May 2025 arXiv papers — page 74

Zherui Zhang, Jiaxin Wu, Changwei Wang, Rongtao Xu

Prompt learning as a parameter-efficient method that has been widely adopted to adapt Vision-Language Models (VLMs) to downstream tasks. While hard-prompt design requires domain expertise and iterative optimization, soft-prompt methods rely heavily on task-specific hard labels, limiting their generalization to unseen categories. Recent popular distillation-b

Zhihua Liu, Lei Tong, Xilin He, Che Liu

Existed echocardiography segmentation methods often suffer from anatomical inconsistency challenge caused by shape variation, partial observation and region ambiguity with similar intensity across 2D echocardiographic sequences, resulting in false positive segmentation with anatomical defeated structures in challenging low signal-to-noise ratio conditions. T

Anthony Fuller, Yousef Yassin, Junfeng Wen, Daniel G. Kyrollos

Vision transformers are ever larger, more accurate, and more expensive to compute. The expense is even more extreme at high resolution as the number of tokens grows quadratically with the image size. We turn to adaptive computation to cope with this cost by learning to predict where to compute. Our LookWhere method divides the computation between a low-resol

Diana Manvelyan-Stroot, Yevgeniya Filanova, Igor Pontes Duff, Peter Benner

In this paper, we propose an operator-inference-based reduction approach for contact problems, leveraging snapshots from simulations without active contact. Contact problems are solved using adjoint methods, by switching to the dual system, where the corresponding Lagrange multipliers represent the contact pressure. The Craig-Bampton-like substructuring meth

Gaole Dai, Menghang Dong, Rongyu Zhang, Ruichuan An

The process through which humans perceive and learn visual representations in dynamic environments is highly complex. From a structural perspective, the human eye decouples the functions of cone and rod cells: cones are primarily responsible for color perception, while rods are specialized in detecting motion, particularly variations in light intensity. Thes

Simon Malzard, Nitish Mital, Richard Walters, Victoria Nockles

Computer vision (CV) models for detection, prediction or classification tasks operate on video data-streams that are often degraded in the real world, due to deployment in real-time or on resource-constrained hardware. It is therefore critical that these models are robust to degraded data, but state of the art (SoTA) models are often insufficiently assessed

Sudarshan Rajagopalan, Kartik Narayan, Vishal M. Patel

The use of latent diffusion models (LDMs) such as Stable Diffusion has significantly improved the perceptual quality of All-in-One image Restoration (AiOR) methods, while also enhancing their generalization capabilities. However, these LDM-based frameworks suffer from slow inference due to their iterative denoising process, rendering them impractical for tim

H. Y. Yuan, Rembert A. Duine

A key challenge in spintronics is to efficiently generate and manipulate spin current for information processing. Here we study ultrashort spin transport and associated terahertz (THz) emission in a hybrid structure comprising gold nanoparticles, a ferromagnet (FM) and a normal metal (NM) and show that plasmon excitation in the nanoparticles strongly enhance

Youssouf Sidibé, Julia Gersey

In-home elderly monitoring requires systems that can detect emergency events - such as falls or prolonged inactivity - while preserving privacy and requiring no user input. These systems must be embedded into the surrounding environment, capable of capturing activity, and responding promptly. This paper presents a low-cost, privacy-preserving solution using

Zhishuai Liu, Pan Xu

Many real-world decision-making problems face the off-dynamics challenge: the agent learns a policy in a source domain and deploys it in a target domain with different state transitions. The distributionally robust Markov decision process (DRMDP) addresses this challenge by finding a robust policy that performs well under the worst-case environment within a

Changyeol Lee, Yongho Shin, Hyung-Chan An

Clustering is a fundamental task in both machine learning and data mining. Among various methods, edge-colored clustering (ECC) has emerged as a useful approach for handling categorical data. Given a hypergraph with (hyper)edges labeled by colors, ECC aims to assign vertex colors to minimize the number of edges where the vertex color differs from the edge's

Shuai Su, Xiurong Yan, Qian Zhang

We propose a novel parameter-free and locking-free enriched Galerkin (EG) method for solving the linear elasticity problem in both two and three dimensions. Unlike existing locking-free EG methods, our method enriches the first-order continuous Galerkin (CG) space with piecewise constants along edges in two dimensions or faces in three dimensions. This enric

Monica Barnard, Ankur Ghosh, Jagdish C. Joshi, Soebur Razzaque

The recent detection of very high energy (VHE, $\gtrsim$ 100 GeV) $\gamma$-ray emission from gamma-ray bursts (GRBs) has provided new insights into afterglow physics. Understanding the temporal and spectral evolution of VHE GRBs requires detailed modelling of multiwavelength observations spanning radio to VHE $\gamma$ rays. Previous studies interpreted after

Minxue Niu, Emily Mower Provost

The ability to handle various emotion labels without dedicated training is crucial for building adaptable Emotion Recognition (ER) systems. Conventional ER models rely on training using fixed label sets and struggle to generalize beyond them. On the other hand, Large Language Models (LLMs) have shown strong zero-shot ER performance across diverse label space

Li Zhong, Ahmed Ghazal, Jun-Jun Wan, Frederik Zilly

Foundation models like CLIP (Contrastive Language-Image Pretraining) have revolutionized vision-language tasks by enabling zero-shot and few-shot learning through cross-modal alignment. However, their computational complexity and large memory footprint make them unsuitable for deployment on resource-constrained edge devices, such as in-car cameras used for i

Vincent Jeanselme, Marco Palma, Jessica K Barrett

Linear Mixed Model (LMM) is a common statistical approach to model the relation between exposure and outcome while capturing individual variability through random effects. However, this model assumes the homogeneity of the error term's variance. Breaking this assumption, known as homoscedasticity, can bias estimates and, consequently, may change a study's co

Khanh-Hung Giang-Tran, Soroosh Shafiee, Nam Ho-Nguyen

We propose efficient methods for solving stochastic simple bilevel optimization problems with convex inner levels, where the goal is to minimize an outer stochastic objective function subject to the solution set of an inner stochastic optimization problem. Existing methods often rely on costly projection or linear optimization oracles over complex sets, limi

Annabel L Davies

Network meta-analysis (NMA) combines evidence from multiple trials comparing treatment options for the same condition. The method derives its name from a graphical representation of the data where nodes are treatments, and edges represent comparisons between treatments in trials. However, edges in this graph are limited to pairwise comparisons and fail to re

Naseem Khan, Tuan Nguyen, Amine Bermak, Issa Khalil

The proliferation of sophisticated AI-generated deepfakes poses critical challenges for digital media authentication and societal security. While existing detection methods perform well within specific generative domains, they exhibit significant performance degradation when applied to manipulations produced by unseen architectures--a fundamental limitation

Wentao Sun, João Paulo Nogueira, Alonso Silva

Despite remarkable advances in the field, LLMs remain unreliable in distinguishing causation from correlation. Recent results from the Corr2Cause dataset benchmark reveal that state-of-the-art LLMs -- such as GPT-4 (F1 score: 29.08) -- only marginally outperform random baselines (Random Uniform, F1 score: 20.38), indicating limited capacity of generalization

Christof Schötz, Jan Hassel, Christian Otto

We assess empirical models in climate econometrics using modern statistical learning techniques. Existing approaches are prone to outliers, ignore sample dependencies, and lack principled model selection. To address these issues, we implement robust preprocessing, nonparametric time-trend controls, and out-of-sample validation across 700+ climate variables.

Maximilian Mueller, Matthias Hein

Detecting out-of-distribution (OOD) examples is an important task for deploying reliable machine learning models in safety-critial applications. While post-hoc methods based on the Mahalanobis distance applied to pre-logit features are among the most effective for ImageNet-scale OOD detection, their performance varies significantly across models. We connect

Julia A. Kunzelmann, Anton Trushechkin, Nikolai Wyderka, Hermann Kampermann

Multipartite quantum repeaters play an important role in quantum communication networks enabling the transmission of quantum information over larger distances. To increase the rates for multipartite entanglement distribution, multiplexing of quantum memories is included. Understanding the limitations of achievable rates in the stationary regime for different

Hazhar Rahmani, Jie Fu

Many preference elicitation algorithms consider preference over propositional logic formulas or items with different attributes. In sequential decision making, a user's preference can be a preorder over possible outcomes, each of which is a temporal sequence of events. This paper considers a class of preference inference problems where the user's unknown pre

Matthew T. Littlehales, Max T. Birch, Akiko Kikkawa, Yasujiro Taguchi

The interaction between conduction electrons and spin textures gives rise to remarkable phenomena associated with the Berry phase. The Berry phase acquired by conduction electrons acts as an emergent electromagnetic field, facilitating phenomena analogous to classical electromagnetism, such as the Lorentz force and electromagnetic induction. Magnetic skyrmio

Zizhao Chen, Yoav Artzi

We propose KnotGym, an interactive environment for complex, spatial reasoning and manipulation. KnotGym includes goal-oriented rope manipulation tasks with varying levels of complexity, all requiring acting from pure image observations. Tasks are defined along a clear and quantifiable axis of complexity based on the number of knot crossings, creating a natur

Dongkeun Lee, Hoon Ryu

We propose a cost-efficient measurement scheme of the variational quantum eigensolver (VQE) for atomistic simulations of electronic structures based on a tight-binding (TB) theory. Leveraging the lattice geometry of a material domain, the sparse TB Hamiltonian is constructed in a bottom-up manner and is represented as a linear combination of the standard-bas

R. Lukyanenko, O. Pastor, V. C. Storey

All aspects of our society, including the life sciences, need a mechanism for people working within them to represent the concepts they employ to carry out their research. For the information systems being designed and developed to support researchers and scientists in conducting their work, conceptual models of the relevant domains are usually designed as b

Alexander Yakunin, Andrey Kupavskii, Alexander Sushin, Stanislav Moiseev

We study the problem of edge partitioning, where the goal is to partition the edge set of a graph into $k$ parts. The replication factor of a vertex $v$ is the number of parts that contain edges incident to $v$. The goal is to minimize the average replication factor of the vertices while keeping the sizes of the parts nearly equal. We study the regime where

Evangelos Sariyanidi, Claudio Ferrari, Federico Nocentini, Stefano Berretti

Computing the standard benchmark metric for 3D face reconstruction, namely geometric error, requires a number of steps, such as mesh cropping, rigid alignment, or point correspondence. Current benchmark tools are monolithic (they implement a specific combination of these steps), even though there is no consensus on the best way to measure error. We present a

Xiaobao Wei, Jiawei Liu, Dongbo Yang, Junda Cheng

We find that the EPE evaluation metrics of RAFT-stereo converge inconsistently in the low and high frequency regions, resulting high frequency degradation (e.g., edges and thin objects) during the iterative process. The underlying reason for the limited performance of current iterative methods is that it optimizes all frequency components together without di

Duc Anh Nguyen, Ernesto Araya, Adalbert Fono, Gitta Kutyniok

Recent years have seen significant progress in developing spiking neural networks (SNNs) as a potential solution to the energy challenges posed by conventional artificial neural networks (ANNs). However, our theoretical understanding of SNNs remains relatively limited compared to the ever-growing body of literature on ANNs. In this paper, we study a discrete

Liang Yao, Fan Liu, Delong Chen, Chuanyi Zhang

We aim to develop a robust yet flexible visual foundation model for Earth observation. It should possess strong capabilities in recognizing and localizing diverse visual targets while providing compatibility with various input-output interfaces required across different task scenarios. Current systems cannot meet these requirements, as they typically utilize

Yao Sun, Sining Chen, Yifan Tian, Xiao Xiang Zhu

Accurate information on the number of building floors, or above-ground storeys, is essential for household estimation, utility provision, risk assessment, evacuation planning, and energy modeling. Yet large-scale floor-count data are rarely available in cadastral and 3D city databases. This study proposes an end-to-end deep learning framework that infers flo

Xuhang Chen, Zhuo Li, Yanyan Shen, Mufti Mahmud

Functional ultrasound (fUS) imaging provides exceptional spatiotemporal resolution for neurovascular mapping, yet its practical application is significantly hampered by critical challenges. Foremost among these are data scarcity, arising from ethical considerations and signal degradation through the cranium, which collectively limit dataset diversity and com

Fulvio Missoni, Katarina Poole, Lorenzo Picinali, Andrea Canessa

Spatial hearing, the brain's ability to use auditory cues to identify the origin of sounds, is crucial for everyday listening. While simplified paradigms have advanced the understanding of spatial hearing, their lack of ecological validity limits their applicability to real-life conditions. This study aims to address this gap by investigating the effects of

Rashmi Gupta, Aditya K Gupta, Aarav Jain, Avinash C Pandey

Like any other discipline, Large Language Models (LLMs) have significantly impacted software engineering by helping developers generate the required artifacts across various phases of software development. This paper presents a case study comparing the performance of popular LLMs GPT, Claude, Gemini, and DeepSeek in generating functional specifications that

Lijiang Liu, Junyu Shi, Yong Sun, Zhiyuan Zhang

Current exoskeleton control methods often face challenges in delivering personalized treatment. Standardized walking gaits can lead to patient discomfort or even injury. Therefore, personalized gait is essential for the effectiveness of exoskeleton robots, as it directly impacts their adaptability, comfort, and rehabilitation outcomes for individual users. T

Matthieu Blanke, Yongquan Qu, Sara Shamekh, Pierre Gentine

Deep generative models hold great promise for representing complex physical systems, but their deployment is currently limited by the lack of guarantees on the physical plausibility of the generated outputs. Ensuring that known physical constraints are enforced is therefore critical when applying generative models to scientific and engineering problems. We a

Piyush Sakrikar, Bin Shen, Eduardo H. T. Poldi, Faranak Bahrami

Magnetic exchange interactions are mediated via orbital overlaps across chemical bonds. Thus, modifying the bond angles by physical pressure or strain can tune the relative strength of competing interactions. Here we present a remarkable case of such tuning between the Heisenberg (J) and Kitaev (K) exchange, which respectively establish magnetically ordered

Shashank Agnihotri, David Schader, Jonas Jakubassa, Nico Sharei

Reliability and generalization in deep learning are predominantly studied in the context of image classification. Yet, real-world applications in safety-critical domains involve a broader set of semantic tasks, such as semantic segmentation and object detection, which come with a diverse set of dedicated model architectures. To facilitate research towards ro

Benedikt Hahn, Bettina Klinz, Birgit Vogtenhuber

We study the \emph{geometric $k$-colored crossing number} of complete graphs $\overline{\overline{\text{cr}}}_k(K_n)$, which is the smallest number of monochromatic crossings in any $k$-edge colored straight-line drawing of $K_n$. We substantially improve asymptotic upper bounds on $\overline{\overline{\text{cr}}}_k(K_n)$ for $k=2,\ldots, 10$ by developing a

Hanze Zhang, Kaiming Wang, Rong Chen, Xingda Wei

The disaggregated memory (DM) architecture offers high resource elasticity at the cost of data access performance. While caching frequently accessed data in compute nodes (CNs) reduces access overhead, it requires costly centralized maintenance of cache coherence across CNs. This paper presents DiFache, an efficient, scalable, and coherent CN-side caching fr

Miguel Neves, Pedro Neto

The classification of human-performed assembly tasks is essential in collaborative robotics to ensure safety, anticipate robot actions, and facilitate robot learning. However, achieving reliable classification is challenging when segmenting tasks into smaller primitive actions is unfeasible, requiring us to classify long assembly tasks that encompass multipl

Prince Kumar, Rudra Murthy, Riyaz Bhat, Danish Contractor

Despite rapid advances in the capabilities of Large Language Models (LLMs), they continue to struggle with following relatively simple and unambiguous instructions, particularly when compositional structure is involved. Recent work suggests that models may follow instructions more effectively when they are expressed in pseudo-code rather than natural languag

Jens De Winne, Siri Willems, Siri Luthman, Danilo Babin

Accurate, real-time monitoring of tissue ischemia is crucial to understand tissue health and guide surgery. Spectral imaging shows great potential for contactless and intraoperative monitoring of tissue oxygenation. Due to the difficulty of obtaining direct reference oxygenation values, conventional methods are based on linear unmixing techniques. These are

Simin Shen, Zaiwu Gong, Bin Zhou, Roman Słowiński

Challenges, such as a lack of information for emergency decision-making, time pressure, and limited knowledge of experts acting as decision-makers (DMs), can result in the generation of poor or inconsistent indirect information regarding DMs' preferences. Simultaneously, the empathic relationship represents a tangible social connection within the context of

Joseph Maffetone, Julia Gersey, Pei Zhang

ZV-Sim is an open-source, modular Python framework for probabilistic simulation and analysis of pre-emergent novel zoonotic diseases using pervasive sensing data. It incorporates customizable Human and Animal Presence agents that leverage known and simulated location data, contact networks, and illness reports to assess and predict disease origins and spread

Thomas Oliver de Jong, Khemraj Shukla, Mircea Lazar

In this paper, we consider the design of model predictive control (MPC) algorithms based on deep operator neural networks (DeepONets). These neural networks are capable of accurately approximating real and complex valued solutions of continuous time nonlinear systems without relying on recurrent architectures. The DeepONet architecture is made up of two feed

Arjun Bhalla, Qi Huang

Intent classification is an important component of a functional Information Retrieval ecosystem. Many current approaches to intent classification, typically framed as a classification problem, can be problematic as intents are often hard to define and thus data can be difficult and expensive to annotate. The problem is exacerbated when we need to extend the

Sergio Luigi Cacciatori, Fabrizio Canfora, Evangelo Delgado, Federica Muscolino

Through the Hamilton-Jacobi equation of classical mechanics, BPS magnetized Baryonic layers (possessing both baryonic charge and magnetic flux) have been constructed in the gauged non-linear sigma model (G-NLSM) minimally coupled to Maxwell theory, which is one of the most relevant effective theories for Quantum Chromodynamics (QCD) in the strongly interacti

Sergio Calo, Anders Jonsson, Gergely Neu, Ludovic Schwartz

Bisimulation metrics are powerful tools for measuring similarities between stochastic processes, and specifically Markov chains. Recent advances have uncovered that bisimulation metrics are, in fact, optimal-transport distances, which has enabled the development of fast algorithms for computing such metrics with provable accuracy and runtime guarantees. Howe

BESIII Collaboration, M. Ablikim, M. N. Achasov, P. Adlarson

By analyzing $e^+e^-$ collision data taken at center-of-mass energies $\sqrt{s}$ between 4.600 and 4.699 GeV with the BESIII detector at the BEPCII collider, corresponding to an integrated luminosity of $\rm 4.5~fb^{-1}$, we study the hadronic decays $\Lambda_{c}^{+} \rightarrow \Sigma^{+} \eta$ and $\Lambda_{c}^{+} \rightarrow \Sigma^{+} \eta^{\prime}$ usin

Joshua Clymer, Jonah Weinbaum, Robert Kirk, Kimberly Mai

Existing evaluations of AI misuse safeguards provide a patchwork of evidence that is often difficult to connect to real-world decisions. To bridge this gap, we describe an end-to-end argument (a "safety case") that misuse safeguards reduce the risk posed by an AI assistant to low levels. We first describe how a hypothetical developer red teams safeguards, es

Di Jin, Jingyi Cao, Xiaobao Wang, Bingdao Feng

Graph anomaly detection aims to identify unusual patterns in graph-based data, with wide applications in fields such as web security and financial fraud detection. Existing methods typically rely on contrastive learning, assuming that a lower similarity between a node and its local subgraph indicates abnormality. However, these approaches overlook a crucial

Yuta Shigedomi, Tsuneya Yoshida

Liouvillian skin effects, manifested as the localization of Liouvillian eigenstates around the boundary, are distinctive features of non-Hermitian systems and are particularly notable for their impact on system dynamics. Despite their significance, Liouvillian skin effects have not been sufficiently explored in electron systems. In this work, we demonstrate

Valentin Kilian, Stefano Cortinovis, François Caron

Given a large pool of unlabelled data and a smaller amount of labels, prediction-powered inference (PPI) leverages machine learning predictions to increase the statistical efficiency of confidence interval procedures based solely on labelled data, while preserving fixed-time validity. In this paper, we extend the PPI framework to the sequential setting, wher

Honghao Li, Yiwen Zhang, Yi Zhang, Lei Sang

Hadamard Product (HP) has long been a cornerstone in click-through rate (CTR) prediction tasks due to its simplicity, effectiveness, and ability to capture feature interactions without additional parameters. However, the underlying reasons for its effectiveness remain unclear. In this paper, we revisit HP from the perspective of Quadratic Neural Networks (QN

Xuhang Chen, Michael Kwok-Po Ng, Kim-Fung Tsang, Chi-Man Pun

Brain network analysis plays a crucial role in diagnosing and monitoring neurodegenerative disorders such as Alzheimer's disease (AD). Existing approaches for constructing structural brain networks from diffusion tensor imaging (DTI) often rely on specialized toolkits that suffer from inherent limitations: operator subjectivity, labor-intensive workflows, an

Nura Aljaafari, Danilo S. Carvalho, André Freitas

Modern transformer models exhibit phase transitions during training, distinct shifts from memorisation to abstraction, but the mechanisms underlying these transitions remain poorly understood. Prior work has often focused on endpoint representations or isolated signals like curvature or mutual information, typically in symbolic or arithmetic domains, overloo

Lukas Schulthess, Silvano Cortesi, Michele Magno

Large-scale Internet of Things (IoT) applications, such as asset tracking and remote sensing, demand multi-year battery lifetimes to minimize maintenance and operational costs. Traditional wireless protocols often employ duty cycling, introducing a tradeoff between latency and idle consumption - both unsuitable for event-driven and ultra-low power systems. A

Jintian Shao, Yiming Cheng, Hongyi Huang, Beiwen Zhang

The VAPO framework has demonstrated significant empirical success in enhancing the efficiency and reliability of reinforcement learning for long chain-of-thought (CoT) reasoning tasks with large language models (LLMs). By systematically addressing challenges such as value model bias, heterogeneous sequence lengths, and sparse reward signals, VAPO achieves st

Mokai Pan, Kaizhen Zhu, Yuexin Ma, Yanwei Fu

Recent advances in diffusion bridge models leverage Doob's $h$-transform to establish fixed endpoints between distributions, demonstrating promising results in image translation and restoration tasks. However, these approaches often produce blurred or excessively smoothed image details and lack a comprehensive theoretical foundation to explain these shortcom

Pedro Jose Bauza-Ruiz, Oscar Blanch, Paolo G. Calisse, Anna Campoy-Ordaz

The Barcelona Raman LIDAR (BRL) will provide continuous monitoring of the aerosol extinction profile along the line of sight of the Cherenkov Telescope Array Observatory (CTAO). It will be located at its Northern site (CTAO-N) on the Observatorio del Roque de Los Muchachos. This article presents the performance of the pathfinder Barcelona Raman LIDAR (pBRL),

DMC Collaboration, N. Darvishi, S. Autti, L. Bloomfield

We present the projected sensitivity to non-relativistic Effective Field Theory (EFT) operators for Dark Matter (DM) direct detection using the QUEST-DMC experiment. QUEST-DMC employs superfluid Helium-3 as a target medium and measures energy deposition via nanomechanical resonators with SQUID-based readout to probe DM interactions. The experiment aims to ex

Zhihua Liu, Amrutha Saseendran, Lei Tong, Xilin He

Open-set image segmentation poses a significant challenge because existing methods often demand extensive training or fine-tuning and generally struggle to segment unified objects consistently across diverse text reference expressions. Motivated by this, we propose Segment Anyword, a novel training-free visual concept prompt learning approach for open-set la

Jungho Ahn, Tala Eagling-Vose, Felicia Lucke, Daniël Paulusma

The Matching Cut problem is to decide if the vertex set of a connected graph can be partitioned into two non-empty sets $B$ and $R$ such that the edges between $B$ and $R$ form a matching, that is, every vertex in $B$ has at most one neighbour in $R$, and vice versa. If for some integer $d\geq 1$, we allow every neighbour in $B$ to have at most $d$ neighbour

Fahd Alhamazani, Yu-Kun Lai, Paul L. Rosin

3D reconstruction from 2D inputs, especially for non-rigid objects like humans, presents unique challenges due to the significant range of possible deformations. Traditional methods often struggle with non-rigid shapes, which require extensive training data to cover the entire deformation space. This study addresses these limitations by proposing a canonical

Anna Grünebohm, Matous Mrovec, Maxim N. Popov, Lan-Tien Hsu

Barium titanate (BTO) is a representative perovskite oxide that undergoes three first-order ferroelectric phase transitions related to exceptional functional properties. In this work, we develop two atomic cluster expansion (ACE) models for BTO to reproduce fundamental properties of bulk as well as defective BTO phases. The two ACE models do not target full

Shuhang Zhang, Bryan Olmos

Registers are primary storage elements in System-on-chip~(SoC) designs and play an important role in maintaining state information and processing data in digital systems. With respect to the ISO26262 standard, these registers require high levels of reliability and fault tolerance. For this reason, safety-critical applications require that normal registers ar

Benjamin Turtel, Danny Franklin, Kris Skotheim, Luke Hewitt

Reinforcement Learning with Verifiable Rewards (RLVR) has been an effective approach for improving Large Language Models' reasoning in domains such as coding and mathematics. Here, we apply RLVR methods towards forecasting future real-world events - a challenging task for RL due to the very noisy (and delayed) outcomes involved. Using a novel dataset of rece

Yutong Chen, Jiandong Gao, Ji Wu

R1-style Reinforcement Learning (RL) significantly enhances Large Language Models' reasoning capabilities, yet the mechanism behind rule-based RL remains unclear. We found that small-scale SFT has substantial influence on RL but shows poor efficiency. To explain our observations, we propose an analytical framework and compare the efficiency of SFT and RL by

Weihang You, Hanqi Jiang, Zishuai Liu, Zihang Xie

Real world collection of Activities of Daily Living data is challenging due to privacy concerns, costly deployment and labeling, and the inherent sparsity and imbalance of human behavior. We present ADLGen, a generative framework specifically designed to synthesize realistic, event triggered, and symbolic sensor sequences for ambient assistive environments.

Rodrigo Bernal, Federico Milano

This paper introduces the concept of Transient Slack Capability (TSC), a set of three necessary device-level conditions to ensure stability under sustained power perturbations. TSC states that a device must (1) possess sufficient stored energy; (2) a controlled input power; and (3) maintain internal energy balance and synchronization. The paper shows that th

Roland Schäfer, Manuel Neubauer, Klaus Meerholz, Klas Lindfors

We couple a templated layer of merocyanine molecules with surface lattice resonances in a plasmonic grating. The templating of the molecular layer is achieved using a layer of aligned graphene nanoribbons, resulting in anisotropic optical properties. The anisotropy manifests itself in polarization-dependent coupling between excitons in the organic layer and

Bryan Wong, Jong Woo Kim, Huazhu Fu, Mun Yong Yi

Vision-language models (VLMs) have recently been integrated into multiple instance learning (MIL) frameworks to address the challenge of few-shot, weakly supervised classification of whole slide images (WSIs). A key trend involves leveraging multi-scale information to better represent hierarchical tissue structures. However, existing methods often face two k

Richard Mycroft, Camila Zárate-Guerén

We give, for each $k \geq 3$, the precise best possible minimum positive codegree condition for a perfect matching in a large $k$-uniform hypergraph $H$ on $n$ vertices. Specifically we show that, if $n$ is sufficiently large and divisible by $k$, and $H$ has minimum positive codegree $\delta^+(H) \geq \frac{k-1}{k}n - (k-2)$ and no isolated vertices, then $

Radoslaw Klimek, Jakub Semczyszyn

This study empirically validates automated logical specification methods for behavioural models, focusing on their robustness, scalability, and reproducibility. By the systematic reproduction and extension of prior results, we confirm key trends, while identifying performance irregularities that suggest the need for adaptive heuristics in automated reasoning

Rui Cao, Zifeng Ding, Zhijiang Guo, Michael Schlichtkrull

Textual claims are often accompanied by images to enhance their credibility and spread on social media, but this also raises concerns about the spread of misinformation. Existing datasets for automated verification of image-text claims remain limited, as they often consist of synthetic claims and lack evidence annotations to capture the reasoning behind the

Farbod Daneshyan, Runzhi He, Jianyu Wu, Minghui Zhou

The release note is a crucial document outlining changes in new software versions. Yet, many developers view the process of writing software release notes as a tedious and dreadful task. Consequently, numerous tools have been developed by researchers and practitioners to automate the generation of software release notes. However, these tools fail to consider

Osama Abuzaid

This paper studies when a sequence of probability measures on a metric space admit subsequential weak limits. A sufficient condition called sequential tightness is formulated, which relaxes some assumptions for asymptotic tightness used in the Prokhorov -- Le Cam theorem. The proof only uses elementary tools from probability theory. Sequential tightness give

Yahya Naveed, Julia Gersey, Pei Zhang

Before aerosols can be sensed, sampling technologies must capture the particulate matter of interest. To that end, for systems deployed in open environments where the location of the aerosol is unknown, extending the reach of the sampler could lessen the precision required in sensor placement or reduce the number of sensors required for full spatial coverage

Simone Gaisbauer, Prabin Gyawali, Qilin Zhang, Olaf Wysocki

Feature matching is a necessary step for many computer vision and photogrammetry applications such as image registration, structure-from-motion, and visual localization. Classical handcrafted methods such as SIFT feature detection and description combined with nearest neighbour matching and RANSAC outlier removal have been state-of-the-art for mobile mapping

Kazi Mahmudul Hassan, Xuyang Zhao, Hidenori Sugano, Toshihisa Tanaka

Feature engineering for generalized seizure detection models remains a significant challenge. Recently proposed models show variable performance depending on the training data and remain ineffective at accurately distinguishing artifacts from seizure data. In this study, we propose a novel end-to-end model, "Multiresolutional EEGWaveNet (MR-EEGWaveNet)," whi

Danial Khan, Zohaib Salahuddin, Yumeng Zhang, Sheng Kuang

We present a fully automated, anatomically guided deep learning pipeline for prostate cancer (PCa) risk stratification using routine MRI. The pipeline integrates three key components: an nnU-Net module for segmenting the prostate gland and its zones on axial T2-weighted MRI; a classification module based on the UMedPT Swin Transformer foundation model, fine-

Lu Wang, Gui Zhou, Changheng Li, Luis F. Abanto-Leon

This work investigates a practical reconfigurable intelligent surface (RIS)-aided integrated sensing and communication (ISAC) system, where a subset of RIS elements fail to function properly and reflect incident signals randomly towards unintended directions, thereby degrading system performance. To date, no study has addressed such impairments caused by fau

Lorenzo Poggioni, Didier Clamond, Yves D'Angelo

We introduce a novel class of finite difference approximations, termed zigzag schemes, that employ a hybrid stencil that is neither symmetrical, nor fully one-sided. These zigzag schemes often enjoy more permissive stability constraints and see their coefficients vanish as the order tends to infinity. This property permits the formulation of higher order sch

Jiayi Geng, Howard Chen, Dilip Arumugam, Thomas L. Griffiths

Using AI to create autonomous researchers has the potential to accelerate scientific discovery. A prerequisite for this vision is understanding how well an AI model can identify the underlying structure of a black-box system from its behavior. In this paper, we explore how well a large language model (LLM) learns to identify a black-box function from passive

Ionut-Vlad Modoranu, Mher Safaryan, Erik Schultheis, Max Ryabinin

Low-rank optimization has emerged as a promising direction in training large language models (LLMs) to improve running time and reduce the memory usage of adaptive optimizers by constraining learning to a lower-dimensional space. Prior work typically projects gradients of linear layers using approaches based on Singular Value Decomposition (SVD) or QR-decomp

Frederik Nolte, Andreas Geiger, Bernhard Schölkopf, Ingmar Posner

This paper evaluates single-view mesh reconstruction models for their potential in enabling instant digital twin creation for real-time planning and dynamics prediction using physics simulators for robotic manipulation. Recent single-view 3D reconstruction advances offer a promising avenue toward an automated real-to-sim pipeline: directly mapping a single o

Daniel Cortild, Lucas Ketels, Juan Peypouquet, Guillaume Garrigos

The non-asymptotic analysis of Stochastic Gradient Descent (SGD) typically yields bounds that decompose into a bias term and a variance term. In this work, we focus on the bias component and study the extent to which SGD can match the optimal convergence behavior of deterministic gradient descent. Assuming only (strong) convexity and smoothness of the object

Fabian Bach

New and efficient methods based on noise addition to protect the confidentiality in population statistics have been developed, tested and applied in census production by various members of the European Statistical System over the past years. Basic demographic statistics - such as population stocks, live births and deaths by age, sex and region - may be prote

James A. Walker, Moein Khajehnejad, Adeel Razi

We propose a Bayesian framework for training binary and spiking neural networks that achieves state-of-the-art performance without normalisation layers. Unlike commonly used surrogate gradient methods -- often heuristic and sensitive to hyperparameter choices -- our approach is grounded in a probabilistic model of noisy binary networks, enabling fully end-to

Lorenz Halbeisen, Silvan Horvath, Saharon Shelah

We show that in the model obtained by iteratively pseudo-intersecting a Ramsey ultrafilter via a length-$\omega_2$ countable support iteration of restricted Mathias forcing over a ground model satisfying $\textsf{CH}$, there is a unique $Q$-point up to isomorphism. In particular, it is consistent that there is only one $Q$-point while there are $2^{\mathfrak

Nguyen Duc, Yan-Ling Lai, Patrick Madlindl, Xinyuan Zhu

Owing to the typical long-tail data distribution issues, simulating domain-gap-free synthetic data is crucial in robotics, photogrammetry, and computer vision research. The fundamental challenge pertains to credibly measuring the difference between real and simulated data. Such a measure is vital for safety-critical applications, such as automated driving, w

Leo Maas

An "upward-driven disk" is a novel mechanical device built from LEGO parts. A circular disk is suspended from the point where it is sandwiched between two wheels, making it free to oscillate as a pendulum, but the location of that suspension point on the disk changes with time due to a steady upward driving force applied by rotation of one of the wheels. The

Jihoon Kim, Roopendra Rajawat, Polina Blinova, Andrey Mironov

Attaining practical Inertial Fusion Energy (IFE) depends on how efficiently one can couple the driver energy to the nuclear fusion fuel for compression and ignition. While the excimer lasers provide an efficient alternative compared to existing laser technology, it is unclear how the lasers can be harnessed to form a pulse with desired pulse shape and intens

Yujin Jeong, Arnas Uselis, Seong Joon Oh, Anna Rohrbach

Understanding visual scenes is fundamental to human intelligence. While discriminative models have significantly advanced computer vision, they often struggle with compositional understanding. In contrast, recent generative text-to-image diffusion models excel at synthesizing complex scenes, suggesting inherent compositional capabilities. Building on this, z

Masahiro Watari

Piontkowski proved the existence of affine cell decompositions of Jacobian factors of plane curve singularities with a single Puiseux pair. He also provided a combinatorial description of the Euler numbers and Betti numbers of these Jacobian factors. Following his results, Oblomkov, Rasmussen, and Shende demonstrated the existence of affine cell decompositio

Pramod N. Achar, Gurbir Dhillon, Simon Riche

We compute the dimension of the cohomology of stalks of intersection cohomology complexes on Zastava schemes and Drinfeld compactifications associated with a connected reductive algebraic group $G$, in case the characteristic of the coefficients field $\Bbbk$ is good for $G$. In particular, we show that these dimensions do not depend on the choice of $\Bbbk$