March 2026 arXiv papers — page 113

Yuxiang Mei, Delai Qiu, Shengping Liu, Jiaen Liang

Speech Large Language Models (Speech-LLMs) have emerged as a powerful approach for automatic speech recognition (ASR) by aligning speech encoders with large language models. However, adapting these systems to multilingual settings with imbalanced data distributions remains challenging. In such scenarios, a stability-plasticity dilemma often arises: fully sha

Haoran Nie, Chaoran Jiang, Xiangying Shen, Lei Xu

Non-Hermitian physics traditionally relies on active gain--loss modulation or non-reciprocal couplings, which often introduce significant complexity, compromise stability, and offer very limited scalability in conservative systems. Here we propose an attenuation-gauge duality paradigm in which non-Hermitian topology emerges within fully passive, conservative

Hugo Caselles-Dupré, Mathis Koroglu, Guillaume Jeanneret, Arnaud Dapogny

Diffusion-based image-to-video (I2V) models are increasingly effective, yet they struggle to scale to ultra-high-resolution inputs (e.g., 4K). Generating videos at the model's native resolution often loses fine-grained structure, whereas high-resolution tiled denoising preserves local detail but breaks global layout consistency. This failure mode is part

Qihong Tang, Changhan Liu, Shaofeng Zhang, Wenbin Li

Identifying potential objects is critical for object recognition and analysis across various computer vision applications. Existing methods typically localize potential objects by relying on exemplar images, predefined categories, or textual descriptions. However, their reliance on image and text prompts often limits flexibility, restricting adaptability in

A. I. Komech, E. A. Kopylova

The article addresses the damped driven Jaynes-Cummings for quantised one-mode Maxwell field coupled to a two-level molecule. We consider a broad class of damping and pumping which are polynomial in the creation and annihilation operators. Our main result is the construction of a contraction dynamical semigroup in the Hilbert space of Hermitian Hilbert-Schmi

Leonardo Pedroso, Andrea Agazzi, W. P. M. H. Heemels, Mauro Salazar

Self-interested behavior in sharing economies often leads to inefficient aggregate outcomes compared to a centrally coordinated allocation, ultimately harming users. Yet, centralized coordination removes individual decision power. This issue can be addressed by designing rules that align individual preferences with system-level objectives. Unfortunately, rul

Assaf Goldberger, Radel Ben-Av, Giora Dula, Yoseph Strassler

Integer weighing matrices (IW-matrices for short) are integer valued orthogonal square matrices. One usecase of these is to create classical weighing matrices with various block structures. In this paper we study and classify the space $IW(n,k)$ of the integer weighing matrices of small size $n\times n$ and weight $k$. Our classification includes a full list

Caren Hasler

We study the consistency of the $k$-nearest neighbor regressor under complex survey designs. While consistency results for this algorithm are well established for independent and identically distributed data, corresponding results for complex survey data are lacking. We show that the $k$-nearest neighbor regressor is consistent under regularity conditions on

Rui Guo, Cai-Na Hao, Xiaoyang Xia, Yong Shi

To understand the complicated formation processes of disk galaxies, we carry out a comparative study for NUV-r blue and red spiral galaxies drawn from a parent sample of u-r red spirals with $M_{*} > 10^{10.5} M_{\odot}$ at 0.02 < z < 0.07, based on the optical data from the Sloan Digital Sky Survey (SDSS) and the ultraviolet (UV) data from the Galaxy Evolut

Reza Sedghi, Robin Schiewer, Anand Subramoney, David Kappel

At typical context lengths, the feed-forward MLP block accounts for a large share of a transformer's compute budget, motivating sparse alternatives to dense MLP blocks. We study sparse, tree-structured feed-forward layers as drop-in replacements for MLP blocks in deep transformer architectures, enabling conditional computation via hard hierarchical routing w

Lanlan Yu, Quan-Hui Liu, Haoyue Zheng, Xinfu Yang

Estimating time-varying reproduction numbers from epidemic incidence data is a central task in infectious disease surveillance, yet it poses an inherently ill-posed inverse problem. Existing approaches often rely on strong structural assumptions derived from epidemiological models, which can limit their ability to adapt to non-stationary transmission dynamic

Isabella Marasco, Davide Evangelista, Elena Loli Piccolomini, Michele Colajanni

Artificial intelligence systems predominantly rely on static data distributions, making them ineffective in dynamic real-world environments, such as cybersecurity, autonomous transportation, or finance, where data shifts frequently. Continual learning offers a potential solution by enabling models to learn from sequential data while retaining prior knowledge

Sirong Piao, Ying Ming, Ruijie Zhao, Jiaru Wang

To characterize lobar and segmental airway volume differences between systemic lupus erythematosus (SLE) patients with interstitial lung disease (ILD) and those without ILD (non-ILD) using a deep learning-based approach on non-contrast chest high-resolution CT (HRCT). Methods: A retrospective analysis was conducted on 106 SLE patients (27 SLE-ILD, 79 SLE-non

Daowen Li, Ruixiao Dong, Kai Li, Ying Chen

Perceptual video compression leverages generative priors to reconstruct realistic textures and motions at low bitrates. However, existing perceptual codecs often lack native support for variable bitrate and progressive delivery, and their generative modules are weakly coupled with entropy coding, limiting bitrate reduction. Inspired by the next-scale predict

Margarita A. Guerrero, Henrik Sandberg, Cristian R. Rojas

Quantifying model mismatch in a control-relevant manner is fundamental in robust control. A well-known metric for this purpose is the $\nu$-gap, or Vinnicombe metric, which measures the discrepancy between a nominal model and the real system from a closed-loop viewpoint. However, its computation typically requires explicit knowledge of the true system. In th

Nicola J. Müller, Moritz Oster, Isabel Valera, Jörg Hoffmann

Learning per-domain generalizing policies is a key challenge in learning for planning. Standard approaches learn state-value functions represented as graph neural networks using supervised learning on optimal plans generated by a teacher planner. In this work, we advocate for learning Q-value functions instead. Such policies are drastically cheaper to evalua

Patrycja Strycharczuk, Sam Kirkham

This paper outlines the conceptual and computational foundations of the AURORA (Acoustic Understanding and Real-time Observation of Resonant Articulations) model. AURORA predicts tongue displacement and shape in vowel sounds based on the first two formant values. It is intended as a didactic aid helping to explain the relationship between formants and the un

Anupam Gupta, Haim Kaplan, Alexander Lindermayr, Jens Schlöter

We simplify the proof of the optimality of the Shortest Lower-Bound First (SLF) algorithm, introduced by Gupta, Kaplan, Lindermayr, Schl\"oter, and Yingchareonthawornchai [FOCS'25], for minimizing the total flow time in the $\varepsilon$-clairvoyant setting.

Linghao Zhang, Jungang Li, Yonghua Hei, Sicheng Tao

Multimodal large language models (MLLMs) are typically trained in multiple stages, with video-based supervised fine-tuning (Video-SFT) serving as a key step for improving visual understanding. Yet its effect on the fine-grained evolution of visual capabilities, particularly the balance between spatial and temporal understanding, remains poorly understood. In

Edoardo D'Amico, Marco De Nadai, Praveen Chandar, Divita Vohra

Podcast listening is often grounded in a set of favorite shows, while listener intent can evolve over time. This combination of stable preferences and changing intent motivates recommendation approaches that support both familiarity and exploration. Traditional recommender systems typically emphasize long-term interaction patterns, and are less explicitly de

Agustín Muñoz González

This work builds on the theoretical frameworks presented in "Liquidity pools as mean field games: A new framework" and "Liquidity pools as mean field games with transaction costs" by the same author, where the strategic interactions among traders in a constant-product market-making protocol were modelled using mean field games (MFG), first without transactio

Jaein Kim, Hee Bin Yoo, Dong-Sig Han, Byoung-Tak Zhang

A symmetry on rigid motion is one of the salient factors in efficient learning of 3D point cloud problems. Group convolution has been a representative method to extract equivariant features, but its realizations have struggled to retain both rigorous symmetry and scalability simultaneously. We advocate utilizing the intertwiner framework to resolve this trad

Pietro Negri, Manuel Sica, Rocco Zaccagnino, Rosalba Zizza

The Lyndon array stores, at each position of a word, the length of the longest maximal Lyndon subword starting at that position, and plays an important role in combinatorics on words, for example in the construction of fundamental data structures such as the suffix array. In this paper, we introduce the Inverse Lyndon Array, the analogous structure for inver

Alexander Köhler, Michael Breuß

In many CAD-based applications, complex geometries are defined by a high number of design parameters. This leads to high-dimensional design spaces that are challenging for downstream engineering processes like simulations, optimization, and design exploration tasks. Therefore, dimension reduction methods such as principal component analysis (PCA) are used. T

Saugat Aryal, Mark T. Keane

Recently, in eXplainable AI (XAI), $\textit{even if}$ explanations -- so-called semi-factuals -- have emerged as a popular strategy that explains how a predicted outcome $\textit{can remain the same}$ even when certain input-features are altered. For example, in the commonly-used banking app scenario, a semi-factual explanation could inform customers about b

Marco De Nadai, Edoardo D'Amico, Max Lefarov, Alexandre Tamborrino

LLMs are increasingly applied to recommendation, retrieval, and reasoning, yet deploying a single end-to-end model that can jointly support these behaviors over large, heterogeneous catalogs remains challenging. Such systems must generate unambiguous references to real items, handle multiple entity types, and operate under strict latency and reliability cons

Mohammad Nooraiepour

Accurate prediction of permeability tensors from pore-scale microstructure images is essential for subsurface flow modeling, yet direct numerical simulation requires hours per sample, fundamentally limiting large-scale uncertainty quantification and reservoir optimization workflows. A physics-informed deep learning framework is presented that resolves this b

Manuel Barusco, Davide Dalle Pezze, Francesco Borsatti, Gian Antonio Susto

Visual Anomaly Detection (VAD) is crucial for industrial inspection, yet most existing methods are limited to single-category scenarios, failing to address the multi-class and continual learning demands of real-world environments. While Teacher-Student (TS) architectures are efficient, they remain unexplored for the Continual Setting. To bridge this gap, we

Binqing Wu, Zongjiang Shang, Shiyu Liu, Jianlong Huang

Accurate air quality forecasting is essential for public health and environmental sustainability, but remains challenging due to the complex pollutant dynamics. Existing deep learning methods often model pollutant dynamics as an instantaneous process, overlooking the intrinsic delays in pollutant propagation. Thus, we propose AirDDE, the first neural delay d

Shunqi Mao, Wei Guo, Dingxin Zhang, Chaoyi Zhang

Large language models are increasingly used to support organizational decisions from hiring to governance, raising fairness concerns in AI-assisted evaluation. Prior work has focused mainly on demographic bias and broader preference effects, rather than on whether evaluators reward expressed trust in AI itself. We study this phenomenon as LLM Nepotism, an at

Yimin Wei, Aoran Xiao, Hongruixuan Chen, Junshi Xia

Open-vocabulary segmentation enables pixel-level recognition from an open set of textual categories, allowing generalization beyond fixed classes. Despite great potential in remote sensing, progress in this area remains largely limited to clear-sky optical data and struggles under cloudy or haze-contaminated conditions. We present MM-OVSeg, a multimodal Opti

Haoxiang Rao, Zhao Wang, Chenyang Si, Yan Lyu

Industrial anomaly detection (AD) is characterized by an abundance of normal images but a scarcity of anomalous ones. Although numerous few-shot anomaly synthesis methods have been proposed to augment anomalous data for downstream AD tasks, most existing approaches require time-consuming training and struggle to learn distributions that are faithful to real

Evangelos Vassos, Thomas Whittaker, Abdul Jabbar, Aakash Bansal

This article presents a fully 3D-printed wideband metasurface folded reflectarray antenna (MFRA) operating in the millimeter-wave n257 band. The proposed MFRA integrates a novel polarization-rotating reflective metasurface (RMS), a compact embedded horn feed, and a polarization-selective metasurface polarization grid (MPG), all fabricated using a low-cost in

F. A. Rasulova, A. A. Kuznetsov, V. O. Nesterenko, J. H. Khushvaktov

This work used the {\gamma}-activation approach to conduct tests at bremsstrahlung end-point energies of 10-23 MeV utilising the MT-25 microtron beam. The experimental values of relative yields and cross sections per equivalent quantum of photonuclear reactions on stable isotopes of cadmium and tellurium were compared to theoretical calculations obtained fro

Gaoge Han, Zhengqing Gao, Ziwen Li, Jiaxin Huang

In this paper, we introduce a novel kinematics-rich vision-language-action (VLA) task, in which language commands densely encode diverse kinematic attributes (such as direction, trajectory, orientation, and relative displacement) from initiation through completion, at key moments, unlike existing action instructions that capture kinematics only coarsely or p

Luca Pellegrini

Neural operators (NOs) are powerful deep learning frameworks designed to learn solution operators of partial differential equations. This study evaluates the ability of NOs&#39; to capture the stiff spatio-temporal dynamics of the FitzHugh-Nagumo model. A key contribution of this study is the assessment of the translation invariance using a novel training st

Pedro Duarte, André Coelho, Francisco Ribeiro, Filipe B. Teixeira

This paper proposes a vision-based framework for the intelligent control of mobile Open Radio Access Network (O-RAN) base stations (gNBs) operating in dynamic wireless environments. The framework comprises three innovative components. The first is the introduction of novel Service Models (SMs) within a vision-enabled O-RAN architecture, termed VisionRAN. The

Madhav S. Baidya, S. S. Baidya, Chirag Chawla

The rapid proliferation of large language models (LLMs) has created an urgent need for robust and generalizable detectors of machine-generated text. Existing benchmarks typically evaluate a single detector on a single dataset under ideal conditions, leaving open questions about cross-domain transfer, cross-LLM generalization, and adversarial robustness. We p

Masafumi Hattori, Theodoros Stylianos Papazachariou, Aline Zanardini

A general net of quadric surfaces, together with a choice of a base point, defines a net of plane cubics via the Gale transformation of the remaining seven base points. To both nets, one can also naturally associate the same smooth plane quartic. In this paper, we generalize the cycle of correspondences arising from nets of quadrics that define rational elli

Jianjian Yin, Tao Chen, Yi Chen, Gensheng Pei

Recent advances in vision-language models (VLMs) have garnered substantial attention in open-vocabulary semantic and part segmentation (OSPS). However, existing methods extract image-text alignment cues from cost volumes through a serial structure of spatial and class aggregations, leading to knowledge interference between class-level semantics and spatial c

Guibiao Liao, Qian Ren, Kaimin Liao, Hua Wang

Semantic-aware 3D reconstruction from sparse, unposed images remains challenging for feed-forward 3D Gaussian Splatting (3DGS). Existing methods often predict an over-complete set of Gaussian primitives under sparse-view supervision, leading to unstable geometry and inferior depth quality. Meanwhile, they rely solely on 2D segmenter features for semantic lif

Wasif H. Syed, Juan E. Machado, Hans Würfel, Ekrem Hanli

To reduce CO2 emissions and tackle increasing fuel costs, the aviation industry is swiftly moving towards the electrification of aircraft. From the viewpoint of systems and control, a key challenge brought by this transition corresponds to the management and safe operation of the propulsion system's onboard electrical power distribution network. In this work

Eric Diehl, Adem Tosun, Dimitrios Loukrezis

We propose a data-efficient workflow to optimize the efficiency of a radial turbine design under a strict budget of high-fidelity computational fluid dynamics simulations. Assuming anisotropic parameter impact, we use a maximum-projection initial experimental design to ensure space-filling and strong projection properties on low-dimensional subspaces. Bayesi

İsmail Alperen Öğüt

Motivated by the simplification of decomposition formulas for fibred bisets, we study the homomorphism extension problem for subdirect products of finite groups when the codomain is an abelian group satisfying certain hypothesis. We prove that every homomorphism of subdirect products whose Goursat quotients have trivial Schur multipliers is extensible. We al

Qijie Wei, Hailan Lin, Xirong Li

Current methods for multimodal medical imaging based disease recognition face two major challenges. First, the prevailing "fusion after unimodal image embedding" paradigm cannot fully leverage the complementary and correlated information in the multimodal data. Second, the scarcity of labeled multimodal medical images, coupled with their significant domain s

De Zhang Lee, Han Fang, Ee-Chien Chang

Recent advancements in AI-generated content (AIGC) have introduced new challenges in intellectual property protection and the authentication of generated objects. We focus on scenarios in which an author seeks to assert authorship of an object generated using latent diffusion models (LDMs), in the presence of adversaries who attempt to falsely claim authorsh

Mengyu Bu, Yang Feng

Large language models (LLMs) exhibit strong general intelligence, yet their multilingual performance remains highly imbalanced. Although LLMs encode substantial cross-lingual knowledge in a unified semantic space, they often struggle to reliably interface this knowledge with low-resource or unseen languages. Fortunately, pretrained encoder-decoder translatio

Chen Zhao, Zhuoran Wang, Haoyang Li, Shifeng Bao

Vision-Language-Action (VLA) models have recently demonstrated strong performance across embodied tasks. Modern VLAs commonly employ diffusion action experts to efficiently generate high-precision continuous action chunks, while auto-regressive generation can be slower and less accurate at low-level control. Yet auto-regressive paradigms still provide comple

Aleksandra Edwards, Thomas Edwards, Jose Camacho-Collados, Alun Preece

Large Language Models (LLMs) are extensively used in text generation tasks. These generative capabilities bring us to a point where LLMs could potentially provide useful insights in policy making or agency operations. In this paper, we introduce a new task consisting of generating recommendations which can be used to inform future actions and improvements of

Syamjith KS, Alan Ranjit Jacob

The utility of non functionalized poly(N-isopropylacrylamide) (pNIPAM) microgels in physiological and environmental applications is strictly dependent on their reversible thermoresponsiveness and stability in saline media. Despite their importance, a unified understanding of how network topology specifically crosslinker concentration and distribution regulat

Tharun Sethuraman, Subham Agrawal, Nils Dengler, Jorge de Heuvel

Robots operating in human-shared environments must not only achieve task-level navigation objectives such as safety and efficiency, but also adapt their behavior to human preferences. However, as human preferences are typically expressed in natural language and depend on environmental context, it is difficult to directly integrate them into low-level robot c

Klaus Pinn

I propose and investigate the use of continuous functional equations for the study of meta-Fibonacci integer sequences. This exploratory study includes three sequences with quite different behavior: Conway's famous sequence $A(n)= A(A(n-1))+A(n-A(n-1))$, the sequence $D(n)= D(D(n-1))+D(n-1-D(n-2))$ introduced by the present author more than 25 years ago, and

Jiawei Zhou, Chi Zhang, Xiang Feng, Qiming Zhang

We present Omni-I2C, a comprehensive benchmark designed to evaluate the capability of Large Multimodal Models (LMMs) in converting complex, structured digital graphics into executable code. We argue that this task represents a non-trivial challenge for the current generation of LMMs: it demands an unprecedented synergy between high-fidelity visual perception

Jiarui Guo, Yuemeng Xu, Zongwei Lv, Yangyujia Wang

Retrieval-Augmented Generation (RAG) enables Large Language Models (LLMs) to extend their existing knowledge by dynamically incorporating external information. However, practical deployment is fundamentally constrained by the LLM's finite context window, forcing a trade-off between information sufficiency and token consumption. Standard pipelines address thi

Charuka Herath, Yogachandran Rahulamathavan, Varuna De Silva, Sangarapillai Lambotharan

Federated Learning (FL) enables privacy-preserving intelligence on Internet of Things (IoT) devices but incurs a significant carbon footprint due to the high energy cost of frequent uplink transmission. While pre-trained models are increasingly available on edge devices, their potential to reduce the energy overhead of fine-tuning remains underexplored. In t

Yitian Gong, Botian Jiang, Yiwei Zhao, Yucheng Yuan

This technical report presents MOSS-TTS, a speech generation foundation model built on a scalable recipe: discrete audio tokens, autoregressive modeling, and large-scale pretraining. Built on MOSS-Audio-Tokenizer, a causal Transformer tokenizer that compresses 24 kHz audio to 12.5 fps with variable-bitrate RVQ and unified semantic-acoustic representations, w

M. M. Ridha Fathima, Alexandre M. R. Almeida, Mattia Bulla, Jaziel G. Coelho

Discoveries of ultra-massive magnetic white dwarfs (WDs) and peculiar pulsars have been proposed to originate in double white dwarf (DWD) mergers. There are three possible post-merger central remnants of non-explosive mergers: 1) a stable sub-Chandrasekhar WD; 2) a rapidly rotating super-Chandrasekhar WD; 3) a neutron star (NS). In this work, we explore the

Cem Uluoglakci, Tugba Taskaya Temizel

Large language models (LLMs) often hallucinate, producing fluent but false information, partly because supervised fine-tuning (SFT) implicitly rewards always responding. We introduce $\textit{HypoTermInstruct}$, an SFT dataset (31,487 responses for 11,151 questions) designed to teach models epistemological humility-the ability to recognize the limits of thei

H. -R. Zhang, Z. -F. Cui, J. Segovia

We study the pseudoscalar ($J^{PC}=0^{-+}$) and vector ($1^{--}$) top--antitop (toponium) systems within the rainbow--ladder truncation of the Dyson--Schwinger and Bethe--Salpeter equations, employing the Qin--Chang effective interaction. After validating the framework in the charmonium and bottomonium sectors, we extend it consistently to the top sector, in

Dáire Healy, Isadora Antoniano-Villalobos, Claudia Collarin, Nathan Huet

This article summarises the methods used by the team ``Ca' Foscari" for the EVA 2025 Data Challenge. The questions of the challenge concern the estimation of exceedance probabilities across several locations. Rather than modelling the spatial dependence structure, we reduce the problems to univariate ones by considering relevant spatial order statistics acro

Arvin Rasoulzadeh

Alignable nets are grid structures that can collapse to a planar strip, which is in fact the real-world counterpart of a curve. This property simplifies on-site assembly and enables compact transport and storage. These grid structures can then be deployed by a scissor motion at each vertex in a desired location. In this article, we classify all surfaces supp

Wei Hong, Luca Izzo, Massimo Della Valle, Orlando Luongo

Context. Gamma-ray bursts (GRBs) reach redshifts beyond Type Ia supernovae (SNe Ia) and can extend distance measurements into the early Universe, but their use as distance indicators is limited by the circularity problem in calibrating empirical luminosity relations. Aims. We present a model-independent methodology to overcome this circularity by combining P

Huansheng Ning, Jianguo Ding

Human communication is undergoing a fundamental paradigm shift. Physical space, social relations, mental states, and digital information are converging into a unified cyber-physical-social-thinking (CPST) fusion space, rendering them no longer separable domains. However, all existing communication systems, including natural and programming languages, as well

Jianghong Dong, Chunying Yang, Mengchi Cai, Chaoyi Chen

Sufficient testing under corner cases is critical for the long-term operation of vehicle-infrastructure cooperation systems (VICS). However, existing corner-case generation methods are primarily AI-driven, and VICS testing under corner cases is typically limited to simulation. In this paper, we introduce an L5 ''Interactable'' level to the VICS digital twin

Thomas Duboudin, Xavier Fontaine, Etienne Andrier, Lionel Guillou

Computational pathology has made significant progress in recent years, fueling advances in both fundamental disease understanding and clinically ready tools. This evolution is driven by the availability of large amounts of digitized slides and specialized deep learning methods and models. Multiple self-supervised foundation feature extractors have been devel

J. M. Morgado-Chávez, J. Segovia, F. de Soto, J. Rodríguez-Quintero

Generalised Parton Distributions (GPDs) provide multidimensional insight into hadron structure and are particularly relevant for the pion, whose dynamics are intimately linked to chiral symmetry breaking. We introduce a novel modelling strategy for pion GPDs that satisfies all QCD constraints by construction: support, polynomiality, positivity, and the soft-

Thiago Marzagao

Classical equations for predicting one-repetition maximum (1RM) from submaximal performance were derived from small samples performing a single exercise, yet are routinely applied to hundreds of exercises. All use a fixed conversion factor relating repetitions to estimated 1RM, regardless of exercise or load. We used large-scale observational data from a con

Yi Qin, Yee Sin Ang, Linhu Li, Ching Hua Lee

We show that anyonic statistics fundamentally reshapes non-Hermitian many-body physics by intrinsically breaking pseudo-Hermiticity, leading to a unique real-complex spectral transition with characteristically dense states in Im$E$. This anyon-induced transition occurs even when bosonic and pseudofermionic counterparts remain entirely real, revealing a form

Zhehui Wang, Justin C. Burton, Niklas Dormagen, Cheng-Ran Du

Dusty plasmas are ubiquitous throughout the universe, spanning laboratory and industrial plasmas, fusion devices, planetary environments, cometary comae, and interstellar media. Despite decades of research, many aspects of their behavior remain poorly understood within a unified framework. While numerous theoretical and numerical models describe specific phe

Shenghui Huang, Menghao Hu, Longkun Zou, Hongyu Chi

Detecting Unmanned Aerial Vehicles (UAVs) in low-altitude environments is essential for perception and defense systems but remains highly challenging due to complex backgrounds, camouflage, and multimodal interference. In real-world scenarios, UAVs are frequently visually blended with surrounding structures such as buildings, vegetation, and power lines, res

Pascal Auscher, Lukas Niebel

We define and study homogeneous kinetic Sobolev spaces adapted to the Kolmogorov equation. We consider both local and non-local diffusion. The spaces are built from the Lebesgue spaces L p for all integrability exponents p $\in$ (1, $\infty$) with regularity assumptions in the transport and diffusive directions according to the scaling of the Kolmogorov equa

Christophe Gomez, Knut Sølna

In this paper, we consider the reflection and transmission problem of waves by a rapidly oscillating rough interface that exhibits general mixing properties. Using an asymptotic analysis based on a separation of scales, corresponding to a paraxial (parabolic) scaling regime, we precisely characterize the specular and speckle (diffusive) components of the ref

Qingguang Xie, Paolo Malgaretti, Othmane Aouane, Simon Thiele

Bubble nucleation at catalyst surfaces plays a critical role in the operation of electrolyzers. However, achieving controlled bubble nucleation remains challenging due to limited understanding of the underlying mechanisms. Here, we present a free-energy model that quantitatively predicts both the activation energy and critical nucleus size of bubbles at give

Kenta Hashizume

We establish the generalized canonical bundle formula for generalized lc-trivial fibrations with irrational coefficients over non-compact bases in the complex analytic setting, and we show that the discriminant b-divisor and moduli b-divisor are compatible with restriction to arbitrary open subsets. We also discuss the boundedness of complements in this sett

Amijit Bhattacharjee, Himanshu Bora, Prabwal Phukon

We study the thermodynamic geometry of the one-dimensional Blume--Capel model within the Tsallis nonextensive framework to understand how generalized statistics modify correlation structure and pseudo-critical behaviour. Using the transfer matrix method, we construct the Tsallis entropy based thermodynamic metric as its negative Hessian on the parameter spac

L. Drouglazet, E. Alecian, A. Sousa, P. I. Cristofari

Magnetic fields play a crucial role throughout stellar evolution, regulating angular momentum, channelling accretion, and launching jets and outflows. While the magnetic properties of Classical T Tauri Stars (CTTS) are well characterised, those of their progenitors, Class I and Flat-Spectrum (FS) protostars, remain poorly constrained due to observational cha

Andrea Montessori, Maria Rosa Lisboa, Marco Lauricella, Sauro Succi

Unresolved thin-film dynamics often causes spurious coalescence in diffuse-interface simulations of multiphase flows. We address this issue by introducing a fully local repulsive near-contact flux in a conservative Allen--Cahn phase-field model coupled to lattice Boltzmann hydrodynamics. The interaction activates only for oppositely oriented nearby interface

Chen Liu, Hongxin Kong, Lang Feng, Wenchao Qian

As VLSI designs grow in complexity, partitioning is widely adopted to accelerate physical design through parallel computing. However, traditional hypergraph partitioning methods often degrade in performance when applied to 2D layouts due to spatial constraints. For routers with post-placement locations, a spatial-aware partitioning method fully utilizing pla

Magalie Bénéfice, Michel Bonnefont, Marc Arnaudon, Delphine Féral

We establish an integration by parts formula for the semi-group in time $T > 0$ of the kinetic Brownian motion in the Euclidean plane together with its speed in the circle. The stochastic differential equation of our kinetic Brownian motion is driven here by one real-valued Brownian motion constructed with an orthonormal basis of $L^2([0,T],\mathbb R)$ and a

Fritz Körmann, Yuji Ikeda, Konstantin Glazyrin, Maxim Bykov

Hydrogen uptake in complex multicomponent alloys, including high-entropy alloys (HEAs), governs both hydrogen storage capacity and resistance to hydrogen-induced degradation. We combine high-pressure experiments, density-functional theory (DFT), and a GRACE universal interatomic potential to investigate hydrogen absorption in Al$_{0.3}$CoCrFeNi and Al$_3$CoC

Hao Ma, Zhiqiang Pu, Yang Liu, Xiaolin Ai

Constraints are essential for stabilizing reinforcement learning fine-tuning (RFT) and preventing degenerate outputs, yet they inherently conflict with the optimization objective because stronger constraints limit the ability of a fine-tuned model to discover better solutions. We propose \textit{dynamic constraints} that resolve this tension by adapting to t

Przemyslaw Chojecki

We prove that every sufficiently large integer $n$ can be written in the form $n=x^2+y^2-z^2$ with $\textrm{max}(x^2,y^2,z^2)\le n$. The proof converts the problem into finding a primitive binary quadratic form of positive discriminant $4n$ inside a fixed relatively compact open patch of the real hyperboloid $b^2-4ac=4n$. This is then supplied by Duke's theo

Ahmet Kaplan

This study explores the combination of automated machine learning (AutoML) with model-based deep unfolding (DU) for optimizing wireless beamforming and waveforms. We convert the iterative proximal gradient descent (PGD) algorithm into a deep neural network, wherein the parameters of each layer are learned instead of being predetermined. Additionally, we enha

Changjian Xie

In this paper, we propose a structure preserving method using a Crank-Nicolson's type method with an implicit Gauss-Seidel fractional iteration. Such a method is of first-order accuracy in time and second-order accuracy in space, stable and length preserving. Such a proposed method brings great benefits for the theoretical analysis. The numerical accuracy, n

Segyu Lee, Boryeong Cho, Hojung Jung, Seokhyun An

Unified Multimodal Models (UMMs) offer powerful cross-modality capabilities but introduce new safety risks not observed in single-task models. Despite their emergence, existing safety benchmarks remain fragmented across tasks and modalities, limiting the comprehensive evaluation of complex system-level vulnerabilities. To address this gap, we introduce UniSA

Jasper Jian, Christopher D. Manning

Categorization is a core component of human linguistic competence. We investigate how a transformer-based language model (LM) learns linguistic categories by comparing its behaviour over the course of training to behaviours which characterize abstract feature-based and concrete exemplar-based accounts of human language acquisition. We investigate how lexical

Zelin Zang, Yehui Yang, Fei Wang, Liangyu Li

Unsupervised Domain Adaptation (UDA) seeks to transfer knowledge from a labeled source domain to an unlabeled target domain but often suffers from severe domain and scale gaps that degrade performance. Existing cross-attention-based transformers can align features across domains, yet they struggle to preserve content semantics under large appearance and scal

Bin Bin Han, Wen Ting Zhang, Yan Feng Luo

Let $\mathsf{mSt}_n$ be the plactic-like monoid obtained by factoring the free monoid over a finite alphabet $\mathcal{A}_n$ by the meet of the stalactic congruence and its dual. In this paper, we prove that $\mathsf{mSt}_n$ can be equipped with multiple involutions, and divide these involutions into $\lfloor\frac{n}{2}\rfloor+1$ types. A faithful representa

Wei Tang, Xuejing Liu, Yanpeng Sun, Zechao Li

The Segment Anything Model (SAM) excels at general image segmentation but has limited ability to understand natural language, which restricts its direct application in Referring Expression Segmentation (RES). Toward this end, we propose SSP-SAM, a framework that fully utilizes SAM's segmentation capabilities by integrating a Semantic-Spatial Prompt (SSP) enc

Xin Wei Chia, Swee Liang Wong, Jonathan Pan

Recent incidents have highlighted alarming cases where human-AI interactions led to negative psychological outcomes, including mental health crises and even user harm. As LLMs serve as sources of guidance, emotional support, and even informal therapy, these risks are poised to escalate. However, studying the mechanisms underlying harmful human-AI interaction

Camille Grange, Quentin Bourgeais, Rodolphe Charrier, Géraldine Del Mondo

Team sports represent complex phenomena characterized by both spatial and temporal dimensions, making their analysis inherently challenging. In this study, we examine team sports as complex systems, specifically focusing on the tactical aspects influenced by external constraints. To this end, we introduce a new generic graph-based model to analyze these phen

Chupeng Liu, Jiyong Rao, Shangquan Sun, Runkai Zhao

Monocular 3D object detection typically relies on pseudo-labeling techniques to reduce dependency on real-world annotations. Recent advances demonstrate that deterministic linguistic cues can serve as effective auxiliary weak supervision signals, providing complementary semantic context. However, hand-crafted textual descriptions struggle to capture the inhe

Jan-Ole Fischer

Hidden Markov models (HMMs) are powerful tools for analysing time series data that depend on discrete underlying but unobserved states. As such, they have gained prominence across numerous empirical disciplines, in particular ecology, medicine, and economics. However, the increasing complexity of empirical data is often accompanied by additional latent struc

Hao Ma, Zhiqiang Pu, Xiaolin Ai, Huimu Wang

We present GuidedSAC, a novel reinforcement learning (RL) algorithm that facilitates efficient exploration in vast state-action spaces. GuidedSAC leverages large language models (LLMs) as intelligent supervisors that provide action-level guidance for the Soft Actor-Critic (SAC) algorithm. The LLM-based supervisor analyzes the most recent trajectory using sta

Jens Markus Melenk, David Wörgötter

We consider the time-harmonic Maxwell equations with impedance boundary conditions on a bounded Lipschitz domain $\Omega$ with analytic boundary $\Gamma$. We suppose that $\Omega$ consists of multiple subdomains, and that the permeability and permittivity tensors are analytic on every subdomain, but may jump across subdomain interfaces. Under these condition

Diego Cruces, Minxi He, Shi Pi, Jianing Wang

In this letter, we propose a single-field inflation model that realizes a slow-roll-to-ultra-slow-roll transition while keeping the second slow-roll parameter smoothly varying throughout. The model is built through a minimal modification by introducing a simple time dependence in the effective mass term of the Mukhanov-Sasaki equation. We obtain fully analyt

Ya-Wen Xiao, Jian-Fu Zhang, Alex Lazarian

Based on multi-frequency radio polarization survey datasets, we investigate the spectral characteristics of the Galactic interstellar medium (ISM) using the polarization frequency analysis (PFA) method, referred to as polarization variance. By comparing this novel PFA technique with the traditional power spectrum approach, and by cross-examining data from tw

Daisuke Yasui, Toshitaka Matsuki, Hiroshi Sato

In locomotion control tasks, Deep Reinforcement Learning (DRL) has demonstrated high performance; however, the decision-making process of the learned policy remains a black box, making it difficult for humans to understand. On the other hand, in periodic motions such as walking, it is well known that implicit motion phases exist, such as the stance phase and

Massimiliano Caporin, Daniele Girolimetto, Emanuele Lopetuso

We assess the advantage of combining univariate and multivariate portfolio risk forecasts with the aid of forecast reconciliation techniques. In our analyzes, we assume knowledge of portfolio weights, a standard for portfolio risk management applications. With an extensive simulation experiment, we show that, if the true covariance is known, forecast reconci

Denis S. Goldobin, Yuriy L. Raikher

A fluid mechanics problem is solved which technological prototype is a fluid acoustic contact that is an inherent element of ultrasonic non-destructive testing procedures. It is well known that the acoustic contact established with an ordinary fluid suffers from essential disadvantage that is the loss of stability due to the gravity-induced fluid leakage in