May 2025 arXiv papers — page 111

Cailan Li

We investigate the structure of reduced triply graded link homology $\overline{\mathrm{HHH}}$ in the top/bottom three $T-$degrees for links arising as closures of positive/negative braids. Using a diagrammatic approach to the Hochschild cohomology of Soergel bimodules, we provide explicit computations of $\overline{\mathrm{HHH}}$ as $R-$modules in these degr

Yunlong Liang, Fandong Meng, Jiaan Wang, Jie Zhou

The challenge of slang translation lies in capturing context-dependent semantic extensions, as slang terms often convey meanings beyond their literal interpretation. While slang detection, explanation, and translation have been studied as isolated tasks in the era of large language models (LLMs), their intrinsic interdependence remains underexplored. The mai

Songhao Wu, Quan Tu, Mingjie Zhong, Hong Liu

In the realm of information retrieval, users often engage in multi-turn interactions with search engines to acquire information, leading to the formation of sequences of user feedback behaviors. Leveraging the session context has proven to be beneficial for inferring user search intent and document ranking. A multitude of approaches have been proposed to exp

Tong Bao, Heng Zhang, Chengzhi Zhang

Abstractive summarization of scientific papers has always been a research focus, yet existing methods face two main challenges. First, most summarization models rely on Encoder-Decoder architectures that treat papers as sequences of words, thus fail to fully capture the structured information inherent in scientific papers. Second, existing research often use

Xiang Zhang, Juntai Cao, Jiaqi Wei, Yiwei Xu

Tokenization is the first - and often underappreciated - layer of computation in language models. While Chain-of-Thought (CoT) prompting enables transformer models to approximate recurrent computation by externalizing intermediate steps, we show that the success of such reasoning is fundamentally bounded by the structure of tokenized inputs. This work presen

Marien Renaud, Valentin De Bortoli, Arthur Leclaire, Nicolas Papadakis

We consider the problem of sampling distributions stemming from non-convex potentials with Unadjusted Langevin Algorithm (ULA). We prove the stability of the discrete-time ULA to drift approximations under the assumption that the potential is strongly convex at infinity. In many context, e.g. imaging inverse problems, potentials are non-convex and non-smooth

Tyrone Fernando, Mohamed Darouach

This paper introduces the new concepts of Functional Controllability and Functional Stabilizability, and establishes their duality with Functional Observability and Functional Detectability, respectively. A Generalized Separation Principle is presented, under which the classical Separation Principle emerges as a special case. Conditions for the existence of

Xiangyu Hui, Samuel Karumba, Sid Chi-Kin Chau, Mohiuddin Ahmed

Cyberattacks on smart inverters and distributed PV are becoming an imminent threat, because of the recent well-documented vulnerabilities and attack incidents. Particularly, the long lifespan of inverter devices, users' oblivion of cybersecurity compliance, and the lack of cyber regulatory frameworks exacerbate the prospect of cyberattacks on smart inverters

Yusuf Denizay Dönder, Derek Hommel, Andrea W Wen-Yi, David Mimno

LLMs are effective at code generation tasks like text-to-SQL, but is it worth the cost? Many state-of-the-art approaches use non-task-specific LLM techniques including Chain-of-Thought (CoT), self-consistency, and fine-tuning. These methods can be costly at inference time, sometimes requiring over a hundred LLM calls with reasoning, incurring average costs o

Yunlong Liang, Fandong Meng, Jie Zhou

The sparse Mixture-of-Experts (MoE) has achieved significant progress for neural machine translation (NMT). However, there exist two limitations in current MoE solutions which may lead to sub-optimal performance: 1) they directly use the task knowledge of NMT into MoE (\emph{e.g.}, domain/linguistics-specific knowledge), which are generally unavailable at pr

Adrian Cosma, Stefan Ruseti, Emilian Radoi, Mihai Dascalu

Despite their remarkable progress across diverse domains, Large Language Models (LLMs) consistently fail at simple character-level tasks, such as counting letters in words, due to a fundamental limitation: tokenization. In this work, we frame this limitation as a problem of low mutual information and analyze it in terms of concept emergence. Using a suite of

Samuel W. Coles, Amir Hajibabaei, Venkat Kapil, Xavier R. Advincula

Superionic ice, where water molecules dissociate into a lattice of oxygen ions and a rapidly diffusing 'gas' of protons, represents an exotic state of matter with broad implications for planetary interiors and energy applications. Recently, a nanoconfined superionic state of water has been predicted which, in sharp contrast to bulk ice, is comprised of intac

Chen Zhang, Weixin Bu, Zeyi Ren, Zhengwu Liu

Inferring properties of graph-structured data, e.g., the solubility of molecules, essentially involves learning the implicit mapping from graphs to their properties. This learning process is often costly for graph property learners like Graph Convolutional Networks (GCNs). To address this, we propose a paradigm called Graph Neural Teaching (GraNT) that reint

Miriam Doh, Aditya Gulati, Matei Mancas, Nuria Oliver

This paper examines how synthetically generated faces and machine learning-based gender classification algorithms are affected by algorithmic lookism, the preferential treatment based on appearance. In experiments with 13,200 synthetically generated faces, we find that: (1) text-to-image (T2I) systems tend to associate facial attractiveness to unrelated posi

Rodrigo A. González, Maarten van der Hulst, Koen Classens, Tom Oomen

Many applications in mechanical, acoustic, and electronic engineering require estimating complex dynamical models, often represented as additive multi-input multi-output (MIMO) transfer functions with structural constraints. This paper introduces a two-stage procedure for estimating structured additive MIMO models, where structural constraints are enforced t

Zongyan Lv, Xiaoyu Zeng, Huan-Song Zhou

In this paper, we consider the following Br\'{e}zis-Nirenberg problem with prescribed $ L^2$-norm (mass) constraint: \begin{equation*} \begin{cases} -\Delta u=|u|^{2^*-2} u +\lambda_\rho u\quad \text { in } \Omega, u>0, \quad u \in H_0^1(\Omega), \quad \int_{\Omega} u^2dx=\rho, \end{cases} \end{equation*} where $N \geqslant 6$, $2^*=2 N /(N-2)$ is the critic

Changgu Chen, Xiaoyan Yang, Junwei Shu, Changbo Wang

In recent years, large-scale pre-trained diffusion transformer models have made significant progress in video generation. While current DiT models can produce high-definition, high-frame-rate, and highly diverse videos, there is a lack of fine-grained control over the video content. Controlling the motion of subjects in videos using only prompts is challengi

Anna Ottavia Schulte, Samar Alqatari, Saverio Rossi, Francesco Zamponi

Protein fitness landscapes frequently exhibit epistasis, where the effect of a mutation depends on the genetic context in which it occurs, i.e., the rest of the protein sequence. Epistasis increases landscape complexity, often resulting in multiple fitness peaks. In its simplest form, known as global epistasis, fitness is modeled as a non-linear function of

Zhenkai Qin, Jiajing He, Qiao Fang

Fine-grained sentiment analysis (FGSA) aims to identify sentiment polarity toward specific aspects within a text, enabling more precise opinion mining in domains such as product reviews and social media. However, traditional FGSA approaches often require task-specific architectures and extensive annotated data, limiting their generalization and scalability.

Marcel Arpogaus, Thomas Kneib, Thomas Nagler, David Rügamer

Density regression models allow a comprehensive understanding of data by modeling the complete conditional probability distribution. While flexible estimation approaches such as normalizing flows (NF) work particularly well in multiple dimensions, interpreting the input-output relationship of such models is often difficult, due to the black-box character of

He Wang, Alexander Hanbo Li, Yiqun Hu, Sheng Zhang

Large language model (LLM) agents have shown promising performance in generating code for solving complex data science problems. Recent studies primarily focus on enhancing in-context learning through improved search, sampling, and planning techniques, while overlooking the importance of the order in which problems are tackled during inference. In this work,

Jaime S. Buruaga, Augustine Bugler, Juan P. Brito, Vicente Martin

Advancements in quantum computing pose a significant threat to most of the cryptography currently deployed. Fortunately, cryptographic building blocks to mitigate the threat are already available; mostly based on post-quantum and quantum cryptography, but also on symmetric cryptography techniques. Notably, quantum-safe building blocks must be deployed as soo

Ting Wei, Biao Mei, Junliang Lyu, Renquan Zhang

Personalized Bayesian federated learning (PBFL) handles non-i.i.d. client data and quantifies uncertainty by combining personalization with Bayesian inference. However, existing PBFL methods face two limitations: restrictive parametric assumptions in client posterior inference and naive parameter averaging for server aggregation. To overcome these issues, we

Zahraa Al Sahili, Ioannis Patras, Matthew Purver

Multilingual vision-language models (VLMs) promise universal image-text retrieval, yet their social biases remain underexplored. We perform the first systematic audit of four public multilingual CLIP variants: M-CLIP, NLLB-CLIP, CAPIVARA-CLIP, and the debiased SigLIP-2, covering ten languages that differ in resource availability and morphological gender mark

Junjie Li, Jiawei Wang, Miyu Li, Yu Liu

Depth estimation plays a great potential role in obstacle avoidance and navigation for further Mars exploration missions. Compared to traditional stereo matching, learning-based stereo depth estimation provides a data-driven approach to infer dense and precise depth maps from stereo image pairs. However, these methods always suffer performance degradation in

Sanjay Govindan, Maurice Pagnucco, Yang Song

Large Language Models (LLMs) are trained on diverse and often conflicting knowledge spanning multiple domains and time periods. Some of this knowledge is only valid within specific temporal contexts, such as answering the question, "Who is the President of the United States in 2022?" Ensuring LLMs generate time appropriate responses is crucial for maintainin

Pittawat Taveekitworachai, Potsawee Manakul, Sarana Nutanong, Kunat Pipatanakul

This paper investigates prior prompt engineering (pPE) in the context of reinforcement fine-tuning (RFT), where language models (LMs) are incentivized to exhibit behaviors that maximize performance through reward signals. While existing RFT research has primarily focused on algorithms, reward shaping, and data curation, the design of the prior prompt--the in

Songhao Wu, Quan Tu, Hong Liu, Jia Xu

Session search involves a series of interactive queries and actions to fulfill user's complex information need. Current strategies typically prioritize sequential modeling for deep semantic understanding, overlooking the graph structure in interactions. While some approaches focus on capturing structural information, they use a generalized representation for

Kai Frye-Arndt, Matthew Glaysher, Marius Glaeser, Matthias Koch

Ultracold atomic gases with uniform density can be created by flat-bottom optical traps. These gases provide an ideal platform to study many-body physics in a system that allows for simple connections with theoretical models and emulation of numerous effects from a wide range of fields of physics. In Earth-bound laboratories the trap sizes, number of species

I. Yu. Chestnov, A. Kudlis, A. V. Nalitov, I. A. Shelykh

We theoretically investigate the interplay between Zeeman splitting and TE-TM-induced spin-flip tunneling in coupled exciton-polariton condensates systems and its impact on the spin-Meissner effect. We demonstrate that although a single condensate exhibits the effect of full paramagnetic screening via spin-anisotropic interactions, the inter-site spin-flip t

Thrassos K. Oikonomou, George K. Karagiannidis

This paper introduces Elliptic Curve Modulation (ECM), a novel modulation scheme that can be leveraged to effectively shuffle transmitted data while maintaining symbol error probability (SEP) performance equivalent to unencrypted systems. By utilizing the well-distributed elliptic curve points over the field of large primes, ECM enhances symbol obfuscation,

L. L. Salcedo

Using an infinitesimal approach, this work addresses the renormalization problem to deal with the ultraviolet divergences arising in quantum field theory. Under the assumption that the action has already been renormalized to yield an ultraviolet-finite effective action that satisfies a certain set of renormalization conditions, we analyze how the action must

Jiaming Li, Sheng Wang, Xin Wang, Yitao Zhu

Given the audio-visual clip of the speaker, facial reaction generation aims to predict the listener's facial reactions. The challenge lies in capturing the relevance between video and audio while balancing appropriateness, realism, and diversity. While prior works have mostly focused on uni-modal inputs or simplified reaction mappings, recent approaches such

Lucía Rossi

Consider $\alpha \in \Q(i)$ satisfying $|\alpha| >1$. Let $\D = \{0,1,\ldots,|a_0|-1\}$, where $a_0$ is the independent coefficient of the minimal primitive polynomial of $\alpha$. We introduce a way of expanding complex numbers in base $\alpha$ with digits in $\D$ that we call $\alpha$-expansions, which generalize rational base number systems introduced by

Chengzhi Zhang, Xinyi Yan, Lei Zhao, Yingyi Zhang

The exponential increase in academic papers has significantly increased the time required for researchers to access relevant literature. Keyphrase Extraction (KPE) offers a solution to this situation by enabling researchers to efficiently retrieve relevant literature. The current study on KPE from academic articles aims to improve the performance of extracti

Fan Liu, Zherui Yang, Cancheng Liu, Tianrui Song

Mathematical modeling is a cornerstone of scientific discovery and engineering practice, enabling the translation of real-world problems into formal systems across domains such as physics, biology, and economics. Unlike mathematical reasoning, which assumes a predefined formulation, modeling requires open-ended problem analysis, abstraction, and principled f

Xiong Jun Wu, Zhenduo Zhang, ZuJie Wen, Zhiqiang Zhang

Training large reasoning models (LRMs) with reinforcement learning in STEM domains is hindered by the scarcity of high-quality, diverse, and verifiable problem sets. Existing synthesis methods, such as Chain-of-Thought prompting, often generate oversimplified or uncheckable data, limiting model advancement on complex tasks. To address these challenges, we in

Pengcheng Jiang, Xueqiang Xu, Jiacheng Lin, Jinfeng Xiao

Retrieval-augmented generation (RAG) systems empower large language models (LLMs) to access external knowledge during inference. Recent advances have enabled LLMs to act as search agents via reinforcement learning (RL), improving information acquisition through multi-turn interactions with retrieval engines. However, existing approaches either optimize retri

Thomas Franosch, Cristiano De Michele, Rolf Schilling

For a fluid of convex hard particles, characterized by a length scale $\sigma_\text{min}$ and an anisotropy parameter $\epsilon$, we develop a formalism allowing one to relate thermodynamic quantities to the body's shape. In a first step its thermodynamics is reduced to that of spherical particles. The latter have a hard core of diameter $\sigma_\text{min }$

Xizhe Xie, Wengu Chen, Zheng Ma, Han Wang

The Gray Radiative Transfer Equations (GRTEs) are high-dimensional, multiscale problems that pose significant computational challenges for traditional numerical methods. Current deep learning approaches, including Physics-Informed Neural Networks (PINNs) and Asymptotically Preserving Neural Networks (APNNs), are largely restricted to low-dimensional or linea

Shuo Zhang, Jinsong Zhang, Zhejun Zhang, Lei Li

Multi-task learning (MTL) enables the efficient transfer of extra knowledge acquired from other tasks. The high correlation between multimodal sentiment analysis (MSA) and multimodal emotion recognition (MER) supports their joint training. However, existing methods primarily employ hard parameter sharing, ignoring parameter conflicts caused by complex task c

Gijs Wijngaard, Elia Formisano, Michele Esposito, Michel Dumontier

Audio-language models have shown promising results in various sound understanding tasks, yet they remain limited in their ability to reason over the fine-grained semantics of sound. In this paper, we present AudSemThinker, a model whose reasoning is structured around a framework of auditory semantics inspired by human cognition. To support this, we introduce

Fanglin Mo, Junzhe Chen, Haoxuan Zhu, Xuming Hu

Mobile GUI agents execute user commands by directly interacting with the graphical user interface (GUI) of mobile devices, demonstrating significant potential to enhance user convenience. However, these agents face considerable challenges in task planning, as they must continuously analyze the GUI and generate operation instructions step by step. This proces

Qianyue Hao, Sibo Li, Jian Yuan, Yong Li

Despite rapid advancements in large language models (LLMs), the token-level autoregressive nature constrains their complex reasoning capabilities. To enhance LLM reasoning, inference-time techniques, including Chain/Tree/Graph-of-Thought(s), successfully improve the performance, as they are fairly cost-effective by guiding reasoning through sophisticated log

Marvin Alles, Nutan Chen, Patrick van der Smagt, Botond Cseke

The use of guidance to steer sampling toward desired outcomes has been widely explored within diffusion models, especially in applications such as image and trajectory generation. However, incorporating guidance during training remains relatively underexplored. In this work, we introduce energy-guided flow matching, a novel approach that enhances the trainin

Dong Huang, Pengkun Yang

Correlation analysis is a fundamental step in uncovering meaningful insights from complex datasets. In this paper, we study the problem of detecting correlations between two random graphs following the Gaussian Wigner model with unlabeled vertices. Specifically, the task is formulated as a hypothesis testing problem: under the null hypothesis, the two graphs

Vojtěch Kůr, Vít Musil, Vojtěch Řehák

Adversarial Patrolling games form a subclass of Security games where a Defender moves between locations, guarding vulnerable targets. The main algorithmic problem is constructing a strategy for the Defender that minimizes the worst damage an Attacker can cause. We focus on the class of finite-memory (also known as regular) Defender's strategies that experime

Ryo Bertolissi, Jonas Hübotter, Ido Hakimi, Andreas Krause

Mixture of expert (MoE) models are a promising approach to increasing model capacity without increasing inference cost, and are core components of many state-of-the-art language models. However, current MoE models typically use only few experts due to prohibitive training and inference cost. We propose Test-Time Model Merging (TTMM) which scales the MoE para

Ruihuang Li, Caijin Zhou, Shoujian Zheng, Jianxiang Lu

Intelligent game creation represents a transformative advancement in game development, utilizing generative artificial intelligence to dynamically generate and enhance game content. Despite notable progress in generative models, the comprehensive synthesis of high-quality game assets, including both images and videos, remains a challenging frontier. To creat

Andrea Mammola, Quentin Schaeverbeke, Giuseppe Di Molfetta

In this work we map NISQ-friendly implementations of the non-interacting QCA to a circuit Quantum Electrodynamics (cQED) hardware. We perform both noiseless and noisy simulations of the QCA one particle sector, namely the Quantum Walk, on $N$-cycles and $N \times N$ torus graphs. Moreover, within this framework, we also investigate the search problem and pre

Seyed Soroush Karimi Madahi, Kenneth Bruninx, Bert Claessens, Chris Develder

Transmission System Operators (TSOs) rely on balancing energy provided by Balancing Service Providers (BSPs) to maintain the supply-demand balance in real time. Balance Responsible Parties (BRPs) can simultaneously deviate from their day-ahead schedules in response to imbalance prices, e.g., by controlling flexible assets such as batteries. According to the

Markus Haase, Henrik Kreidler

This paper is a continuation of our work on the functional-analytic core of the classical Furstenberg-Zimmer theory. We introduce and study (in the framework of lattice-ordered spaces) the notions of total order-boundedness and uniform total order-boundedness. Either one generalizes the concept of ordinary precompactness known from metric space theory. These

Wei Zhou, Mohsen Mesgar, Heike Adel, Annemarie Friedrich

In table question answering (TQA), tables are encoded as either texts or images. Prior work suggests that passing images of tables to multi-modal large language models (MLLMs) performs comparably to or even better than using textual input with large language models (LLMs). However, the lack of controlled setups limits fine-grained distinctions between these

Filip Miletić, Aaron Schmid, Sabine Schulte im Walde

This paper investigates the extent to which pretrained German BERT encodes knowledge of noun compound semantics. We comprehensively vary combinations of target tokens, layers, and cased vs. uncased models, and evaluate them by predicting the compositionality of 868 gold standard compounds. Looking at representational patterns within the transformer architect

Manuel Calixto, Alberto Mayorgas, Julio Guerrero

Lieb-Mattis theorem orders the lowest-energy states of total spin $s$ of a system of $P$ interacting fermions. We generalize these predictions to fermionic mixtures of $P$ particles with more than $N=2$ spinor components/species in the thermodynamic limit $P\to\infty$. The lowest-energy state inside each permutation symmetry sector $h$, arising in the $P$-fo

Jed Muff, Keiichi Ito, Elijah H. W. Ang, Karine Miras

Evolution and learning have historically been interrelated topics, and their interplay is attracting increased interest lately. The emerging new factor in this trend is morphological evolution, the evolution of physical forms within embodied AI systems such as robots. In this study, we investigate a system of hexacopter-type drones with evolvable morphologie

Yihang Du, Jiaying Hu, Suyang Hou, Yueyang Ding

Spatial labeling assigns labels to specific spatial locations to characterize their spatial properties and relationships, with broad applications in scientific research and practice. Measuring the similarity between two spatial labelings is essential for understanding their differences and the contributing factors, such as changes in location properties or l

Roberto Passante, Lucia Rizzuto, Peter Schall, Emanuele Marino

Fluctuation-induced forces, primarily represented by quantum and critical Casimir effects, play a pivotal role at the nanoscale. This review explores the theoretical and experimental landscapes of these forces, offering a comprehensive analysis of their similarities and distinctions. We emphasize the effects of material properties, geometry, and temperature

Yuan-Hao Jiang, Kezong Tang, Zi-Wei Chen, Yuang Wei

Knowledge components (KCs) are the fundamental units of knowledge in the field of education. A KC graph illustrates the relationships and dependencies between KCs. An accurate KC graph can assist educators in identifying the root causes of learners' poor performance on specific KCs, thereby enabling targeted instructional interventions. To achieve this, we h

Viet Anh Khoa Tran, Emre Neftci, Willem A. M. Wybo

Biological brains learn continually from a stream of unlabeled data, while integrating specialized information from sparsely labeled examples without compromising their ability to generalize. Meanwhile, machine learning methods are susceptible to catastrophic forgetting in this natural learning setting, as supervised specialist fine-tuning degrades performan

Hongjun Choi, Eun Som Jeon, Ankita Shukla, Pavan Turaga

Knowledge distillation (KD) is a valuable technique for compressing large deep learning models into smaller, edge-suitable networks. However, conventional KD frameworks rely on pre-trained high-capacity teacher networks, which introduce significant challenges such as increased memory/storage requirements, additional training costs, and ambiguity in selecting

Kensuke Arakawa, Bastiaan Cnossen

We give a concise, conceptual proof of the universality of the relative Rezk nerve, due to Mazel-Gee.

Ehsan Masoudian, Ali Mirzaei, Hossein Bagheri

This study investigates the multifaceted factors influencing wildfire risk in Iran, focusing on the interplay between climatic conditions and human activities. Utilizing advanced remote sensing, geospatial information system (GIS) processing techniques such as cloud computing, and machine learning algorithms, this research analyzed the impact of climatic par

Jason D. Lotay, Jakob Stein

Nearly $G_2$-structures define positive Einstein metrics in $7$ dimensions and are critical points, up to scale, for a geometric flow of co-closed $G_2$-structures with good analytic properties called the modified $G_2$-Laplacian co-flow. We introduce a suitable normalization of this flow so that nearly $G_2$-structures are stable under rescaling. However, w

Kathleen Barsse, Paolo Perinotti, Alessandro Tosini, Leonardo Vaglini

The causal effects activated by a quantum interaction are studied, modelling the last one as a bipartite unitary channel. The two parties, say Alice and Bob, can use the channel to exchange messages -- i.e. to signal. On the other hand, the most general form of causal influence includes also the possibility for Alice, via a local operation on her system, to

Yuki Sagawa, Jonte R. Hance, Holger F. Hofmann, Takafumi Ono

Quantum contextuality, where measurement outcomes depend on the measurement context, implies a failure of classical realism in quantum systems. As recently shown, the transition between measurement contexts can be mapped onto the path that a quantum particle takes through an interferometer. Here, we investigate the relation between contextuality and the coun

Abdollah Masoud Darya, Saeed Abdallah

Massive multiple-input multiple-output low-Earth-orbit communication channels are highly time-varying due to severe Doppler shifts and propagation delays. While satellite-mobility-induced Doppler shifts can be compensated using known ephemeris data, those caused by user mobility require accurate user positioning information; the absence of such information c

Xinyi Shang, Peng Sun, Fengyuan Liu, Tao Lin

This paper pioneers a novel data-centric paradigm to maximize the utility of unlabeled data, tackling a critical question: How can we enhance the efficiency and sustainability of deep learning training by optimizing the data itself? We begin by identifying three key limitations in existing model-centric approaches, all rooted in a shared bottleneck: knowledg

Hongru Wang, Deng Cai, Wanjun Zhong, Shijue Huang

Inference-time scaling has attracted much attention which significantly enhance the performance of Large Language Models (LLMs) in complex reasoning tasks by increasing the length of Chain-of-Thought. These longer intermediate reasoning rationales embody various meta-reasoning skills in human cognition, such as reflection and decomposition, being difficult t

Jonas Neumeyer, Michael Wolfgang Kaiser, Thomas-Peter Fries

A novel mixed-hybrid method for Kirchhoff-Love shells is proposed that enables the use of classical, possibly higher-order Lagrange elements in numerical analyses. In contrast to purely displacement-based formulations that require higher continuity of shape functions as in IGA, the mixed formulation features displacements and moments as primary unknowns. The

Francesco Sala

A. Vistoli proved a decomposition theorem for the rational equivariant algebraic K-theory of a variety under the action of a finite group $G$. We generalize his result to more general algebraic (co)homology theories having the Mackey property and admitting localization long exact sequences. In general, the pieces are indexed by conjugacy classes of subgroups

Bruno Viti, Elias Karabelas, Martin Holler

Most machine learning-based image segmentation models produce pixel-wise confidence scores that represent the model's predicted probability for each class label at every pixel. While this information can be particularly valuable in high-stakes domains such as medical imaging, these scores are heuristic in nature and do not constitute rigorous quantitative un

Tianle Gu, Zongqi Wang, Kexin Huang, Yuanqi Yao

Logit-based LLM watermarking traces and verifies AI-generated content by maintaining green and red token lists and increasing the likelihood of green tokens during generation. However, it fails in low-entropy scenarios, where predictable outputs make green token selection difficult without disrupting natural text flow. Existing approaches address this by ass

Massimo Ferrario, Mauro Migliorati, Luigi Palumbo

As a charged particle beam moves through perfectly conducting structures with varying cross-sectional boundaries - such as RF cavities, tapers, bellows, kickers, ... - it induces both longitudinal and transverse electromagnetic fields called wakefields. In this lecture, we explore the fundamental characteristics of wakefields and illustrate key concepts usin

Maarten Derickx, Michael Stoll

We study the asymptotics of the set $S(d)$ of possible prime orders of $K$-rational points on elliptic curves over number fields $K$ of degree $d$ as $d$ tends to infinity. Assuming some conjectures on the sparsity of newforms of weight $2$ and prime level with unexpectedly high analytic rank, we show that $\max S(d) \le 3d + 1$ for sufficiently large even $

Tobias Chemnitz, Christian Reiter, Florian Kraus, David Novog

This paper provides a unique and to the best of our knowledge first-of-a-kind attempt to develop chemical processes that may contribute to the volume reduction of SMR TRISO-based fuels and aims at the eventual ability to reprocess the spent fuel. To this end, the etching behavior of two materials, silicon carbide, SiC and pyrolytic carbon, PyC, that are gene

Minhyuk Seo, Taeheon Kim, Hankook Lee, Jonghyun Choi

As AI becomes more personal, e.g., Agentic AI, there is an increasing need for personalizing models for various use cases. Personalized federated learning (PFL) enables each client to collaboratively leverage other clients' knowledge for better adaptation to the task of interest, without privacy risks. Despite its potential, existing PFL methods remain confi

Yakun Zhu, Zhongzhen Huang, Linjie Mu, Yutong Huang

The emergence of groundbreaking large language models capable of performing complex reasoning tasks holds significant promise for addressing various scientific challenges, including those arising in complex clinical scenarios. To enable their safe and effective deployment in real-world healthcare settings, it is urgently necessary to benchmark the diagnostic

Li Li, Peilin Cai, Ryan A. Rossi, Franck Dernoncourt

We present PersonaConvBench, a large-scale benchmark for evaluating personalized reasoning and generation in multi-turn conversations with large language models (LLMs). Unlike existing work that focuses on either personalization or conversational structure in isolation, PersonaConvBench integrates both, offering three core tasks: sentence classification, imp

Zaifa Xue, Tao Zhang, Tuo Xu, Huaixin Liang

GraphSAGE is a widely used graph neural network. The introduction of causal inference has improved its robust performance and named as Causal GraphSAGE. However, Causal GraphSAGE focuses on measuring causal weighting among individual nodes, but neglecting the cooperative relationships among sampling nodes as a whole. To address this issue, this paper propose

Zsófia Molnár, Gergely Szabó, András Horváth

Vision models pretrained on large-scale RGB natural image datasets are widely reused for electron microscopy image segmentation. In electron microscopy, volumetric data are acquired as serial sections and processed as stacks of adjacent grayscale slices, where neighboring slices provide symmetric contextual information for identifying features on the central

Fatemeh Chajaei, Hossein Bagheri

Three-dimensional reconstruction of buildings, particularly at Level of Detail 1 (LOD1), plays a crucial role in various applications such as urban planning, urban environmental studies, and designing optimized transportation networks. This study focuses on assessing the potential of LiDAR data for accurate 3D building reconstruction at LOD1 and extracting m

Wei Fan, Tianshi Zheng, Yiran Hu, Zheye Deng

Legal rules encompass not only codified statutes but also implicit adjudicatory principles derived from precedents that contain discretionary norms, social morality, and policy. While computational legal research has advanced in applying established rules to cases, inducing legal rules from judicial decisions remains understudied across jurisdictions. The ad

Guangke Chen, Fu Song, Zhe Zhao, Xiaojun Jia

Jailbreak attacks to Large audio-language models (LALMs) are studied recently, but they exclusively focused on the attack scenario where the adversary can fully manipulate user prompts (named strong adversary) and limited in effectiveness, applicability, and practicability. In this work, we first conduct an extensive evaluation showing that advanced text jai

Shogo Iwazaki, Junpei Komiyama, Masaaki Imaizumi

We consider the kernelized contextual bandit problem with a large feature space. This problem involves $K$ arms, and the goal of the forecaster is to maximize the cumulative rewards through learning the relationship between the contexts and the rewards. It serves as a general framework for various decision-making scenarios, such as personalized online advert

Ambre Marie, Marine Garnier, Thomas Bertin, Laura Machart

Suicide remains a public health challenge, necessitating improved detection methods to facilitate timely intervention and treatment. This systematic review evaluates the role of Artificial Intelligence (AI) and Machine Learning (ML) in assessing suicide risk through acoustic analysis of speech. Following PRISMA guidelines, we analyzed 33 articles selected fr

Ernests Lavrinovics, Russa Biswas, Katja Hose, Johannes Bjerva

Large Language Models (LLMs) have inherent limitations of faithfulness and factuality, commonly referred to as hallucinations. Several benchmarks have been developed that provide a test bed for factuality evaluation within the context of English-centric datasets, while relying on supplementary informative context like web links or text passages but ignoring

Qianxiong Xu, Lanyun Zhu, Xuanyi Liu, Guosheng Lin

Few-Shot Segmentation (FSS) aims to learn class-agnostic segmentation on few classes to segment arbitrary classes, but at the risk of overfitting. To address this, some methods use the well-learned knowledge of foundation models (e.g., SAM) to simplify the learning process. Recently, SAM 2 has extended SAM by supporting video segmentation, whose class-agnost

Yihua Zhu, Qianying Liu, Akiko Aizawa, Hidetoshi Shimodaira

Knowledge Base Question Answering (KBQA) aims to answer natural language questions using structured knowledge from KBs. While LLM-only approaches offer generalization, they suffer from outdated knowledge, hallucinations, and lack of transparency. Chain-based KG-RAG methods address these issues by incorporating external KBs, but are limited to simple chain-st

Yan Wang, Feng Shu, Xianpeng Wang, Minghao Chen

In this paper, channel estimation (CE) for uplink hybrid-field communications involving multiple Internet of Things (IoT) devices assisted by an active intelligent reflecting surface (IRS) is investigated. Firstly, to reduce the complexity of near-field (NF) channel modeling and estimation between IoT devices and active IRS, a sub-blocking strategy for activ

Juan J. Omiste, Lars Bojer Madsen

We study the impact of two-color ($\omega$ and $2\omega$) co- and counter-rotating ultrashort attosecond laser pulses on non-isotropic atomic targets through the one- and two-photon interference pattern of the photoelectron spectrum. Specifically, we take the ground state of atomic carbon, i. e., $(1s^22s^22p^2,{}^3\text{P}^\text{e})$ as a prototype. We obse

Ladislav Straka, Marek Vronka, Jan Maňák, Petr Veřtát

The atomic topology and magnetic microstructure of individual, highly mobile Type I and Type II twin boundaries in 10M Ni-Mn-Ga martensite were investigated by transmission electron microscopy (TEM). The twin boundaries established in a bulk single crystal showed twinning stresses of ~1 MPa for Type I and ~0.1 MPa for Type II twin boundaries. TEM lamellae wi

Soheil Hataminia, Tania Khosravi

In this study, the impact of research and development (R&D) expenditures on the value added of the agricultural sector in Iran was investigated for the period 1971-2021. For data analysis, the researchers utilized the ARDL econometric model and EViews software. The results indicated that R&D expenditures, both in the short and long run, have a significant po

Sat Paul, Jitendra Kumar, S. K. Maurya

The main aim of this study is to examine the behaviour of physical parameters of an anisotropic compact star model demonstrating spherical symmetry in F(Q) modified gravity. To evaluate the behaviour and the stability of an anisotropic compact star model, we utilise the measured mass and radius of an anisotropic compact star model. This study obtained an ani

Sinorul Haque, Shweta R. Keshri, G. Ganesh, Kaustuv Chatterjee

Stabilizing gold nanoparticles with tunable surface composition via reactive metal support interactions under ambient conditions remains a significant challenge. We discovered that a reactive glass metal interaction (RGMI) under ambient conditions, driven by the intrinsic catalytic activity of gold nanoislands (GNIs) and the unique properties of sodium alumi

Takayuki Harada, Zuin Ping Lily Ang, Yuki Sakakibara, Takuro Nagai

As integrated circuits continue to scale down, the search for new metals is becoming increasingly important due to the rising resistivity of traditional copper-based interconnects. A layered oxide PdCoO$_2$ is one of the candidate materials for interconnects, having bulk ab-plane conductivity exceeding that of elemental Al. Despite its potential, wafer-scale

M. Bordag, I. G. Pirozhenko

We compute the vacuum energy of a scalar field rotating with angular velocity $\Omega$ on a disk of radius $R$ and with Dirichlet boundary conditions. The rotation is introduced by a metric obtained by a Galilean transformation from a rest frame. The constraint $\Omega R<c$ must be obeyed to maintain causality. To compute the vacuum energy, we use an imagina

Marco Faella, Gennaro Parlato

Programs that manipulate tree-shaped data structures often require complex, specialized proofs that are difficult to generalize and automate. This paper introduces a unified, foundational approach to verifying such programs. Central to our approach is the knitted-tree encoding, modeling each program execution as a tree structure capturing input, output, and

Varun Raaghav, Dimitrios Bikos, Antonio Rago, Francesca Toni

Composites are amongst the most important materials manufactured today, as evidenced by their use in countless applications. In order to establish the suitability of composites in specific applications, finite element (FE) modelling, a numerical method based on partial differential equations, is the industry standard for assessing their mechanical properties

Massimo Ferrario, Mauro Migliorati, Luigi Palumbo

Space charge forces, which arise directly from the beam's charge distribution and include the influence of image charges and currents induced by interactions with a perfectly conducting, smooth pipe, are very important in high-intensity, low-energy accelerators. These forces play a key role under various beam dynamics regimes, leading to effects such as ener