May 2025 arXiv papers — page 44

Jorge Fariña-Asategui

The author introduced recently a new natural construction which associates a measure-preserving dynamical system to any fractal profinite group. Here, we investigate these measure-preserving dynamical systems under the extra assumption on the groups to be branch. First, we compute their $f$-invariant, a measure-conjugacy invariant introduced by Bowen, and sh

Marshal Arijona Sinaga, Julien Martinelli, Samuel Kaski

Gaussian processes (GPs) are widely used for regression and optimization tasks such as Bayesian optimization (BO) due to their expressiveness and principled uncertainty estimates. However, in settings with large datasets corrupted by outliers, standard GPs and their sparse approximations struggle with computational tractability and robustness. We introduce R

Samiran Ghosh, Malay Banerjee, Amit K Chattopadhyay

In this study, we present an immuno-epidemic model to understand mitigation options during an epidemic break. The model incorporates comorbidity and multiple-vaccine doses through a system of coupled integro-differential equations to analyze the epidemic rate and intensity from a knowledge of the basic reproduction number and time-distributed rate functions.

Guang-Chen He, Zhao-Xian Chen, Xiao-Meng Zhang, Ze-Guo Chen

Topological phases are states of matter defined by global topological invariants that remain invariant under adiabatic parameter variations, provided no topological phase transition occurs. This endows them with intrinsic robustness against local perturbations. Experimentally, these phases are often identified indirectly by observing robust boundary states,

Jia Li, Jiacheng Shen, Yuxin Su, Michael R. Lyu

Directed fuzzing is a critical technique in cybersecurity, targeting specific sections of a program. This approach is essential in various security-related domains such as crash reproduction, patch testing, and vulnerability detection. Despite its importance, current directed fuzzing methods exhibit a trade-off between efficiency and effectiveness. For insta

Hannes Wallimann, Widar von Arx, Ann Hesse

The opening of the Gotthard Base Tunnel in 2017, the longest railway tunnel in the world, marked a milestone in Swiss transport policy. The tunnel, a part of the New Rail Link through the Alps, serves as a key instrument of the so-called "modal shift policy," which aims to transfer transalpine freight traffic from road to rail. The reduction in travel time b

Jan Greve, Lukas Sablica

This work introduces an anonymization scheme for a corpus of texts to safeguard metadata from disclosure. It specifically aims to prevent large language models from identifying metadata associated with texts, thereby avoiding their influence on query responses. The core mechanism is called named entity swapping, a technique inspired by data swapping in stati

Eva Sextl, Rolf-Peter Kudritzki, Fabio Bresolin, Kathryn Grasha

We apply population synthesis techniques to analyze TYPHOON long slit spectra of the starburst barred spiral galaxy M83. The analysis covers a central square of 5 arcmin side length. We determine the spatial distribution of dust through the analysis of reddening and extinction, together with star formation rates, ages, and metallicities of young and old stel

Salman N. Salman, Sergey A. Shteingolts, Ron Levie, Dan Mendels

Machine learning models often require large datasets and struggle to generalize beyond their training distribution. These limitations pose significant challenges in scientific and engineering contexts, where generating exhaustive datasets is often impractical and the goal is frequently to discover novel solutions outside the training domain. In this work, we

Rohit Sunil Kanchi, Benjamin Melanson, Nithin Somasekharan, Shaowu Pan

We present UniFoil, a large publicly available universal airfoil dataset based on Reynolds-averaged Navier-Stokes (RANS) simulations. It contains over 500,000 samples spanning a wide range of Reynolds and Mach numbers, capturing both transitional and fully turbulent flows across incompressible to compressible regimes. UniFoil is designed to support machine l

M. Laura Arias, Maximiliano Contino, Stefania Marcantognini

We study the class of those linear relations that can be factorized as products of idempotent relations. We provide several characterizations of this class, extending known factorization results for operators to the more general setting of linear relations.

Jean-Christophe Pain

A few years ago, Kurzweil and Hazak developed the Configurationally Resolved Super-Transition-Arrays (CRSTA) method for the computation of hot-plasma radiative opacity. Their approach, based on a temporal integration, is an important refinement of the standard Super-Transition-Arrays (STA) approach, which enables one to recover the underlying structure of th

Côme Tabary

We derive a weak-strong uniqueness and stability principle for the Landau equation in the soft potentials case (including Coulomb interactions). The distance between two solutions is measured by their relative entropy, which to our knowledge was never used before in stability estimates. The logarithm of the strong solution is required to have polynomial grow

Moritz A. Zanger, Max Weltevrede, Yaniv Oren, Pascal R. Van der Vaart

Estimating epistemic uncertainty in value functions is a crucial challenge for many aspects of reinforcement learning (RL), including efficient exploration, safe decision-making, and offline RL. While deep ensembles provide a robust method for quantifying value uncertainty, they come with significant computational overhead. Single-model methods, while comput

Donald Marolf, Bruno Missoni

Previous studies of linearized stability of asymptotically flat Euclidean axion wormholes found that symmetric modes suffered from divergences. We show that such divergences were an artifact of a particular way of solving the constraints, and that a full treatment leads to finite actions for such modes. The modes must thus be included in a stability analysis

Adeela Islam, Stefano Fiorini, Stuart James, Pietro Morerio

The task of reassembly is a significant challenge across multiple domains, including archaeology, genomics, and molecular docking, requiring the precise placement and orientation of elements to reconstruct an original structure. In this work, we address key limitations in state-of-the-art Deep Learning methods for reassembly, namely i) scalability; ii) multi

Sergey Pletenev, Maria Marina, Nikolay Ivanov, Daria Galimzianova

Large Language Models (LLMs) often hallucinate in question answering (QA) tasks. A key yet underexplored factor contributing to this is the temporality of questions -- whether they are evergreen (answers remain stable over time) or mutable (answers change). In this work, we introduce EverGreenQA, the first multilingual QA dataset with evergreen labels, suppo

Shuai Wang, Zexian Li, Qipeng zhang, Tianhui Song

Diffusion models have demonstrated remarkable generation quality but at the cost of numerous function evaluations. Recently, advanced ODE-based solvers have been developed to mitigate the substantial computational demands of reverse-diffusion solving under limited sampling steps. However, these solvers, heavily inspired by Adams-like multistep methods, rely

John A. Baldwin, Steven Sivek, Jonathan Zung

We prove that for each $n\in\mathbb{N}$ there is a hyperbolic L-space with $n$ pseudo-Anosov flows, no two of which are orbit equivalent. These flows have no perfect fits and are thus quasigeodesic. In addition, our flows admit positive Birkhoff sections, which we argue implies that they give rise to $n$ universally tight contact structures whose lifts to an

Hazem Zohny

This paper introduces ADEPT, a system using Large Language Model (LLM) personas to simulate multi-perspective ethical debates. ADEPT assembles panels of 'AI personas', each embodying a distinct ethical framework or stakeholder perspective (like a deontologist, consequentialist, or disability rights advocate), to deliberate on complex moral issues. Its applic

Shishuo Fu, James Sellers

In 2002, Andrews, Lewis, and Lovejoy introduced the combinatorial objects which they called partitions with designated summands. These are constructed by taking unrestricted integer partitions and designating exactly one of each occurrence of a part. In the same work, they also considered the restricted partitions with designated summands wherein all parts m

Masahiro N. Machida, Shingo Hirano, Shantanu Basu

We examine the impact of the magnetic field on Population III star formation by varying the magnetic field strength. We perform simulations with magnetic field strengths ranging from $10^{-20}$ G to $10^{-4}$ G, in addition to a model without a magnetic field. The simulations are run for $>1000-1400$ yr after the first protostar forms. In weak-field models,

Bogdan Bogachov, Yaoyao Fiona Zhao

Despite recent advancements in domain adaptation techniques for large language models, these methods remain computationally intensive, and the resulting models can still exhibit hallucination issues. Most existing adaptation methods do not prioritize reducing the computational resources required for fine-tuning and inference of language models. Hallucination

Abdulla Rakhimov, Sanathon Tukhtasinova, Vyacheslav I. Yukalov

In 2015 Dmitry Petrov theoretically suggested that, in a binary mixture of bosons a quantum liquid droplet may arise due to the competition between attractive intercomponent and repulsive intracomponent forces. Although this prediction has been confirmed experimentally, the model by itself suffers from a serious conceptual problem: The low - lying excitation

Bianca M. Poggianti, Benedetta Vulcani, Neven Tomicic, Alessia Moretti

We present the full sample of 76 galaxies in 39 galaxy cluster fields at z=0.04-0.07 observed with VLT/MUSE by the GASP survey. Most of them (64) were observed as possible ram pressure stripped galaxies (stripping candidates) based on optical B-band images, while the remaining 12 were a control sample of both star-forming and passive galaxies. Based on spati

Zhengyang Ji, Yifan Jia, Shang Gao, Yutao Yue

Large Vision Language Models (LVLMs) have achieved remarkable progress in multimodal tasks, yet they also exhibit notable social biases. These biases often manifest as unintended associations between neutral concepts and sensitive human attributes, leading to disparate model behaviors across demographic groups. While existing studies primarily focus on detec

Rui Zhang, Xiaotong Guo, Qiusheng Gu, Guanwen Fang

Compton-thick active galactic nuclei (CT-AGNs), which are defined by column density $\mathrm{N_H} \geqslant 1.5 \times 10^{24} \ \mathrm{cm}^{-2}$, emit feeble X-ray radiation, even undetectable by X-ray instruments. Despite this, the X-ray emissions from CT-AGNs are believed to be a substantial contributor to the cosmic X-ray background (CXB). According to

Fotis Farakos, George Tringas, Thomas Van Riet

Flux compactifications of IIA supergravity on orientifolded G$_2$-manifolds have been argued to allow for classical Minkowski$_3$ vacua with moduli and scale-separated AdS$_3$ vacua with full moduli stabilisation. To further uplift these vacua to meta-stable dS$_3$ vacua using anti-D2 branes, warped throats are desirable. We study the flux-stabilisation of l

Franz Baader, Stefan Borgwardt, Filippo De Bortoli, Patrick Koopmann

Standard Description Logics (DLs) can encode quantitative aspects of an application domain through either number restrictions, which constrain the number of individuals that are in a certain relationship with an individual, or concrete domains, which can be used to assign concrete values to individuals using so-called features. These two mechanisms have been

Leighton P. Barnes, Alex Dytso, H. Vincent Poor

We study prior distributions for Poisson parameter estimation under $L^1$ loss. Specifically, we construct a new family of prior distributions whose optimal Bayesian estimators (the conditional medians) can be any prescribed increasing function that satisfies certain regularity conditions. In the case of affine estimators, this family is distinct from the us

Badr Moufad, Yazid Janati, Alain Durmus, Ahmed Ghorbel

Classifier-Free Guidance (CFG) is a widely used technique for improving conditional diffusion models by linearly combining the outputs of conditional and unconditional denoisers. While CFG enhances visual quality and improves alignment with prompts, it often reduces sample diversity, leading to a challenging trade-off between quality and diversity. To addres

Dale S. Kim, Audrey Lu, Qing Zhou

Exploratory factor analysis is often used in the social sciences to estimate potential measurement models. To do this, several important issues need to be addressed: (1) determining the number of factors, (2) learning constraints in the factor loadings, and (3) selecting a solution amongst rotationally equivalent choices. Traditionally, these issues are trea

Tianhao Peng, Ho Man Kwan, Yuxuan Jiang, Ge Gao

Deep learning based Image Super-Resolution (ISR) relies on large training datasets to optimize model generalization; this requires substantial computational and storage resources during training. While dataset condensation (DC) has shown potential in improving data efficiency for high-level computer vision tasks, adopting these methods for ISR is not straigh

Nicole Bäuerle, Tamara Göll, Anna Jaśkiewicz

We consider non-standard Markov Decision Processes (MDPs) where the target function is not only a simple expectation of the accumulated reward. Instead, we consider rather general functionals of the joint distribution of terminal state and accumulated reward which have to be optimized. For finite state and compact action space, we show how to solve these pro

Carina Newen, Luca Hinkamp, Maria Ntonti, Emmanuel Müller

From uncertainty quantification to real-world object detection, we recognize the importance of machine learning algorithms, particularly in safety-critical domains such as autonomous driving or medical diagnostics. In machine learning, ambiguous data plays an important role in various machine learning domains. Optical illusions present a compelling area of s

Stephen Chung, Wenyu Du, Jie Fu

Recent studies show that the reasoning capabilities of Large Language Models (LLMs) can be improved by applying Reinforcement Learning (RL) to question-answering (QA) tasks in areas such as math and coding. With a long context length, LLMs may learn to perform search, as indicated by the self-correction behavior observed in DeepSeek R1. However, this search

Zhong Wang, Chunlong Hu, Changchang Gong, Fuyong Hua

Semimetal molybdenum ditelluride (1T'-MoTe$_2$) possess diverse phase transitions enriching its application prospects. The structural response during these transitions is crucial to understanding the underlying mechanisms, but the desired details of pathway and time span are still insufficient. Here, we investigate the lattice evolution in few-layer 1T'-MoTe

Kei Takemura, Ryuta Matsuno, Keita Sakuma

A central goal in online learning is to achieve adaptivity to unknown problem characteristics, such as environmental changes captured by gradient variation (GV), function curvature (universal online learning, UOL), and gradient scales (Lipschitz adaptivity, LA). Simultaneously achieving these with optimal performance is a major challenge, partly due to limit

Amit K Chattopadhyay, Aimee Pascaline N Unkundiye, Gillian Pearce

This is a Machine Learning guided study towards zone-specific ray therapy. Combining Machine Learning (Extreme Gradient Boosting) with continuum modeling (exponential and logistic growth), we find that while fluorodeoxyglucose-coated (mNP-FDG) can control cancerous tumor progression within 2 days compared to 18 days by Superparamagnetic Iron Oxide Nanopartic

Francesco Pierotti, Andrea Bandini

The analysis of speech in individuals with amyotrophic lateral sclerosis is a powerful tool to support clinicians in the assessment of bulbar dysfunction. However, current methods used in clinical practice consist of subjective evaluations or expensive instrumentation. This study investigates different approaches combining audio-visual analysis and machine l

Daeen Kabir, Minhajur Rahman Chowdhury Mahim, Sheikh Shafayat, Adnan Sadik

In this work, we introduce BLUCK, a new dataset designed to measure the performance of Large Language Models (LLMs) in Bengali linguistic understanding and cultural knowledge. Our dataset comprises 2366 multiple-choice questions (MCQs) carefully curated from compiled collections of several college and job level examinations and spans 23 categories covering k

Anna Neumann, Elisabeth Kirsten, Muhammad Bilal Zafar, Jatinder Singh

System prompts in Large Language Models (LLMs) are predefined directives that guide model behaviour, taking precedence over user inputs in text processing and generation. LLM deployers increasingly use them to ensure consistent responses across contexts. While model providers set a foundation of system prompts, deployers and third-party developers can append

Jonas Deré, Joren Matthys

Given a finitely generated residually finite group $G$, the residual finiteness growth $\text{RF}_G: \mathbb{N} \to \mathbb{N}$ bounds the size of a finite group $Q$ needed to detect an element of norm at most $r$. More specifically, if $g\in G$ is a non-trivial element with $\|g\|_G \leq r$, so $g$ can be written as a product of at most $r$ generators or th

Junjue Wang, Weihao Xuan, Heli Qi, Zhihao Liu

Large vision-language models (VLMs) have made great achievements in Earth vision. However, complex disaster scenes with diverse disaster types, geographic regions, and satellite sensors have posed new challenges for VLM applications. To fill this gap, we curate a remote sensing vision-language dataset (DisasterM3) for global-scale disaster assessment and res

Navojit Dhali Pallab

This paper investigates synchronization phenomena in networks of coupled oscillators governed by three-time-scale dynamical systems exhibiting canard dynamics. A mathematical framework has been developed to analyze the synchronization of fast variables across heterogeneous systems, deriving a sufficient condition for the synchronization error to fall below a

Marta Grobelna, Jan Křetínský, Maximilian Weininger

We consider two-player zero-sum concurrent stochastic games (CSGs) played on graphs with reachability and safety objectives. These include degenerate classes such as Markov decision processes or turn-based stochastic games, which can be solved by linear or quadratic programming; however, in practice, value iteration (VI) outperforms the other approaches and

Young-Min Cho, Sharath Chandra Guntuku, Lyle Ungar

Recent advancements in Large Language Models (LLMs) have enabled the emergence of multi-agent systems where LLMs interact, collaborate, and make decisions in shared environments. While individual model behavior has been extensively studied, the dynamics of peer influence in such systems remain underexplored. In this paper, we investigate herd behavior, the t

Long Cheng, Dalong Qi, Jiali Yao, Ning Xu

Snapshot ultrafast optical imaging (SUOI) plays a vital role in capturing complex transient events in real time, with significant implications for both fundamental science and practical applications. As an outstanding talent in SUOI, compressed ultrafast photography (CUP) has demonstrated remarkable frame rate reaching trillions of frames per second and hund

Risa Ogino, Shun Uchino

We show that supercurrent properties in a superfluid or superconducting junction are significantly modified due to single-particle losses present in a conduction channel. In the presence of a spin-independent particle loss, we find regimes where the Josephson current $I_N(\phi)$ takes zero at a position in between $\phi= 0$ and $\phi=\pi$, and the direction

Prabal Dasgupta, Debashis Gangopadhyay

New convenient thumbrules are obtained to test entanglement of wavefunctions for bipartite qubit and qutrit systems. All results are analytic. The new results are: (a) For bipartite qubit systems there exists a matrix $A$ for which $\det A = 0$ implies unentanglement while $\det A \ne 0$ implies entanglement. There is an underlying SU(2) algebra. (2) Telepor

Giuseppe Pipoli, Joao Paulo dos Santos, Giuseppe Tinaglia

In this work, we study complete properly immersed translators in the product space $\mathbb H^2\times\mathbb R$, focusing on their asymptotic behavior at infinity. We classify the asymptotic boundary components of these translators under suitable continuity assumptions. Specifically, we prove that if a boundary component lies in the vertical asymptotic bound

Jieyong Kim, Tongyoung Kim, Soojin Yoon, Jaehyung Kim

While black-box large language models are widely deployed, they produce generic outputs that overlook individual user preferences. Current personalization methods are fundamentally limited to response-level personalization; they only match final outputs, failing to model the underlying reasoning that connects user behavior to responses. To address this, this

A. Errehymy, O. Donmez, A. Syzdykova, K. Myrzakulov

In the late 1980s, Morris and Thorne led in theoretical physics by creating solutions to wormholes and formulating the crucial requirements for safe traversability of wormholes. They found that exotic matter must meet the requirement $P_r + \rho < 0$, where $P_r$ is radial pressure and $\rho$ is energy density. This is a rudimentary grasp of our understandin

Santanu Maity, Mousa Alrubayan, Prabhakar Pradhan

We explored the fractal and multifractal characteristics of breast mammogram micrographs to identify quantitative biomarkers associated with breast cancer progression. In addition to conventional fractal and multifractal analyses, we employed a recently developed fractal-functional distribution method, which transforms fractal measures into Gaussian distribu

Yue Zhang, Yingzhao Jian, Hehe Fan, Yi Yang

Recent advancements in multimodal large language models (MLLMs) have demonstrated considerable potential for comprehensive 3D scene understanding. However, existing approaches typically utilize only one or a limited subset of 3D modalities, resulting in incomplete representations of 3D scenes and reduced interpretive accuracy. Furthermore, different types of

Enrico Bernardi, Tatsuo Nishitani

In this paper we study a class of non-effectively hyperbolic operators vanishing of order 2 on a manifold, on a sub-region of which the spectral structure of the Hamilton map changes type. Suitable normal symplectic coordinates are found together with an analysis of the Hamilton system associated to the principal symbol and a factorization result, preparing

Mete Erdogan, Francesco Tonin, Volkan Cevher

The high inference demands of transformer-based Large Language Models (LLMs) pose substantial challenges in their deployment. To this end, we introduce Neural Block Linearization (NBL), a novel framework for accelerating transformer model inference by replacing self-attention layers with linear approximations derived from Linear Minimum Mean Squared Error es

Weihao Xuan, Junjue Wang, Heli Qi, Zihang Chen

Multimodal large language models (MLLMs) have demonstrated remarkable capabilities in visual understanding, but their application to long-term Earth observation analysis remains limited, primarily focusing on single-temporal or bi-temporal imagery. To address this gap, we introduce DVL-Suite, a comprehensive framework for analyzing long-term urban dynamics t

Zehua Yu, Yongshun Ren, Kang Liu

Water evaporation is critically important for hydrogels in open-air applications, but theoretically modeling is difficult due to the complicated intermolecular interactions and sustained deformation. In this work, we construct a simplified model to describe the state of water inside the hydrogel by only considering mechanical stretching. We employ "negative

Yichuan Cao, Yibo Miao, Xiao-Shan Gao, Yinpeng Dong

Text-to-image (T2I) models raise ethical and safety concerns due to their potential to generate inappropriate or harmful images. Evaluating these models' security through red-teaming is vital, yet white-box approaches are limited by their need for internal access, complicating their use with closed-source models. Moreover, existing black-box methods often as

Pierre Houedry, Nicolas Courty, Florestan Martin-Baillon, Laetitia Chapel

Trees and the associated shortest-path tree metrics provide a powerful framework for representing hierarchical and combinatorial structures in data. Given an arbitrary metric space, its deviation from a tree metric can be quantified by Gromov's $\delta$-hyperbolicity. Nonetheless, designing algorithms that bridge an arbitrary metric to its closest tree metri

Ekaterina Fadeeva, Aleksandr Rubashevskii, Dzianis Piatrashyn, Roman Vashurin

Large Language Models (LLMs) enhanced with retrieval, an approach known as Retrieval-Augmented Generation (RAG), have achieved strong performance in open-domain question answering. However, RAG remains prone to hallucinations: factually incorrect outputs may arise from inaccuracies in the model's internal knowledge and the retrieved context. Existing approac

Kai Pfeiffer

Hierarchical least-squares programming (HLSP) is an important tool in optimization as it enables the stacking of any number of priority levels in order to reflect complex constraint relationships, for example in physical systems like robots. Existing solvers typically address the primal formulation of HLSP's, which is computationally efficient due to sequent

Zeqing Wang, Bowen Zheng, Xingyi Yang, Zhenxiong Tan

Diffusion Transformer (DiT)-based video diffusion models generate high-quality videos at scale but incur prohibitive processing latency and memory costs for long videos. To address this, we propose a novel distributed inference strategy, termed DualParal. The core idea is that, instead of generating an entire video on a single GPU, we parallelize both tempor

Zhengmin Yu, Yuan Zhang, Ming Wen, Yinan Nie

Project building is pivotal to support various program analysis tasks, such as generating intermediate rep- resentation code for static analysis and preparing binary code for vulnerability reproduction. However, automating the building process for C/C++ projects is a highly complex endeavor, involving tremendous technical challenges, such as intricate depend

Anil Batra, Laura Sevilla-Lara, Marcus Rohrbach, Frank Keller

Procedural texts help AI enhance reasoning about context and action sequences. Transforming these into Semantic Role Labeling (SRL) improves understanding of individual steps by identifying predicate-argument structure like {verb,what,where/with}. Procedural instructions are highly elliptic, for instance, (i) add cucumber to the bowl and (ii) add sliced toma

Xiao Hu, Xingyu Lu, Liyuan Mao, YiFan Zhang

Reinforcement learning (RL) has played an important role in improving the reasoning ability of large language models (LLMs). Some studies apply RL directly to \textit{smaller} base models (known as zero-RL) and also achieve notable progress. However, in this paper, we show that using only 920 examples, a simple distillation method based on the base model can

Vojtěch Kala, Casper A. Breum, Mikkel V. Larsen, Ulrik L. Andersen

In continuous variable optical platforms, large-scale Gaussian cluster states have already been demonstrated, but non-Gaussian resources are essential to achieve universality and fault tolerance in measurement-based quantum computation. However, characterizing and certifying non-Gaussian cluster states remains an outstanding challenge. Here, we introduce a g

Pan Dong, Nathan C. Keim, Joseph D. Paulsen

When a ribbon or tube is twisted far enough it forms buckles and wrinkles. Its new geometry can be strikingly ordered, or hopelessly disordered. Here we study this process in a tube with hybrid boundary conditions: one end a cylinder, and the other end crimped flat like a ribbon, so that the sample resembles a toothpaste tube. The resulting irregular structu

M. O. Ajeesh, Joe D. Thompson, Eric D. Bauer, Filip Ronning

We report the temperature-pressure phase diagram of a UTe$_2$ single crystal that does not undergo a bulk superconducting transition but shows filamentary superconductivity with a critical transition temperature of 1 K at ambient pressure. Electrical-resistivity measurements reveal that the evolution of the filamentary superconducting state under pressure re

Federico Echenique, Michael Olabisi

This paper introduces a novel revealed-preference approach to ranking colleges and professional schools based on applicants' choices and standardized test scores. Unlike traditional rankings that rely on data supplied by institutions or expert opinions, our methodology leverages the decentralized beliefs of potential students, as revealed through their appli

Davide Lobba, Fulvio Sanguigni, Bin Ren, Marcella Cornia

Virtual try-on (VTON) has been widely explored for rendering garments onto person images, while its inverse task, virtual try-off (VTOFF), remains largely overlooked. VTOFF aims to recover standardized product images of garments directly from photos of clothed individuals. This capability is of great practical importance for e-commerce platforms, large-scale

Fatemeh Pesaran Zadeh, Yoojin Oh, Gunhee Kim

Aligning large VLMs with human preferences is a challenging task, as methods like RLHF and DPO often overfit to textual information or exacerbate hallucinations. Although augmenting negative image samples partially addresses these pitfalls, no prior work has employed listwise preference optimization for VLMs, due to the complexity and cost of constructing li

Peng Wang, Xiang Liu, Peidong Liu

Stylizing 3D scenes instantly while maintaining multi-view consistency and faithfully resembling a style image remains a significant challenge. Current state-of-the-art 3D stylization methods typically involve computationally intensive test-time optimization to transfer artistic features into a pretrained 3D representation, often requiring dense posed input

Masayuki Hamada, Masahiro Haze, Junya Okazaki, Yukio Hasegawa

Investigation of transport properties is fundamental for characterizing electronic properties and phase transitions. However, most of the transport measurements on conductive layers have been performed at macroscopic scales, and thus the development of microscopic methods to measure transport is important. Scanning tunneling potentiometry (STP) is a powerful

Andrew Parry, Debasis Ganguly, Sean MacAvaney

The training process of ranking models involves two key data selection decisions: a sampling strategy, and a labeling strategy. Modern ranking systems, especially those for performing semantic search, typically use a ``hard negative'' sampling strategy to identify challenging items using heuristics and a distillation labeling strategy to transfer ranking "kn

Haixin Zhao, Nilesh Madhu

In speech enhancement, achieving state-of-the-art (SotA) performance while adhering to the computational constraints on edge devices remains a formidable challenge. Networks integrating stacked temporal and spectral modelling effectively leverage improved architectures such as transformers; however, they inevitably incur substantial computational complexity

Alina Mreńca-Kolasińska, Christophe De Beule, Jia-Tong Shi, Aitor Garcia-Ruiz

In graphene, long-wavelength deformations that result in elastic shear strain couple to the low-energy Dirac electrons as pseudogauge fields. Using a scalable tight-binding model, we consider analogs to magnetotransport in mesoscopic strained graphene devices with nearly uniform pseudomagnetic fields. In particular, we consider transverse pseudomagnetic focu

Kaiming Liu, Xuanyu Lei, Ziyue Wang, Peng Li

Large language model (LLM) agents have shown impressive reasoning capabilities in interactive decision-making tasks. These agents interact with environment through intermediate interfaces, such as predefined action spaces and interaction rules, which mediate the perception and action. However, mismatches often happen between the internal expectations of the

Tomislav Miškić, Juraj Krsnik, Stefano Ragni, Andrey S. Mishchenko

We present an iterative method for generating the complete set of self-energy Feynman diagrams at arbitrary order for the single-polaron problem with arbitrary linear coupling to the lattice. The approach combines a combinatorial representation of noncrossing diagrams, based on Dyck paths associated with Stieltjes-Rogers polynomials, with the constraints of

Goran B. Scharmer, Bo Lindberg

We describe a proposal for the optical design of three dual Fabry-Perot based narrowband filter systems for for the future European Solar Telescope (EST). These are intended to constitute the core elements of three imaging spectropolarimeters, foreseen to become amongst the most important science instruments for EST. The designs proposed here rely heavily on

Yiping Liu, Yi Zhou, Zhenxiang Xu, Mingyu Xiao

The Generalized Independent Set (GIS) problem extends the classical maximum independent set problem by incorporating profits for vertices and penalties for edges. This generalized problem has been identified in diverse applications in fields such as forest harvest planning, competitive facility location, social network analysis, and even machine learning. Ho

Jianmin Liu, Li Yan, Borui Li, Lei Yu

Federated fine-tuning is critical for improving the performance of large language models (LLMs) in handling domain-specific tasks while keeping training data decentralized and private. However, prior work has shown that clients' private data can actually be recovered via gradient inversion attacks. Existing privacy preservation techniques against such attack

Xin Yang, Jiantao Lin, Yingjie Xu, Haodong Li

Despite the availability of large-scale 3D datasets and advancements in 3D generative models, the complexity and uneven quality of 3D geometry and texture data continue to hinder the performance of 3D generation techniques. In most existing approaches, 3D geometry and texture are generated in separate stages using different models and non-unified representat

Adebola K. Ojo, Ifechukwude Jude Okafor

Investors and stock market analysts face major challenges in predicting stock returns and making wise investment decisions. The predictability of equity stock returns can boost investor confidence, but it remains a difficult task. To address this issue, a study was conducted using a Long Short-term Memory (LSTM) model to predict future stock market movements

Dehao Wang, Haohang Zhu, Yiwen Xu, Kaiqi Liu

Road potholes pose a serious threat to driving safety and comfort, making their detection and assessment a critical task in fields such as autonomous driving. When driving vehicles, the operators usually avoid large potholes and approach smaller ones at reduced speeds to ensure safety. Therefore, accurately estimating pothole area is of vital importance. Mos

Sanskriti Agrawal, Raktim Abir

We present the complete eigenvalue spectrum of the partially transposed density matrix for a pure bipartite quantum state acting on a generic $2 \otimes n$ Hilbert space. The spectrum contains four non-zero eigenvalues, as, \begin{eqnarray} \lambda_{1,2}=\pm \sqrt{A}, ~~~ \lambda_{3,4}= \frac{1}{2}(1\pm\sqrt{1-4 A}), \nonumber \end{eqnarray} where $A$ is the

Elias R. Udnæs, Tiago M. D. Pereira

Understanding energy transfer through the chromosphere is paramount to solving the coronal heating problem. We investigated the energy dissipation of acoustic waves in the chromosphere of the quiet Sun using 3D radiative magnetohydrodynamic (rMHD) simulations. We analysed the characteristics of acoustic-wave heating and its dependence on height and magnetic

Zhenling Chen, Haiwei Fu, Zhiguo Zeng

Digital twins offer a promising solution to the lack of sufficient labeled data in deep learning-based fault diagnosis by generating simulated data for model training. However, discrepancies between simulation and real-world systems can lead to a significant drop in performance when models are applied in real scenarios. To address this issue, we propose a fa

Lingyi Cai, Ruichen Zhang, Changyuan Zhao, Yu Zhang

Low-Altitude Economic Networking (LAENet) aims to support diverse flying applications below 1,000 meters by deploying various aerial vehicles for flexible and cost-effective aerial networking. However, complex decision-making, resource constraints, and environmental uncertainty pose significant challenges to the development of the LAENet. Reinforcement learn

Yi-Qing Zhang, Liang-Jun He, Han Pu, Zheng-Wei Zhou

We propose generating long-range and nonreciprocal three-body interactions in quantum gases via optical feedback. By placing a quasi-two-dimensional Bose-Einstein condensate (BEC) in front of two reflecting mirrors and illuminating it with dichromatic laser beams, these driving optical fields traverse the BEC twice, thereby inducing a feedback effect on the

Sam O'Connor Russell, Naomi Harte

Turn-taking is richly multimodal. Predictive turn-taking models (PTTMs) facilitate naturalistic human-robot interaction, yet most rely solely on speech. We introduce MM-VAP, a multimodal PTTM which combines speech with visual cues including facial expression, head pose and gaze. We find that it outperforms the state-of-the-art audio-only in videoconferencing

Martin Erdmann, Niklas Langner, Josina Schulte, Dominik Wirtz

Transformer networks excel in scientific applications. We explore two scenarios in ultra-high-energy cosmic ray simulations to examine what these network architectures learn. First, we investigate the trained positional encodings in air showers which are azimuthally symmetric. Second, we visualize the attention values assigned to cosmic particles originating

Weihang Liu, Yuhui Zhong, Yuke Li, Xi Chen

Accurate and efficient modeling of large-scale urban scenes is critical for applications such as AR navigation, UAV based inspection, and smart city digital twins. While aerial imagery offers broad coverage and complements limitations of ground-based data, reconstructing city-scale environments from such views remains challenging due to occlusions, incomplet

Wei Chen, Zhao Zhang, Meng Yuan, Kepeng Xu

In this paper, we address the task of targeted sentiment analysis (TSA), which involves two sub-tasks, i.e., identifying specific aspects from reviews and determining their corresponding sentiments. Aspect extraction forms the foundation for sentiment prediction, highlighting the critical dependency between these two tasks for effective cross-task knowledge

Louis Allain, Sébastien da Veiga, Brian Staber

Conformal Prediction (CP) is a popular framework for constructing prediction bands with valid coverage in finite samples, while being free of any distributional assumption. A well-known limitation of conformal prediction is the lack of adaptivity, although several works introduced practically efficient alternate procedures. In this work, we build upon recent

Faruk Alpay

This paper presents a formal categorical proof that exam-driven educational systems obstruct identity emergence and block creative convergence. Using the framework of Alpay Algebra II and III, we define Exam-Grade Collapse Systems (EGCS) as functorial constructs where learning dynamics $\varphi$ are recursively collapsed by evaluative morphisms $E$. We prove

Bin Qin, Qirui Ji, Jiangmeng Li, Yupeng Wang

Self-supervised topological deep learning (TDL) represents a nascent but underexplored area with significant potential for modeling higher-order interactions in simplicial complexes and cellular complexes to derive representations of unlabeled graphs. Compared to simplicial complexes, cellular complexes exhibit greater expressive power. However, the advancem

Palle E. T. Jorgensen, James Tian

We introduce and study a class $\mathcal{M}$ of generalized positive definite kernels of the form $K\colon X\times X\to L(\mathfrak{A},L(H))$, where $\mathfrak{A}$ is a unital $C^{*}$-algebra and $H$ a Hilbert space. These kernels encode operator-valued correlations governed by the algebraic structure of $\mathfrak{A}$, and generalize classical scalar-valued

Aiyue Chen, Bin Dong, Jingru Li, Jing Lin

Video generation using diffusion models is highly computationally intensive, with 3D attention in Diffusion Transformer (DiT) models accounting for over 80\% of the total computational resources. In this work, we introduce {\bf RainFusion}, a novel training-free sparse attention method that exploits inherent sparsity nature in visual data to accelerate atten