April 2026 arXiv papers — page 91

I. Henum, D. Hobbs, Ó. Jiménez-Arranz, P. J. McMillan

The proposed GaiaNIR mission would extend Gaia's astrometric capabilities into the near-infrared, improving astrometric precision and enabling observations in heavily dust-obscured regions. In this work, we investigate the impact of GaiaNIR on the detectability of dark matter halos in globular clusters by comparing its performance with that of Gaia. Expected

Fei Wang, Li Shen, Liang Ding, Chao Xue

Zeroth-Order optimization presents a promising memory-efficient paradigm for fine-tuning Large Language Models by relying solely on forward passes. However, its practical adoption is severely constrained by slow wall-clock convergence and high estimation variance. In this work, we dissect the runtime characteristics of ZO algorithms and identify a critical s

Rupert L. Frank, Simon Larson

We review our recent results on the problem of optimizing Riesz means of Laplace eigenvalues among convex sets of given measure in the regime where the cut-off parameter in the definition of the Riesz means tends to infinity. We show that for a certain range of Riesz exponents, the optimizing sets converge to a ball. We also present some new results where we

Chih-Kang Huang, Ludovick Gagnon, Miha Založnik, Benoît Appolaire

The multi-scale and non-linear nature of phase-field models of solidification requires fine spatial and temporal discretization, leading to long computation times. This could be overcome with artificial-intelligence approaches. Surrogate models based on neural operators could have a lower computational cost than conventional numerical discretization methods.

Han Li, Zehao Huang, Jiahui Fu, Naiyan Wang

Multimodal large language models have demonstrated remarkable capabilities in 2D vision, motivating their extension to 3D scene understanding. Recent studies represent 3D scenes as 3D spatial videos composed of image sequences with depth and camera pose information, enabling pre-trained video-language models to perform 3D reasoning tasks. However, the large

Jen-Yuan Huang, Tong Lin, Yilun Du

While modern text-to-image (T2I) models excel at generating images from intricate prompts, they struggle to capture the key details when the inputs are descriptive paragraphs. This limitation stems from the prevalence of concise captions that shape their training distributions. Existing methods attempt to bridge this gap by either fine-tuning T2I models on l

Rahul Mehta, Kavin R, Indrajit Pal, Tushar Abhishek

Query auto-completion (QAC) has been widely studied in the context of web search, yet remains underexplored for in-document search, which we term DocQAC. DocQAC aims to enhance search productivity within long documents by helping users craft faster, more precise queries, even for complex or hard-to-spell terms. While global historical queries are available t

Svetlana Pavlitska, Malte Stüven, Beyza Keskin, J. Marius Zöllner

Mixture-of-Experts (MoE) models provide a structured approach to combining specialized neural networks and offer greater interpretability than conventional ensembles. While MoEs have been successfully applied to image classification and semantic segmentation, their use in object detection remains limited due to challenges in merging dense and structured pred

Xuanyu Liu, Shijian Gao, Boxun Liu, Xiang Cheng

Current learning-based wireless methods struggle with generalization due to the fragmented processing of communication and sensing data. WiFo-MiSAC addresses this as a task-agnostic foundation model that tokenizes heterogeneous signals into a unified space for self-supervised pre-training. A shared-specific disentangled mixture-of-experts (SS-DMoE) architect

Salmane Chafik, Saad Ezzini, Ismail Berrada

Recently, code-oriented large language models (LLMs) have demonstrated strong capabilities in translating natural language into executable code. Text-to-SQL is a significant application of this ability, enabling non-technical users to interact with relational databases using natural language. However, state-of-the-art models continue to struggle with highly

Simone Catanzaro, Elvira Di Nardo

The first passage time problem is considered for stochastic logistic growth model with constant harvesting and multiplicative environmental noise. Explicit expressions for the moments and cumulants of both upcrossing and downcrossing FPTs in the presence of constant thresholds are obtained through a power-series expansion of the Laplace transform. Then a clo

Kangrou Guo, Xiumin Huang, Dong Lai

A key feature of close-in, multiple super-Earth (SE) systems is the tendency for adjacent planet pairs to lie just wide of low-order mean-motion resonances (MMR). This period ratio distribution has motivated numerous theoretical studies, particularly those invoking post-disk processes that perturb initially resonant architectures. We investigate whether orbi

Hamed Ouattara, Pascal Houssam Salmane, Pierre Duthon, Frédéric Bernardin

In this paper, we present three neural network architectures designed for real-time classification of weather conditions (sunny, rain, snow, fog) from images. These models, inspired by recent advances in style transfer, aim to capture the stylistic elements present in images. One model, called "Multi-PatchGAN", is based on PatchGANs used in well-known archit

Xixi Liu, Jorge Lazo, Andreas Hallqvist, Mikael Johansson

Accurate prognostication and risk estimation are essential for guiding clinical decision-making and optimizing patient management. While radiologist-assessed features from CT scans provide valuable indicators of disease severity and outcomes, interpreting such images requires expert knowledge, and translating rich visual information into textual summaries in

Felix Herron, Maja Hjuler, Solange Rossato, Alexandre Allauzen

Speech encoder models are known to model members of some speaker groups (SGs) better than others. However, there has been little work in establishing why this occurs on a technological level. To our knowledge, we present the first layerwise fairness analysis of pretrained self-supervised speech encoder models (S3Ms), probing each embedding layer for speaker

Alessio Baldelli, Marco Baldi, Davide De Zuane, Paolo Santini

Bit-Flipping (BF) decoders are a family of decoders widely employed in post-quantum cryptographic schemes based on Quasi-Cyclic Moderate-Density Parity-Check (QC-MDPC) codes, such as BIKE. BF decoders suffer from trapping sets, corresponding to low-weight error patterns that likely lead to decoding failures. For QC-MDPC codes, the most relevant family of tra

A. Errehymy, S. K. Maurya, M. Govender, K. N. Singh

In this work, we explore for the first time slowly rotating traversable wormholes embedded in holographic dark energy. We focus on three representative holographic dark energy models -- R\'{e}nyi, mixed, and Moradpour -- and construct the wormhole shape functions directly from these energy density profiles using a Teo-type rotating wormhole metric. This allo

Fernando Reitich

Large language models operate in protocols containing multiple calls, yet added calls are usually evaluated only by their net effect. That summary cannot distinguish correcting unsuccessful outputs from corrupting initially successful ones. We develop a paired audit recording success before and after a specified operation on the same tasks under one binary r

Luca Capizzi, Benoît Ferté

Quantum many-body scars are rare exceptions to thermalization: they sustain non-thermal stationary states without the protection of any local conservation law, and are generally expected to be fragile. Here we construct an analytically tractable random unitary circuit hosting a single scar, and derive from first principles the thermalization mechanism govern

Simon Hochgerner, Jonas Ingmanns, Nicole Kastanek

We are concerned with the market-consistent valuation of lifelong health insurance products, which are subject to adjustments derived from the actuarial equivalence principle and driven by (medical) inflation. Such products are well-established in the European national markets, and the dynamics of the adjustment mechanism is well-understood from an actuarial

Kilian Rausch

In this paper, we calculate an exact formula for the number of partitions of a natural number $n$, where the largest part is even and no odd parts appears more than two times. The generating functions of the number of these partitions is a mixed mock modular form of weight 0. In order to obtain the formula we apply an extended version of the circle method, d

Wentao Shi, Yu Wang, Yuyang Zhao, Yuxin Chen

As reinforcement learning continues to scale the training of large language model-based agents, reliably verifying agent behaviors in complex environments has become increasingly challenging. Existing approaches rely on rule-based verifiers or LLM-as-a-Judge models, which struggle to generalize beyond narrow domains. Agent-as-a-Judge addresses this limitatio

Zhuojun Tian, Chaouki Ben Issaid, Mehdi Bennis

In large-scale distributed scenarios, increasingly complex tasks demand more intelligent collaboration across networks, requiring the joint extraction of structural representations from data samples. However, conventional task-specific approaches often result in nonstructural embeddings, leading to collapsed variability among data samples within the same cla

Alex Mitrevski, Ayush Salunke

In the context of robot learning for manipulation, curated datasets are an important resource for advancing the state of the art; however, available datasets typically only include successful executions or are focused on one particular type of skill. In this short paper, we briefly describe a dataset of various skills performed in the context of coffee prepa

Lorenz Brehme, Thomas Ströhle, Ruth Breu

Retrieval-augmented generation (RAG) enhances large language models (LLMs) with external knowledge to answer questions more accurately. However, research on evaluating RAG systems-particularly the retriever component-remains limited, as most existing work focuses on single-context retrieval rather than multi-hop queries, where individual contexts may appear

Jordan Auge, Sam Betts, Giovanna Carofiglio, Giulio Grassi

Network change validation remains a critical yet predominantly manual, time-consuming, and error-prone process in modern network operations. While formal network verification has made substantial progress in proving correctness properties, it is typically applied in offline, pre-deployment settings and faces challenges in accommodating continuous changes and

Javad Maheri, K. K. Krishnan Namboodiri, Petros Elia

We consider a distributed computing system in which a master node coordinates $N$ workers to evaluate a function over $n$ input files, where this function accepts general decomposition. In particular, we focus on the general case where the requested function admits a $d$-uniform decomposition, meaning that it can be decomposed into a set of subfunctions that

Sina Abdollahi, Mohammad M Maheri, Javad Forough, Amir Al Sadi

Large Language Model (LLM) agents provide powerful automation capabilities, but they also create a substantially broader attack surface than traditional applications due to their tight integration with non-deterministic models and third-party services. While current deployments primarily rely on cloud-hosted services, emerging designs increasingly execute ag

Risa Togo, Sara Cardona, Irène Nagle, Gijsje H. Koenderink

Fibrous networks are ubiquitous structural components in biology, spanning cellulose in plant cell walls, fibrin in blood clots, and collagen in the extracellular matrix of animal tissues. Theoretical models predict that network connectivity critically influences their mechanical behavior. However, accurately reconstructing network topology from 3D image dat

Ulysse Naepels, Victor M. Panaretos

Smoothness has long been the dominant form of parsimony in functional data analysis, to the point of occasionally being conflated with the very notion of functional data. However, many core inferential tasks depend on the inverse covariance, where sparsity--rather than smoothness--emerges as the more natural structural constraint. In this paper, we explore M

Anaïs Besnard, Valentin Sauvage, Bruno Maffei

After the Planck mission's launch in 2009, bolometers of its High Frequency Instrument (HFI) were considerably affected by cosmic rays, which necessitated several years of post-treatment to clean the data. To study the susceptibility of high sensitivity cryogenic detectors to particle impacts, IAS has developed the DRACuLA facility. We present the results of

Alberto Tagliaferro, Bruno Guindani, Livia Lestingi, Matteo Rossi

Early-stage specifications of safety-critical systems are typically expressed in natural language, making it difficult to derive formal properties suitable for verification and needed to guarantee safety. While recent Large Language Model (LLM)-based approaches can generate formal artifacts from text, they mainly focus on syntactic correctness and do not ens

Pierre-Carl Langlais, Pavel Chizhov, Yannick Detrois, Carlos Rosas Hinostroza

Automated analysis of customer feedback on social media is hindered by three challenges: the high cost of annotated training data, the scarcity of evaluation sets, especially in multilingual settings, and privacy concerns that prevent data sharing and reproducibility. We address these issues by developing a generalizable synthetic data generation pipeline ap

Xinping Lei, Xinyu Che, Junqi Xiong, Chenchen Zhang

Large language models are rapidly evolving into interactive coding agents capable of end-to-end web coding, yet existing benchmarks evaluate only narrow slices of this capability, typically text-conditioned generation with static-correctness metrics, leaving visual fidelity, interaction quality, and codebase-level reasoning largely unmeasured. We introduce W

Zhen Liu, Yuhan Liu, Jinjun Wang, Jianyi Liu

Vision-and-Language Navigation requires agents to follow natural-language instructions in visually changing environments. A central challenge is the dynamic entanglement between language and observations: the meaning of instruction shifts as the agent's field of view and spatial context evolve. However, many existing models encode the instruction as a static

David Kivarkis, Waleed Mouhali, Sadruddin Benkadda, Kai Schneider

We develop two deep learning surrogate autoregressive models for the prediction of the temporal evolution of two-dimensional ideal magnetohydrodynamic (MHD) Kelvin-Helmholtz instabilities across a range of magnetic field strengths. Using two neural network architectures, a Koopman-based Transformer model and a ConvLSTM-UNet, our approach enables simultaneous

M Reza Salarian

This note provides an introduction to selected topics in algebraic graph theory, including strongly regular graphs, Steiner systems, and automorphism groups. We describe constructions and properties of notable graphs such as the Petersen graph, Paley graphs, Hamming graphs, and the Hoffman-Singleton graph, with emphasis on their symmetry and combinatorial st

Zikun Zhou, Wenshuo Wang, Wenzhuo Liu, Hui Yao

Electroencephalography (EEG) signals have been promising for long-term braking intensity prediction but are prone to various artifacts that limit their reliability. Here, we propose a novel framework that models EEG signals as mixtures of independent blind sources and identifies those strongly correlated with braking action. Our method employs independent co

Sandip Halder, Moshe Schechter

We investigate the magnetotransport properties of the one-band Hubbard-Holstein model at half-filling in three dimensions (3D) using exact diagonalization based semi-classical Monte Carlo simulations with phonons treated in the adiabatic limit. The low-temperature electronic correlation $U$ vs electron-phonon coupling $V$ phase diagram reveals two insulating

Shang-Wei Lin, Ji-Qing Yan, Yean-Ru Chen, Zhe Hou

Satisfiability (SAT) is a central problem in computer science, and advances in SAT-solving algorithms have a far-reaching impact across many fields. Recent works have proposed quantum SAT solvers based on Grover's algorithm, a quantum search technique. However, Grover-based approaches face a key limitation: they typically require prior knowledge of the numbe

Xiamiao Zhao, Xin Cheng, Dániel Gerbner

Let $\mathcal{H}$ be a hypergraph and $F$ be a graph. If there exists a bijection between the hyperedges of $\mathcal{H}$ and the edges of $F$ such that each hyperedge contains its image, then we say that $\mathcal{H}$ is a \textit{Berge copy} of $F$, and the collection of Berge copies of $F$ is denoted by Berge-$F$. Given $r$-graphs $\mathcal{F}$ and $\math

Woojin Lee, Jin-Xia Huang

State-of-the-art code generation frameworks rely on mental simulation, where LLMs internally trace execution to verify correctness. We expose a fundamental limitation: the Mental-Reality Gap -- where models hallucinate execution traces and confidently validate buggy code. This gap manifests along two orthogonal dimensions: the Specification Gap (overlooking

Yanjun Guo, Zhengqiang Zhang, Pengfei Wang, Xinyue Liang

Spatially consistent long-horizon video generation aims to maintain temporal and spatial consistency along predefined camera trajectories. Existing methods mostly entangle memory modeling with video generation, leading to inconsistent content during scene revisits and diminished generative capacity when exploring novel regions, even trained on extensive anno

The STAR Collaboration

The STAR Collaboration reports measurements of the collision energy dependence of hypertriton (${}^{3}_{\Lambda}$H) transverse momentum spectra and $p_{\rm T}$-integrated yields at mid-rapidity ($|y|<$0.5) in Au+Au collisions at 11 collision energies between 3.2 and 27\,GeV. The measured ${}^{3}_{\Lambda}$H yields and ${}^{3}_{\Lambda}$H/$\Lambda$ yields rat

Yikang Hou, Junjie Huang, Yijun Ran, Tao Jia

Dynamic signed networks (DSNs) are common in online platforms, where time-stamped positive and negative relations evolve over time. A core task in DSNs is dynamic edge prediction, which forecasts future relations by jointly modeling edge existence and polarity (positive, negative, or non-existent). However, existing dynamic signed network embedding (DSNE) me

Robert Lasarzik, Elisabetta Rocca, Giulio Schimperna

We prove existence of weak solutions and weak-strong uniqueness for a mathematical model which couples the evolution of a phase-parameter $\varphi$ satisfying a Cahn-Hilliard type relation with the one of an additional variable $\sigma$ influencing the phase separation process. The main application of the model refers to cancer growth processes, where $\sigm

Sheng Xu, Guiliang Liu, Tarak Kharrat, Yudong Luo

Success in association football relies on both individual skill and coordinated tactics. While recent advancements in spatio-temporal data and deep learning have enabled predictive analyses like trajectory forecasting, the development of tactical design remains limited. Bridging this gap is essential, as prediction reveals what is likely to occur, whereas ta

Benjamin Hertzsch, Job Feldbrugge, Rien van de Weygaert

The caustic skeleton is a parameter-free and mathematically rigorous formalism for tracing the hierarchical formation history of the multiscale cosmic web from the singularities in the underlying dark matter flow. In the present study, we explicitly use the multistreaming nature of the cosmic mass distribution to address the influence of the weblike embeddin

Andreas Kriegler, Csaba Beleznai, Margrit Gelautz

Symmetric objects are common in daily life and industry, yet their inherent orientation ambiguities that impede the training of deep learning networks for pose estimation are rarely discussed in the literature. To cope with these ambiguities, existing solutions typically require the design of specific loss functions and network architectures or resort to sym

Eylem Erdogan, Mohammed Elamassie, Ibrahim Altunbas, Gunes Karabulut Kurt

Free-space optical (FSO) communication is emerging as a key backhaul technology for next-generation vertical heterogeneous networks (VHetNets), whose architecture spans satellites, high-altitude platform stations (HAPS), unmanned aerial vehicles (UAVs), and terrestrial nodes. Along these vertical and slant paths, optical beams traverse successive atmospheric

Yanzhen Lu, Muchen Jiang, Zhicheng Qian, Xingyu Zhou

Prompt-injected memory can improve reasoning without updating model weights, but it also creates a control problem: retrieved content helps only when it is applied in the right state. We study this problem in a strict training-free setting and formulate it as applicability control: when to trigger a memory-assisted second pass, when to trust it, and how to m

Mikolaj Zielinski, Eryk Vykysaly, Bartlomiej Biesiada, Jan Baturo

Recent advances in neural rendering have introduced numerous 3D scene representations. Although standard computer vision metrics evaluate the visual quality of generated images, they often overlook the fidelity of surface geometry. This limitation is particularly critical in robotics, where accurate geometry is essential for tasks such as grasping and object

V. S. D. S. Mahesh Akavarapu, Michael Daniel, Gerhard Jäger

We present a phoneme-level analysis of automatic speech recognition (ASR) for two low-resourced and phonologically complex East Caucasian languages, Archi and Rutul, based on curated and standardized speech-transcript resources totaling approximately 50 minutes and 1 hour 20 minutes of audio, respectively. Existing recordings and transcriptions are consolida

Serena Suriano, Shamil Biktimirov, Dmitry Pritykin, Anton Ivanov

Asteroid mining can become an enabling technology to establish a sustainable manned colony on Mars, which requires metallic materials more often than they are readily available in shipments from Earth. This paper describes a feasibility study of a supply chain that delivers metals extracted from metallic asteroids to Mars. The asteroids are selected to respe

Samuel G. Balter, Ethan Jerzak, Connor T. Jerzak

Multimodal LLMs can accurately perceive numerical content across modalities yet fail to perform exact multi-digit multiplication when the identical underlying arithmetic problem is presented as numerals, number words, images, or in audio form. Because existing benchmarks often lack systematically paired instances across modalities, it remains difficult to co

Lachlan Ewen MacDonald

We prove a centre manifold theorem for a map along a manifold-with-boundary of fixed points, and provide an application to the study of gradient descent with large step size on two-layer matrix factorisation problems.

Geet Sethi, Panav Shah, Ashutosh Gandhe, Soumitra Darshan Nayak

Diffusion models have emerged as powerful tools for a wide range of vision tasks, including text-guided image generation and editing. In this work, we explore their potential for object grounding in remote sensing imagery. We propose a hybrid pipeline that integrates diffusion-based localization cues with state-of-the-art segmentation models such as RemoteSA

Wuhan Chen, Min Gao, Xin Xia, Zongwei Wang

Large language models (LLMs) have recently shown promise in recommendation by providing rich semantic knowledge. While most existing approaches rely on external textual corpora to align LLMs with recommender systems, we revisit a more fundamental yet underexplored question: Can recommendation benefit from LLM token embeddings alone without textual input? Thr

Catalin-Viorel Dinu, Diederick Vermetten, Carola Doerr

In black-box optimization, a central question is which algorithm to use to solve a given, previously unseen, problem. Selecting a single algorithm, however, entails inherent risks: inaccuracies in the selector may lead to poor choices, and even well-performing algorithms with high variance can yield unsatisfactory results in a single run. A natural remedy is

Marco Gallo, Luigi Vezzoni

We establish a general result ensuring a $C^1$ a priori bound for smooth curves of Hermitian metrics. As a main application, we obtain a new regularity result for Hermitian curvature flows, and in particular for the second Chern-Ricci flow.

Arkadii Kazanskii, Tatiana Petrova, Andrey Ustyuzhanin, Konstantin Bagrianskii

Single-step generators promise high-fidelity synthesis at a fraction of the inference and training cost of ordinary differential equation (ODE)-based flow models, a central concern when compute is limited. Drifting Models (DMs) train a one-step generator by evolving samples under a kernel-based drift field, avoiding ODE integration entirely, but a two-partic

Armin Nurkanović

Mathematical programs with complementarity constraints (MPCCs) are a challenging class of nonlinear optimization problems, because their nonlinear programming reformulations violate standard constraint qualifications at every feasible point. This paper analyzes sequential quadratic programming with complementarity constraints (SQPCC). In this method, the com

Tim Goppelsroeder, Rasmus Jensen

We propose MADDPG-K, a scalable extension to Multi-Agent Deep Deterministic Policy Gradient (MADDPG) that addresses the computational limitations of centralized critic approaches. Centralized critics, which condition on the observations and actions of all agents, have demonstrated significant performance gains in cooperative and competitive multi-agent setti

Chiara Ghinami, Igor Pontes Tresolavy, Luis Seibt, Nils Bosbach

The increasing complexity of embedded software has made comprehensive manual testing impractical, motivating the use of automated techniques such as fuzzing. Coverage-guided fuzzers like AFL++ have shown strong results for conventional software but remain challenging to apply effectively in embedded contexts, where peripheral behaviors play critical roles. E

Abril Sahade, Cecilia Mac Cormack, Angelos Vourlidas, Teresa Nieves-Chinchilla

The morphology and heliospheric impact of coronal mass ejections (CMEs) are strongly shaped by their preeruptive magnetic configuration and surrounding coronal environment, yet these influences remain difficult to constrain observationally. We analyze a complex CME that erupted on 2024 October 26 using multiviewpoint remote sensing observations and in situ m

Xiang He, Chenxing Li, Jinting Wang, Yan Rong

Large Audio-Language Models (LALMs) have made significant progress in audio understanding, yet they primarily operate as perception-and-answer systems without explicit reasoning processes. Existing methods for enhancing audio reasoning rely either on supervised chain-of-thought (CoT) fine-tuning, which is limited by training data quality, or on reinforcement

Roson Nongthombam, Urmimala Dewan, Amarendra K. Sarma

Superconducting qubits, realized by incorporating Josephson junctions into superconducting circuits, behave as artificial atoms with anharmonic energy spectra and can be precisely controlled and measured using microwave cavities within the framework of circuit quantum electrodynamics (cQED). Since its emergence in the early 2000s, cQED has established superc

Xu Wang, Shengeng Tang, Wan Jiang, Yaxiong Wang

Continuous Sign Language Recognition (CSLR) has achieved remarkable progress in recent years; however, most existing methods are developed under single-view settings and thus remain insufficiently robust to viewpoint variations in real-world scenarios. To address this limitation, we propose CanonSLR, a canonical-view guided framework for multi-view CSLR. Spe

Lidia N. Makarova, Denis G. Purytin, R. Brent Tully, Gagandeep S. Anand

We present a homogeneous metallicity analysis of old stellar populations in Local Volume (LV) galaxies using data from the CMDs/TRGB catalog of the Extragalactic Distance Database (EDD; http://edd.ifa.hawaii.edu), which provides uniformly measured TRGB distances and PSF photometry for resolved stars in over 500 nearby galaxies observed with the Hubble Space

Anuj Malik

We study the large-$N$ limit of $U(N)$ and $SU(N)$ unitary matrix models inspired by QCD. The model is analyzed in two cases: $\mu = 0$, where the potential is real, and finite $\mu$, where it becomes complex. The complex action drives the eigenvalues into the complex plane, leading to $\langle U \rangle \neq \langle U^{-1} \rangle$. In the ungapped phase, w

Ben Deitmar

This paper introduces the separable covariance mixture model, which assumes a data-matrix $Y$ to be of the form $$ \sum\limits_{r=1}^R A_r X B_r $$ for one random $(d \times n)$-matrix $X$ with independent centered variance-one entries, and for two families of deterministic matrices $A_1,\dots,A_R \in \mathbb{C}^{d \times d}$ and $B_1,\dots,B_R \in \mathbb{C

Thijs Juan van Eeden

The IceCube Neutrino Observatory has opened a new window into the high-energy Universe, providing measurements of neutrinos over a broad energy range. This contribution presents recent results, including a follow-up on the first identification of a steady neutrino source NGC 1068, measurements of the flavor composition of the diffuse astrophysical flux, limi

Sungeun An, Swanand Ravindra Kadhe, Shailja Thakur, Chad DeLuca

Benchmarks are often used as a standard to understand LLM capabilities in different domains. However, aggregate benchmark scores provide limited insight into compositional skill gaps of LLMs and how to improve them. To make these weaknesses visible, we propose Scaffolded Task Design (STaD) framework. STaD generates controlled variations of benchmark tasks ba

Wentao Zhang, Yan Zhuang, ZhuHang Zheng, Mingfei Zhang

Existing jamming attacks on Retrieval-Augmented Generation (RAG) systems typically induce explicit refusals or denial-of-service behaviors, which are conspicuous and easy to detect. In this work, we formalize a subtler availability threat, termed soft failure, which degrades system utility by inducing fluent and coherent yet non-informative responses rather

Songxin Qu, Tai-Ping Sun, Yun-Jie Wang, Huan-Yu Liu

Large language models (LLMs) show strong capabilities in general reasoning but typically lack reliability in scientific domains like quantum mechanics, which demand strict adherence to physical constraints. This limitation arises from the scarcity of verifiable training resources and the inadequacy of coarse feedback signals in standard alignment paradigms.

Andrea Moiola, Nicola Galante, Emile Parolin

Classical Trefftz methods approximate Helmholtz solutions using propagative plane waves and are subject to strong numerical instabilities. Evanescent plane wave bases can substantially mitigate this phenomenon. We propose a simple recipe to select such basis functions. We show that the numerical results obtained by the Ultraweak Variational Formulation (UWVF

Nazanin Davari, Rosa Valiante, Alessandro Trinca, Raffaella Schneider

Next generation detectors, such as LISA, LGWA, and ET will, for the first time, probe the high redshift Universe, offering unique insight into the birth, growth, and dynamics of the first black holes (BHs) during their earliest stages formation. We aim to predict merger rates and gravitational wave (GW) signatures of "cosmic" binary BHs, forming as a result

Nicholas A. Scoville

The complex of discrete Morse matchings $\M(K)$, introduced by Chari and Joswig, is a simplicial complex whose simplices are the acyclic matchings on the Hasse diagram of $K$. Its homotopy type is known in only a handful of cases. In this paper, we compute the homotopy types of $\M(\Delta^3)$ and $\M(\partial\Delta^3)$, the corresponding pure complexes $\M_{

Peinuan Qin, Justin Peng, Zhengtao Xu, Jiting Cheng

Non-native speakers (NNSs) frequently encounter speaking difficulties in multilingual communication, where existing approaches have shown promise in facilitating NNSs' comprehension and participation in real-time communication. However, they often overlook providing direct speaking support, where anxiety stemming from linguistic inadequacy and uncertain comm

Ran Zhang, Steffen Eger, Arda Tezcan, Wei Zhao

Large language models (LLMs) are increasingly used for creative tasks such as literary translation. Yet translational creativity remains underexplored and is rarely evaluated at scale, while source-text comprehension is typically studied in isolation, despite the fact that, in professional translation, comprehension and creativity are tightly intertwined. We

Chenxi Zhao, Chen Zhu, Xiaokun Feng, Aiming Hao

Few-step generation has been a long-standing goal, with recent one-step generation methods exemplified by MeanFlow achieving remarkable results. Existing research on MeanFlow primarily focuses on class-to-image generation. However, an intuitive yet unexplored direction is to extend the condition from fixed class labels to flexible text inputs, enabling riche

Venkatesh Thirugnana Sambandham, Torsten Schön

Modern text-to-image (T2I) models amplify harmful societal biases, challenging their ethical deployment. We introduce an inference-time method that reliably mitigates social bias while keeping prompt semantics and visual context (background, layout, and style) intact. This ensures context persistency and provides a controllable parameter to adjust mitigation

Zongyao Zhao, Zhaolin Wang, Lincong Han, Liang Xu

Near-field integrated sensing and communication (ISAC) requires target models beyond the point-target abstraction when the target has a non-negligible spatial extent. In this letter, a geometry-aware transmit design is developed for a parametric extended target (ET) described by its center, orientation, and size under spherical-wave propagation. The CRB for

Sua Lee, Sanghee Park, Jinbae Im

Multimodal Large Language Models (MLLMs) have been increasingly used as automatic evaluators-a paradigm known as MLLM-as-a-Judge. However, their reliability and vulnerabilities to biases remain underexplored. We find that many MLLM judges fail to reliably integrate key visual or textual cues, yielding unreliable evaluations when evidence is missing or mismat

Aman Rojjha, Gaurang Tandon, Varul Srivastava, Kannan Srinathan

Electronic voting systems must balance public verifiability with voter privacy and coercion resistance. Existing cryptographic protocols typically achieve end-to-end verifiability by revealing vote distributions, relying on trusted clients, or enabling transferable receipts - design choices that often compromise trust or privacy in real-world deployments. We

Yan Tan, Tong Liu, Xiangchen Meng, Yangdi Lyu

Large Language Models (LLMs) have recently achieved strong performance in software code generation. However, applying them to hardware description languages (HDLs), such as Verilog, remains challenging because high-quality training data are relatively scarce. In practice, LLM-generated Verilog often contains syntactic or structural errors that either cause c

Ku Onoda, Paavo Parmas, Manato Yaguchi, Yutaka Matsuo

In policy gradient reinforcement learning, access to a differentiable model enables 1st-order gradient estimation that accelerates learning compared to relying solely on derivative-free 0th-order estimators. However, discontinuous dynamics cause bias and undermine the effectiveness of 1st-order estimators. Prior work addressed this bias by constructing a con

Jacek Gosciniak

General-purpose programmable photonic processors are considered a crucial technology because they combine the ultra high-speed, massive bandwidth, and energy efficiency of light-based computing with the flexibility of software-defined hardware. Unlike application-specific photonic integrated circuits (ASPIC) designed for one task, these processors use reconf

Yun Hong, Yan Zhou, Yang Feng

Empathy is essential for fostering natural interactions in spoken dialogue systems, as it enables machines to recognize the emotional tone of human speech and deliver empathetic responses. Recent research has made significant progress in developing empathetic spoken chatbots based on large language models (LLMs). However, several challenges still exist when

Yanzhen Lu, Zhicheng Qian, Muchen Jiang, Xingyu Zhou

Prompt-based interventions can change model behavior, but trained success alone does not identify where the behaviorally relevant state is represented. We study this question in controlled routing tasks using interfaces chosen on support data, held-out query evaluation, and matched necessity, sufficiency, and wrong-interface controls. On GPT-2 triop, an earl

Štěpán Ondřej, Jiří Spurný

Let X be an L1-predual and E,F be Banach spaces. We use the fact that an unconditionally converging operator T from the injective tensor product of X and E to F is strongly bounded and extend T to an operator S on continuous F-valued functions on the dual unit ball of X with the preservation of properties of T. This procedure provides a unified approach for

Heyou Liu, Bang Huang, Mohamed-Slim Alouini

Space--air--ground integrated networks (SAGINs) are emerging as a key foundation for future non-terrestrial networks (NTNs) and low-altitude economy services. However, their performance is increasingly limited not only by communication resources, but by the inability to adapt to rapidly changing spatial geometry. Here, spatial geometry refers to the relative

Chuan Zhang, Eduard Boos, Roland Boehm, Ramy Cherif

The development of Ultra-High Frequency (UHF) linear accelerators via Metal Additive Manufacturing (MAM) is a strategic research focus of the RACERS team at GSI. The 704.4 MHz Crossbar H-mode (CH) cavity, proposed in 2021 to facilitate efficient frequency jumps and downsize accelerator footprints, represents both the highest-frequency CH structure to date an

Peinuan Qin, Jiting Cheng, Jungup Lee, Junti Zhang

Cyberbullying is a pervasive problem in online environments, causing substantial psychological harm to victims. Although bystander intervention has proven effective in mitigating its impact, motivating bystanders to engage in direct intervention remains a persistent challenge. Studies have suggested that difficulties in intervention skills and defending self

Sebastian Fischer, Lukas Burk, Carson Zhang, Bernd Bischl

Deep learning (DL) has become a cornerstone of modern machine learning (ML) praxis. We introduce the R package mlr3torch, which is an extensible DL framework for the mlr3 ecosystem. It is built upon the torch package, and simplifies the definition, training, and evaluation of neural networks for both tabular data and generic tensors (e.g., images) for classi

Steffen Knoblauch, Levi Szamek, Iddy Chazua, Benedcto Adamu

Urban flooding is a growing climate change-related hazard in rapidly expanding African cities, where inadequate waste management often blocks drainage systems and amplifies flood risks. This study introduces an AI-powered urban waste mapping workflow that leverages openly available aerial and street-view imagery to detect municipal solid waste at high resolu

Murilo Macedo, Carlos Alexandre Wuensche, Oswaldo Duarte Miranda

This article uses an adapted version of the semi-analytical model of cosmic chemical enrichment developed by \citet{Corazza_2022} to reproduce the observed abundances of C, N, and O in absorption systems of quasar spectra (ASQS) at $z \gtrsim 3-6$, addressing an overproduction issue of the abovementioned elements. We address this discrepancy by updating the

Ikumi Banno

In the past few years, data informativity with prior knowledge has attracted increasing attention. This line of research aims to characterize a dataset on a dynamical system that enables system analysis or design only by the dataset and given prior knowledge on the system. In this paper, we investigate such a characterization for the data-driven problem of c

Ammar Bhilwarawala, Mainak Bandyopadhyay

Automated fetal head segmentation in ultrasound images is critical for accurate biometric measurements in prenatal care. While existing deep learning approaches have achieved a reasonable performance, they struggle with issues like low contrast, noise, and complex anatomical boundaries which are inherent to ultrasound imaging. This paper presents Attention-R

Michael T. M. Emmerich

We investigate \emph{magnitude} as a new unary and strictly Pareto-compliant quality indicator for finite approximation sets to the Pareto front in multiobjective optimization. Magnitude originates in enriched category theory and metric geometry, where it is a notion of size or point content for compact metric spaces and a generalization of cardinality. For