp. By modelling the three orthogonal dipole components at each location as a single coherent group, we demonstrate that depth bias and orientatio

Qingyuan Cai, Linxin Zhang, Xuecai Hu, Saihui Hou

Recent approaches for monocular 3D human pose estimation (3D HPE) have achieved leading performance by directly regressing 3D poses from 2D keypoint sequences. Despite the rapid progress in 3D HPE, existing methods are typically trained and evaluated under disparate frameworks, lacking a unified framework for fair comparison. To address these limitations, we

Keiichi Ishikawa, Yuma Matsumoto, Taro Yaoyama, Sangwon Lee

In a performance based earthquake engineering (PBEE) framework, nonlinear time-history response analysis (NLTHA) for numerous ground motions are required to assess the seismic risk of buildings or civil engineering structures. However, such numerical simulations are computationally expensive, limiting the real-world practical application of the framework. To

Christophe Sun, Pierre-Olivier Michel, François David, Nathalie Rouach

Objectives. Accurately predicting transitions to anesthetic drugs overdosage is a critical challenge in general anesthesia as it requires the identification of EEG indicators relevant for anticipating the evolution of the depth of anesthesia. Methods. In this study, we introduce a real-time, data-driven framework based on alpha spindle dynamics extracted fro

Wei-Kang Xie, Guo-Fu Yu

We derive a novel variant of the Blaszak-Szum lattice equation by introducing a new class of trigonometric-type bilinear operators. By employing Hirota's bilinear method, we obtain the Gram-type determinant solution of the variant Blaszak-Szum lattice equation. One-soliton and two-soliton solutions are constructed, with a detailed analysis of the asymptotic

Zixun Luo, Yuhang Fan, Yufei Li, Youzhi Zhang

The rapid expansion of AI-driven applications powered by large language models has led to a surge in AI interaction data, raising urgent challenges in security, accountability, and risk traceability. This paper presents AiAuditTrack (AAT), a blockchain-based framework for AI usage traffic recording and governance. AAT leverages decentralized identity (DID) a

Shuxin Zhao, Bo Lang, Nan Xiao, Yilang Zhang

Object detection models deployed in real-world applications such as autonomous driving face serious threats from backdoor attacks. Despite their practical effectiveness,existing methods are inherently limited in both capability and robustness due to their dependence on single-trigger-single-object mappings and fragile pixel-level cues. We propose CIS-BA, a n

Daiki Watarai, Kent Yagi, Shammi Tahura

Binary black hole mergers with asymmetric component masses are key targets for both third-generation ground-based and future space-based gravitational-wave (GW) detectors, offering unique access to the strong-field dynamics of gravity. The evolution is commonly divided into three stages: the adiabatic inspiral, the transition, and the plunge. To date, constr

Precious Ogbonda Amadi, Paphon Pewkhom, Pruet Kalasuwan, Norshamsuri Ali

We examine how the informational properties of a confined single ion response in a Paul trap modified by optical-lattice. We focus on the ground and first excited motional states and show that Fisher information, Shannon entropy, and Fisher-Shannon complexity track the effective frequency $ω_{\mathrm{eff}}=ω\sqrt{1-κ}$ of the potential. We show that the Fish

Michael Mandl, Erhard Seiler, Dénes Sexty

We discuss recent developments regarding the use of kernels in complex Langevin simulations. In particular, we outline how a kernel can be used to solve the problem of wrong convergence in a simple toy model. Since conventional correctness criteria for complex Langevin results are only necessary but not sufficient, the correct convergence of complex Langevin

Frank Aurzada, Kilian Raschel

We study the persistence probabilities of a moving average process of order one with innovations that follow a Laplace distribution. The persistence probabilities can be computed fully explicitly in terms of classical combinatorial quantities like certain $q$-Pochhammer symbols or $q$-deformed analogues of Euler's zigzag numbers, respectively. Similarly, the

Songze Liu, Hongkun Du, Shaowen Wang

Large Language Models (LLMs), such as GPT and LLaMA, introduce unique memory access characteristics during inference due to frequent token sequence lookups and embedding vector retrievals. These workloads generate highly irregular and bursty access patterns, causing traditional prefetching and replacement policies to mispredict and trigger severe cache pollu

Zhijie Zhong, Zhiwen Yu, Pengyu Li, Jianming Lv

Radio path loss prediction (RPP) is critical for optimizing 5G networks and enabling IoT, smart city, and similar applications. However, current deep learning-based RPP methods lack proactive environmental modeling, struggle with realistic multi-transmitter scenarios, and generalize poorly under distribution shifts, particularly when training/testing environ

Shuo Lu, Yinuo Xu, Jianjie Cheng, Lingxiao He

Deep Research agents predominantly optimize search policies to maximize retrieval probability. However, we identify a critical bottleneck: the retrieval-utilization gap, where models fail to use gold evidence even after it is retrieved, due to context blindness in noisy environments. To bridge this gap, we propose DeepResearch-Slice, a simple yet effective n

Vadim Alekseev, Andreas Thom

We show that any isometric action of a residually finite group admits approximate local finite models. As a consequence, if $G$ is residually finite, every isometric $G$-action embeds isometrically into a metric ultraproduct of finite isometric $G$-actions.

K. Aditya, A. Mangalam

We construct mass models of Milky Way dwarf spheroidal galaxies to place constraints on the central black hole (BH) masses they can host. We model the galaxies as a three-component system consisting of the stars, dark matter halo, and a central black hole, using the Osipkov--Merritt--Cuddeford class of anisotropic distribution function. The posterior distrib

Min Zhang, Yudong Luo, Akira Dohi, Xing Xu

The recent mass measurement of $^{70}$Kr using the $B\rho$-defined isochronous mass spectrometry yields a mass excess of $-41320(140)$ keV, indicating a 220-keV increase in binding energy compared to the AME2020 prediction. We utilize this experimental mass -- the last piece of information needed -- to model the potential waiting point $^{68}$Se in $rp$-proc

Shanshan Ding, Jose Antonio Valerrama Botia, Aleksi Julku, Zhigang Wu

The rapidly growing class of atomically thin and tunable van der Waals materials is intensely investigated both in the context of fundamental science and for new technologies. There is in this connection a widespread need for new ways to probe the electronic properties of these layered materials, since their two-dimensional (2D) character make conventional p

Giuliana Cosentino, Izaskun Jiménez-Serra, Laura Colzi, Víctor Rivilla

There is growing evidence that the Sun might have formed within a nebula impacted by at least one SNR. In this scenario, ejecta and shocks from SNRs may have provided the elements on which life as we know it is based. Investigating the chemical complexity of molecular clouds impacted by SNRs is therefore essential to unveil the star formation process and how

Hongqiu Ni, Jiabao Zhang, Guopeng Li, Zilong Wang

Large Language Models (LLMs) are increasingly being deployed as intelligent agents. Their multi-stage workflows, which alternate between local computation and calls to external network services like Web APIs, introduce a mismatch in their execution pattern and the scheduling granularity of existing inference systems such as vLLM. Existing systems typically f

Hanning Chen, Keyu Man, Kevin Zhu, Chenguang Zhu

Identifying and addressing performance anti-patterns in machine learning (ML) models is critical for efficient training and inference, but it typically demands deep expertise spanning system infrastructure, ML models and kernel development. While large tech companies rely on dedicated ML infrastructure engineers to analyze torch traces and benchmarks, such r

Julian Speith, Jörn Langheinrich, Marc Fyrbiak, Max Hoffmann

HAL is an open-source framework for gate-level netlist analysis, an integral step in hardware reverse engineering. It provides analysts with an interactive GUI, an extensible plugin system, and APIs in both C++ and Python for rapid prototyping and automation. In addition, HAL ships with plugins for word-level modularization, cryptographic analysis, simulatio

So Kuroki, Manami Nakagawa, Shigeo Yoshida, Yuki Koyama

Many real-world tasks, such as trip planning or meal planning, can be formulated as combinatorial optimization problems. However, using optimization solvers is difficult for end users because it requires problem instantiation: defining candidate items, assigning preference scores, and specifying constraints. We introduce LAPPI (LLM-Assisted Preference-based

Ashish Mishra, Tarun Kumar, Gyanaranjan Nayak, Arpit Shah

We introduce a novel, closed-form approach for selective unlearning in multimodal models, specifically targeting pretrained models such as CLIP. Our method leverages nullspace projection to erase the target class information embedded in the final projection layer, without requiring any retraining or the use of images from the forget set. By computing an orth

Xiaojie Tao, Rajit Gadh

High renewable penetration has significantly reduced system inertia in modern power grids, increasing the need for fast frequency response (FFR) from distributed and non-traditional resources. While electric vehicles (EVs), data centers, and battery energy storage systems (BESS) have each demonstrated the capability to provide sub-second active power support

Sami Lakka

We develop a geometric framework for irreversible transport phenomena in which macroscopic evolution equations arise from the combined structure of a thermodynamic state metric and an Onsager-based dissipation metric. The construction begins by defining a pseudo-Riemannian manifold from the Hessian of an appropriate thermodynamic potential. When the enthalpy

Yijun Chen, Shanpu Shen, Tianrui Qiao, Hongyu Li

This paper investigates antenna coding based on pixel antennas as a new degree of freedom for enhancing multiple-input multiple-output (MIMO) wireless power transfer (WPT) systems. Antenna coding is closely related to the Fluid Antenna System (FAS) concept and further generalizes the radiation pattern reconfigurability. We first introduce a beamspace channel

Sungwoo Kang

Chae and Kang (2019, \textit{Pacific-Basin Finance Journal}) documented a puzzling Low Volume Return Premium (LVRP) in Korea -- contradicting global High Volume Return Premium (HVRP) evidence. We resolve this puzzle. Using Korean market data (2020-2024), we demonstrate that HVRP exists in Korea but is masked by (1) pooling heterogeneous investor types and (2

Tianyi Xie, Yunuo Chen, Yaowei Guo, Yin Yang

Creating realistic 3D animation remains a time-consuming and expertise-dependent process, requiring manual rigging, keyframing, and fine-tuning of complex motions. Meanwhile, video diffusion models have recently demonstrated remarkable motion imagination in 2D, generating dynamic and visually coherent motion from text or image prompts. However, their results

Prasanjit Dubey, Xiaoming Huo

Identifying the most powerful test in multiple hypothesis testing under strong family-wise error rate (FWER) control is a fundamental problem in statistical methodology. State-of-the-art approaches formulate this as a constrained optimisation problem, for which a dual problem with strong duality has been established in a general sense. However, a constructiv

Amirmohammad Pasdar, Toby Murray, Van-Thuan Pham

We introduce UIXPOSE, a source-code-agnostic framework that operates on both compiled and open-source apps. This framework applies Intention Behaviour Alignment (IBA) to mobile malware analysis, aligning UI-inferred intent with runtime semantics. Previous work either infers intent statically, e.g., permission-centric, or widget-level or monitors coarse dynam

Weida Yin, Masato Miyakawa, Satoshi Semboshi, Noriharu Yodoshi

To optimize the processing conditions for the (Cr,Fe)S non-equilibrium phase with a pyrrhotite-type structure, the phase states and magnetic properties of the specimens obtained at various sintering temperatures were investigated. A slightly off-stoichiometric composition of Cr23Fe23S54 (approximately (Cr,Fe)7S8) sintered and quenched from 1323 K indicates a

Xiaojie Tao, Rajit Gadh

The rapid growth of renewable energy sources has significantly reduced system inertia and increased the need for fast frequency response (FFR) in modern power systems. Data centers, as large and flexible electrical consumers, hold great potential to contribute to frequency stabilization due to their controllable IT workloads and on-site uninterruptible power

Junyi Wu, Van Nguyen Nguyen, Benjamin Planche, Jiachen Tao

We introduce Consistent Instance Field, a continuous and probabilistic spatio-temporal representation for dynamic scene understanding. Unlike prior methods that rely on discrete tracking or view-dependent features, our approach disentangles visibility from persistent object identity by modeling each space-time point with an occupancy probability and a condit

Dongzhuoran Zhou, Yuqicheng Zhu, Xiaxia Wang, Hongkuan Zhou

Large language models (LLMs) achieve strong results on knowledge graph question answering (KGQA), but most benchmarks assume complete knowledge graphs (KGs) where direct supporting triples exist. This reduces evaluation to shallow retrieval and overlooks the reality of incomplete KGs, where many facts are missing and answers must be inferred from existing fa

Boris S. Mordukhovich, Peipei Tang, Chengjing Wang

This paper characterizes the well-posedness of Karush-Kuhn-Tucker system for perturbed composite optimization. Using the parabolic regularity, we introduce a novel second-order variational function, shown to be the pivotal object governing second-order behavior. This foundational result yields the strong second-order sufficient condition introduced here for

Hongguang Liang, Meiling Wang, Juan Wang, Yan Li

This paper investigates the formation of Spin Faraday waves in spin-orbit-coupled Bose-Einstein condensate under the stripe phase and explores the dispersion relation under three different phases. We discover that the SFW exhibit temporal and spatial patterns when the interaction is modulated periodically, and appear with resonant waves and higher order harm

Christopher A. Fuchs, Blake C. Stacey

QBism pursues the real by first eliminating the elements of quantum theory too fragile to be ontologies on their own. Thereafter, it seeks an "ontological lesson" from whatever remains. Here, we explore this program by highlighting three tenets of QBism. First, the Born Rule is a normative statement. It is about the decision-making behavior any individual ag

Wenbo Tian, Ruting Lin, Hongxian Zheng, Yaodong Yang

Existing intelligent sports analysis systems mainly focus on "scoring and visualization," often lacking automatic performance diagnosis and interpretable training guidance. Recent advances in Large Language Models (LLMs) and motion analysis techniques provide new opportunities to address the above limitations. In this paper, we propose SportsGPT, an LLM-driv

Boris Fomin, Mikhail Razumovskiy

The FKDM $k$-distribution technique is applied to parameterize absorption of thermal radiation in the lower and middle atmosphere of Venus, targeting modeling scenarios where the cost of full radiative transfer calculations necessitates efficient parameterizations (e.g. climate modeling). Line-by-line reference modeling based on a Monte Carlo method for radi

Peter Cox, Rafael E. Pérez, Raymond R. Volkas

Asymmetric dark matter is a well-motivated approach to explain the apparent coincidence between the relic densities of visible and dark matter, $\Omega_D \simeq 5.4\Omega_b$. A complete explanation requires two components, a relation between the particle masses of the dark and visible matter, and a second relation between the number densities in each sector.

Yiran Zhang, Jincheng Hu, Mark Dras, Usman Naseem

Large language models (LLMs) excel at single-turn reasoning but often lose accuracy and coherence over extended, multi-turn interactions. Recent evaluations such as TurnBench highlight recurring failure modes-reasoning bias, task drift, hallucination, overconfidence, and memory decay. Current approaches typically append full conversational histories, causing

E. Epelbaum, A. M. Gasparyan, J. Gegelia, D. Hog

In a recent publication, Cirigliano {\it et al.} [Phys. Rev. Lett. 135, 022501 (2025)] argue that three-nucleon forces (3NFs) involving short-range operators that couple two pions with two nucleons are enhanced beyond what is expected in chiral effective field theory based on naive dimensional analysis. Here, we scrutinize the arguments and conclusions of th

Ruohai Yang, Shuangyang Li, Han Yu, Zhiqiang Wei

Orthogonal time frequency space (OTFS) modulation has emerged as a robust solution for high-mobility wireless communications. However, conventional detection algorithms, such as linear equalizers and message passing (MP) methods, either suffer from noise enhancement or fail under complex doubly-selective channels, especially in the presence of fractional del

Ramesh Gundluru, Shubham Gupta, Sri Rama Murty K

Acoustic Word Embeddings (AWEs) improve the efficiency of speech retrieval tasks such as Spoken Term Detection (STD) and Keyword Spotting (KWS). However, existing approaches suffer from limitations, including unimodal supervision, disjoint optimization of audio-audio and audio-text alignment, and the need for task-specific models. To address these shortcomin

Juncheng Zhou, Peter V. Danchev, Hongfeng Wu

We show that there are only finitely many finite fields whose members are the sum of an $n$-potent element and a $5$-potent element. Combining this with the algorithmic results provided by S.D. Cohen {\it et al.}, we confirm in the affirmative the conjecture in \cite{Cohen} concerning all finite fields satisfying this condition. Furthermore, we obtain severa

Rui-Yang Ju, KokSheik Wong, Yanlin Jin, Jen-Shiun Chiang

Document image enhancement and binarization are commonly performed prior to document analysis and recognition tasks for improving the efficiency and accuracy of optical character recognition (OCR) systems. This is because directly recognizing text in degraded documents, particularly in color images, often results in unsatisfactory recognition performance. To

Ashish Mishra, Gyanaranjan Nayak, Tarun Kumar, Arpit Shah

Pretrained models like CLIP have demonstrated impressive zero-shot classification capabilities across diverse visual domains, spanning natural images, artistic renderings, and abstract representations. However, real-world applications often demand the removal (or "unlearning") of specific object classes without requiring additional data or retraining, or aff

Chunan Tong

Supply chain management (SCM) faces significant challenges like demand fluctuations and the bullwhip effect. Traditional methods and even state-of-the-art LLMs struggle with benchmarks like the Vending Machine Test, failing to handle SCM's complex continuous time-series data. While ML approaches like LSTM and XGBoost offer solutions, they are often limited b

Chenzui Li, Yiming Chen, Xi Wu, Tao Teng

Industrial human-robot collaboration requires motion planning that is collision-free, responsive, and ergonomically safe to reduce fatigue and musculoskeletal risk. We propose the Configuration Space Ergonomic Field (CSEF), a continuous and differentiable field over the human joint space that quantifies ergonomic quality and provides gradients for real-time

Peter Athron, Csaba Balázs, Jon Butterworth, Christopher Chang

In this work, we review the history and current role of global fits in the search for physics beyond the Standard Model~(BSM), including precision tests of the Standard Model (SM). Although BSM global fits were initially focused on minimal supersymmetric models, we describe how fits have evolved in response to new data from the Large Hadron Collider (LHC) an

Hisakazu Matsuki, Rustem Khasanov, Jonas A. Krieger, Thomas J. Hicken

Muon spin rotation ($\mu$SR) measurements of internal magnetic field shifts, known as the muon Knight shift, is used for determining pairing symmetries in superconductors. While this technique has been especially effective for $f$-electron-based heavy-fermion superconductors, it remains challenging in $d$-electron-based superconductors such as Sr$_2$RuO$_4$,

Ijaz Ul Haq, Byung Suk Lee, Julia N. Perdrial, David Baude

Advances in sensor networks have enabled real-time stream discharge monitoring, yet persistent sensor malfunctions limit data utility. Manual quality control by expert hydrologists cannot scale with networks generating millions of measurements annually. We introduce HydroGEM, a foundation model for continental-scale streamflow quality control designed to sup

Nithin V. Sabu, Kaushlendra Pandey, Abhishek K. Gupta, Sameer S. M

This work focuses on the development of an analytical framework to study a diffusion-assisted molecular communication-based network of nano-machines (NMs) with a clustered initial deployment to detect a target in a three-dimensional (3D) medium. Leveraging the Poisson cluster process to model the initial locations of clustered NMs, we derive the analytical e

Karthikeyan Sankaralingam

Top-tier academic conferences are failing under the strain of two irreconcilable roles: (1) rapid dissemination of all sound research and (2) scarce credentialing for prestige and career advancement. This conflict has created a reviewer roulette and anonymous tribunal model - a zero-cost attack system - characterized by high-stakes subjectivity, turf wars, a

Neetu Tiwari, Ajit Kumar Dubey

Pressure field estimation from PIV data has been a well-established technique. However, time-resolved pressure estimation strongly depends on the temporal resolution of the PIV measurements. Generally, PIV data has limited time resolution creating challenges to understand high Reynolds number flows. To overcome this challenge, sensor data measured at few opt

Emanuele Mezzi, Gertjan Burghouts, Maarten Kruithof

Text-to-image retrieval in remote sensing (RS) has advanced rapidly with the rise of large vision-language models (LVLMs) tailored for aerial and satellite imagery, culminating in remote sensing large vision-language models (RS-LVLMS). However, limited explainability and poor handling of complex spatial relations remain key challenges for real-world use. To

Curtis T. McMullen, Tina Torkaman

We show the map $\sigma : T_g \to C_g$ sending a compact hyperbolic surface $X$ to a random simple closed geodesic on $X$ determines a proper embedding of Teichm\"uller space into the space of geodesic currents. The proof depends on a formula for the intersection number $i(C,C')$ of a pair of multicurves, expressed in terms of Dehn coordinates on $ML_g(\math

Chunjin Jian, Xinhua Zhu

Reinforcement Learning from Human Feedback (RLHF) plays a crucial role in aligning large language models (LLMs) with human values and preferences. However, the quality and stability of the trained reward model largely determine the final alignment performance. Existing approaches such as Proximal Policy Optimization (PPO) rely heavily on reward models to gui

Jeff J. Ma, Jae-Won Chung, Jisang Ahn, Yizhuo Liang

Any-to-Any models are an emerging class of multimodal models that accept combinations of text and multimodal data as input and generate them as output, introducing heterogeneous computation paths and component scaling characteristics. There are existing mechanisms for deploying Any-to-Any models--or special cases of them--for inference serving, but they eith

Yunfei Yang, Xiaojun Chen, Zhendong Zhao, Yu Zhou

The rapid advancement of deep learning has turned models into highly valuable assets due to their reliance on massive data and costly training processes. However, these models are increasingly vulnerable to leakage and theft, highlighting the critical need for robust intellectual property protection. Model watermarking has emerged as an effective solution, w

Sisi Dai, Kai Xu

Despite significant progress in text-driven 4D human-object interaction (HOI) generation with supervised methods, the scalability remains limited by the scarcity of large-scale 4D HOI datasets. To overcome this, recent approaches attempt zero-shot 4D HOI generation with pre-trained image diffusion models. However, interaction cues are minimally distilled dur

Wei Yi Oon, Yuchen Tang, Baiqian Qi, Wei-Ning Lee

Acoustoelectric (AE) imaging provides electro-anatomical contrast by mapping the distribution of electric fields in biological tissues, by delivering ultrasound waves which spatially modulate the medium resistivity via the AE effect. The conventional method in AE imaging is to transmit focused ultrasound (FUS) beams; however, the depth-of-field (DOF) of FUS-

Nhi Nguyen, Constantino Álvarez Casado, Le Nguyen, Manuel Lage Cañellas

Video-based respiratory rate (RR) estimation is often unreliable due to inconsistent signal quality across extraction methods. We present a predictive, quality-aware framework that integrates heterogeneous signal sources with dynamic assessment of reliability. Ten signals are extracted from facial remote photoplethysmography (rPPG), upper-body motion, and de

Felix Holm, Ghazal Ghazaei, Nassir Navab

Purpose: Detailed surgical recognition is critical for advancing AI-assisted surgery, yet progress is hampered by high annotation costs, data scarcity, and a lack of interpretable models. While scene graphs offer a structured abstraction of surgical events, their full potential remains untapped. In this work, we introduce ProtoFlow, a novel framework that le

James Daniel Whitfield

Few, if any, applications of quantum technology are as widely known as the quantum simulation of quantum matter. Consequently, many interesting questions have been sparked at the intersection of condensed matter, quantum chemistry, and quantum computing. Given the common mathematical foundation of these subjects, we walk through the necessary group theory an

Taig Singh, Shreshth Rajan, Nikhil Jain

As modern neural networks become increasingly memory-bound, inference throughput is limited by DRAM bandwidth rather than compute. We present Arithmetic-Intensity-Aware Quantization (AIQ), a mixed precision quantization framework that chooses per-layer bit-widths to maximize arithmetic intensity (AI) while minimizing accuracy loss. AIQ is a post-training qua

Taylan Demir, Atakan Koçyiğit

We present an adaptive wavelet Galerkin method for transient heat conduction in heterogeneous composite materials. The approach combines multiresolution wavelet bases with an implicit time discretization to efficiently resolve sharp temperature gradients near material interfaces and boundary layers. Adaptive refinement is driven by wavelet coefficients, sign

David Roberts, Trevor McCourt, Geremia Massarelli, Jeremy Rothschild

Diffusive approximations of Markov jump processes often fail to accurately capture large fluctuations. This is confounding, as the rare events triggered by these large fluctuations, such as the failure of electronic memories, are often the object of interest. In this paper we present an improved diffusive approximation, extending a method previously limited

Yang Yang, Risa Shinoda, Hiroaki Santo, Fumio Okura

We present a method for jointly recovering the appearance and internal structure of botanical plants from multi-view images based on 3D Gaussian Splatting (3DGS). While 3DGS exhibits robust reconstruction of scene appearance for novel-view synthesis, it lacks structural representations underlying those appearances (e.g., branching patterns of plants), which

Boyuan Yao, Dingcheng Luo, Lianghao Cao, Nikola Kovachki

We present approximation theories and efficient training methods for derivative-informed Fourier neural operators (DIFNOs) with applications to PDE-constrained optimization. A DIFNO is an FNO trained by minimizing its prediction error jointly on output and Fr\'echet derivative samples of a high-fidelity operator (e.g., a parametric PDE solution operator). As

Koji Inoue, Mikey Elmers, Yahui Fu, Zi Haur Pang

We present a multilingual, continuous backchannel prediction model for Japanese, English, and Chinese, and use it to investigate cross-linguistic timing behavior. The model is Transformer-based and operates at the frame level, jointly trained with auxiliary tasks on approximately 300 hours of dyadic conversations. Across all three languages, the multilingual