December 2025 arXiv papers — page 52

Jinwei Chi, Ke Wang, Yu Chen, Xuanye Lin

Automated essay scoring (AES) is a challenging task in cross-prompt settings due to the diversity of scoring criteria. While previous studies have focused on the output of large language models (LLMs) to improve scoring accuracy, we believe activations from intermediate layers may also provide valuable information. To explore this possibility, we evaluated t

Thittipat Pairatsuppawat, Abhibhu Tachaapornchai, Paweekorn Kusolsomboon, Chutikan Chaiwong

Open-weights large language models remain difficult to deploy for Thai due to unstable generation under complex instructions, despite strong English performance. To mitigate these limitations, We present SiamGPT-32B, an open-weights model based on Qwen3-32B, fine-tuned with a Quality-First strategy emphasizing curated supervision over data scale. The fine-tu

David Fajman, Maciej Maliborski, Maximilian Ofner, Todd Oliynyk

We consider the gravitational Euler-Poisson system with a linear equation of state on an expanding cosmological model of the Universe. The expansion of the spatial sections introduces an additional dissipating effect in the Euler equation. We prescribe the expansion rate of space by a scale factor $a(t)=t^\alpha$ with $\alpha\in(0,1)$, which describes the gr

Zhenglong Guo, Yiming Zhao, Feng Jiang, Heng Jin

Embodied robotic AI systems designed to manage complex daily tasks rely on a task planner to understand and decompose high-level tasks. While most research focuses on enhancing the task-understanding abilities of LLMs/VLMs through fine-tuning or chain-of-thought prompting, this paper argues that defining the planned skill set is equally crucial. To handle th

Nitin Kumar Singh, Arie Rachmad Syulistyo, Yuichiro Tanaka, Hakaru Tamukoh

Sign language recognition (SLR) facilitates communication between deaf and hearing communities. Deep learning based SLR models are commonly used but require extensive computational resources, making them unsuitable for deployment on edge devices. To address these limitations, we propose a lightweight SLR system that combines parallel bidirectional reservoir

Leo D'Amato, Davide Nuzzi, Alberto Testolin, Ivilin Peev Stoianov

Number sense is a core cognitive ability supporting various adaptive behaviors and is foundational for mathematical learning. Here, we study its emergence in unsupervised generative models through the lens of rate-distortion theory (RDT), a normative framework for understanding information processing under limited resources. We train $\beta$-Variational Auto

Naseem Machlovi, Maryam Saleki, Ruhul Amin, Mohamed Rahouti

As large language models (LLMs) become deeply embedded in daily life, the urgent need for safer moderation systems that distinguish between naive and harmful requests while upholding appropriate censorship boundaries has never been greater. While existing LLMs can detect dangerous or unsafe content, they often struggle with nuanced cases such as implicit off

Ibrahim Al Kathemi, Vishnu Arumughan, Mohamed Zbiri, Marcel Kröger

Aqueous zinc (Zn) batteries (AZBs) face significant challenges due to the limited compatibility of Zn anodes with conventional separators, leading to dendrite growth, hydrogen evolution reaction (HER), and poor cycling stability. While separator design is crucial for optimizing battery performance, its potential remains underexplored. The commonly used glass

Huan Xiao

Pausinger recently investigated a special determinant involving prime numbers. In this short note we point out that this type of determinants was already known in linear algebra and its computation is unrelated to prime numbers.

Daniel Casini, Jian-Jia Chen, Jing Li, Federico Reghenzani

The Robot Operating System 2 (ROS~2) has emerged as a relevant middleware framework for robotic applications, offering modularity, distributed execution, and communication. In the last six years, ROS~2 has drawn increasing attention from the real-time systems community and industry. This survey presents a comprehensive overview of research efforts that analy

Erdem Sucu, Izzet Sakallı, Orhan Donmez, G. Mustafa

We investigate a static, spherically symmetric black hole solution arising from Einstein gravity coupled to a confining nonlinear electrodynamics model that reproduces Maxwell theory in the strong-field regime while introducing confinement-like corrections at large distances. The resulting metric function is asymptotically Schwarzschild but carries a charact

Yuankun Chen, Zifei Nie, Xun Gong, Yunfeng Hu

Differentiable optimal control, particularly differentiable nonlinear model predictive control (NMPC), provides a powerful framework that enjoys the complementary benefits of machine learning and control theory. A key enabler of differentiable optimal control is the computation of derivatives of the optimal trajectory with respect to problem parameters, i.e.

Guan-Cheng Chen, Chieh-Lin Tsai, Pei-Hsuan Tsai, Yuan-Hao Chang

Compute-In-Memory (CIM) systems, particularly those utilizing ReRAM and memristive technologies, offer a promising path toward energy-efficient neural network computation. However, conventional quantization and compression techniques often fail to fully optimize performance and efficiency in these architectures. In this work, we present a structured quantiza

Borja Anguiano, David Valls-Gabaud, Andrés del Pino, Guillaume F. Thomas

\textsc{GALATEA} (the \emph{Galactic Archaeology and Local-group Astrophysics Telescope for Extended Areas}) is a concept for a dedicated 15-m, wide-field, 10,000-fibre spectroscopic survey facility in the northern hemisphere, optimized for degree-scale, multi-object spectroscopy. With a $\sim 1~\mathrm{deg}^2$ corrected field-of-view and both medium- ($R \s

Evelyn Zhang, Fufu Yu, Aoqi Wu, Zichen Wen

Processing long visual token sequences poses a significant computational burden on Multimodal Large Language Models (MLLMs). While token pruning offers a path to acceleration, we find that current methods, while adequate for general understanding, catastrophically fail on fine-grained localization tasks. We attribute this failure to the inherent flaws of the

Simon Welker, Bunlong Lay, Maris Hillemann, Tal Peer

Diffusion-based generative models have greatly impacted the speech processing field in recent years, exhibiting high speech naturalness and spawning a new research direction. Their application in real-time communication is, however, still lagging behind due to their computation-heavy nature involving multiple calls of large DNNs. Here, we present Stream$.$FM

Benjamin Bonnefont, Hermanni Rajamäki, Vincent Vargas

We study the high-frequency Fourier asymptotics of imaginary Gaussian multiplicative chaos on the unit circle, a complex-valued random distribution formally given by $\mathrm M_{\mathrm i\beta}=\exp(\mathrm i\beta X)$, where $X$ is a log-correlated Gaussian field. In the subcritical phase $\beta\in(0,1)$, we prove that its Fourier dimension, defined by the o

Antonio Consolo, Andrea Manno, Edoardo Amaldi

Kernel logistic regression (KLR) is a widely used supervised learning method for binary and multi-class classification, which provides estimates of the conditional probabilities of class membership for the data points. Unlike other kernel methods such as Support Vector Machines (SVMs), KLRs are generally not sparse. Previous attempts to deal with sparsity in

Rixin Yu

Learning accurate and stable time-advancement operators for nonlinear partial differential equations (PDEs) remains challenging, particularly for chaotic, stiff, and long-horizon dynamical systems. While neural operator methods such as the Fourier Neural Operator (FNO) and Koopman-inspired extensions achieve good short-term accuracy, their long-term stabilit

Jose-Maria Fernandez-Batanero, Pedro Roman-Gravan

In contemporary society, marked by rapid technological evolution, education faces the challenge and the opportunity of incorporating new digital tools that transform learning, making it more inclusive, flexible, and meaningful. This book aligns with this commitment to educational innovation and equity, focusing on a group that requires specialized and sensit

Fei Ge, Ying Huang, Jie Liu, Guixuan Zhang

Existing deep image watermarking methods follow a fixed embedding-distortion-extraction pipeline, where the embedder and extractor are weakly coupled through a final loss and optimized in isolation. This design lacks explicit collaboration, leaving no structured mechanism for the embedder to incorporate decoding-aware cues or for the extractor to guide embed

R. Ammendola, A. Apponi, G. Benato, M. G. Betti

The PTOLEMY project is prototyping a novel electromagnetic filter for high-precision $\beta$ spectroscopy, with the ultimate and ambitious long-term goal of detecting the cosmic neutrino background through electron capture on tritium bound to graphene. Intermediate small-scale prototypes can achieve competitive sensitivity to the effective neutrino mass, eve

H. A. Harutyunian, E. H. Nikogosyan

It is shown that, owing to the interaction of baryonic matter with the carrier of dark energy, all configurations of baryonic matter acquire energy and inevitably must expand. This conclusion applies to all hierarchical levels of the baryonic universe, including galaxy clusters. We propose a simple statistical method for identifying possible radial motions o

Rafal Bogaczewicz, Pawel Machnikowski

We present a theory of resonance fluorescence (RF) of a solid-state quantum emitter in the regime of weak optical excitation. The emitter is coupled to phonon modes of the surrounding bulk semiconductor, described by a super-Ohmic spectral density. We show that the RF spectrum of this system consists of a central elastic line, a broad phonon sideband known f

Md. Tanvirul Islam

Cockcroft-Walton (CW) voltage multipliers suffer from output ripple that classical analytical models underestimate due to neglected non-idealities like diode drops and capacitor ESR, particularly in high-stage, low-frequency and heavy-load regimes. This paper proposes a hybrid framework that generates a comprehensive 324-case MATLAB/Simulink dataset varying

Yi Xin, Siqi Luo, Tianxiang Xu, Qi Qin

Diffusion Multi-modal Large Language Models (dMLLMs) have recently emerged as a novel architecture unifying image generation and understanding. However, developing effective and efficient Test-Time Scaling (TTS) methods to unlock their full generative potential remains an underexplored challenge. To address this, we propose dMLLM-TTS, a novel framework opera

Quyu Kong, Xu Zhang, Zhenyu Yang, Nolan Gao

Among existing online mobile-use benchmarks, AndroidWorld has emerged as the dominant benchmark due to its reproducible environment and deterministic evaluation; however, recent agents achieving over 90% success rates indicate its saturation and motivate the need for a more challenging benchmark. In addition, its environment lacks key application categories,

D. Mancelli, G. Andrianaki, I. Tazes, C. Vlachos

We demonstrate a high-energy, high-charge, electron source produced by the irradiation of a novel gaseous target by an ultra-intense femtosecond laser pulse. By exploiting a nonsymmetrical nozzle, we increased the total charge of the electron beam by at least an order of magnitude with respect to our previous experiments using symmetrical nozzles. In additio

Juliana Xavier

We survey recent results and current challenges concerning the growth rate inequality for sphere endomorphisms, and present a number of open problems and conjectures arising in this context.

Zhou Rui, Zhi-Tian Zou, Ya Li, Ying Li

We perform a detailed analysis of semileptonic $\Xi_b$ decays mediated by flavor-changing neutral currents ($b\to s$ and $b\to d$) with dilepton or dineutrino final states within the perturbative QCD framework. All independent form factors including vector, axial-vector, tensor, and pseudotensor currents are calculated and are used to analyze the decay branc

Zhang Chong

We revisit a basic question in sequence modeling: is explicit self-attention actually necessary for strong performance and reasoning? We argue that standard multi-head attention is best seen as a form of tensor lifting: hidden vectors are mapped into a high-dimensional space of pairwise interactions, and learning proceeds by constraining this lifted tensor t

Yue Zeng, Rong-Hua Li, Qiangqiang Dai, Guoren Wang

Community detection, which uncovers closely connected vertex groups in networks, is vital for applications in social networks, recommendation systems, and beyond. Real-world networks often have bipartite structures (vertices in two disjoint sets with inter-set connections), creating unique challenges on specialized community detection methods. Biclique perco

Tillmann Bühler, Anna Gusakova, Konstantin Recke

We show that tessellations of hyperbolic space by isometry-invariant Poisson processes of $(d-1)$-dimensional hyperplanes do not have an unbounded cell at the critical intensity. This extends a result by Porret-Blanc for the hyperbolic plane (C. R. Acad. Sci. Paris, Ser. I, Vol. 344 (2007)) to dimensions $d\ge3$. We also show that for intensities strictly be

Jian Yang, Wei Zhang, Yizhi Li, Shawn Guo

Large language models (LLMs) have made significant strides in code generation, achieving impressive capabilities in synthesizing code snippets from natural language instructions. However, a critical challenge remains in ensuring LLMs generate factually accurate responses about programming concepts, technical implementations, etc. Most previous code-related b

Clara L. Del Pio, Francesco P. Ucci

The absolute machine luminosity is a key quantity to achieve the high-precision physics program of future $e^+e^-$ collider. It is determined by measuring a theoretically well-known process, which, ideally, can be computed with arbitrary precision in the perturbation theory. However, yet undiscovered new physics could give a non-negligible contribution to th

Muhammad Mansur Zubairu, Abdullahi Umar, Fatma Salim Al-Kharousi

Let $[n]$ be a finite $n-$chain $\{1, 2, \dots, n\}$, and let $\mathcal{LS}_{n}$ be the Schr\"{o}der monoid, consisting of all isotone and order-decreasing partial transformations on $[n]$. Furthermore, let $\mathcal{SS}^{\prime}_{n} = \{\alpha \in \mathcal{LS}_{n} : \, 1\not\in \text{ Dom } \alpha\}$ be the subsemigroup of $\mathcal{LS}_{n}$, consisting of

Karthik Adimurthi, Mitesh Modasiya

We study local H\"older regularity of bounded, weak solutions for the nonlocal quasilinear equations of the form \[ (|u|^{q-2}u)_t + \text{P.V.} \int_{\mathbb{R}^n} \frac{|u(x,t) - u(y,t)|^{p-2}(u(x,t)-u(y,t))}{|x-y|^{n+sp}} dy = 0, \] with $p\in (1,\infty)$, $q\in (1,\infty)$ and $s \in (0,1)$. Analogous H\"older continuity result in the local case is known

Ondřej Kubů, Piergiulio Tempesta

In a recent preprint, we discussed geometric lifts, and in particular the notion of St\"ackel lift. In this note, we wish to clarify several aspects raised in the comment arXiv:2511.05765.

Changwon Lee, Daniel K. Park

Predicting ground state energies of quantum many-body systems is one of the central computational challenges in quantum chemistry, physics, and materials science. Krylov subspace methods, such as Krylov Quantum Diagonalization and Sample-based Krylov Quantum Diagonalization, are promising approaches for this task on near-term quantum computers. However, both

Tomáš Pikálek, Miroslav Stibůrek, Tereza Tučková, Petra Kolbábková

Confocal and multi-photon microscopy are widely used for in-vivo fluorescence imaging of biological tissues such as the brain, offering non-invasive access up to ~1 mm depth without major loss in performance. A recently-developed alternative is holographic endoscopy, which exploits controlled light transport through hair-thin optical fibres. With minimal inv

Glenn Barnich, Laurent Baulieu, Marc Henneaux, Tom Wetzstein

The BRST Noether theorem, or ``Noether's 1.5 theorem'', asserts the triviality of the BRST Noether current. We provide two proofs of this theorem that are both valid without restriction on the structure of the gauge theory, extending thereby previous proofs holding in the case of gauge theories for which the solution of the master equation is linear in the a

Johann Coraux, Nicolas Rougemaille, Cedric Robert, Clément Faugeras

Magnetic two-dimensional (2D) crystals were isolated about a decade ago, triggering a tremendous research activity worldwide. This colloquium raises a stiff question: what is really new about them? At first sight, they seem to be purer implementations of 2D spin models than traditional systems such as ultra-thin films. Yet, they partly realized their promise

Mamoru Saita, Yutaka Hori

Temperature is a fundamental regulator of chemical and biochemical kinetics, yet capturing nonlinear thermal effects directly from experimental data remains a major challenge due to limited throughput and model flexibility. Recent advances in machine learning have enabled flexible modeling beyond conventional physical laws, but most existing strategies remai

Xiaoming Zhang, Chunli Li, Jiacheng Hao, Yuan Gao

Esophageal varices (EV) represent a critical complication of portal hypertension, affecting approximately 60% of cirrhosis patients with a significant bleeding risk of ~30%. While traditionally diagnosed through invasive endoscopy, non-contrast computed tomography (NCCT) presents a potential non-invasive alternative that has yet to be fully utilized in clini

Jiaren Peng, Hongda Sun, Xuan Tian, Cheng Huang

The automation of Cyber Threat Intelligence (CTI) relies heavily on Named Entity Recognition (NER) to extract critical entities from unstructured text. Currently, Large Language Models (LLMs) primarily address this task through retrieval-based In-Context Learning (ICL). This paper analyzes this mainstream paradigm, revealing a fundamental flaw: its success s

Camilo Granados, Bálint Kiss, Eric Cormier, Bikash Kumar Das

The ability to sculpt light in space, time, and polarization has revolutionized studies of light-matter interaction and enabled breakthroughs in optical communication, imaging, and ultrafast science. Among the many degrees of freedom of light, orbital angular momentum (OAM) further expands these capabilities by unlocking new regimes of control in information

Himadri Halder

This article introduces the notion of arithmetic Bohr radius for operator valued pluriharmonic functions on complete Reinhardt domains in $\mathbb{C}^n$. Using tools from local Banach space theory, we determine its asymptotic behavior in both finite and infinite dimensions. Asymptotic estimates for this constant are derived for both convex and non-convex com

Elad Hazan, Shai Shalev Shwartz, Nathan Srebro

Modern learning systems increasingly interact with data that evolve over time and depend on hidden internal state. We ask a basic question: when is such a dynamical system learnable from observations alone? This paper proposes a research program for understanding learnability in dynamical systems through the lens of next-token prediction. We argue that learn

Philipp L. Kinon, Simon R. Eugster, Peter Betsch

An energy-based modeling framework for the nonlinear dynamics of spatial Cosserat rods undergoing large displacements and rotations is proposed. The mixed formulation features independent displacement, velocity and stress variables and is further objective and locking-free. Finite rotations are represented using a director formulation that avoids singulariti

Arno Hoefnagels

The study of cosmological correlators, and more generally Feynman integrals, is greatly aided by considering them as solutions to differential equations. Often, such systems of differential equations are reducible, which, broadly speaking, implies that the differential system is composed of various subsystems. Studying such decompositions and subsystems can

Louis Libat, Can Selçuk, Eric Chénier, Vincent Le Chenadec

We present a Cartesian cut-cell finite-volume method for sharp-interface two-phase diffusion problems in static geometries. The formulation follows a two-fluid approach: independent diffusion equations are discretized in each phase on a fixed Cartesian grid, while the phases are coupled through embedded interface conditions enforcing continuity of diffusive

Jussi Keppo, Yingkai Li

This paper studies optimal contract design in private market investing, focusing on internal decision making in venture capital and private equity firms. A principal relies on an agent who privately exerts costly due diligence effort and then recommends whether to invest. Outcomes are observable ex post even when an opportunity is declined, allowing compensa

Madhumitha K, Harshitha A, Swati Nayak, Sabitha D'Souza

Let $\Gamma$ be a simple graph on $n$ vertices. Lanzhou index is defined as $Lz(\Gamma)=\sum\limits_{u \in V(\Gamma)}d_\Gamma(u)^2d_{\overline{\Gamma}}(u).$ In this manuscript, the Lanzhou matrix, denoted by $A_{Lz}(\Gamma)$, has been defined, and its spectral properties are studied. The $uv^{th}$ entry in $A_{Lz}(\Gamma)$ is $d_\Gamma(u)d_{\overline{\Gamma}

Yujie Zhao, Hongwei Fan, Di Chen, Shengcong Chen

Recent progress in robot learning has been driven by large-scale datasets and powerful visuomotor policy architectures, yet policy robustness remains limited by the substantial cost of collecting diverse demonstrations, particularly for spatial generalization in manipulation tasks. To reduce repetitive data collection, we present Real2Edit2Real, a framework

Anupam Das, Abhishek De, Stepan L. Kuznetsov

In this work we prove the undecidability (and $\Sigma^0_1$-completeness) of several theories of semirings with fixed points. The generality of our results stems from recursion theoretic methods, namely the technique of effective inseperability. Our result applies to many theories proposed in the literature, including Conway $\mu$-semirings, Park $\mu$-semiri

Yacouba Diarra, Panga Azazia Kamate, Nouhoum Souleymane Coulibaly, Michael Leventhal

We present Kunkado, a 160-hour Bambara ASR dataset compiled from Malian radio archives to capture present-day spontaneous speech across a wide range of topics. It includes code-switching, disfluencies, background noise, and overlapping speakers that practical ASR systems encounter in real-world use. We finetuned Parakeet-based models on a 33.47-hour human-re

Sandro Andric

Interpretability methods for large language models (LLMs) typically derive directions from textual supervision, which can lack external grounding. We propose using human brain activity not as a training signal but as a coordinate system for reading and steering LLM states. Using the SMN4Lang MEG dataset, we construct a word-level brain atlas of phase-locking

Jiahua Jiang, Joseph Marsh, Rowland G Seymour

Comparative judgement studies elicit quality assessments through pairwise comparisons, typically analysed using the Bradley-Terry model. A challenge in these studies is experimental design, specifically, determining the optimal pairs to compare to maximize statistical efficiency. Constructing static experimental designs for these studies requires spectral de

Giuseppe Mario Rago

Using Gegenbauer polynomials and the zonal harmonic functions we build an explicit representation formula for the Green function with Neumann boundary conditions in the annulus.

Runze Li, Yuwen Zhai, Bo Xu, LiWu Xu

Contemporary GUI agents, while increasingly capable due to advances in Large Vision-Language Models (VLMs), often operate with a critical limitation: they treat each task in isolation, lacking a mechanism to systematically learn from past successes. This digital ''amnesia'' results in sub-optimal performance, repeated errors, and poor generalization to novel

Yuantian Yu, Huihui Zhang, Minjie Zhang

The edge-spectral-Tur\'an type problem is also called the Brualdi-Hoffman-Tur\'an type problem, which is a central topic in spectral graph theory, seeking to determine the maximum spectral radius $\lambda(G)$ of an $F$-free graph $G$ with $m$ edges. This problem has attracted significant attention in recent years. In this paper, we will sort out several clos

Manuel G. Satué, Fernando Castaño, Manuel G. Ortega, Francisco R. Rubio

This paper presents a control strategy for sun trackers which adapts continuously to different sources of error, avoiding the necessity of any kind of calibration by analyzing the produced electric power to sense the position of the Sun. The proposed strategy is able to meet the strict specifications for HCPV sun trackers despite of mechanical uncertainties

Eloise Lardet, Letian Chen, Thibault Bertrand

The theoretical understanding of pattern formation in active systems remains a central problem of interest. Heterogeneous flocks made up of multiple species can exhibit a remarkable diversity of collective states that cannot be obtained from single-species models. In this paper, we derive a kinetic theory for multi-species systems of self-propelled particles

David Turton, Alexander Tyukov

Supergravity calculations of holographic four-point correlation functions are a powerful tool for deriving dynamical information about the dual field theory at strong coupling. Recently, a number of such computations have been performed by studying light probes of smooth horizonless supergravity solutions, including certain asymptotically AdS$_5 \times$S$^5$

Hieu Minh Nguyen, Tam Le-Thanh Dang, Kiet Van Nguyen

Understanding signboard text in natural scenes is essential for real-world applications of Visual Question Answering (VQA), yet remains underexplored, particularly in low-resource languages. We introduce ViSignVQA, the first large-scale Vietnamese dataset designed for signboard-oriented VQA, which comprises 10,762 images and 25,573 question-answer pairs. The

Jordan Chapman, Arindom Nag, Thang Pham, Vsevolod Ivanov

Defects in atomically thin van der Waals materials have recently been investigated as sources of spin-photon entanglement with sensitivity to strain tuning. Unlike many two-dimensional materials, quasi-one-dimensional materials such as transition metal trichalcogenides exhibit in-plane anisotropy resulting in axis-dependent responses to compressive and tensi

Hongwei Fan, Hang Dai, Jiyao Zhang, Jinzhou Li

The robotics field is evolving towards data-driven, end-to-end learning, inspired by multimodal large models. However, reliance on expensive real-world data limits progress. Simulators offer cost-effective alternatives, but the gap between simulation and reality challenges effective policy transfer. This paper introduces TwinAligner, a novel Real2Sim2Real sy

Lars Humpert, Dante M. Kennes, Jan-Niklas Herre

We investigate the disordered spin-$\frac12$Heisenberg model in two dimensions and employ tree tensor networks (TTNs) with a physics-informed structural optimization of the tree layout, to simulate dynamics in the many-body localization problem. We find that TTNs are able to capture two-dimensional entanglement patterns more effectively than matrix product s

Matteo Castiglioni, Junjie Chen, Yingkai Li

A principal selects a team of agents for collaborating on a joint project. The principal aims to design a revenue-optimal contract that incentivizes the team of agents to exert costly effort while satisfying fairness constraints. We show that the optimal fair contract ensures that there is a minimum share, and every agent receives a linear contract weakly hi

Yueyao Chen, Kai-Ni Wang, Dario Tayupo, Arnaud Huaulm'e

Purpose: Surgical workflow recognition enables context-aware assistance and skill assessment in computer-assisted interventions. Despite recent advances, current methods suffer from two critical challenges: prediction jitter across consecutive frames and poor discrimination of ambiguous phases. This paper aims to develop a stable framework by selectively pro

Laura Cotter, Antonio Martin-Carrillo, Joseph Fisher, Gabriel Finneran

Science is currently at an age where there is more data than we know how to deal with. Machine learning (ML) is an emerging tool that is useful for drawing valuable science out of incomprehensibly large datasets and identifying complex trends in data that may otherwise be overlooked. Moreover, ML can potentially enhance the quality and quantity of scientific

Przemysław Ohrysko, Michał Wojciechowski

We introduce Nevanlinna--Pick norms associated with finite families of characters on a commutative semisimple Banach algebra $A$ and study the class $NP_\infty$ of algebras for which all these norms coincide with the $\ell^\infty$ norm. Our positive result is topological: if $A\in NP_\infty$ and $K\subset Δ(A)$ is compact scattered Hausdorff, then the restri

Valentin Yu. Irkhin, Zhehong Liu, Danil A. Myakotnikov, Evgenia V. Komleva

Experimental data on the specific heat $C_p$ of quadruple perovskites ACu$_3$Fe$_2$Re$_2$O$_{12}$ (A = Mn, Cu, La, Ce, Dy) are presented, demonstrating an anomalous concave-down $C_p/T$ vs. $T^2$ curve and a bell-shaped feature in $\beta(T) = (C_p - \gamma T)/T^3$ plotted against $T$ on a logarithmic scale. This feature is most pronounced for A = Cu and Mn.

Zihe Zhang, Can Zhang, Yanheng Xu, Xin Hu

This paper presents PediaMind-R1, a domain-specialized large language model designed to achieve active personalization in intelligent parenting scenarios. Unlike conventional systems that provide generic suggestions, PediaMind-R1 draws on insights from developmental psychology. It introduces temperament theory from the Thomas-Chess framework and builds a tem

Abdelmadjid Benmachiche, Khadija Rais, Hamda Slimi

The spread of a resource-constrained Internet of Things (IoT) environment and embedded devices has put pressure on the real-time detection of anomalies occurring at the edge. This survey presents an overview of machine-learning methods aimed specifically at on-device anomaly detection with extremely strict constraints for latency, memory, and power consumpti

Xiansheng Cai, Tao Wang, Shuai Zhang, Tiantian Zhang

More than a century after discovery, the theory of conventional superconductivity remains incomplete. While the importance of electron-phonon coupling is understood, a controlled first-principles treatment of Coulomb interaction is lacking. Current ab initio calculations of superconductivity rely on a phenomenological downfolding approximation, replacing Cou

Zikang Wang, Xiaomeng Xu

In this paper, we study the isomonodromy deformation equations for the $n\times n$ system of first order meromorphic linear ordinary differential equations with two second order poles. We analyze the asymptotic behaviour of the solutions at a boundary point of the isomonodromic deformation space, and derive a parameterization of the solutions via asymptotic

J. W. F. Venderbos

Condensed matter physics is often concerned with determining the response of a solid to an external stimulus. This paper revisits and extends the microscopic formalism for calculating response coefficients -- here referred to as (generalized) polarizabilities -- in crystalline electronic insulators. The main focus is on the Berry phase polarization and orbit

Xueming Yan, Boyan Xu, Yaochu Jin, Lixian Xiao

Indonesian, spoken by over 200 million people, remains underserved in multimodal emotion recognition research despite its dominant presence on Southeast Asian social media platforms. We introduce IndoMER, the first multimodal emotion recognition benchmark for Indonesian, comprising 1,944 video segments from 203 speakers with temporally aligned text, audio, a

Zhiqing Hu, Chenxu Zhao, Jiazhong Lu, Xiaolei Liu

Misuse of LLM-generated text can be curbed by watermarking techniques that embed implicit signals into the output. We propose a watermark that partitions the vocabulary at each decoding step into three sets (Green/Yellow/Red) with fixed ratios and restricts sampling to the Green and Yellow sets. At detection time, we replay the same partitions, compute Green

Eneko Lazpita, Andrés Bell-Navas, Jesús Garicano-Mena, Petros Koumoutsakos

Understanding intraventricular hemodynamics requires compact and physically interpretable representations of the underlying flow structures, as characteristic flow patterns are closely associated with cardiovascular conditions and can support early detection of cardiac deterioration. Conventional visualization of velocity or pressure fields, however, provide

Shifan Cui

A DFT benchmark on water including more than 50 functionals from GGA to double-hybrid levels is reported. The main metric is the accuracy of forces, allowing better structural coverage, higher statistical confidence, and fewer error sources compared to conventional benchmarks. The input structures include water clusters of 4~128 molecules, with highly varied

Wenyu Luo, Jinhui Chen

Speech intelligibility assessment is essential for many speech-related applications. However, most objective intelligibility metrics are intrusive, as they require clean reference speech in addition to the degraded or processed signal for evaluation. Furthermore, existing metrics such as STOI are primarily designed for normal hearing listeners, and their pre

Valdemar Bille-Lauridsen, Rasmus Ellebæk Christiansen, Yi Yu, Jesper Mørk

We introduce a novel light-matter interface that integrates a nanoscale buried heterostructure emitter into a dielectric bowtie cavity, co-localising the optical hotspot and the electronic wavefunction. This platform enables strong light-matter interaction through deep subwavelength confinement while remaining compatible with scalable fabrication. We show th

Bocong Chen, Jing Huang, Hao Wu

Let $q=2^m$ with $m\ge 3$ and set $n:=q+1$. We investigate $(q+1)$-arcs $\mathcal A\subset \mathrm{PG}(3,q)$ that admit a regular cyclic subgroup $C\le \mathrm{PGL}(4,q)$ of order $n$. Over $K=\mathbb{F}_{q^2}$, such an action can be conjugated to a diagonal one, producing explicit cyclic monomial models \[ \mathcal M_a = \{[1:t:t^a:t^{a+1}]:t\in U_n\}\subse

Li Lin Yang, Yiyang Zhang

We propose '$\mathrm{d} \mathcal{E}$-forms' as fundamental building blocks of canonical integrands for elliptic Feynman integrals, which lead to Kronecker-Eisenstein $\omega$-form symbol letters. Built upon pure elliptic multiple polylogarithms, they provide a natural extension of the '$\mathrm{d} \! \log$-form' integrands and $\mathrm{d} \! \log$ letters fo

Christian Hägg, Kathlén Kohn, Giovanni Luca Marchetti, Boris Shapiro

We introduce Sprecher Networks (SNs), a family of trainable architectures derived from David Sprecher's 1965 constructive form of the Kolmogorov-Arnold representation. Each SN block implements a "sum of shifted univariate functions" using only two shared learnable splines per block, a monotone inner spline $\phi$ and a general outer spline $\Phi$, together w

Simon Ståhlberg, Blai Bonet, Hector Geffner

While reinforcement learning methods have delivered remarkable results in a number of settings, generalization, i.e., the ability to produce policies that generalize in a reliable and systematic way, has remained a challenge. The problem of generalization has been addressed formally in classical planning where provable correct policies that generalize over a

Zhongwei Chen, Hai-Jun Rong, Zhao-Xu Yang, Guoqi Li

Traditional drone-view geo-localization (DVGL) methods based on artificial neural networks (ANNs) have achieved remarkable performance. However, ANNs rely on dense computation, which results in high power consumption. In contrast, spiking neural networks (SNNs), which benefit from spike-driven computation, inherently provide low power consumption. Regrettabl

Massimo Iuliani, Blake Sawyer, Marco Fontani, David Spreadborough

The need to estimate the speed of road vehicles has become increasingly important in the field of video forensics, particularly with the widespread deployment of CCTV cameras worldwide. Despite the development of various approaches, the accuracy of forensic speed estimation from real-world footage remains highly dependent on several factors, including camera

Canran Xiao, Jiabao Dou, Zhiming Lin, Zong Ke

How should we quantify the value of each training example when datasets are large, heterogeneous, and geometrically structured? Classical Data-Shapley answers in principle, but its O(n!) complexity and point-wise perspective are ill-suited to modern scales. We propose Hierarchical Contrastive Data Valuation (HCDV), a three-stage framework that (i) learns a c

Jinyeop Lee, Kunlun Qi

We study the semiclassical limit of the two-dimensional Dirac--Hartree equation in the presence of a periodic external potential. The spinor dynamics are formulated using the matrix-valued Wigner transform together with spectral projectors onto the positive and negative energy bands. Under suitable assumptions on the initial data and the potentials, we rigor

Isshaan Singh, Divyansh Chawla, Anshu Garg, Shivin Mangal

The need for an intelligent, real-time spoilage prediction system has become critical in modern IoT-driven food supply chains, where perishable goods are highly susceptible to environmental conditions. Existing methods often lack adaptability to dynamic conditions and fail to optimize decision making in real time. To address these challenges, we propose a hy

Markus Ekvall, Ludvig Bergenstråhle, Patrick Truong, Ben Murrell

Retrieval-augmented generation improves large language models by grounding outputs in external knowledge sources, reducing hallucinations and addressing knowledge cutoffs. However, standard embedding-based retrieval fails to capture the complexity of multi-concept queries, particularly in domains like biomedicine, where biological data are inherently high-di

Francesco P. Ucci

The pion form factor plays a crucial role in the determination of the contribution of the hadronic vacuum polarisation to the muon anomalous magnetic moment. In order to measure this quantity, energy-scan experiments rely on Monte Carlo generator to simulate the $e^+e^- \to \pi^+\pi^-(\gamma)$ process. For the theoretical accuracy to match the experimental p

Ronghao Yin, Yugang Ren, Deok Young Seo, Anoushka Sinha

Mesoscopic particles levitated by optical, electrical or magnetic fields act as mechanical oscillators with a range of surprising properties, such as tuneable oscillation frequencies, access to rotational motion, and remarkable quality factors. Coupled levitated particles display rich dynamics and non-reciprocal interactions, with applications in sensing and

Philipp Bringmann, Maximilian Brunner, Dirk Praetorius

This paper concerns the inclusion of Newton's method into an adaptive finite element method (FEM) for the solution of nonlinear partial differential equations (PDEs). It features an adaptive choice of the damping parameter in the Newton iteration for the discretized nonlinear problems on each level ensuring both global linear and local quadratic convergence.

Thilo Hartel, Lucas Picasarri-Arrieta, Dieter Rautenbach

Let $G$ be a $K_4$-free graph of order $n$ and let $k$ be an integer with $0\leq k\leq n$. We show the existence of positive constants $\eta$ and $\nu$ such that $G$ has at most $(4-\eta)^{(5-\eta)k-n}(5-\eta)^{n-(4-\eta)k}$ maximal independent sets of order $k$ and at most $O\left((12-\nu)^{\frac{n}{4}}\right)$ maximal induced bipartite subgraphs.

Simon Ståhlberg, Hector Geffner

First-order relational languages have been used in MDP planning and reinforcement learning (RL) for two main purposes: specifying MDPs in compact form, and representing and learning policies that are general and not tied to specific instances or state spaces. In this work, we instead consider the use of first-order languages in goal-conditioned RL and genera

Zhenyang Huang, Xiao Yu, Yi Zhang, Decheng Wang

Remote sensing image change detection is one of the fundamental tasks in remote sensing intelligent interpretation. Its core objective is to identify changes within change regions of interest (CRoI). Current multimodal large models encode rich human semantic knowledge, which is utilized for guidance in tasks such as remote sensing change detection. However,