May 2025 arXiv papers — page 22

Rahul Kumar, Devranjan Samanta, Srikant S. Padhee

Drawing inspiration from the adaptive wing shape of birds in flight, this study introduces a bio-inspired concept for shape adaptation utilizing bend-twist coupling (BTC) in composite laminates. The primary aim of the design optimization is to identify the optimal fibre orientation angles needed to produce the required bending and twisting deformations, whic

Ehtesham Anwar, Ujjwal Kumar Nandi, Palak Patel, Sanket Kumawat

Understanding the connection between thermodynamics and dynamics in glass-forming liquids remains a central challenge in condensed matter physics. In this study, we investigate a novel model system that enables a continuous crossover from a standard three dimensional liquid to a fully connected mean field like system by introducing pseudo neighbours. These p

Jacek Miȩkisz, Javad Mohamadichamgavi

We examine the impact of the maintenance cost of social links on cooperative behavior in the Prisoner's Dilemma game on the Barab\'asi-Albert scale-free network with a pairwise stochastic imitation. We show by means of Monte Carlo simulations and pair approximation that the cooperation frequency changes abruptly from an almost full cooperation to a much smal

Mannmohan Muthuraman

Dynamic Spectral Backpropagation (DSBP) enhances neural network training under resource constraints by projecting gradients onto principal eigenvectors, reducing complexity and promoting flat minima. Five extensions are proposed, dynamic spectral inference, spectral architecture optimization, spectral meta learning, spectral transfer regularization, and Lie

Beiduo Chen, Yang Janet Liu, Anna Korhonen, Barbara Plank

The recent rise of reasoning-tuned Large Language Models (LLMs)--which generate chains of thought (CoTs) before giving the final answer--has attracted significant attention and offers new opportunities for gaining insights into human label variation, which refers to plausible differences in how multiple annotators label the same data instance. Prior work has

Jeonghyeok Do, Sungpyo Kim, Geunhyuk Youk, Jaehyup Lee

PAN-sharpening aims to fuse high-resolution panchromatic (PAN) images with low-resolution multi-spectral (MS) images to generate high-resolution multi-spectral (HRMS) outputs. However, cross-modality misalignment -- caused by sensor placement, acquisition timing, and resolution disparity -- induces a fundamental challenge. Conventional deep learning methods

Ahmad Ghodsi, Elias Kiritsis, Parisa Mashayekhi, Francesco Nitti

We analyze linear fluctuations of five-dimensional Einstein-Dilaton theories dual to holographic quantum field theories defined on four-dimensional de Sitter and Anti-de Sitter space-times. We identify the physical propagating scalar and tensor degrees of freedom. For these, we write the linearized bulk field equations as eigenvalue equations. In the dual QF

Yang Qiao, Xiaoyu Zhong, Xiaofeng Gu, Zhiguo Yu

Multimodal information processing has become increasingly important for enhancing image classification performance. However, the intricate and implicit dependencies across different modalities often hinder conventional methods from effectively capturing fine-grained semantic interactions, thereby limiting their applicability in high-precision classification

Dongming Hua

We study the rigidity of the volume entropy for weighted word metrics on hyperbolic groups, building on a recent convexity result due to Cantrell-Tanaka. Using ideas from small cancellation theory, we give conditions under which a hyperbolic group admits a unique normalized weight minimizing the entropy. Moreover, we show that these conditions are generic fo

Hongzhan Chen, Tao Yang, Shiping Gao, Ruijun Chen

Process reward models (PRMs) provide more nuanced supervision compared to outcome reward models (ORMs) for optimizing policy models, positioning them as a promising approach to enhancing the capabilities of LLMs in complex reasoning tasks. Recent efforts have advanced PRMs from step-level to token-level granularity by integrating reward modeling into the tra

Liu Yang, Ryo Sugano, Ryomei Takabayashi, Hidetoshi Kanzawa

AMTIR-1 chalcogenide glass has shown its potential for use in thermal imaging systems owing to its low refractive index, thermal resistance and high transparency across the infrared wavelength regime. Here we report a millimeter-scale high-Q whispering gallery mode microresonator made of AMTIR-1. The recorded Q-factor has reached $1.2\times10^7$ at 1550 nm,

Zeyuan Liu, Zhihe Yang, Jiawei Xu, Rui Yang

Real-world datasets collected from sensors or human inputs are prone to noise and errors, posing significant challenges for applying offline reinforcement learning (RL). While existing methods have made progress in addressing corrupted actions and rewards, they remain insufficient for handling corruption in high-dimensional state spaces and for cases where m

Pavan Kumar Singeetham, Deepak Madival, Piyush Garg, Ganesh Subramanian

We investigate the orientation dynamics of a neutrally buoyant spheroid, of an arbitrary aspect ratio ($\kappa$), freely rotating in a weakly viscoelastic fluid undergoing simple shear flow. Weak elasticity is characterized by a small but finite Deborah number ($De$), and the suspending fluid rheology is therefore modeled as a second-order fluid, with the co

Wiktoria Mieleszczenko-Kowszewicz, Beata Bajcar, Aleksander Szczęsny, Maciej Markiewicz

In this work we present the Social Influence Technique Taxonomy (SITT), a comprehensive framework of 58 empirically grounded techniques organized into nine categories, designed to detect subtle forms of social influence in textual content. We also investigate the LLMs ability to identify various forms of social influence. Building on interdisciplinary founda

Fereshte Shahbeigi, M. Hamed Mohammady

In order to determine what quantum operations and measurements are consistent with the laws of thermodynamics, one must start by allowing all processes allowed by the framework of quantum theory, and then impose the laws of thermodynamics as a set of constraints. Here, we consider a hierarchy of quantum operations and measurements that are consistent with ($

Yuanxin Liu, Kun Ouyang, Haoning Wu, Yi Liu

Recent studies have shown that long chain-of-thought (CoT) reasoning can significantly enhance the performance of large language models (LLMs) on complex tasks. However, this benefit is yet to be demonstrated in the domain of video understanding, since most existing benchmarks lack the reasoning depth required to demonstrate the advantages of extended CoT ch

Reem AlJunaid, Muzammil Behzad

Generating informative and knowledge-rich image captions remains a challenge for many existing captioning models, which often produce generic descriptions that lack specificity and contextual depth. To address this limitation, we propose KRCapVLM, a knowledge replay-based novel image captioning framework using vision-language model. We incorporate beam searc

Cristina-Elena Popa, Cristian Damian, Daniela Coltuc

The paper proposes a method to secure the Compressive Sensing (CS) streams. It consists in protecting part of the measurements by a secret key and inserting the code into the rest. The secret key is generated via a cryptographically secure pseudo-random number generator (CSPRNG) and XORed with the measurements to be inserted. For insertion, we use a reversib

Pritish Joshi, Abhishek Bera, Niladri Patra

Diseases caused by bacteria have been a threat to human civilisation for centuries. Despite the availability of numerous antibacterial drugs today, bacterial diseases continue to pose life-threatening challenges. The credit for this goes to Gram-Negative bacteria, which have developed multi-drug resistant properties towards \b{eta}-lactams, chloramphenicols,

Maxiu Xiao, Jianglin Lan, Jingxin Yu, Weihong Ma

Climate control is crucial for greenhouse production as it directly affects crop growth and resource use. Reinforcement learning (RL) has received increasing attention in this field, but still faces challenges, including limited training efficiency and high reliance on initial learning conditions. Interactive RL, which combines human (grower) input with the

Meital Bojan, Sanketh Vedula, Advaith Maddipatla, Nadav Bojan Sellam

The local structure of a protein strongly impacts its function and interactions with other molecules. Therefore, a concise, informative representation of a local protein environment is essential for modeling and designing proteins and biomolecular interactions. However, these environments' extensive structural and chemical variability makes them challenging

Sina Amirrajab, Volker Vehof, Michael Bietenbeck, Ali Yilmaz

Purpose: We investigated the utilization of privacy-preserving, locally-deployed, open-source Large Language Models (LLMs) to extract diagnostic information from free-text cardiovascular magnetic resonance (CMR) reports. Materials and Methods: We evaluated nine open-source LLMs on their ability to identify diagnoses and classify patients into various cardiac

Alexandra G. Roberts, Ha M. Luu, Mert Şişman, Alexey V. Dimov

Quantitative susceptibility maps from magnetic resonance images can provide both prognostic and diagnostic information in multiple sclerosis, a neurodegenerative disease characterized by the formation of lesions in white matter brain tissue. In particular, susceptibility maps provide adequate contrast to distinguish between "rim" lesions, surrounded by depos

Xu Shen, Yixin Liu, Yiwei Dai, Yili Wang

The communication topology in large language model-based multi-agent systems fundamentally governs inter-agent collaboration patterns, critically shaping both the efficiency and effectiveness of collective decision-making. While recent studies for communication topology automated design tend to construct sparse structures for efficiency, they often overlook

Sudam M. Wasala, Jurre Wolff, Yixian Shen, Anuj Pathania

Optimizing performance and energy efficiency in many-core processors, especially within Non-Uniform Cache Access (NUCA) architectures, remains a critical challenge. The performance heterogeneity inherent in S-NUCA systems complicates task scheduling due to varying cache access latencies across cores. This paper introduces a novel QoS management policy to mai

Nunzia Gavitone, Alba Lia Masiello, Gloria Paoli, Giorgio Poggesi

It is well known that there is a deep connection between Serrin's symmetry result -- dealing with overdetermined problems involving the Laplacian -- and the celebrated Alexandrov's Soap Bubble Theorem (SBT) -- stating that, if the mean curvature $H$ of the boundary of a smooth bounded connected open set $\Om$ is constant, then $\Om$ must be a ball. One of th

Sheng Ouyang, Yulan Hu, Ge Chen, Qingyang Li

Rewards serve as proxies for human preferences and play a crucial role in Reinforcement Learning from Human Feedback (RLHF). However, if these rewards are inherently imperfect, exhibiting various biases, they can adversely affect the alignment of large language models (LLMs). In this paper, we collectively define the various biases present in rewards as the

Marco Caroccia, Nicolas Van Goethem

We provide a structural analysis of the space of functions of bounded deviatoric deformation, $\mathrm{BD}_{dev}$, which arises in models of plasticity and fluid mechanics. The main result is the identification of the annihilator and a rigidity theorem for $\mathrm{BD}_{dev}$-maps with constant polar vector in the wave cone characterizing the structure of si

Isarun Chamveha, Supphanut Chaiyungyuen, Sasinun Worakriangkrai, Nattawadee Prasawang

This study presents a deep learning system for breast cancer detection in mammography, developed using a modified EfficientNetV2 architecture with enhanced attention mechanisms. The model was trained on mammograms from a major Thai medical center and validated on three distinct datasets: an in-domain test set (9,421 cases), a biopsy-confirmed set (883 cases)

Yuting Li, Shaoyuan Huang, Tengwen Zhang, Cheng Zhang

With the rapid growth of live streaming services, Crowdsourced Cloud-edge service Platforms (CCPs) are playing an increasingly important role in meeting the increasing demand. Although stream scheduling plays a critical role in optimizing CCPs' revenue, most optimization strategies struggle to achieve practical results due to various anomalies in unstable CC

Yexiong Lin, Yu Yao, Tongliang Liu

Flow Matching (FM) is an effective framework for training a model to learn a vector field that transports samples from a source distribution to a target distribution. To train the model, early FM methods use random couplings, which often result in crossing paths and lead the model to learn non-straight trajectories that require many integration steps to gene

Yang Liu, Deyu Bo, Wenxuan Cao, Yuan Fang

Graph self-supervised learning seeks to learn effective graph representations without relying on labeled data. Among various approaches, graph autoencoders (GAEs) have gained significant attention for their efficiency and scalability. Typically, GAEs take incomplete graphs as input and predict missing elements, such as masked nodes or edges. While effective,

Jakub Martinka, Lina Zhang, Yi-Fan Hou, Mikołaj Martyka

Nonadiabatic couplings (NACs) play a crucial role in modeling photochemical and photophysical processes with methods such as the widely used fewest-switches surface hopping (FSSH). There is therefore a strong incentive to machine learn NACs for accelerating simulations. However, this is challenging due to NACs' vectorial, double-valued character and the sing

Sixian Wang, Zhiwei Tang, Tsung-Hui Chang

Diffusion models often exhibit inconsistent sample quality due to stochastic variations inherent in their sampling trajectories. Although training-based fine-tuning (e.g. DDPO [1]) and inference-time alignment techniques[2] aim to improve sample fidelity, they typically necessitate full denoising processes and external reward signals. This incurs substantial

Haeun Lee, Omin Kwon, Yeonhong Park, Jae W. Lee

Meeting service-level objectives (SLOs) in Large Language Models (LLMs) serving is critical, but managing the high variability in load presents a significant challenge. Recent advancements in FP8 inference, backed by native hardware support, offer a potential solution: executing FP16 models by default, while switching to FP8 models during sudden load surges

Maya Dewhurst, Jack Collins, Justin J. H. Lo, Roy Alderton

We introduce Nosey (Nasalance Open Source Estimation sYstem), a low-cost, customizable, 3D-printed system for recording acoustic nasalance data that we have made available as open-source hardware (http://github.com/phoneticslab/nosey). We first outline the motivations and design principles behind our hardware nasalance system, and then present a comparison b

Zahoor Malik, Sikandar Akbar, Zahir Shah, Ranjeev Misra

This study examines the flux and photon index distributions of 11 Very High Energy (VHE) Flat Spectrum Radio Quasars (FSRQs) using over 16 years of Fermi-LAT $\gamma$-ray data. The distributions reveal double lognormal profiles in both flux and index, primarily in the 3-day and 7-day binnings, supporting the ``two-flux-state hypothesis" for blazars. These pr

Yixian Shen, Qi Bi, Jia-Hong Huang, Hongyi Zhu

We present a new adaptation method MaCP, Minimal yet Mighty adaptive Cosine Projection, that achieves exceptional performance while requiring minimal parameters and memory for fine-tuning large foundation models. Its general idea is to exploit the superior energy compaction and decorrelation properties of cosine projection to improve both model efficiency an

Chetan Verma, Aditya Srinivas Timmaraju, Cho-Jui Hsieh, Suyash Damle

Industry-grade ML models are carefully designed to meet rapidly evolving serving constraints, which requires significant resources for model development. In this paper, we propose MatTA, a framework for training multiple accurate Student models using a novel Teacher-TA-Student recipe. TA models are larger versions of the Student models with higher capacity,

Monika Wesołowska, Sławomir Kuźmar, Bartosz Totleben, Dawid Piątek

This article addresses the origins of income inequality in post-socialist countries from Central and Eastern Europe and Central Asia, from 1991 to 2016. The aim is to analyze the relationship between democracy and income inequality. In previous studies, this topic has led to ambiguous findings, especially in the context of the group of countries we are focus

Zhihan Jin, Matthew Kwan, Lisa Sauermann, Yiting Wang

Consider a degree-$d$ polynomial $f(\xi_1,\dots,\xi_n)$ of independent Rademacher random variables $\xi_1,\dots,\xi_n$. To what extent can $f(\xi_1,\dots,\xi_n)$ concentrate on a single point? This is the so-called polynomial Littlewood-Offord problem. A nearly optimal bound was proved by Meka, Nguyen and Vu: the point probabilities are always at most about

Tu Bui, Mohamed Suliman, Aparajita Haldar, Mohammed Amer

Despite the transformative impact of deep learning on text, audio, and image datasets, its dominance in tabular data, especially in the medical domain where data are often scarce, remains less clear. In this paper, we propose X2Graph, a novel deep learning method that achieves strong performance on small biological tabular datasets. X2Graph leverages externa

Lukas Bauer

This paper provides comprehensive simulation results on the finite sample properties of the Diebold-Mariano (DM) test by Diebold and Mariano (1995) and the model confidence set (MCS) testing procedure by Hansen et al. (2011) applied to the asymmetric loss functions specific to financial tail risk forecasts, such as Value-at-Risk (VaR) and Expected Shortfall

S. A. Avdonin, V. S. Mikhaylov, A. V. Rybkin

We link the Boundary Control Theory and the Titchmarsh-Weyl Theory. This provides a natural interpretation of the $A-$amplitude due to Simon and yields a new efficient method to evaluate the Titchmarsh-Weyl $m-$function associated with the Schr\"{o}dinger operator $H=-\partial _{x}^{2}+q\left( x\right) $ on $L_{2}\left( 0,\infty \right) $ with Dirichlet boun

Matteo Gallici, Haitz Sáez de Ocáriz Borde

Fine-tuning pre-trained generative models with Reinforcement Learning (RL) has emerged as an effective approach for aligning outputs more closely with nuanced human preferences. In this paper, we investigate the application of Group Relative Policy Optimization (GRPO) to fine-tune next-scale visual autoregressive (VAR) models. Our empirical results demonstra

Thomas Goodwin, Arthur Guillaumin, Matias Quiroz, Mattias Villani

Bayesian inference for stationary random fields is computationally demanding. Whittle-type likelihoods in the frequency domain based on the fast Fourier Transform (FFT) have several appealing features: i) low computational complexity of only $\mathcal{O}(n \log n)$, where $n$ is the number of spatial locations, ii) robustness to assumptions of the data-gener

S. A. Avdonin, V. S. Mikhaylov

We establish connections between different approaches to inverse spectral problems: the classical Gelfand--Levitan theory, the Krein method, the Simon theory, the approach proposed by Remling and the Boundary Control method. We show that the Boundary Control approach provides simple and physically motivated proofs of the central results of other theories. We

Roberta Vieliute, Juan V. Hernández Santisteban, Keith Horne, Hannah Cornfield

Intensive reverberation mapping monitoring programs combine ground-based photometric observations from different telescopes, requiring intercalibration of lightcurves to reduce systematic instrumental differences. We present a new iterative algorithm to calibrate photometric time-series data of active galactic nuclei (AGN) using 100s of comparison stars on t

Ramesh Mete

In this paper, we study the existence of twisted constant scalar curvature K\"{a}hler (cscK) metrics and non-existence of coupled cscK metrics on minimal ruled surfaces over a Riemann surface of genus $2$. Moreover, we give a bound for the Chen-Cheng invariant related to Chen's continuity path for cscK problem on these ruled surfaces.

Junhua Zhu, Lan Luo

Entrepreneurship education equips students to transform innovative ideas into actionable entrepreneurship plans, yet traditional approaches often struggle to provide the personalized guidance and practical alignment needed for success. Focusing on the business plan as a key learning tool and evaluation method, this study investigates the design needs for an

Zhaokun Wang, Jinyu Guo, Jingwen Pu, Lingfeng Chen

Current parameter-efficient fine-tuning methods for adapting pre-trained language models to downstream tasks are susceptible to interference from noisy data. Conventional noise-handling approaches either rely on laborious data pre-processing or employ model architecture modifications prone to error accumulation. In contrast to existing noise-process paradigm

Hengyuan Cao, Yutong Feng, Biao Gong, Yijing Tian

Video generative models can be regarded as world simulators due to their ability to capture dynamic, continuous changes inherent in real-world environments. These models integrate high-dimensional information across visual, temporal, spatial, and causal dimensions, enabling predictions of subjects in various status. A natural and valuable research direction

Annesha Deb, Minerva Mukhopadhyay, Subhajit Dutta

This paper investigates the effectiveness of using the Random Projection Ensemble (RPE) approach in Quadratic Discriminant Analysis (QDA) for ultrahigh-dimensional classification problems. Classical methods such as Linear Discriminant Analysis (LDA) and QDA are used widely, but face significant challenges in their implementation when the data dimension (say,

Zeyu Liu, Zhitian Hou, Guanghao Zhu, Zhijie Sang

Multimodal Large Language Models (MLLMs) have achieved remarkable progress in domains such as visual understanding and mathematical reasoning. However, their application in the medical domain is constrained by two key challenges: (1) multimodal medical datasets are scarce and often contain sparse information, limiting reasoning depth; and (2) Reinforcement L

Harish Tayyar Madabushi, Melissa Torgbi, Claire Bonial

In this position paper we raise critical awareness of a realistic view of LLM capabilities that eschews extreme alternative views that LLMs are either 'stochastic parrots' or in possession of 'emergent' advanced reasoning capabilities, which, due to their unpredictable emergence, constitute an existential threat. Our middle-ground view is that LLMs extrapola

Eleftherios Chatzitheodoridis

In Quillen's paper on rational homotopy theory, the category of 1-reduced simplicial sets is endowed with a family of model structures, the most prominent of which is the one in which the weak equivalences are the rational homotopy equivalences and the fibrant objects are the rational Kan complexes. In this paper, we give a modern approach to this family of

A. S. Mikhaylov, V. S. Mikhaylov

We show the equivalence of inverse problems for different dynamical systems and corresponding canonical systems. For canonical system with general Hamiltonian we outline the strategy of studying the dynamic inverse problem and procedure of construction of corresponding de Branges space.

Connor Cooper, Geoffrey I. Webb, Daniel F. Schmidt

Bayesian network classifiers (BNCs) possess a number of properties desirable for a modern classifier: They are easily interpretable, highly scalable, and offer adaptable complexity. However, traditional methods for learning BNCs have historically underperformed when compared to leading classification methods such as random forests. Recent parameter smoothing

Jian Wang, Yong Wang

The spectral torsion is defined by three vector fields and Dirac operators and the noncommutative residue. Motivated by the spectral torsion and the one form rescaled Dirac operator, we give some new spectral torsion which is the extension of spectral torsion for Dirac operators, and compute the spectral torsion for the one form rescaled Dirac operator on ev

Ioana-Claudia Lazar

We investigate the structure of the minimal displacement set in CAT(0) cubical complexes. We show that such set is convex, it is locally endowed with a CAT(0) metric and it is simply connected.

Woojin Shin, Donghwa Kang, Byeongyun Park, Brent Byunghoon Kang

Detection Transformers (DETR) are increasingly adopted in autonomous vehicle (AV) perception systems due to their superior accuracy over convolutional networks. However, concurrently executing multiple DETR tasks presents significant challenges in meeting firm real-time deadlines (R1) and high accuracy requirements (R2), particularly for safety-critical obje

Kaiyang Guo, Yinchuan Li, Zhitang Chen

Direct alignment methods typically train large language models (LLMs) by contrasting the likelihoods of preferred and dispreferred responses. While effective at capturing relative preferences, these methods are widely observed to suppress the absolute likelihoods of example responses. As a result, aligned models can deviate from expected patterns, exhibiting

Abhirup Chakravarty, Mark Brenchley, Trevor Breakspear, Ian Lewin

A key ethical challenge in Automated Essay Scoring (AES) is ensuring that scores are only released when they meet high reliability standards. Confidence modelling addresses this by assigning a reliability estimate measure, in the form of a confidence score, to each automated score. In this study, we frame confidence estimation as a classification task: predi

Rong-Gen Cai, Li Hu, Shao-Jiang Wang

It is well-known that the unitary nature of black hole evaporation enforces its entropy evolution to follow the Page curve. In this Letter, we find that the quantum speed limit on how fast a quantum system can evolve unitarily, when the maximal information transmission bound is saturated, will slow down the would-be divergent decreasing rate of dynamical bla

Weizhe Kong, Xiao Wang, Ruichong Gao, Chenglong Li

Pedestrian Attribute Recognition (PAR) is an indispensable task in human-centered research and has made great progress in recent years with the development of deep neural networks. However, the potential vulnerability and anti-interference ability have still not been fully explored. To bridge this gap, this paper proposes the first adversarial attack and def

Finn Carter

Text-to-image diffusion models have shown unprecedented generative capability, but their ability to produce undesirable concepts (e.g.~pornographic content, sensitive identities, copyrighted styles) poses serious concerns for privacy, fairness, and safety. {Concept erasure} aims to remove or suppress specific concept information in a generative model. In thi

Bin Wang, Yongqi Han, Minbo Ma, Tianrui Li

Deep learning-based approaches have demonstrated significant advancements in time series forecasting. Despite these ongoing developments, the complex dynamics of time series make it challenging to establish the rule of thumb for designing the golden model architecture. In this study, we argue that refining existing advanced models through a universal calibra

Rolf Drechsler

Verification is one of the central tasks in circuit and system design. While simulation and emulation are widely used, complete correctness can only be ensured based on formal proof techniques. But these approaches often have very high run time and memory requirements. Recently, Polynomial Formal Verification (PFV) has been introduced showing that for many i

Xiaoye Michael Wang, Matthew Prenevost, Aneesh Tarun, Ian Robinson

While virtual reality (VR) holds significant potential to revolutionize digital user interaction, how visual information is presented through VR head-mounted displays (HMDs) differs from naturalistic viewing and interactions in physical environments, leading to performance decrements. One critical challenge in VR development is the vergence-accommodation con

Yixin Ren, Chenghou Jin, Yewei Xia, Li Ke

Determining conditional independence (CI) relationships between random variables is a fundamental yet challenging task in machine learning and statistics, especially in high-dimensional settings. Existing generative model-based CI testing methods, such as those utilizing generative adversarial networks (GANs), often struggle with undesirable modeling of cond

Nimrod Shabtay, Zvi Kons, Avihu Dekel, Hagai Aronowitz

Question answering (QA) systems are designed to answer natural language questions. Visual QA (VQA) and Spoken QA (SQA) systems extend the textual QA system to accept visual and spoken input respectively. This work aims to create a system that enables user interaction through both speech and images. That is achieved through the fusion of text, speech, and ima

Ankit Jain, Yagyank Srivastava, Amey G. Gokhale, Nidheesh Virakante

The phonon thermal conductivity of semiconducting periodic solids can be obtained using the lattice dynamics calculations along with the Boltzmann transport equation and with input from density functional theory calculations. These calculations have resulted in an excellent agreement with experiments without requiring any fitting parameters. However, over th

Marcel van der Westhuizen, David Figueruelo, Rethabile Thubisi, Shambel Sahlu

We consider a non-linear interaction between the dark matter and dark energy components of the universe. In particular, within the FLRW geometry, where dark matter is described by a dust fluid and dark energy by an ideal gas with a constant equation of state parameter, we introduce energy transfer between the two fluids. The effective cosmological fluid lead

Yunkee Chae, Kyogu Lee

We present MGE-LDM, a unified latent diffusion framework for simultaneous music generation, source imputation, and query-driven source separation. Unlike prior approaches constrained to fixed instrument classes, MGE-LDM learns a joint distribution over full mixtures, submixtures, and individual stems within a single compact latent diffusion model. At inferen

Yi Luo, Qiwen Wang, Junqi Yang, Luyao Tang

Generalized Category Discovery (GCD) aims to classify both known and novel categories using partially labeled data that contains only known classes. Despite achieving strong performance on existing benchmarks, current textual GCD methods lack sufficient validation in realistic settings. We introduce Event-Centric GCD (EC-GCD), characterized by long, complex

Ashokkumar P, Sathish Aravindh M, Venkatesan A, Lakshmanan M

The Chua's circuit is examined using a State Controlled-Cellular Neural Network (SC-CNN) framework with two logical square wave input signals. We illustrate, in particular, that this nonlinear circuit can generate all the basic logic operations, including OR/NOR, AND/NAND, and XOR/XNOR gates, by making use of the hopping of attractors which this circuit prod

Tim Hargreaves, Qing Li, Charles Knipp, Frederic Wantiez

State space models (SSMs) are a powerful and widely-used class of probabilistic models for analysing time-series data across various fields, from econometrics to robotics. Despite their prevalence, existing software frameworks for SSMs often lack compositionality and scalability, hindering experimentation and making it difficult to leverage advanced inferenc

Rim Rekik, Stefanie Wuhrer, Ludovic Hoyet, Katja Zibrek

Virtual human animations have a wide range of applications in virtual and augmented reality. While automatic generation methods of animated virtual humans have been developed, assessing their quality remains challenging. Recently, approaches introducing task-oriented evaluation metrics have been proposed, leveraging neural network training. However, quality

Stefanos Lappas, Tuomas Oikari

Given two variable exponent Muckenhoupt weights $w\in \mathcal{A}_{p(\cdot)}$ and $w_1\in \mathcal{A}_{p_1(\cdot)}$, we prove that for all small enough $θ>0,$ there holds that $w_0\in \mathcal{A}_{p_0(\cdot)},$ where the weight is determined by $w = w_0^{1-θ}w_1^θ$ and exponent of the weight class by $1/p(\cdot) = (1-θ)/p_0(\cdot) + θ/p_1(\cdot).$ The proof

Julia Belikova, Konstantin Polev, Rauf Parchiev, Dmitry Simakov

Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) systems are increasingly deployed in industry applications, yet their reliability remains hampered by challenges in detecting hallucinations. While supervised state-of-the-art (SOTA) methods that leverage LLM hidden states -- such as activation tracing and representation analysis -- show p

Xiaofeng Pan, Jing Chen, Haitong Zhang, Menglin Xing

Recent works of music representation learning mainly focus on learning acoustic music representations with unlabeled audios or further attempt to acquire multi-modal music representations with scarce annotated audio-text pairs. They either ignore the language semantics or rely on labeled audio datasets that are difficult and expensive to create. Moreover, me

Daryna Dementieva, Nikolay Babakov, Alexander Fraser

While Ukrainian NLP has seen progress in many texts processing tasks, emotion classification remains an underexplored area with no publicly available benchmark to date. In this work, we introduce EmoBench-UA, the first annotated dataset for emotion detection in Ukrainian texts. Our annotation schema is adapted from the previous English-centric works on emoti

Philip Heltweg, Dirk Riehle, Georg-Daniel Schwarz

Selecting a subset of cells is a common task in data engineering, for example, to remove errors or select only specific parts of a table. Multiple approaches to express this selection exist. One option is numeric indexing, commonly found in general programming languages, where a tuple of numbers identifies the cell. Alternatively, the separate dimensions can

James Xu Zhao, Jimmy Z. J. Liu, Bryan Hooi, See-Kiong Ng

Large language models (LLMs) are widely used for long-form text generation. However, factual errors in the responses would undermine their reliability. Despite growing attention to LLM factuality, the effect of response length on factuality remains underexplored. In this work, we systematically investigate this relationship by first introducing an automatic

Ling Liang, Shujun Bi

This study focuses on solving group zero-norm regularized robust loss minimization problems. We propose a proximal Majorization-Minimization (PMM) algorithm to address a class of equivalent Difference-of-Convex (DC) surrogate optimization problems. First, we present the core principles and iterative framework of the PMM method. Under the Kurdyka-{\L}ojasiewi

Evangelos Charalampakis, Vasileios Mygdalis, Ioannis Pitas

This work explores the application of Federated Learning (FL) to Unsupervised Semantic image Segmentation (USS). Recent USS methods extract pixel-level features using frozen visual foundation models and refine them through self-supervised objectives that encourage semantic grouping. These features are then grouped to semantic clusters to produce segmentation

Xinye Li, Zunwen Zheng, Qian Zhang, Dekai Zhuang

Knowledge Editing (KE) has gained increasing attention, yet current KE tasks remain relatively simple. Under current evaluation frameworks, many editing methods achieve exceptionally high scores, sometimes nearing perfection. However, few studies integrate KE into real-world application scenarios (e.g., recent interest in LLM-as-agent). To support our analys

Hao Li, Ju Dai, Xin Zhao, Feng Zhou

In 3D speech-driven facial animation generation, existing methods commonly employ pre-trained self-supervised audio models as encoders. However, due to the prevalence of phonetically similar syllables with distinct lip shapes in language, these near-homophone syllables tend to exhibit significant coupling in self-supervised audio feature spaces, leading to t

Dingding Li, Chao Zhang

We investigate the interior Sobolev regularity of weak solutions to the nonlocal $(1, p)$-Laplace equations in the superquadratic case $p\ge 2$. As a product, the explicit H\"{o}lder continuity estimates of weak solutions are derived. The proof relies on a detailed analysis of the structural characteristics of $(1, p)$-growth in the nonlocal setting, combine

Emanuele Masiero, Vito Trianni, Giuseppe Vizzari, Dimitri Ognibene

Precision agriculture requires efficient autonomous systems for crop monitoring, where agents must explore large-scale environments while minimizing resource consumption. This work addresses the problem as an active exploration task in a grid environment representing an agricultural field. Each cell may contain targets (e.g., damaged crops) observable from n

Chikaha Tsuji, Enrique Flores Medina, Harshit Gupta, Md Ferdous Alam

With the advancement of generative AI, research on its application to 3D model generation has gained traction, particularly in automating the creation of Computer-Aided Design (CAD) files from images. GenCAD is a notable model in this domain, leveraging an autoregressive transformer-based architecture with a contrastive learning framework to generate CAD pro

Alexander Aycock

A Thesis about Euler discussing the possibilities and limits of his method of work in Mathematics.

Muhammad Shafi, Syed Mohsin Bokhari

Land cover and land use (LULC) changes are key applications of satellite imagery, and they have critical roles in resource management, urbanization, protection of soils and the environment, and enhancing sustainable development. The literature has heavily utilized multispectral spatiotemporal satellite data alongside advanced machine learning algorithms to m

Valeria Banica, Renato Lucà, Nikolay Tzvetkov, Luis Vega

We focus on a class of solutions of the binormal flow, model of the evolution of vortex filaments, that generate several corner singularities in finite time. This phenomenon has been studied earlier in the regular case, which in this context is in terms of the summability of the angles of the corners generated. Our goal here is to investigate the lower regul

Zhihong Tan, Jiayi Wang, Huiying Shi, Binyuan Huang

Detecting forged remote sensing images is becoming increasingly critical, as such imagery plays a vital role in environmental monitoring, urban planning, and national security. While diffusion models have emerged as the dominant paradigm for image generation, their impact on remote sensing forgery detection remains underexplored. Existing benchmarks primaril

Seong Jun Kwon, Kyusung Hwang, Suk Bum Chung

It is now well understood that non-Kitaev spin interactions can be added to the Kitaev quantum spin liquid by applying external fields. Recent years have seen intensive discussion on the possible phase transitions that these spin interactions induce. In this paper, we will show through the perturbation theory the possibility of accessing a gapped spin liquid

Mislav Balunović, Jasper Dekoninck, Ivo Petrov, Nikola Jovanović

The rapid advancement of reasoning capabilities in large language models (LLMs) has led to notable improvements on mathematical benchmarks. However, many of the most commonly used evaluation datasets (e.g., AIME 2024) are widely available online, making it difficult to disentangle genuine reasoning from potential memorization. Furthermore, these benchmarks d

Dániel Keliger, Illés Horváth

We study the deterministic Susceptible-Infected-Susceptible (SIS) epidemic model on weighted graphs. In their numerical study [10] van Mieghem et al. have shown that it is possible to learn an estimated network from a finite time sample of the trajectories of the dynamics that in turn can give an accurate prediction beyond the sample time range, even though

Chuandong Liu, Huijiao Wang, Lei Yu, Gui-Song Xia

Recent advances in 3D Gaussian Splatting have shown remarkable potential for novel view synthesis. However, most existing large-scale scene reconstruction methods rely on the divide-and-conquer paradigm, which often leads to the loss of global scene information and requires complex parameter tuning due to scene partitioning and local optimization. To address

Thea Kozakis, João M. Mendonça, Lars A. Buchhave, Luisa M. Lara

Molecular oxygen (O2) will be an important molecule in the search for biosignatures in terrestrial planetary atmospheres in the coming decades. In particular, O2 combined with a reducing gas is thought to be strong evidence for disequilibrium caused by surface life. However, there are circumstances where it would be very difficult or impossible to detect O2,