October 2025 arXiv papers — page 214

Corey O'Meara

This thesis focuses on the Lie-theoretic foundations of controlled open quantum systems. We describe Markovian open quantum system evolutions by Lie semigroups, whose corresponding infinitesimal generators lie in a special type of convex cone - a Lie wedge. The Lie wedge associated to a given control system therefore consists of all generators of the quantum

Sarel Duanis, Asnat Greenstein-Messica, Eliya Habba

Large language models (LLMs) can modify JSON documents through natural language commands, but current approaches regenerate entire structures for each edit, resulting in computational inefficiency. We present JSON Whisperer, a framework that enables LLMs to generate RFC 6902 diff patches-expressing only the necessary modifications-rather than complete docume

Maximilian R. P. von Liechtenstein

Curved Boolean Logic (CBL) generalizes propositional logic by allowing local truth assignments that do not extend to a single global valuation, analogous to curvature in geometry. We give equivalent sheaf and exclusivity-graph semantics and a context-aware proof calculus that is conservative in the flat limit. We formalize CBL-SAT and basic complexity (NP-co

P. J. Robert, S. Manzoor, M. Chiarotti, N. Poli

We present a comprehensive study of the electronic transitions in neutral cadmium (Cd I) with a focus on forbidden transitions, motivated by recent advances in laser technology and the growing relevance of cadmium in quantum gas research, precision metrology, and atom trapping. General analytic expressions are derived for transition matrix elements of all mu

KunHo Heo, GiHyun Kim, SuYeon Kim, MyeongAh Cho

3D Semantic Scene Graph Prediction aims to detect objects and their semantic relationships in 3D scenes, and has emerged as a crucial technology for robotics and AR/VR applications. While previous research has addressed dataset limitations and explored various approaches including Open-Vocabulary settings, they frequently fail to optimize the representationa

Evgeni Dimitrov, Zongrui Yang

In this note we provide a short proof of the distributional equality between last passage percolation with geometric weights along a general down-right path and Schur processes. We do this in both the full-space and half-space settings, and for general parameters. The main inputs for our arguments are generalizations of the Robinson-Schensted-Knuth correspon

Luo Cheng, Song Siyang, Yan Siyuan, Yu Zhen

The automatic generation of diverse and human-like facial reactions in dyadic dialogue remains a critical challenge for human-computer interaction systems. Existing methods fail to model the stochasticity and dynamics inherent in real human reactions. To address this, we propose ReactDiff, a novel temporal diffusion framework for generating diverse facial re

Aoyang Fang, Songhan Zhang, Yifan Yang, Haotong Wu

While cloud-native microservice architectures have revolutionized software development, their inherent operational complexity makes failure Root Cause Analysis (RCA) a critical yet challenging task. Numerous data-driven RCA models have been proposed to address this challenge. However, we find that the benchmarks used to evaluate these models are often too si

Zexin Wang, Changhua Pei, Yang Liu, Hengyue Jiang

Web service administrators must ensure the stability of multiple systems by promptly detecting anomalies in Key Performance Indicators (KPIs). Achieving the goal of "train once, infer across scenarios" remains a fundamental challenge for time series anomaly detection models. Beyond improving zero-shot generalization, such models must also flexibly handle seq

Saeed Fakhry, Maryam Shiravand, Antonino Del Popolo

The James Webb Space Telescope (JWST) has unveiled a population of unexpectedly massive and luminous galaxies at redshifts $z \gtrsim 7$, posing a significant challenge to the standard $\Lambda$CDM cosmological paradigm. In this work, we address the tension between early JWST observations of luminous high-redshift galaxies and predictions of the standard $\L

Kilian Rausch

In this paper, the generating functions of Garvans so-called $k$-ranks are used, to define a family of mock Eisenstein series. The $k$-rank moments are then expressed as partition traces of these functions. We explore the modular properties of this new family, give recursive formulas for them involving divisor like sums, and prove that their Fourier coeffici

Jaime Calderón-Figueroa, David Seery

In the "stochastic $\delta N$ formalism", the statistics of the inflationary density perturbation are obtained from the first passage distribution of a stochastic process. We develop a general framework in which to evaluate the rare tail of this distribution, based on an instanton approximation to a path integral representation for the transition probability

Foivos Paraperas Papantoniou, Stefanos Zafeiriou

Human-centric generative models designed for AI-driven storytelling must bring together two core capabilities: identity consistency and precise control over human performance. While recent diffusion-based approaches have made significant progress in maintaining facial identity, achieving fine-grained expression control without compromising identity remains c

Quang-Khai Bui-Tran, Minh-Toan Dinh, Thanh-Huy Nguyen, Ba-Thinh Lam

Accurate liver segmentation in multi-phase MRI is vital for liver fibrosis assessment, yet labeled data is often scarce and unevenly distributed across imaging modalities and vendor systems. We propose a label-efficient segmentation approach that promotes cross-modality generalization under real-world conditions, where GED4 hepatobiliary-phase annotations ar

Zhe Zhao, Swarnim Shashank, Debtroy Das, Cosimo Bambi

Gravitational wave astronomy has opened an unprecedented window onto tests of gravity and fundamental physics in the strong-field regime. In this study, we examine a series of well-motivated deviations from the classical Kerr solution of General Relativity and employ gravitational wave data to place constraints on possible deviations from the Kerr geometry.

Atsuhide Ishida

In the previous study (Ishida, 2025), the author proved the uniqueness of short-range potential functions using the Enss-Weder time-dependent method (Enss and Weder, 1995) for a two-body quantum system described by time-decaying harmonic oscillators. In this study, we extend the result of Ishida (2025) to the $N$-body case. We use the approaches developed in

Jesper Lykke Jacobsen, Rongvoram Nivesvivat, Sylvain Ribault, Paul Roux

In two-dimensional models of critical non-intersecting loops, there are $\ell$-leg fields that insert $\ell\in\mathbb{N}^*$ open loop segments, and diagonal fields that change the weights of closed loops. We conjecture an exact formula for 3-point functions of such fields on the sphere. In the cases of diagonal or spinless 2-leg fields, the conjecture agrees

T. P. C. Klaver, R. Gabbrielli, V. Tynianska, A. Iorio

We demonstrate the high stability of simulated graphene hyperbolic pseudospheres under large externally imposed deformations and high temperature annealing. Hyperbolic pseudospheres are produced in a two-step Molecular Dynamics simulation process. First, carbon atoms are forced down a thin three-dimensional volume of a chosen shape. During this extrusion pro

Laura Estivalez, Adriano Moura

We express the outer multiplicities in the tensor products of two fundamental simple modules for an affine Kac-Moody algebra of type $A$ in terms of counting certain sets of multipartitions by exploring the stabilizing limits of certain excellent filtrations. This extends for all ranks a previously obtained result by Jakeli\'c and the second author for rank

Alexander L. Mitchell, Joe Watson, Ingmar Posner

There are many challenges in bimanual assembly, including high-level sequencing, multi-robot coordination, and low-level, contact-rich operations such as component mating. Task and motion planning (TAMP) methods, while effective in this domain, may be prohibitively slow to converge when adapting to disturbances that require new task sequencing and optimisati

Yiding Wang, Zhepei Wei, Xinyu Zhu, Yu Meng

Enabling large language models (LLMs) to utilize search tools offers a promising path to overcoming fundamental limitations such as knowledge cutoffs and hallucinations. Recent work has explored reinforcement learning (RL) for training search-augmented agents that interleave reasoning and retrieval before answering. These approaches usually rely on outcome-b

Lucas Bandarkar, Chenyuan Yang, Mohsen Fayyaz, Junlin Hu

Mixture-of-Experts (MoE) architectures have become the key to scaling modern LLMs, yet little is understood about how their sparse routing dynamics respond to multilingual data. In this work, we analyze expert routing patterns using parallel multilingual datasets and present highly interpretable layer-wise phenomena. We find that MoE models route tokens in l

Na Liu, Jakub Wiktor Both, Geir Ersland, Jan Martin Nordbotten

Understanding processes in porous media is fundamental to a broad spectrum of environmental, energy, and geoscience applications. These processes include multiphase fluid transport, interfacial dynamics, reactive transformations, and interactions with solids or microbial components, all governed by wettability, capillarity, and reactive transport at fluid-fl

P. N. Vabishchevich

The inverse potential problem consists in determining the density of the volume potential from measurements outside the sources. Its ill-posedness is due both to the non-uniqueness of the solution and to the instability of the solution with respect to measurement errors. The inverse problem is solved under additional assumptions about the sources using regul

Lyes Saad Saoud, Irfan Hussain

Biomimetic intelligence and robotics are transforming field ecology by enabling lifelike robotic surrogates that interact naturally with animals under real world conditions. Studying avian behavior in the wild remains challenging due to the need for highly realistic morphology, durable outdoor operation, and intelligent perception that can adapt to uncontrol

Fumio Hiai

In this paper, for $\alpha\in(0,\infty)\setminus\{1\}$, $p>0$ and positive semidefinite matrices $A$ and $B$, we consider the quasi-extension $\mathcal{M}_{\alpha,p}(A,B):=\mathcal{M}_\alpha(A^p,B^p)^{1/p}$ of several $\alpha$-weighted geometric type matrix means $\mathcal{M}_\alpha(A,B)$ such as the $\alpha$-weighted geometric mean in Kubo--Ando's sense, th

João P. da Cruz

Shor's factoring algorithm guarantees a success probability of at least one half for any fixed modulus N = pq with distinct primes p and q. We show that this guarantee does not extend to the asymptotic regime. As N -> infinity, the multiplicative groups Omega_N = (Z/NZ)^x form a non-tight family of probability spaces, and the probability weight associated wi

Christian Berg, Ryszard Szwarc

We consider the Jacobi operator (T,D(T)) associated with an indeterminate Hamburger moment problem, and present countable subsets S of the domain D(T) such that span(S) is dense in \ell^2. As an example we have S={(p_n(u))+B(u)(p_n(0)):D(u)=0}, where (p_n) denotes the orthonormal polynomials of the moment problem and B,D are two of the Nevanlinna functions.

Chengwei Liu, Wenbo Guo, Yuxin Zhang, Limin Wang

In recent years, the landscape of software threats has become significantly more dynamic and distributed. Security vulnerabilities are no longer discovered and shared only through formal channels such as public vulnerability databases or vendor advisories. Increasingly, criti- cal threat information emerges informally through blogs, social media, developer f

Joann Ching, Gerhard Widmer

Music Emotion Recognition (MER) is a task deeply connected to human perception, relying heavily on subjective annotations collected from contributors. Prior studies tend to focus on specific musical styles rather than incorporating a diverse range of genres, such as rock and classical, within a single framework. In this paper, we address the task of recogniz

Tekin Dereli, Yorgo Senikoglu

Linearized Einstein-Weyl equations are solved precisely in the context of sandwich gravitational waves. The neutrino's energy-momentum depends on the geometry and composition of the gravitational pulse when it is scattered. Since the background remains unchanged at the test field level, the neutrino's energy density will exhibit fluctuations between positive

Chenxiang Zhang, Alexander Theus, Damien Teney, Antonio Orvieto

Model merging combines independent solutions with different capabilities into a single one while maintaining the same inference cost. Two popular approaches are linear interpolation, which simply averages multiple model weights, and task arithmetic, which combines task vectors obtained by the difference between finetuned and base models. While useful in prac

Mélodie Andrieu, Shalom Eliahou, Léo Vivion

The rational base number system, introduced by Akiyama, Frougny, and Sakarovitch in 2008, is a generalization of the classical integer base number system. Within this framework two interesting families of infinite words emerge, called minimal and maximal words. We conjecture that every minimal and maximal word is normal over an appropriate subalphabet. To su

Shuche Wang, Adarsh Barik, Peng Zhao, Vincent Y. F. Tan

We develop the first parameter-free algorithms for the Stochastically Extended Adversarial (SEA) model, a framework that bridges adversarial and stochastic online convex optimization. Existing approaches for the SEA model require prior knowledge of problem-specific parameters, such as the diameter of the domain $D$ and the Lipschitz constant of the loss func

Shao-Ping Li, Jens Chluba

Extra radiation injection after neutrino decoupling in the early Universe contributes to the effective number of neutrino species that can be constrained by the cosmic microwave background (CMB). However, any effective neutrino number itself cannot uniquely determine the underlying source. We argue that the degeneracy can be relaxed by CMB spectral distortio

Sayan Ranjan Bhowal, Debashis Paul, Gopal K Basak, Samarjit Das

Debiasing group graphical lasso estimates enables statistical inference when multiple Gaussian graphical models share a common sparsity pattern. We analyze the estimation properties of group graphical lasso, establishing convergence rates and model selection consistency under irrepresentability conditions. Based on these results, we construct debiased estima

Xiaohua Yang, Carlos Peralta, Bjarne Amstrup, Kasper Stener Hintz

The DANish regional atmospheric ReAnalysis (DANRA) is a novel high-resolution (2.5 km) reanalysis dataset covering Denmark and its surrounding regions over a 34-year period (1990-2023). Denmark's complex coastline, with over 400 islands and an extensive 7,400 km coastline, means that most municipalities experience mixed land-sea variability. This complexity

James Stewart, Philippa K. Browning, Mykola Gordovskyy

Twisted coronal loops in the solar atmosphere may become kink-unstable when their magnetic field lines are sufficiently twisted. This instability can trigger magnetic reconnection, leading to the emission of electromagnetic radiation, which manifests as a solar flare. Previous research has demonstrated that oscillations in microwave emissions, resembling obs

Zhanfeng Mo, Xingxuan Li, Yuntao Chen, Lidong Bing

Large language models (LLMs) increasingly rely on multi-turn tool-integrated planning for knowledge-intensive and complex reasoning tasks. Existing implementations typically rely on a single agent, but they suffer from limited context length and noisy tool responses. A natural solution is to adopt a multi-agent framework with planner- and worker-agents to ma

C. Foucher, A. Dutrey, V. Piétu, S. Guilloteau

Context. Edge-on disks offer a unique opportunity to directly examine their vertical structure, providing valuable insights into planet formation processes. We investigate the dust properties, as well as the CO and HCO$^+$ gas properties, in the edge-on disk surrounding the T Tauri star 2MASS J04202144+281349 (SSTTau042021). Aims. We estimate the radial and

Qiyu Wei, Haowei Wang, Richard Allmendinger, Mauricio A. Álvarez

Bayesian optimization has emerged as a prominent methodology for optimizing expensive black-box functions by leveraging Gaussian process surrogates, which focus on capturing the global characteristics of the objective function. However, in numerous practical scenarios, the primary objective is not to construct an exhaustive global surrogate, but rather to qu

Sophie Huczynska, Lukas Klawuhn, Maura B. Paterson

Intersection distribution and non-hitting index are concepts introduced recently by Li and Pott as a new way to view the behaviour of a collection of finite field polynomials. With both an algebraic interpretation via the intersection of a polynomial with a set of lines, and a geometric interpretation via a $(q+1)$-set possessing an internal nucleus, the con

Lorenzo Pannacci, Simone Fiorellino, Mario Edoardo Pandolfo, Emilio Calvanese Strinati

Deep joint source-channel coding (DeepJSCC) has emerged as a powerful paradigm for end-to-end semantic communications, jointly learning to compress and protect task-relevant features over noisy channels. However, existing DeepJSCC schemes assume a shared latent space at transmitter (TX) and receiver (RX) - an assumption that fails in multi-vendor deployments

Chan Hee Song, Yiwen Song, Palash Goyal, Yu Su

Computer-using agents (CUAs) must plan task workflows across diverse and evolving applications, yet progress is limited by the lack of large-scale, high-quality training data. Existing datasets are narrow, static, and costly to annotate, while synthetic data often yields oversimplified or misaligned behaviors. We present Watch & Learn (W&L), a framework that

Giacomo Bertazzoni, Petteri Harjulehto, Peter Hästö, Elvira Zappale

We prove a relaxation result for a quasi-convex bulk integral functional with variable exponent growth in a suitable space of bounded variation type. A key tool is a decomposition under mild assumptions of the energy into absolutely continuous and singular parts weighted via a recession function.

Chao Liu, Ling Luo, Tengxiao Lv, Huan Zhuang

With the rapid development of online medical platforms, consumer health questions (CHQs) are inefficient in diagnosis due to redundant information and frequent non-professional terms. The medical question summary (MQS) task aims to transform CHQs into streamlined doctors' frequently asked questions (FAQs), but existing methods still face challenges such as p

Xuanhua Yin, Runkai Zhao, Weidong Cai

Naturalistic fMRI encoding must handle multimodal inputs, shifting fusion styles, and pronounced inter-subject variability. We introduce AFIRE (Agnostic Framework for Multimodal fMRI Response Encoding), an agnostic interface that standardizes time-aligned post-fusion tokens from varied encoders, and MIND, a plug-and-play Mixture-of-Experts decoder with a sub

Philipp Dunkel, Theresa Klütz, Jochen Linßen, Detlef Stolten

The design of the future European energy system depends heavily on how Europe balances its domestic hydrogen production against its reliance on imports. This study reveals that neither-extreme full self-sufficiency nor complete reliance on imports-is economically optimal through 2050. Using a high-resolution energy system model accounting for interannual wea

Habin Lim, Yeongseob Won, Juwon Seo, Gyeong-Moon Park

In recent years, multi-concept personalization for text-to-image (T2I) diffusion models to represent several subjects in an image has gained much more attention. The main challenge of this task is "concept mixing", where multiple learned concepts interfere or blend undesirably in the output image. To address this issue, in this paper, we present ConceptSplit

Fanzhe Fu, Yang Yang

Reversible Instance Normalization (RevIN) is a key technique enabling simple linear models to achieve state-of-the-art performance in time series forecasting. While replacing its non-robust statistics with robust counterparts (termed R$^2$-IN) seems like a straightforward improvement, our findings reveal a far more complex reality. This paper deconstructs th

Zhimin Hou, Jiacheng Hou, Xiao Chen, Hamid Sadeghian

Therapist-in-the-loop robotic rehabilitation has shown great promise in enhancing rehabilitation outcomes by integrating the strengths of therapists and robotic systems. However, its broader adoption remains limited due to insufficient safe interaction and limited adaptation capability. This article proposes a novel telerobotics-mediated framework that enabl

Eva Sciacca, Nicola Tuccari, Umer Arshad, Fabio Pitari

The exponential growth of data in Astrophysics and Cosmology demands scalable computational tools and intuitive interfaces for analysis and visualization. In this work, we present an innovative integration of the VisIVO scientific visualization framework with the InterActive Computing (IAC) service at Cineca, enabling interactive, high-performance visual wor

Jian Xiao, Ji Wang, Qi Sun, Qimei Cui

Next-generation multiple-input multiple-output (MIMO) systems, characterized by extremely large-scale arrays, holographic surfaces, three-dimensional architectures, and flexible antennas, are poised to deliver unprecedented data rates, spectral efficiency and stability. However, these advancements introduce significant challenges for physical layer signal pr

Mihai Pavel, Julius Ross, Matei Toma

We introduce the notion of a Hodge-Riemann pair of cohomology classes that generalizes the classical Hodge-Riemann bilinear relations, and the notion of a Bogomolov pair of cohomology classes that generalizes the Bogomolov inequality for semistable sheaves. We conjecture that every Hodge-Riemann pair is a Bogomolov pair, and prove various cases of this conje

Kirill Boguslavski

A proper description of the non-equilibrium matter preceding the quark-gluon plasma (QGP) in heavy-ion collisions and its observable consequences remain a major theoretical challenge, while at the same time offering new opportunities for experimental exploration. In these proceedings, I provide an overview of studies presented in talks and posters at Quark M

Yuandou Wang, Filip Gunnarsson, Rihan Hai

Tabular data streams are increasingly prevalent in real-time decision-making across healthcare, finance, and the Internet of Things, often generated and processed on resource-constrained edge and mobile devices. Continual learning (CL) enables models to learn sequentially from such streams while retaining previously acquired knowledge. While recent CL advanc

S. Piscitelli, G. Ponti, M. Civitani, D. Spiga

When performing X-ray observations with a Wolter-I telescope, the presence of bright off-axis sources can introduce unfocused rays, known as straylight, which contaminate the detector and compromise the scientific analysis. Among the different components of straylight, single reflections off the hyperboloid section of the mirror shells often produce arc-like

Nian Ran, Zhongzheng Li, Yue Wang, Qingsong Ran

Multi-objective discrete optimization problems, such as molecular design, pose significant challenges due to their vast and unstructured combinatorial spaces. Traditional evolutionary algorithms often get trapped in local optima, while expert knowledge can provide crucial guidance for accelerating convergence. Large language models (LLMs) offer powerful prio

Manthan Verma, Abhishek K. Jha, Shashwat Nirgudkar, Mahendra K. Verma

The two leading models of isotropic magnetohydrodynamic (MHD) turbulence have competing predictions: $k^{-5/3}$ (Kolmogorov) and $k^{-3/2}$ (Iroshnikov-Kraichnan) scalings. This paper identifies the valid MHD turbulence model using high-resolution numerical and diagnostics-structure functions, intermittency exponents, and energy spectra and fluxes of imbalan

Mingzhe Hu, Ziyin Song, Jingwen Cheng, Gexing Qu

Orbital-selective many-body effects, in which electrons occupying different orbitals experience distinct interaction strengths, play a crucial role in correlated multiorbital materials. However, these effects usually manifest in a complex manner, obscuring their microscopic origins. Here, by combining angle-resolved photoemission spectroscopy measurements wi

Tonmoy Sharma, Banajit Barman, Buddhadeb Bhattacharjee

The study of heavy-flavored baryon and meson production in proton-proton (pp) collisions provides crucial insight into the hadronization mechanisms of Quantum Chromodynamics (QCD). In this work an attempt has been made to describe the existing ALICE and LHCb results on the $p_\mathrm{T}$-differential production cross-section of $\Lambda_c^+$, $D^0$, $\Lambda

Yuheng Li, Jiechao Gao, Wei Han, Wenwen Ouyang

Knowledge of the medical decision process, which can be modeled as medical decision trees (MDTs), is critical to building clinical decision support systems. However, current MDT construction methods rely heavily on time-consuming and laborious manual annotation. To address this challenge, we propose PI-LoRA (Path-Integrated LoRA), a novel low-rank adaptation

Andy Cǎtrunǎ, Adrian Cosma, Emilian Rǎdoi

Gait encodes rich biometric and behavioural information, yet leveraging the manner of walking to infer psychological traits remains a challenging and underexplored problem. We introduce a hierarchical Multi-Stage Mixture of Movement Experts (MoME) architecture for multi-task prediction of psychological attributes from gait sequences represented as 2D poses.

Ryota Kojima, Corina Cirstea

We introduce continuation semantics for both fixpoint modal logic (FML) and Computation Tree Logic* (CTL*), parameterised by a choice of branching type and quantitative predicate lifting. Our main contribution is proving that they are equivalent to coalgebraic semantics, for all branching types. Our continuation semantics is defined over coalgebras of the co

Ines Akaichi, Giorgos Flouris, Irini Fundulaki, Sabrina Kirrane

In the digital age, data frequently crosses organizational and jurisdictional boundaries, making effective governance essential. Usage control policies have emerged as a key paradigm for regulating data usage, safeguarding privacy, protecting intellectual property, and ensuring compliance with regulations. A central mechanism for usage control is the handlin

Chemseddine Ananna, Lucia Barbieri, Axel Boeltzig, Matteo Campostrini

Nuclear reactions are responsible for the chemical evolution of stars, galaxies and the Universe. Unfortunately, at temperatures of interest for nuclear astrophysics, the cross-sections of the thermonuclear reactions are in the pico-femto-barn range and thus measuring them in the laboratory is extremely challenging. In this framework, major steps forward wer

Adnan Aijaz, Peizheng Li, Sajida Gufran

Low Earth orbit (LEO) satellites and satellite direct-to-device (D2D) technology are at the heart of the next-generation global connectivity which promises direct access to space-based broadband services for unmodified 3GPP-compliant handsets. With a rapidly evolving ecosystem, it is important to evaluate the feasibility, cost-effectiveness, and profitabilit

Maksym Zavhorodnii, Dmytro Dehtiarov, Anna Konovalenko

This work introduces a novel methodology for the automatic detection of hallucinations generated during large language model (LLM) inference. The proposed approach is based on a systematic taxonomy and controlled reproduction of diverse hallucination types through prompt engineering. A dedicated hallucination dataset is subsequently mapped into a vector spac

Evgeny Ryabkov, Mingzhao Song, Andrey A. Bogdanov, Denis G. Baranov

The progress in understanding the optical and microscopic properties of polaritons relies on various optical cavities to confine electromagnetic radiation, which causes a demand for new platforms with higher $Q$-factors and better fabrication robustness. In this context, so called Mie voids -- spherical cavities inside a dielectric medium, where the light co

Mateo Torres-Ruiz, Robin Piedeleu, Alexandra Silva, Fabio Zanasi

We extend the synthetic theories of discrete and Gaussian categorical probability by introducing a diagrammatic calculus for reasoning about hybrid probabilistic models in which continuous random variables, conditioned on discrete ones, follow a multivariate Gaussian distribution. This setting includes important classes of models such as Gaussian mixture mod

Buyuan Zhu, Shiyu Hu, Yiping Ma, Yuanming Zhang

As large language models are increasingly integrated into education, virtual student agents are becoming vital for classroom simulation and teacher training. Yet their classroom-oriented subjective abilities remain largely unassessed, limiting understanding of model boundaries and hindering trustworthy deployment. We present EduPersona, a large-scale benchma

Jiewen Guan, Bo Jiang, Zhening Li

We study the decomposability and the subdifferential of the tensor nuclear norm. Both concepts are well understood and widely applied in matrices but remain unclear for higher-order tensors. We show that the tensor nuclear norm admits a full decomposability over specific subspaces and determine the largest possible subspaces that allow the full decomposabili

Mario Teixeira Parente

This article offers a motivating travel guide towards the Jordan normal form, one of the highlights in courses on linear algebra or advanced mathematics. Its itinerary is characterized by a focus on core geometric aspects and the avoidance of algebraic tools. In this way, it attempts to encourage academic lecturers from more applied mathematical contexts to

Hugo Resende, Fabio A. Faria, Eduardo B. Neto, Isabela Borlido

Image segmentation is a crucial step in various visual applications, including environmental monitoring through remote sensing. In the context of the ForestEyes project, which combines citizen science and machine learning to detect deforestation in tropical forests, image segments are used for labeling by volunteers and subsequent model training. Traditional

Hirotsugu Kakugawa, Sayaka Kamei, Masahiro Shibata, Fukuhito Ooshita

Self-stabilization is a versatile methodology in the design of fault-tolerant distributed algorithms for transient faults. A self-stabilizing system automatically recovers from any kind and any finite number of transient faults. This property is specifically useful in modern distributed systems with a large number of components. In this paper, we propose a n

Xiangyu Li, Yawen Zeng, Xiaofen Xing, Jin Xu

In this paper, our objective is to develop a multi-agent financial system that incorporates simulated trading, a technique extensively utilized by financial professionals. While current LLM-based agent models demonstrate competitive performance, they still exhibit significant deviations from real-world fund companies. A critical distinction lies in the agent

Marta Bílková, Thomas Ferguson, Daniil Kozhemiachenko

We propose a multi-agent epistemic logic capturing reasoning with degrees of plausibility that agents can assign to a given statement, with $1$ interpreted as "entirely plausible for the agent" and $0$ as "completely implausible" (i.e., the agent knows that the statement is false). We formalise such reasoning in an expansion of G\"{o}del fuzzy logic with an

Ayan Majumdar, Feihao Chen, Jinghui Li, Xiaozhen Wang

Large-scale web-scraped text corpora used to train general-purpose AI models often contain harmful demographic-targeted social biases, creating a regulatory need for data auditing and developing scalable bias-detection methods. Although prior work has investigated biases in text datasets and related detection methods, these studies remain narrow in scope. Th

Ali Asghar, Andreas Becher, Daniel Ziener

The dependence of power-consumption on the processed data is a known vulnerability of CMOS circuits, resulting in side channels which can be exploited by power-based side channel attacks (SCAs). These attacks can extract sensitive information, such as secret keys, from the implementation of cryptographic algorithms. Existing countermeasures against power-bas

F. Iraci, A. Chalumeau, C. Tiburzi, J. P. W. Verbiest

Low-frequency radio data improve the sensitivity of pulsar timing arrays (PTAs) to propagation effects such as dispersion measure (DM) variations, enabling better noise characterization essential for detecting the stochastic gravitational wave background (GWB). We combined LOFAR (100-200 MHz) and NenuFAR (30-90 MHz) observations with the recent European and

Stepan Hudecek

This paper studies the Poisson equation for the $G_2$-Laplacian on 3-forms on the 7-sphere that are invariant under a transitive group action. We establish the existence and uniqueness of $G$-invariant solutions for $G=SU(4),\: Spin(7),\: (Sp(2)\times Sp(1))/\mathbb{Z}_2$. In the case $G=Sp(2)\times U(1)/\mathbb{Z}_2$, we show that the operator does not pres

Zeyi Zhang, Yanju Zhou, Heyuan Yao, Tenglong Ao

We present Social Agent, a novel framework for synthesizing realistic and contextually appropriate co-speech nonverbal behaviors in dyadic conversations. In this framework, we develop an agentic system driven by a Large Language Model (LLM) to direct the conversation flow and determine appropriate interactive behaviors for both participants. Additionally, we

Rémi Busselez

Understanding the relationship between microscopic structure and macroscopic thermodynamic properties is a central challenge in the study of complex fluids. The Kirkwood-Buff (KB) theory offers an elegant and powerful framework for bridging this gap by relating integrals over pair correlation functions to measurable thermodynamic quantities. In multicomponen

Antonio Capanema, Pasquale Blasi, Emanuele Sobacchi

Over the past decades, there has been growing observational and theoretical evidence that cosmic-ray-induced instabilities play an important role in both acceleration and transport of cosmic rays (CRs). For instance, the efficient acceleration of charged particles at supernova remnant shocks requires rapidly growing instabilities, so much so that none of the

Lukas Gienapp, Martin Potthast, Andrew Yates, Harrisen Scells

The unjudged document problem, where systems that did not contribute to the original judgement pool may retrieve documents without a relevance judgement, is a key obstacle to the reuseability of test collections in information retrieval. While the de facto standard to deal with the problem is to treat unjudged documents as non-relevant, many alternatives hav

Marcel David Fabian, Nina Glaser, Gemma C. Solomon

The semi-empirical Pariser-Parr-Pople (PPP) Hamiltonian is reviewed for its ability to provide a minimal model of the chemistry of conjugated $\pi$-electron systems, and its current applications and limitations are discussed. From its inception, the PPP Hamiltonian has helped in the development of new computational approaches in instances where compute is co

Anastasia Zhukova, Jonas Lührs, Christian E. Lobmüller, Bela Gipp

Recent trends in NLP utilize knowledge graphs (KGs) to enhance pretrained language models by incorporating additional knowledge from the graph structures to learn domain-specific terminology or relationships between documents that might otherwise be overlooked. This paper explores how SciNCL, a graph-aware neighborhood contrastive learning methodology origin

Vrushank Ahire, Aniruddh Muley, Shivam Zample, Siddharth Verma

Detecting manipulated media has now become a pressing issue with the recent rise of deepfakes. Most existing approaches fail to generalize across diverse datasets and generation techniques. We thus propose a novel ensemble framework, combining the strengths of transformer-based architectures, such as Swin Transformers and ViTs, and texture-based methods, to

Jiho Noh, Julian Schulz, Wladimir Benalcazar, Christina Jörg

Photonic platforms have emerged as versatile and powerful classical simulators of quantum dynamics, providing clean, controllable optical analogs of extended structured (i.e., crystalline) electronic systems. While most realizations to date have used only the fundamental mode in each site, recent advances in structured light - particularly the use of higher-

Mohamed Ayoub Ben Ayad, Michael Dinzinger, Kanishka Ghosh Dastidar, Jelena Mitrovic

Embedding models are central to dense retrieval, semantic search, and recommendation systems, but their size often makes them impractical to deploy in resource-constrained environments such as browsers or edge devices. While smaller embedding models offer practical advantages, they typically underperform compared to their larger counterparts. To bridge this

Damien Calaque, Stefano Ronchi

These notes are intended to be an introduction to shifted symplectic geometry, targeted to Poisson geometers with a serious background in homological algebra. They are extracted from a mini-course given by the first author at the Poisson 2024 summer school that took place at the Accademia Pontaniana in Napoli.

Eugene Lim, Tzeh Yuan Neoh, Nicholas Teh

We study a sequential decision-making model where a set of items is repeatedly matched to the same set of agents over multiple rounds. The objective is to determine a sequence of matchings that either maximizes the utility of the least advantaged agent at the end of all rounds (optimal) or at the end of every individual round (anytime optimal). We investigat

Shrish Shrinath Vaidya, Gowthamaan Palani, Sidharth Ramesh, Velmurugan Balasubramanian

The deployment of Large Language Models (LLMs) for structuring clinical data is critically hindered by their tendency to hallucinate facts and their inability to follow domain-specific rules. To address this, we introduce MedPAO, a novel agentic framework that ensures accuracy and verifiable reasoning by grounding its operation in established clinical protoc

Youngjoon Lee, Seongmin Cho, Yehhyun Jo, Jinu Gong

The limited data availability due to strict privacy regulations and significant resource demands severely constrains biomedical time-series AI development, which creates a critical gap between data requirements and accessibility. Synthetic data generation presents a promising solution by producing artificial datasets that maintain the statistical properties

Fabien de Montgolfier, Renaud Torfs

There are three usual definitions of a maximum bipartite clique (biclique) in a bipartite graph\,: either maximizing the number of vertices, or of edges, or finding a maximum balanced biclique. The first problem can be solved in polynomial time, the last ones are NP-complete. Here we show how these three problems may be efficiently solved for two classes of

Siu Kei Chung, Francisco Carpio, Andrei Navoichyk, Siarhei Valasovich

The internet faces a sovereignty crisis due to power concentration and data growth among a few hyperscalers, leading to centralization and loss of user control. This consolidation risks censorship and creates single points of failure. While Web3 offers decentralized solutions, they often sacrifice either scalability, decentralization, or security, which are

Ivan Homoliak, Martin Perešíni, Marek Tamaškovič, Timotej Ponek

Proof-of-Stake (PoS) consensus protocols often face a trade-off between performance and security. Protocols that pre-elect leaders for subsequent rounds are vulnerable to Denial-of-Service (DoS) attacks, which can disrupt the network and compromise liveness. In this work, we present PoS-CoPOR, a single-chain PoS consensus protocol that mitigates this vulnera

Qizheng Zhang, Changran Hu, Shubhangi Upasani, Boyuan Ma

Large language model (LLM) applications such as agents and domain-specific reasoning increasingly rely on context adaptation: modifying inputs with instructions, strategies, or evidence, rather than weight updates. Prior approaches improve usability but often suffer from brevity bias, which drops domain insights for concise summaries, and from context collap

Zhejian Lai, Xiang Geng, Zhijun Wang, Yang Bai

Mathematical reasoning is a primary indicator of large language models (LLMs) intelligence. However, existing LLMs exhibit failures in robustness and generalization. This paper attributes these deficiencies to spurious reasoning, wherein generated reasoning traces are driven by superficial correlations, leading models to blindly reproduce memorized patterns

Bohan Cui, Yu Chen, Alessandro Giua, Xiang Yin

In this work, we investigate the problem of synthesizing property-enforcing supervisors for partially-observed discrete-event systems (DES). Unlike most existing approaches, where the enforced property depends solely on the executed behavior of the system, here we consider a more challenging scenario in which the property relies on predicted future behaviors