April 2026 arXiv papers — page 81

Juliusz Wasieleski, Tomasz Służalec, Maciej Woźniak, Marcin Łoś

We develop a wildfire simulation model that evolves the temperature scalar field using an energy balance equation accounting for heat generation, transport, and loss. For these equations, we develop quasi-implicit time integration schemes using direction splitting of the differential operators. We use the Peaceman-Rachford and Strang splitting methods, inclu

Philipp Weigand, Niels Nawrot, Nikolas Ebert, Carsten Hopf

Peak picking is a fundamental preprocessing step in Mass Spectrometry Imaging (MSI), where each sample is represented by hundreds to thousands of ion images. Existing approaches require careful dataset-specific hyperparameter tuning, and often fail to generalize across acquisition protocols. We introduce IonMorphNet, a spatial-structure-aware representation

Ghadah Alosaimi, Hanadi Alhamdan, Wenke E, Stamos Katsigiannis

Predicting driver intention from neurophysiological signals offers a promising pathway for enhancing proactive safety in advanced driver assistance systems, yet remains challenging in real-world driving due to EEG signal non-stationarity and the complexity of cognitive-motor preparation. This study proposes and evaluates an EEG-based driver intention predict

J. Larsson, C. Tegkelidis, C. Fransson, P. Lundqvist

The young supernova remnant SNR 0540-69.3 in the Large Magellanic Cloud offers a detailed view of an energetic pulsar-wind nebula interacting with the surrounding ejecta. We present infrared observations of the central region of SNR 0540-69.3 obtained with the JWST NIRSpec and MRS integral field units. From the observations we reconstruct the 3D morphology o

Konrad Szocik, Abraham Loeb

Recent work on the Loeb Scale has provided astronomy a structured framework for assessing anomalous interstellar objects, including a quantitative mapping of a classification ranking, its evolution with the addition of data, and a broader observational strategy for firming its verdict. What remains unclear is the epistemic and methodological meaning of the t

Mathias Ibsen, Loris Tim Ide, Christian Rathgeb, Christoph Busch

Face recognition systems are often used for biometric authentication. Nevertheless, it is known that without any protective measures, face recognition systems are vulnerable to presentation attacks. To tackle this security problem, methods for detecting presentation attacks have been developed and shown good detection performance on several benchmark dataset

Rafael B. Andrist, Włodzimierz Zwonek

We show that pseudoconvex Reinhardt domains in dimension two with isomorphic semigroups of holomorphic endomorphisms are biholomorphically or anti-biholomorphically equivalent. Moreover, we show that every Stein manifold that retracts to a properly embedded copy of the punctured complex line, is determined (up to biholomorphic or anti-biholomorphic equivalen

Lakshani Manamperi, Disumi Pathirana, Thiwanka Pathirana, Nipun Premarathna

Mobile Crowd Computing (MCdC) leverages the idle computational capacity of consumer smartphones to enable distributed task processing at scale; however, widespread real-world adoption remains constrained by the absence of developer-oriented frameworks capable of transparently managing device heterogeneity, fault tolerance, and connectivity volatility. This p

Remi Blandin, Martin Laabs, Rudolf von Bunau, Bryn Lloyd

This study experimentally validates a numerical model of electromagnetic propagation through the human head during the pronunciation of different vowels, with the goal of improving our understanding of the underlying physical phenomena. A realistic finite element model was created from magnetic resonance images acquired while pronouncing the vowels /a/, /i/,

R. -L. Ballouz, A. J. Ryan, R. J. Macke, O. S. Barnouin

Owing to the extremely low gravity of small near-Earth asteroids (NEAs), it has been assumed that impact-generated rock fragments escape into space and thus do not contribute to the accumulation of regolith. However, centimeter-sized stones returned from the small NEA Bennu by NASA's OSIRIS-REx mission exhibit impact craters up to a few millimeters wide, imp

Stefan Hoffelner

Assuming the consistency of $\mathsf{ZFC}$, we construct a model of set theory in which the boldface $\mathbf{\Sigma}^1_3$-uniformization property holds, yet the lightface $\Sigma^1_4$-uniformization property fails, separating these two principles for the first time. We also indicate how to create a universe where $\Sigma^1_3$-uniformization holds, but $\Sig

Shaun Hargreaves Heap, Mehmet Mars Seven

We study whether zero-sum decision rules, maximin and minimax, harm agents' interests in positive-sum games relative to Nash equilibrium behaviour or, more generally, than best response behaviour. Contrary to an influential evolutionary view, we give illustrations where maximin serves an agent's interests better than Nash equilibrium behaviour. Two n

Daomin Cao, Junhong Fan, Guolin Qin

We prove that for solutions of the Euler equation on the sphere, the vorticity gradient can grow at most double-exponentially in time, and we show that this upper bound is sharp by constructing explicit solutions with odd symmetry that exhibit double-exponential growth in the hemisphere. We also extend the results to the case of a rotating sphere. This seems

Rudolf Debelak

The evaluation of machine learning models typically relies mainly on performance metrics based on loss functions, which risk to overlook changes in performance in relevant subgroups. Auditing tools such as SliceFinder and SliceLine were proposed to detect such groups, but usually have conceptual disadvantages, such as the inability to directly address contin

Mengying Guo, Zhenyu Zhou, Denys Slobodianiuk, Roman Verba

True random number generators (TRNGs) underpin modern cryptography, yet existing implementations face fundamental trade-offs between speed, scalability, and entropy quality. Here, we demonstrate that stochastic switching in the bistable regime of spin-wave dynamics provides a physical entropy source for high-quality random number generation. Our magnonic ran

Ali Al-Kaswan, Maksim Plotnikov, Maxim Hájek, Roland Vízner

Large Language Model (LLM) agents are increasingly proposed for autonomous cybersecurity tasks, but their capabilities in realistic offensive settings remain poorly understood. We present DeepRed, an open-source benchmark for evaluating LLM-based agents on realistic Capture The Flag (CTF) challenges in isolated virtualized environments. DeepRed places an age

Rohit Chaudhuri

This article discusses the application of stochastic intervention to find the optimal treatment distribution yielding a high value of expected potential outcome under the setting where the number of treatments is allowed to vary with $n$. The primary motivation is to obtain a novel summarization of the effect of various treatments which would guide practitio

Jinyu Guo, Zhihan Zhang, Jiehui Xie, Md. Tamim Iqbal

The quadratic computational complexity of the standard attention mechanism constitutes a fundamental bottleneck for large language models in long-context inference. While existing KV cache compression methods alleviate memory pressure, they often sacrifice generation quality and fail to address the high overhead of floating-point arithmetic. This paper intro

Samyak Sanghvi, Piyush Miglani, Sarvesh Shashikumar, Kaustubh R Borgavi

Vision Transformers $(\texttt{ViT})$ have become the architecture of choice for many computer vision tasks, yet their performance in computer-aided diagnostics remains limited. Focusing on breast cancer detection from mammograms, we identify two main causes for this shortfall. First, medical images are high-resolution with small abnormalities, leading to an

Xunpei Sun, Zuoxun Hou, Yi Chang, Gang Chen

Monocular scene flow estimation aims to recover dense 3D motion from image sequences, yet most existing methods are limited to two-frame inputs, restricting temporal modeling and robustness to occlusions. We propose RAFT-MSF++, a self-supervised multi-frame framework that recurrently fuses temporal features to jointly estimate depth and scene flow. Central t

Reinaldo Francener, Victor P. Goncalves

Neutrino trident scattering is a rare process in the Standard Model characterized by two charged leptons in the final state. In this work, we investigate the possibility of probing the neutrino trident process using the Scattering and Neutrino Detector (SND) at the Large Hadron Collider during its high - luminosity run (HL - LHC). In addition, we present, fo

Thai Duong Do, Hoang Hiep Pham

In this paper, we introduce a notion of singularity comparison for plurisubharmonic functions based on the Bedford--Taylor capacity. We establish comparison principles for the complex Monge--Amp\`ere operator on pluripolar sets in the Cegrell classes. As applications, we obtain a characterization of this relation via auxiliary functions in the energy class a

Patrick Koppenburg

The next 20 years will be the golden age of flavour physics, with the operation of the LHCb and Belle II experiments. After that an $e^+e^-$ collider could further improve the precision with sizeable $Z$, $W^+W^-$ and $t\bar{t}$ runs.

Jiamin Chang, Minhui Xue, Ruoxi Sun, Shuchao Pang

Recent advances in embodied Vision-Language Agentic Systems (VLAS), powered by large vision-language models (LVLMs), enable AI systems to perceive and reason over real-world scenes. Within this context, environmental signals such as traffic lights are essential in-band signals that can and should influence agent behavior. However, similar signals could also

Michael Ziegltrum, Jianhao Jiao, Tianhu Peng, Chengxu Zhou

Robotic parkour provides a compelling benchmark for advancing locomotion over highly challenging terrain, including large discontinuities such as elevated steps. Recent approaches have demonstrated impressive capabilities, including dynamic climbing and jumping, but typically rely on sequential multilayer perceptron (MLP) architectures with densely activated

Coşku Can Horuz, Andrea Ceni, Claudio Gallicchio, Sebastian Otte

Memristive devices present a promising foundation for next-generation information processing by combining memory and computation within a single physical substrate. This unique characteristic enables efficient, fast, and adaptive computing, particularly well suited for deep learning applications. Among recent developments, the memristive-friendly echo state

Abdullah Mohammad, Sushant Kumar Ray, Pushkar Arora, Rafiq Ali

Generative AI-powered by Large Language Models (LLMs)-is increasingly deployed in industry across healthcare decision support, financial analytics, enterprise retrieval, and conversational automation, where reliability, efficiency, and cost control are critical. In such settings, models must satisfy strict constraints on energy, latency, and hardware utiliza

Kazuma Iwase, Tomoyuki Takenawa

This study aims to develop and improve machine learning-based post-processing models for precipitation, temperature, and wind speed predictions using the Mesoscale Model (MSM) dataset provided by the Japan Meteorological Agency (JMA) for 18 locations across Japan, including plains, mountainous regions, and islands. By incorporating meteorological variables f

Shijie Wang, Zijian Wang, Yadan Luo, Haojie Li

Ultra-fine-grained visual categorization (Ultra-FGVC) aims to classify highly similar subcategories within fine-grained objects using limited training samples. However, holistic yet discriminative cues, such as leaf contours in extremely similar cultivars, remain under-explored in current studies, thereby limiting recognition performance. Though crucial, mod

Shanshan Zhang, Songjie Yang, Wenxuan Zhang, Youzhi Xiong

This study investigates subarray-level movable antenna (MA) architecture for multi-user MIMO (MU-MIMO) systems. Unlike conventional systems with fixed-position antennas (FPAs), the proposed scheme harnesses the additional positional degrees of freedom (DoFs) of movable subarrays to enhance spatial multiplexing capabilities for both multi-user and multi-strea

Yizhuo Rao, Xingjian Cui, Shangzhi Pang, Jiabin Xie

Particle-in-Cell (PIC) simulations are fundamental to plasma physics but often suffer from limited scalability due to particle-grid interaction bottlenecks and particle redistribution costs. Specifically, the particle-grid interaction computations have not taken full advantage of the emerging Matrix Processing Units (MPUs), the particle motion introduces irr

Harekrushna Sahu, Pratik Jawanpuria, Pranay Sharma

Online federated learning (OFL) has emerged as a popular framework for decentralized decision-making over continuous data streams without compromising client privacy. However, the adversary model assumed in standard OFL typically precludes any potential benefits of parallelization. Further, it fails to adequately capture the different sources of statistical

Simin Yu, Sufia Fathima

The rapid growth of chemical literature has generated vast amounts of unstructured data, where reaction information is particularly valuable for applications such as reaction predictions and drug design. However, the prohibitive cost of expert annotation has led to a scarcity of training data, severely hindering the performance of automatic reaction extracti

Sebastian Fieldhouse, Kea-Tiong Tang

Recent work in the machine learning literature has demonstrated that deep learning can train neural networks made of discrete logic gate functions to perform simple image classification tasks at very high speeds on CPU, GPU and FPGA platforms. By virtue of being formed by discrete logic gates, these Differentiable Logic Gate Networks (DLGNs) lend themselves

Pierpaolo Fontana, Luca Lepori, Andrea Trombettoni

The three-dimensional Hofstadter model exhibits a critical rational flux at which Weyl points emerge in the single-particle spectrum. We study the superconducting regime of the model in the presence of a Hubbard attractive interaction by tuning the magnetic flux across its critical value. We determine the phase diagram in the plane of the coprime pairs param

Jianbo Ma, Richard Cartwright

Recent advances in Text-To-Speech (TTS) synthesis have seen the popularity of multi-stage approaches that first predict semantic tokens and then generate acoustic tokens. In this paper, we extend the coarse-to-fine generation paradigm to the temporal domain and introduce Chain-of-Details (CoD), a novel framework that explicitly models temporal coarse-to-fine

Siyuan Huang, Shuaishuai Sun, Yin Shi, Wentao Wang

The insulator-to-metal transition (IMT) in strongly correlated materials, such as vanadium dioxide (VO2), offers a transformative platform for next-generation adaptive electronics and neuromorphic computing. However, harnessing this non-equilibrium phase transition for deterministic device operation is fundamentally hindered by the inability to disentangle e

Fernando de Juan, Mark H. Fischer

The kagome superconductor CsV$_3$Sb$_5$ is the only one in the AV$_3$Sb$_5$ family (A=K,Rb,Cs) that shows a $2\times2\times4$ charge-density-wave (CDW) ground state competing with the more common $2\times2\times2$. In addition, it is also the only one that shows second-harmonic transport and thus broken inversion symmetry, suggesting these two features are c

Patrick Grenzer, Fabian Lie, Klaus H. Eckstein, Tobias Hertel

Two-dimensional halide perovskites host strongly bound excitons whose fine structure controls polarization selection rules and radiative recombination, yet several spectral features in (PEA)$_2$PbI$_4$ remain controversially assigned. Here, polarization-resolved low-temperature photoluminescence combined with first-principles G$_0$W$_0$+BSE calculations reso

Babak Mohammadikalakoo, Sergio Garcia Villasol, Gabriele Salomone, Marios Kotsonis

An experimental wind-tunnel implementation of a model-free single-step deep reinforcement learning (SDRL) controller is presented for TS wave suppression in a flat plate boundary layer. The controller is deployed in a feedforward layout. The arrangement comprises an upstream reference microphone, a downstream error microphone, and a DBD plasma actuator locat

Tommi Kerola, Yuya Masuda, Takashi Masuko, Toshiki Nakanishi

We introduce PLaMo 2.1-VL, a lightweight Vision Language Model (VLM) for autonomous devices, available in 8B and 2B variants and designed for local and edge deployment with Japanese-language operation. Focusing on Visual Question Answering (VQA) and Visual Grounding as its core capabilities, we develop and evaluate the models for two real-world application s

Gonzalo Nápoles, Isel Grau, Yamisleydi Salgueiro

Concept Bottleneck Models (CBMs) route predictions exclusively through a clinically grounded concept layer, binding interpretability to concept-label consistency. When a dataset contains concept-level inconsistencies, identical concept profiles mapped to conflicting diagnosis labels create an unresolvable bottleneck that imposes a hard ceiling on achievable

Huanchen Cai, Sten Ternström

This study investigates voice mapping as an evaluation framework for text-to-speech (TTS) synthesis quality. The study analyzes six TTS models, including historical and recent ones. The metrics are crest factor, spectrum balance, and cepstral peak prominence (CPPs). We investigated 6 influential TTS models: Merlin, Tacotron 2, Transformer TTS, FastSpeech 2,

Thomas J. Antolin, Jonas Glatthard, Andrew D. Armour

Recent experiments on a range of engineered quantum systems have highlighted the important role of interacting two-level systems (TLSs) in modifying device properties and generating fluctuations. Focusing on the case of an oscillator coupled to a single near-resonant TLS, we explore how interactions between the TLS and lower-frequency thermally activated two

Yusuf Çelebi, Yağız Asker, Özay Ezerceli, Mahmoud ElHussieni

Fine-tuning Large Language Models (LLMs) remains structurally uncertain despite parameter-efficient methods such as Low-Rank Adaptation (LoRA), as the layer-specific roles of internal representations are poorly understood, leading to heuristic decisions about where adaptation should be applied. We model the evolution of hidden states as a high-dimensional ge

Jaemin Seo

Quantum neural networks (QNNs) suffer from a fundamental sampling bottleneck since quantum measurements are probabilistic, requiring many circuit executions to estimate outputs with sufficient accuracy. Conventional Monte-Carlo (MC) inference exhibits an $\mathcal{O}(1/\sqrt{N})$ sampling error, rendering QNN inference and training costly on near-term quantu

Joshua Courtney

Quantum computation is an attractive front for many problems that are intractable for computers today. One such problem is nonadiabatic quantum molecular dynamics, where quantized internal states coupling to parameterized modes result in a Hamiltonian resistant to oracle-based models and spectral decomposition. This dissertation applies diabatic Hamiltonian

Qi Zhang, Jixuan Chen, Kaiyi Zhang, Xinquan Yu

Multi-view crowd tracking estimates each person's tracking trajectories on the ground of the scene. Recent research works mainly rely on CNNs-based multi-view crowd tracking architectures, and most of them are evaluated and compared on relatively small datasets, such as Wildtrack and MultiviewX. Since these two datasets are collected in small scenes and only

Matthew Nicol, Manpreet Singh, Andrew Torok

We investigate the competition between two distinct mechanisms generating stable laws in deterministic dynamical systems: slow mixing of the system and heavy-tailed observables. For heavy-tailed observables on polynomially mixing billiards with cusps we show these two mechanisms interact and there is a transition, depending on the mixing exponent and the ind

Kentaro Kasai, Masahiro Kawasaki, Kai Murai, Shunsuke Neda

Primordial black holes (PBHs) may form an initially clustered population depending on their production mechanism. Motivated by binary black-hole merger events observed by gravitational-wave interferometers, we revisit the evaluation of the merger rate of PBH binaries and extend the formalism to include the effects of clustering. We show that, in the presence

Heng Zhang, Reza Parvaz, Rui Yang

Recovering corrupted images is one of the most challenging problems in image processing. Among various restoration tasks, blind image deblurring has been extensively studied due to its practical importance and inherent difficulty. In this problem, both the point spread function (PSF) and the underlying latent sharp image must be estimated simultaneously. Thi

Drew Heard

For $G$ a finite group and $T$ a $G$-Tambara functor, we construct the frame $\mathop{RadId}_G(T)$ of radical Tambara ideals and show that its points are the Nakaoka primes. We show that this frame is spatial and coherent, and deduce that the Nakaoka spectrum is a spectral space, recovering a recent result of Chan and Spitz.

Robin Young

Neural processes are meta-learning models that map context sets to predictive distributions. While inspired by stochastic processes, NPs do not generally satisfy the Kolmogorov consistency conditions required to define a valid stochastic process. This inconsistency is widely acknowledged but poorly understood. Practitioners note that NPs work well despite th

Salvatore Capozziello, Dario Sauro

We analyze the properties of foliations in presence of non-metricity, deriving the generalized Gauss-Codazzi relations in full generality. These results are employed to study the teleparallel framework of non-metric geometry, obtaining constraints on the extrinsic and intrinsic tensors. In particular, an extrinsic symmetric two-tensor plays the role of the e

Athikash Jeyaganthan, Kai Xu, Franziska Becker, Steffen Koch

Qualitative coding relies on a researcher's application of codes to textual data. As coding proceeds across large datasets, interpretations of codes often shift (temporal drift), reducing the credibility of the analysis. Existing Computer-Assisted Qualitative Data Analysis (CAQDAS) tools provide support for data management but offer no workflow for real-time

Alexander Kalinin, Thilo Meyer-Brandis, Annika Steibel

We present a tractable class of one-dimensional McKean-Vlasov equations that allow for unique strong solutions and extend the dynamics of various SIS epidemic models that are well-established in the literature. While the distribution-dependent drift coefficients are of polynomial type, the diffusion coefficients may involve sums of power functions. Our analy

Mohamad N. Nasser, Oscar Ocampo

We introduce linear representations of the universal virtual braid group $UV_n(c)$, where $n\geq 2$ and $c\geq 1$, which is a unifying framework for braid-type groups with multiple types of crossings. We classify and study its complex homogeneous $2$-local representations for all $n\geq 3$ and $c\geq 1$ (unique up to equivalence) and complex homogeneous $3$-

Alisha Gurung, Subir Mukhopadhyay

We analyze instabilities of an Einstein-Maxwell theory obtained from an N=2, D=5 supergravity. The theory admits a gauge Chern-Simons term in presence of which we consider perturbative instability of an AdS black hole solution. We find that the strength of the gauge Chern-Simons coupling saturates the threshold value for which the instability occurs, which w

Tao Zhang, Hanshu Chen, Ilia Marchevsky, Zhuojia Fu

The aim of this paper is to introduce a Mapping-based Hard-constrained Physics-Informed Neural Network (MH-PINN) for efficiently and accurately solving unbounded wave problems. First, we propose a coordinate mapping technique that compactifies the infinite physical domain into a finite computational space. This effectively resolves the sampling difficulties

Linhao Wu, Yifei Pei, Zhen Yang, Kainan Li

Automated Program Repair (APR) has benefited from the code understanding and generation capabilities of Large Language Models (LLMs). Existing feedback-based APR methods iteratively refine candidate patches using test execution feedback and have shown promising results. However, most rely on outcome-level failure symptoms, such as stack traces, which show ho

Arturo Tozzi

The onset of life is often framed around membrane bound compartments and encoded metabolism, leaving unresolved how spatial organization arose before stable boundaries. In this context, environmental gradients are usually treated as boundary conditions rather than variables structuring chemical dynamics. We ask whether spatial localization and functional cou

Ji Zou, Jelena Klinovaja, Daniel Loss

Magnetic domain walls have long been pursued as carriers of classical information for storage and processing. With the ability to create, control, and probe domain walls at the nanoscale, they are recently recognized as an ideal platform for studying macroscopic quantum effects and provide a natural blueprint for building scalable quantum computing architect

Leonardo F. Cavenaghi, Lino Grama, Ludmil Katzarkov, Pedro Antonio Muniz Martins

This paper investigates the geometric and cohomological properties of non-K\"ahler SYZ mirror symmetry for dual torus fibrations over solvmanifolds in the sense of Lau, Tseng and Yau. We are mainly concerned with three questions: \textbf{(a)} How the Lau-Tseng-Yau notion of non-K\"ahler SYZ is related to the mapping of supersymmetric branes between symplecti

A. Schmaus, N. Marwan, N. Molkenthin

Trajectories of units moving on networks are relevant for nonlinear dynamical systems as diverse as polymers, ocean drifters, and human mobility. Although RQA is a well-researched tool with applications in many areas, it has rarely been used for spatial trajectories on networks. Here, we explore the use of RQA for paths on networks. We find that path dynamic

Mikako Bito, Keita Nishimoto, Kimitaka Asatani, Ichiro Sakata

The conformity bias exhibited by large language models (LLMs) can pose a significant challenge to decision-making in LLM-based multi-agent systems (LLM-MAS). While many prior studies have treated "conformity" simply as a matter of opinion change, this study introduces the social psychological distinction between informational conformity and normative conform

Feiyu Zhao, Yiming Chen, Wenhuan Lu, Daipeng Zhang

Large Audio-Language Models (LALMs) have recently achieved strong performance across various audio-centric tasks. However, hallucination, where models generate responses that are semantically incorrect or acoustically unsupported, remains largely underexplored in the audio domain. Existing hallucination benchmarks mainly focus on text or vision, while the fe

Xinlin Wang, Mats Brorsson

Despite the impressive capabilities of large language models, their substantial computational costs, latency, and privacy risks hinder their widespread deployment in real-world applications. Small Language Models (SLMs) with fewer than 10 billion parameters present a promising alternative; however, their inherent limitations in knowledge and reasoning curtai

Rajveer Singh Pall

We introduce IndiaFinBench, to our knowledge the first publicly available evaluation benchmark for assessing large language model (LLM) performance on Indian financial regulatory text. Existing financial NLP benchmarks draw exclusively from Western financial corpora (SEC filings, US earnings reports, English-language financial news), leaving a significant ga

Hasna Chnafa, Clarence Cortes, David Laroze, Ahmed Jellal

Electron transport in graphene under a laser-modulated barrier is studied in the presence of an energy gap, a scalar potential, and a uniaxial zigzag strain. The transfer-matrix approach is used with the boundary conditions to derive the transmission probabilities as functions of different system parameters. Without strain, raising either the energy gap or t

Konstantin Hess, Dennis Frauen, Niki Kilbertus, Stefan Feuerriegel

Neural operators are widely used to approximate solution maps of complex physical systems. In many applications, however, the goal is not to recover the full solution trajectory, but to summarize the solution trajectory via a scalar target quantity (e.g., a functional such as time spent in a target range, time above a threshold, accumulated cost, or total en

Qingyang Zhang, Xinke Kong, Haitao Wu, Qinghua Hu

Test-time training (TTT) adapts model parameters on unlabeled test instances during inference time, which continuously extends capabilities beyond the reach of offline training. Despite initial gains, existing TTT methods for LRMs plateau quickly and do not benefit from additional test-time compute. Without external calibration, the self-generated reward sig

Brayden Letwin, Mehtaab Sawhney

A Littlewood polynomial is a polynomial of the form \[ f_n(x)=\sum_{k=0}^n \varepsilon_k x^k \] with $\varepsilon_k\in\{-1, 1\}$. Let $(\varepsilon_k)_{k \ge 0}$ be i.i.d. Rademacher coefficients. We show that the lower envelope of $\max_{x\in[-1,1]}|f_n(x)|$ is determined by the small-ball probability of a certain Gaussian process. In particular, almost sur

Chao Qian, Tianheng Ling, Gregor Schiele

Long Short-term Memory Networks (LSTMs) are a vital Deep Learning technique suitable for performing on-device time series analysis on local sensor data streams of embedded devices. In this paper, we propose a new hardware accelerator design for LSTMs specially optimised for resource-scarce embedded Field Programmable Gate Arrays (FPGAs). Our design improves

Guy Mor-Lan, Omer Goldman, Matan Eyal, Adi Mayrav Gilady

Multilingual large language models (LLMs) have minimized the fluency gap between languages. This advancement, however, exposes models to the risk of biased behavior, as knowledge and norms may propagate across languages. In this work, we aim to quantify models' inter- and intra-lingual biases, via their ability to answer locale-ambiguous questions. To this e

Rudy Arthur

Network community detection is usually considered as an unsupervised learning problem. Given a network, the aim is to partition it using some general purpose algorithm. In this paper we instead treat community detection as a hypothesis testing problem. Given a network, we examine the evidence for specific community structure in the observed network compared

Robert Flassig, Emrah Gülay, Daniel Guterding

The Nelson-Siegel-Svensson (NSS) interest rate curve model yields a separable nonlinear least-squares problem whose inner linear block is often ill-conditioned because the basis functions become nearly collinear. We analyze this instability via an exact orthogonal reparametrization of the design matrix. A thin QR decomposition produces orthogonal linear para

Benedetta Fantaci, Alejandro Frechilla, Matteo Frigelli, Philippe Büchler

Accurate assessment of corneal mechanical properties is critical for understanding ocular biomechanics, predicting refractive surgery outcomes, and optimizing cross-linking (CXL) treatments. Conventional uniaxial tensile test is limited by non-physiological boundary conditions and simplified stress distributions. Inflation testing more closely reproduces the

Eric Zimmermann, Stefan Bruckner

The Sierpinski triangle and the Sierpinski arrowhead curve are both defined in dimension 2 and can be used to model the same fractal. While a natural extension of the triangular construction to arbitrary dimensions exists, an analogous extension of the curve representation does not. In this article, we analyze the properties of the two-dimensional Sierpinski

Luca Pennati, Stefano Markidis

Matrix-multiply-accumulate (MMA) units, or tensor cores, are now widespread across modern computing architectures. Yet, their use for particle-grid operators remains limited. In implicit particle methods, mass-matrix assembly is a reduction-dominated kernel in which weighted outer products of interpolation weights are accumulated over particle support. We sh

Benedetta Tessa, Gautam Kishore Shahi, Amaury Trujillo, Stefano Cresci

During major political events, social media platforms encounter increased systemic risks. However, it is still unclear if and how they adjust their moderation practices in response. The Digital Services Act Transparency Database provides-for the first time-an opportunity to systematically examine content moderation at scale, allowing researchers and policyma

Kaoutar Bouaachra, Yvenn Amara-Ouali, Yannig Goude, Raphaël Lachieze-Rey

Accurate forecasting of electric vehicle (EV) charging demand is critical for grid management and infrastructure planning. Yet the field continues to rely on legacy benchmarks; such as the Palo Alto (2020) dataset; that fail to reflect the scale and behavioral diversity of modern charging networks. To address this, we introduce a novel large-scale longitudin

Jussi Behrndt, Petr Siegl, Nicolas Weber

We study the existence of negative eigenvalues for two-dimensional Schr\"odinger operators with real-valued potentials in the weak coupling regime. In his pioneering paper [Simon 1976] from half a century ago, Simon was the first to describe the unique negative eigenvalue emerging from the threshold of the essential spectrum of one- and two-dimensional Schr\

Kavita Yadav, Anoop M Divakaran, Tsunehiro Takeuchi, K. Mukherjee

Full Heusler alloys XInPd2 (X= Zr, Hf and Ti) have recently attracted significant attention owing to their symmetry-driven electronic structure and also due to the interplay between disorder and emergent ground states. Within this family, Pd2ZrIn serves as a unique platform to study the effect of disorder on superconducting pairing. This alloy crystallizes i

L. V. Bogdanov

We construct Orlov-Schulman symmetries for the self-dual conformal structure (SDCS) hierarchy. We provide an explicit proof of compatibility of additional symmetries with the basic Lax-Sato flows of the hierarchy, and consider several simple examples, including Galilean transformations and scalings. We also present a picture of the Orlov-Schulman symmetries

Abu Noman Md Sakib, Md. Main Oddin Chisty, Zijie Zhang

The use of Large Language Models (LLMs) to support patients in addressing medical questions is becoming increasingly prevalent. However, most of the measures currently used to evaluate the performance of these models in this context only measure how closely a model's answers match semantically, and therefore do not provide a true indication of the model's me

Jakob Schwerter, Amer Krivosija, Tim Novak, Katja Ickstadt

Early identification of at risk students in higher education depends on predictive models that maintain accuracy across successive cohorts -- a requirement that single-cohort modeling approaches fail to meet. This study evaluates Bayesian updating with informative priors from a previous cohort to improve cross-cohort prediction robustness using digital trace

Suhaib Abdurahman, Etsuko Ishii, Katerina Margatina, Divya Bhargavi

LLM-based multi-agent systems (MAS) show promise on complex tasks but remain prone to coordination failures such as goal drift, error cascades, and misaligned behaviors. We propose Explicit Trait Inference (ETI), a psychologically grounded method for improving coordination. ETI enables agents to infer and track partner characteristics along two established p

Fernando Carreño-Navas, Siannah Peñaranda, Renato Alvarez-Nodarse, Niurka R. Quintero

We derive an exact solitary wave solution for the $\PTb$-symmetric nonlinear Dirac equation with a scalar-scalar interaction. We consider a power-law nonlinearity of the form $|\bar{\Psi}\,\Psi|^{k}\,\Psi$ for positive values of $k$. The system's energy is conserved despite the presence of a gain-loss term, which is quantified by the parameter $\Lambda$. We

Sora Kang, Youjin Hwang, Joonhwan Lee

Intergenerational linguistic differences pose challenges to effective and intimate family communication. This paper presents GenSync, a chat-based interface that supports intergenerational understanding through different forms of translation visibility. We conducted a controlled within-subjects study with 16 family dyads (32 participants), comparing three co

Joe Asano, Yuto Hama, Hiroki Iimori, Szabolcs Malomsoky

In this letter, we propose a sparsification method for precoding codebooks that reduces the peak-to-average power ratio (PAPR) while preserving the achievable rate. By exploiting the fact that precoder matrices lie on the Grassmann manifold, we formulate a codebook design problem that enables sparsification without modifying the existing feedback mechanism.

Matěj Gajdoš, Ondřej Brichta, Václav Kučera

We study how inexact nonlinear solvers lead to a loss of exact symplecticity in the Symplectic Euler (SE) and Stormer-Verlet (SV) schemes when applied to general nonseparable Hamiltonian systems. These schemes are implicit and require nonlinear solvers in practice. Here, we consider a fixed number $M$ of fixed-point iterations (FPI). While SE is exactly symp

Jan Eube, Kelin Luo, Aneta Neumann, Frank Neumann

The Traveling Thief Problem (TTP) is a multi-component optimization problem that captures the interplay between routing and packing decisions by combining the classical Traveling Salesperson Problem (TSP) and the Knapsack Problem (KP). The TTP has gained significant attention in the evolutionary computation literature and a wide range of approaches have been

Genki Miyauchi, Roderich Groß, Chaona Chen

As groups of robots increasingly collaborate with humans, understanding how humans perceive them is critical for designing effective human-robot teams. While prior research examined how humans interpret and evaluate the abilities and intentions of individual agents, social perception of robot teams remains relatively underexplored. Drawing on the competence-

Lixiang Wang, Shaoyun Shi, Peng Wang, Wenjin Wu

To balance effectiveness and efficiency in recommender systems, multi-stage pipelines commonly use lightweight two-tower models for large-scale candidate retrieval. However, the isolated two-tower architecture restricts representation capacity, embedding-space alignment, and cross-feature interactions. Existing solutions such as late interaction and knowledg

Svetlana Pavlitska, Christopher Gerking, J. Marius Zöllner

Safety and security are essential for the admission and acceptance of automated and autonomous vehicles. Deep neural networks (DNNs) are widely used for perception and further components of the autonomous driving (AD) stack. However, they possess several limitations, including lack of generalization, efficiency, explainability, plausibility, and robustness.

Luis Fernando Cusicanqui Lopez, Ramadan Krasniqi, Florian Feppon, Karl Meerbergen

Density-based topology optimization has become a powerful method for automatically generating optimized designs in a wide variety of applications. However, it comes with a large computational cost when solving the physical model requires large-scale simulations. Here, we investigate the use of model order reduction (MOR) techniques to accelerate the simulati

Louis Dezons, Quentin Picard, Rémi Marsal, François Goulette

Goal-conditioned navigation models for ground robots trained using supervised learning show promising zero-shot transfer, but their collision-avoidance capability nevertheless degrades under distribution shift, i.e. environmental, robot or sensor configuration changes. We propose ViLiNT a multimodal, attention-based policy for goal navigation, trained on het

Ming Shan Loo, Wengen Li, Xudong Jiang, Hailiang Cheng

The reconstruction of ocean subsurface temperature (OST) using satellite remote sensing data holds significant scientific value for advancing the understanding of ocean dynamics and climate variability. However, the scarcity of subsurface observations, combined with the high degree of nonlinearity and spatiotemporal heterogeneity in subsurface processes, pos

Andrei Bulatov, Xiaoyang Gong, Bakh Khoussainov, Xinyao Wang

We study constraint satisfaction problems (CSPs) where the constraint languages are defined by finite automata, giving rise to automata-based CSPs. The key notion is the concept of Automatic Constraint Satisfaction Problem ($AutCSP$), where constraint languages and instances are specified by finite automata. The $AutCSP$ captures infinite yet finitely descri

Shengqin Wang, Wentao Yan, Huichi Zhou, Yihang Chen

Agentic multimodal models have garnered significant attention for their ability to leverage external tools to tackle complex tasks. However, it is observed that such agents often meet premature interaction collapse, caused by two primary reasons: 1) the terminal reward often appending on the last token prevents the advantage from distinguishing trajectories