May 2025 arXiv papers — page 13

Federico Milanesio, Matteo Santoro, Pietro G. Fré, Guido Sanguinetti

Hyperbolic deep learning leverages the metric properties of hyperbolic spaces to develop efficient and informative embeddings of hierarchical data. Here, we focus on the solvable group structure of hyperbolic spaces, which follows naturally from their construction as symmetric spaces. This dual nature of Lie group and Riemannian manifold allows us to propose

Chiara Meroni, Jared Miller, Mauricio Velasco

We introduce polystar bodies: compact starshaped sets whose gauge or radial functions are expressible by polynomials, enabling tractable computations, such as that of intersection bodies. We prove that polystar bodies are uniformly dense in starshaped sets and obtain asymptotically optimal approximation guarantees. We develop tools for the construction of po

Peng Qi, Wenxi Qu, Tianliang Yao, Haonan Ma

Percutaneous Coronary Intervention (PCI) is a minimally invasive procedure that improves coronary blood flow and treats coronary artery disease. Although PCI typically requires 2D X-ray angiography (XRA) to guide catheter placement at real-time, computed tomography angiography (CTA) may substantially improve PCI by providing precise information of 3D vascula

Lorenzo Calibbi, Claudia Hagedorn, Michael A. Schmidt, James Vandeleur

We systematically investigate the possible phenomenological impact of residual flavour groups in the charged lepton sector. We consider all possible flavour charge assignments for abelian residual symmetries up to Z8. The allowed flavour structures of operators in Standard Model Effective Field Theory (up to dimension six) lead to distinctive and observable

G. Sakharova, A. Krokhmal, R. Galimova, A. Khatmullina

Objective: The aim of this study was to evaluate the efficacy of alendronate therapy in improving bone density distribution in skull bones and corresponding ultrasound permeability in patients who had previously experienced unsuccessful transcranial MR-guided focused ultrasound (MRgFUS) ablation. The ability of alendronate treatment to modify skull bone char

Taku Yamazaki, Kaito Watanabe, Tatsuya Kase, Kenta Hasegawa

In this article, we propose a 3D mobile crowdsensing (3D-MCS) framework aimed at sustainable urban digital twins (UDTs). The framework comprises four key mechanisms: (1) the 3D-MCS mechanism, consisting of active and passive models; (2) the Geohash-based spatial information management mechanism; (3) the dynamic point cloud integration mechanism for UDTs; and

Yunze Lin

Mathematical reasoning presents a significant challenge for Large Language Models (LLMs) as it requires ensuring the correctness of each reasoning step. Researchers have been strengthening the mathematical reasoning abilities of LLMs through supervised fine-tuning, but due to the inability to suppress incorrect outputs, illusions can easily arise. Recently,

Yangui Fang, Baixu Chen, Jing Peng, Xu Li

Automatic Speech Recognition (ASR) error correction aims to correct recognition errors while preserving accurate text. Although traditional approaches demonstrate moderate effectiveness, LLMs offer a paradigm that eliminates the need for training and labeled data. However, directly using LLMs will encounter hallucinations problem, which may lead to the modif

Ziyi Wang, Zhi Gao, Boxuan Yu, Zirui Dai

Video understanding has made remarkable progress in recent years, largely driven by advances in deep models and the availability of large-scale annotated datasets. However, existing works typically ignore the inherent domain shifts encountered in real-world video applications, leaving domain generalization (DG) in video understanding underexplored. Hence, we

Jiangnan Xiong

Using a bivariant version of cohomological correspondences, we establish a categorical trace-like formula for the non-acyclicity classes introduced by Yang and Zhao (arXiv:2209.11086). As an application, we prove the additivity for the non-acyclicity classes.

Dominik Erb, Antoine Gérardin, Harvey B. Meyer, Julian Parrino

We compute the isospin-violating part $a_\mu^{\text{HVP}, 38}$ of the hadronic-vacuum-polarization (HVP) contribution to the muon $(g-2)$ in lattice QCD at the SU$(3)_{\rm f}$-symmetric point where $M_\pi=M_K\simeq 416$ MeV. All diagrams involving internal photons are evaluated in coordinate space, employing a Pauli-Villars-regulated photon propagator with a

Cheng Chen, C. J. Nixon

Simulations using the Smoothed Particle Hydrodynamics (SPH) technique typically include numerical viscosity to model shocks and maintain particle order on the kernel scale. This numerical viscosity is composed of linear and quadratic terms, with coefficients $\alpha_{\rm SPH}$ and $\beta_{\rm SPH}$ respectively. Setting these coefficients too high results in

Fanhang Man, Xiaoyue Chen, Huandong Wang, Baining Zhao

Images shared online strongly influence emotions and public well-being. Understanding the emotions an image elicits is therefore vital for fostering healthier and more sustainable digital communities, especially during public crises. We study Visual Emotion Elicitation (VEE), predicting the set of emotions that an image evokes in viewers. We introduce VAEER,

Shujian Yang, Shiyao Cui, Chuanrui Hu, Haicheng Wang

Detecting toxic content using language models is important but challenging. While large language models (LLMs) have demonstrated strong performance in understanding Chinese, recent studies show that simple character substitutions in toxic Chinese text can easily confuse the state-of-the-art (SOTA) LLMs. In this paper, we highlight the multimodal nature of Ch

Gilles Quentin Hacheme, Girmaw Abebe Tadesse, Caleb Robinson, Akram Zaytar

Classifying geospatial imagery remains a major bottleneck for applications such as disaster response and land-use monitoring-particularly in regions where annotated data is scarce or unavailable. Existing tools (e.g., RS-CLIP) that claim zero-shot classification capabilities for satellite imagery nonetheless rely on task-specific pretraining and adaptation t

Yitaek Kim, Christoffer Sloth

This paper investigates how learning can be used to ease the design of high-quality paths for the assembly of deformable objects. Object dynamics plays an important role when manipulating deformable objects; thus, detailed models are often used when conducting motion planning for deformable objects. We propose to use human demonstrations and learning to enab

Rahul Sinha, Thomas E. Browder, N. G. Deshpande, Dibyakrupa Sahoo

The observation of $CP$ violation in the difference of $CP$ asymmetries between $D\to K^+K^-$ and $D\to \pi^+\pi^-$ has raised a debate whether the observed asymmetries can be regarded as a signal of physics beyond the standard model (SM). In this paper we obtain all the topological amplitudes and isospin amplitudes directly from measured observables for $D\

Harrison Nicholls, Raymond Pierrehumbert, Tim Lichtenberg

It is important that we are able to accurately model the atmospheres of (exo)planets. This is because atmospheres play a central role in setting a planet's thermochemical environment at a given point in time, and also in regulating how it evolves over geological timescales. Additionally, it is primarily by observation of their atmospheres that we are able to

Sumit Vohra

Automated Market Makers (AMMs) are decentralized exchange protocols that provide continuous access to token liquidity without the need for order books or traditional market makers. However, this innovation has failed to scale when it comes to cross-chain swaps. Modern cross-chain swaps employ double-sided AMMs, which are not only inefficient due to liquidity

Minsu Kang, Seolhee Lee, Choonghyeon Lee, Namhyun Cho

Human to non-human voice conversion (H2NH-VC) transforms human speech into animal or designed vocalizations. Unlike prior studies focused on dog-sounds and 16 or 22.05kHz audio transformation, this work addresses a broader range of non-speech sounds, including natural sounds (lion-roars, birdsongs) and designed voice (synthetic growls). To accomodate generat

Julius Bohm, Hugo Gerlitz, Christina Jörg, Michael Fleischhauer

One of the hallmarks of topological systems is the robust quantization of particle transport. It is the origin of the integer-valued quantum Hall conductivity and a potential tool for quantum information technology. Recent experiments on topological pumps constructed by using arrays of photonic waveguides and described by the (lattice-translational invariant

Yifu Wang, Lei Ni, Guanchong Cheng, Jialiang Hu

Oscillatory magnetic reconnection is a periodic magnetic reconnection process, during which the current sheet's orientation and the magnetic connections change periodically. This periodic variation is generally considered to originate from the magnetic reconnection itself rather than from external driving processes. We conduct 2.5-dimensional radiative magne

Uzair Khan, Franco Fummi, Luigi Capogrosso

In the era of intelligent manufacturing, anomaly detection has become essential for maintaining quality control on modern production lines. However, while many existing models show promising performance, they are often too large, computationally demanding, and impractical to deploy on resource-constrained embedded devices that can be easily installed on the

Wenxuan Shi, Haochen Tan, Chuqiao Kuang, Xiaoguang Li

Information seeking demands iterative evidence gathering and reflective reasoning, yet large language models (LLMs) still struggle with it in open-web question answering. Existing prompting and supervised fine-tuning (SFT) methods remain fixed by prompt rules or training corpora, and are usually benchmarked only on well-structured wiki sources, limiting real

Fanhang Man, Huandong Wang, Jianjie Fang, Zhaoyi Deng

User sentiment on social media reveals the underlying social trends, crises, and needs. Researchers have analyzed users' past messages to trace the evolution of sentiments and reconstruct sentiment dynamics. However, predicting the imminent sentiment of an ongoing event is rarely studied. In this paper, we address the problem of \textbf{sentiment forecasting

Yingsen Zeng, Zepeng Huang, Yujie Zhong, Chengjian Feng

Despite advances in general video understanding, Video Large Language Models (Video-LLMs) face challenges in precise temporal localization due to discrete time representations and limited temporally aware datasets. Existing methods for temporal expression either conflate time with text-based numerical values, add a series of dedicated temporal tokens, or reg

Fulvio Gesmundo, Alexandros Grosdos, André Uschmajew

We show that one can always identify a point on an algebraic variety $X$ uniquely with $\dim X +1$ generic linear measurements taken themselves from a variety under minimal assumptions. As illustrated by several examples the result is sharp, that is, $\dim X$ measurements are in general not enough for unique identifiability.

Jingjing Liu, Jiashun Jin, Xianchao Xiu, Jianhua Zhang

Remote sensing image (RSI) denoising is an important topic in the field of remote sensing. Despite the impressive denoising performance of RSI denoising methods, most current deep learning-based approaches function as black boxes and lack integration with physical information models, leading to limited interpretability. Additionally, many methods may struggl

Andrei Caragea, Dae Gwan Lee, Romanos Malikiosis, Goetz E. Pfander

Chebotarev's theorem on roots of unity states that all minors of a Fourier matrix are non-zero if and only if the order of the matrix is prime. We establish cases in which all principal minors of Fourier matrices of square-free order are non-zero. In a subsequent paper we discuss the case of composites containing squares.

Surjit Kumar, Milan Kumar Mal, Paramita Pramanick

We provide a description of the Shilov boundary of the classical Cartan domain in terms of Jordan triple determinant. As a consequence, we obtained an intrinsic characterization of Cartan isometries. Further, we obtain (i) invariance of Cartan isometries under the action of the biholomorphic automorphism group, and (ii) a Brown-Halmos type condition for Toep

Ivan Petrukha, Yana Kurliak, Nataliia Stulova

In recent years, large language models (LLMs) have showcased significant advancements in code generation. However, most evaluation benchmarks are primarily oriented towards Python, making it difficult to evaluate other programming languages, such as Swift, with high quality. By examining widely established multilingual benchmarks like HumanEval-XL and MultiP

Yanling Chen, Liyi Gu, Aurora Simionescu, Chunyang Jiang

The galaxy cluster pair 1E2216.0-0401 and 1E2215.7-0404 represents a major cluster merger in its early stages, a phase that has been scarcely explored in previous studies. Within this system, both axial and equatorial merger shocks have been identified. Recent XMM-Newton observations of the southern region of the cluster pair have increased the total exposur

Mikołaj Rogalski, Juan Martinez-Carranza, Bartosz Górski, Piotr Arcab

Fourier ptychographic microscopy (FPM) is a pivotal computational imaging technique that achieves phase and amplitude reconstruction with high resolution and wide field of view, using low numerical aperture objectives and LED array illumination. Despite its unique strengths, FPM remains fundamentally limited in retrieving low spatial frequency phase informat

Yuanyuan Wang, Tianze Wei

In an online fair allocation problem, a sequence of indivisible items arrives online and needs to be allocated to offline agents immediately and irrevocably. In our paper, we study the online allocation of either goods or chores. We employ popular fairness notions, including envy-freeness up to one item (EF1) and maximin share fairness (MMS) to capture fairn

Luca Tognoni, Neil Reichlin, Edoardo Ghignone, Nicolas Baumann

Reactive controllers for autonomous racing avoid the computational overhead of full ee-Think-Act autonomy stacks by directly mapping sensor input to control actions, eliminating the need for localization and planning. A widely used reactive strategy is FTG, which identifies gaps in LiDAR range measurements and steers toward a chosen one. While effective on f

Yuntao Shi, Yi Luo, Yeyun Gong, Chen Lin

Large Language Models (LLMs) have achieved remarkable success in various domains. However, when handling long-form text modification tasks, they still face two major problems: (1) producing undesired modifications by inappropriately altering or summarizing irrelevant content, and (2) missing necessary modifications to implicitly related passages that are cru

Tsan Huang, Zhiyuan Sun

Optical cavities have been widely applied to manipulate the properties of solid state materials inside them. We propose that in systems embedded within optical cavities driven by incident pump light, the pump induces generic phase transitions into new nonequilibrium steady states. This effect arises from the ponderomotive potential, the effective static pote

Yongming Chen, Miner Chen, Liewen Liao, Mingyang Jiang

Reinforcement learning (RL) in autonomous driving employs a trial-and-error mechanism, enhancing robustness in unpredictable environments. However, crafting effective reward functions remains challenging, as conventional approaches rely heavily on manual design and demonstrate limited efficacy in complex scenarios. To address this issue, this study introduce

Paritosh Ghosh, Hemangi Madhusudan Shah, Arindam Bhattacharyya

In this article, we introduce $\omega$-Bach tensor corresponding to one form $\omega$ and correspondingly introduce almost $\omega$-Bach solitons, thereby generalizing the existing notion of Bach tensor and almost Bach solitons. We characterize almost $\omega$-Bach solitons, when the potential vector field of the soliton generates an infinitesimal harmonic t

Jinlu Zhang, Yixin Chen, Zan Wang, Jie Yang

Recent advances in 3D human-aware generation have made significant progress. However, existing methods still struggle with generating novel Human Object Interaction (HOI) from text, particularly for open-set objects. We identify three main challenges of this task: precise human-object relation reasoning, affordance parsing for any object, and detailed human

Peijie Chen, Wenhao Guan, Kaidi Wang, Weijie Wu

Neural speech codecs are essential for advancing text-to-speech (TTS) systems. With the recent success of large language models in text generation, developing high-quality speech tokenizers has become increasingly important. This paper introduces DS-Codec, a novel neural speech codec featuring a dual-stage training framework with mirror and non-mirror archit

Gengze Xu, Wei Yao, Ziqiao Wang, Yong Liu

Weak-to-strong generalization (W2SG) is the phenomenon in which a powerful student model, trained on labels produced by a weaker teacher, ultimately outperforms the teacher on the target task. In this work, we theoretically investigate how W2SG can arise via a generalized bias-variance decomposition under Bregman divergence. We show that the expected populat

Yirui Zhan, Wen Nie, Jun Gao

Accurate cardinality estimation of substring queries, which are commonly expressed using the SQL LIKE predicate, is crucial for query optimization in database systems. While both rule-based methods and machine learning-based methods have been developed to optimize various aspects of cardinality estimation, their absence of error bounds may result in substant

Yi Gu

T-distributed stochastic neighbor embedding (t-SNE) is a well-known algorithm for visualizing high-dimensional data by finding low-dimensional representations. In this paper, we study the convergence of t-SNE with generalized kernels and extend the results of Auffinger and Fletcher in 2023. Our work starts by giving a concrete formulation of generalized inpu

Jiayan Li, Jun Li, Zhourui Zhang, Jianhua Xu

In knowledge distillation (KD), logit distillation (LD) aims to transfer class-level knowledge from a more powerful teacher network to a small student model via accurate teacher-student alignment at the logits level. Since high-confidence object classes usually dominate the distillation process, low-probability classes which also contain discriminating infor

Christian Gang Liu

To alleviate difficulties in writing smart contracts for distributed blockchain applications, as other research, we propose transformation of Business Process Model and Notation (BPMN) models into blockchain smart contracts. Unlike other research, we use Discrete Event Hierarchical State Machine (DE-HSM) multi-modal modeling to identify collaborative trade t

Cheng-An Hsieh, Tomotsugu Goto, Chih-Teng Ling, Seong Jin Kim

This study presents the black hole accretion history (BHAH) of obscured active galactic nuclei (AGNs) identified from the JWST CEERS survey by Chien et al. (2024) using mid-infrared (MIR) SED fitting. We compute black hole accretion rates (BHARs) to estimate the black hole accretion density (BHAD), $\rho_{L_{\mathrm{disk}}}$, across $0 < z < 4.25$. MIR lumin

Weihao Mao, Yang Lu, Yanqing Xu, Bo Ai

Recently, a novel flexible-antenna technology, called pinching antennas, has attracted growing academic interest. By inserting discrete dielectric materials, pinching antennas can be activated at arbitrary points along waveguides, allowing for flexible customization of large-scale path loss. This paper investigates a multi-waveguide pinching-antenna integrat

Kechen Li, Yaotian Tao, Ximing Wen, Quanwei Sun

Recent advancements in Large Language Models (LLMs) have demonstrated their potential in planning and reasoning tasks, offering a flexible alternative to classical pathfinding algorithms. However, most existing studies focus on LLMs' independent reasoning capabilities and overlook the potential synergy between LLMs and traditional algorithms. To fill this ga

Mingxu Zhang, Xiaoqi Li, Jiahui Xu, Kaichen Zhou

Recent advancements in 3D robotic manipulation have improved grasping of everyday objects, but transparent and specular materials remain challenging due to depth sensing limitations. While several 3D reconstruction and depth completion approaches address these challenges, they suffer from setup complexity or limited observation information utilization. To ad

Haopeng Geng, Daisuke Saito, Nobuaki Minematsu

Evaluating L2 speech intelligibility is crucial for effective computer-assisted language learning (CALL). Conventional ASR-based methods often focus on native-likeness, which may fail to capture the actual intelligibility perceived by human listeners. In contrast, our work introduces a novel, perception based L2 speech intelligibility indicator that leverage

Y. Tian, I. Grytsenko, A. Jennings, J. Wang

Electrons floating on a solid neon exhibit long charge coherence times, making them attractive for hybrid quantum systems. When combined with high-quality, high-impedance superconducting resonators and a local magnetic field gradient, this platform enables strong charge--photon and spin--charge coupling-key ingredients for scalable spin qubit architectures.

Enshang Zhang, Zhicheng Zhang, Takashi Hanakawa

Reconstructing visual stimuli from EEG signals is a crucial step in realizing brain-computer interfaces. In this paper, we propose a transformer-based EEG signal encoder integrating the Discrete Wavelet Transform (DWT) and the gating mechanism. Guided by the feature alignment and category-aware fusion losses, this encoder is used to extract features related

Anurag Banerjee, Emile Pangburn, Catherine Pépin, Cristina Bena

In this work, we investigate impurity-induced Friedel oscillations in the doped two-dimensional Hubbard model, focusing on the role of holon and doublon excitations. We show that weak impurities, due to the non-fermionic nature of the underlying quasiparticles, induce Friedel oscillations whose behavior is consistent with an effective non-interacting theory

Wei Fu, Jiaxuan Gao, Xujie Shen, Chen Zhu

Reinforcement learning (RL) has become a dominant paradigm for training large language models (LLMs), particularly for reasoning tasks. Effective RL for LLMs requires massive parallelization and poses an urgent need for efficient training systems. Most existing large-scale RL systems for LLMs are synchronous, alternating generation and training in a batch se

Hyundong Jin, Sicheol Sung, Shinwoo Park, SeungYeop Baik

The reasoning, writing, text-editing, and retrieval capabilities of proprietary large language models (LLMs) have advanced rapidly, providing users with an ever-expanding set of functionalities. However, this growing utility has also led to a serious societal concern: the over-reliance on LLMs. In particular, users increasingly delegate tasks such as homewor

Abiel Costa Macedo, José Francisco de Oliveira, Fábio Sodré Rocha

Let $W^{m,\frac{n}{m}}(\mathbb{R}^n)$ with $1\le m < n$ be the standard higher order derivative Sobolev space in the critical exponential growth threshold. We investigate a new Adams-Adimurthi-Druet type inequality on the whole space $\mathbb{R}^n$ which is strongly influenced by the vanishing phenomenon. Specifically, we prove \begin{equation}\nonumber \sup

Quinn Lanners, Cynthia Rudin, Alexander Volfovsky, Harsh Parikh

Data fusion techniques integrate information from heterogeneous data sources to improve learning, generalization, and decision making across data sciences. In causal inference, these methods leverage rich observational data to improve causal effect estimation, while maintaining the trustworthiness of randomized controlled trials. Existing approaches often re

Yongzhou Chen, Muhammad Taimoor Tariq, Haitham Hassanieh, Radhika Mittal

With increasing density of small cells in modern multi-cell deployments, a given user can have multiple options for its serving cell. The serving cell for each user must be carefully chosen such that the user achieves reasonably high channel quality from it, and the load on each cell is well balanced. It is relatively straightforward to reason about this wit

Yi Zou, Mengjiao Wang, Xinan Zhang, Herbert Ho-Ching Iu

Simulating brain functions using neural networks is an important area of research. Recently, discrete memristor-coupled neurons have attracted significant attention, as memristors effectively mimic synaptic behavior, which is essential for learning and memory. This highlights the biological relevance of such models. This study introduces a discrete memristiv

James R. Golden

Despite significant progress in transformer interpretability, an understanding of the computational mechanisms of large language models (LLMs) remains a fundamental challenge. Many approaches interpret a network's hidden representations but remain agnostic about how those representations are generated. We address this by mapping LLM inference for a given inp

Yueqi Zhang, Peiwen Yuan, Shaoxiong Feng, Yiwei Li

Human-AI conversation frequently relies on quoting earlier text-"check it with the formula I just highlighted"-yet today's large language models (LLMs) lack an explicit mechanism for locating and exploiting such spans. We formalise the challenge as span-conditioned generation, decomposing each turn into the dialogue history, a set of token-offset quotation s

Kaidi Wang, Wenhao Guan, Ziyue Jiang, Hukai Huang

Currently, zero-shot voice conversion systems are capable of synthesizing the voice of unseen speakers. However, most existing approaches struggle to accurately replicate the speaking style of the source speaker or mimic the distinctive speaking style of the target speaker, thereby limiting the controllability of voice conversion. In this work, we propose Di

Kareem Shehata, Han Fangqi, Sri AravindaKrishnan Thyagarajan

Traditionally, threshold secret sharing (TSS) schemes assume all parties have equal weight, yet emerging systems like blockchains reveal disparities in party trustworthiness, such as stake or reputation. Weighted Secret Sharing (WSS) addresses this by assigning varying weights to parties, ensuring security even if adversaries control parties with total weigh

Chun Liu, Guanghui Hu, Tao Yin, Bo Zhang

Consider a time-harmonic elastic point source incident on a bounded obstacle which is embedded in an open space filled with a homogeneous and isotropic elastic medium. This paper is concerned with the inverse problem of recovering the location and shape of the obstacle from near-field data generated by infinitely many incident point source waves at a fixed e

Ege Özsoy, Arda Mamur, Felix Tristram, Chantal Pellegrini

Operating rooms (ORs) demand precise coordination among surgeons, nurses, and equipment in a fast-paced, occlusion-heavy environment, necessitating advanced perception models to enhance safety and efficiency. Existing datasets either provide partial egocentric views or sparse exocentric multi-view context, but do not explore the comprehensive combination of

Jonathan Morgner, Vladimir A. Yerokhin, Charlotte M. Konig, Fabian Heiße

Magnetic moments of bound-electron systems are a sensitive tool for testing fundamental interactions. $g$ factors of lithium-like ions have been rigorously studied in recent years, enabling insights into the relativistic inter-electronic effects. Here, we present the $g$-factor measurement of lithium-like tin, accurate to 0.5 parts per billion, as well as \t

Claudio Sanavio, Fabio Mascherpa, Alessia Marruzzo, Alfonso Amendola

We propose a revisited variational quantum solver for linear systems, designed to circumvent the barren plateau phenomenon by combining two key techniques: adiabatic evolution and warm starts. To this end, we define an initial Hamiltonian with a known ground state which is easily implemented on the quantum circuit, and then "adiabatically" evolve the Hamilto

Gordon Y. Liao, Ziming Zeng, Mira Belenkiy, Jacob Hirshman

This paper introduces a fraud-deterrent access validation system for public blockchains, leveraging two complementary concepts: "Transaction Proximity", which measures the distance between wallets in the transaction graph, and "Easily Attainable Identities (EAIs)", wallets with direct transaction connections to centralized exchanges. Recognizing the limitati

Hang Du, Yanxin Zhou

For any integers $d,q\ge 3$, we consider the $q$-state ferromagnetic Potts model with an external field on a sequence of expander graphs that converges to the $d$-regular tree $\mathtt{T}_d$ in the Benjamini-Schramm sense. We show that along the critical line, any subsequential local weak limit of the Potts measures is a mixture of the free and wired Potts G

Zirui Shang, Xinxiao Wu, Shuo Yang

Language-driven action localization in videos requires not only semantic alignment between language query and video segment, but also prediction of action boundaries. However, the language query primarily describes the main content of an action and usually lacks specific details of action start and end boundaries, which increases the subjectivity of manual b

Yang Sui, Qi Xu, Yang Bai, Annie Qu

Multi-task learning (MTL) has become an essential machine learning tool for addressing multiple learning tasks simultaneously and has been effectively applied across fields such as healthcare, marketing, and biomedical research. However, to enable efficient information sharing across tasks, it is crucial to leverage both shared and heterogeneous information.

Joshua Carlo A. Casapao, Ananda G. Maity, Naphan Benchasattabuse, Michal Hajdušek

With the advent of practical quantum communication networks drawing closer, there is a growing need for reliable estimation protocols that can efficiently characterize quantum resources with minimum resource overhead requirement. A novel approach to this problem is to integrate an estimator into an existing network task, thereby removing the need for an addi

Yu-An Liu, Ruqing Zhang, Jiafeng Guo, Maarten de Rijke

Robustness and Effectiveness are critical aspects of developing dense retrieval models for real-world applications. It is known that there is a trade-off between the two. Recent work has addressed scaling laws of effectiveness in dense retrieval, revealing a power-law relationship between effectiveness and the size of models and data. Does robustness follow

C. Kuckein, M. Collados, A. Asensio Ramos, C. J. Díaz Baso

We study the chromospheric LOS velocities during the GOES M3.2 flare (SOL2013-05-17T08:43) using simultaneous spectroscopic data of the He I 1083.0 nm triplet and Ca II 854.2 nm line. A filament was present in the flaring area. The observational data were acquired with the VTT (Tenerife, Spain) and covered the pre-flare, flare, and post-flare phases. Spectro

Dipanjan Chaudhuri, Eric D'Asaro

Rain affects the buoyancy of the upper ocean in two ways: The freshwater flux in rain makes the water fresher and lighter, stabilizing the ocean (a negative buoyancy flux). The convective systems that produce rain are often accompanied by cold, dry air, often called 'cold pools', and reduced short-wave radiation, which makes the water colder and heavier, des

Haoyang Lin, Wenguo Zhu, Yongchun Zhong, Huihui Lu

We present a lithium niobate multi-functional platform (LN-MFP) that integrates photoacoustic and thermoelastic spectroscopy with photodetection on a single chip. Utilizing LN's piezoelectric and thermoelastic properties and an optimized design, it achieves high sensitivity across visible to long-wave infrared wavelengths. We demonstrate photoacoustic/thermo

Rui Li, Junfeng Kang, Qi Liu, Liyang He

In the pursuit of enhancing software reusability and developer productivity, code search has emerged as a key area, aimed at retrieving code snippets relevant to functionalities based on natural language queries. Despite significant progress in self-supervised code pre-training utilizing the vast amount of code data in repositories, existing methods have pri

Hongyi James Cai, Junlin Wang, Xiaoyin Chen, Bhuwan Dhingra

Recent advancements in large language models (LLMs) suggest that reinforcement learning (RL) effectively internalizes search strategies, yielding significant improvements on challenging reasoning tasks through extended chains of thought. While backtracking is widely viewed as the core mechanism enabling this improvement, its precise dynamics and how SFT and

J. Morgner, B. Tu, M. Moretti, C. M. König

In the ALPHATRAP experiment, the $g$ factor of boron-like $^{118}\mathrm{Sn}^{45+}$ has been measured with a $0.5$ parts-per-billion uncertainty. This is the first high-precision measurement of a heavy boron-like $g$ factor. The measured value of $0.644\,703\,826\,5(4)$ is consistent with the presented \textit{ab initio} state-of-the-art theory calculations,

Tadahiro Oh, Yuzhao Wang

In a seminal paper (1996), Bourgain proved invariance of the Gibbs measure for the defocusing cubic nonlinear Schr\"odinger equation on the two-dimensional torus by constructing local-in-time solutions in a probabilistic manner. In this note, we revisit and streamline his argument, using the random tensor estimate developed by Deng, Nahmod, and Yue (2022).

Jiayu Yao, Shenghua Liu, Yiwei Wang, Lingrui Mei

Multimodal Retrieval-Augmented Generation (RAG) systems have become essential in knowledge-intensive and open-domain tasks. As retrieval complexity increases, ensuring the robustness of these systems is critical. However, current RAG models are highly sensitive to the order in which evidence is presented, often resulting in unstable performance and biased re

Massimiliano Gubinelli, Guopeng Li, Jiawei Li, Tadahiro Oh

We study the modulated Korteweg-de~Vries equation (KdV) on the circle with a time non-homogeneous modulation acting on the linear dispersion term. By adapting the normal form approach to the modulated setting, we prove sharp unconditional uniqueness of solutions to the modulated KdV in $L^2(\mathbb T)$ if a modulation is sufficiently irregular. For example,

Aleksandr Algazinov, Joydeep Chandra, Matt Laing

In-network computation represents a transformative approach to addressing the escalating demands of Artificial Intelligence (AI) workloads on network infrastructure. By leveraging the processing capabilities of network devices such as switches, routers, and Network Interface Cards (NICs), this paradigm enables AI computations to be performed directly within

Shenchao Jin, Xiayang Fan, Xin Wang, Yi Song

Recently, the rapid progress of quantum sensing research reveals that the Rydberg atoms have great potentials in becoming high-precision centimeter-scale antenna of low-frequency fields. In order to facilitate efficient and reliable detection of low-frequency fields via Rydberg atoms, we design, implement and analyze a special but low-cost and scalable metho

Liancheng Fang, Aiwei Liu, Henry Peng Zou, Yankai Chen

We introduce MUSE, a watermarking algorithm for tabular generative models. Previous approaches typically leverage DDIM invertibility to watermark tabular diffusion models, but tabular diffusion models exhibit significantly poorer invertibility compared to other modalities, compromising performance. Simultaneously, tabular diffusion models require substantial

Banseok Lee, Dongkyu Kim, Youngcheon You, Youngmin Kim

The deployment of large language models (LLMs) is frequently hindered by prohibitive memory and computational requirements. While quantization mitigates these bottlenecks, maintaining model fidelity in the sub-1-bit regime remains a persistent challenge. In this paper, we introduce LittleBit, a novel framework for extreme LLM compression. We target quantizat

Guanren Qiao, Sixu Lin, Ronglai Zuo, Zhizheng Wu

Sign language is a natural and visual form of language that uses movements and expressions to convey meaning, serving as a crucial means of communication for individuals who are deaf or hard-of-hearing (DHH). However, the number of people proficient in sign language remains limited, highlighting the need for technological advancements to bridge communication

Harsh Goel, Mohammad Omama, Behdad Chalaki, Vaishnav Tadiparthi

Multi-agent reinforcement learning (MARL) has achieved significant progress in large-scale traffic control, autonomous vehicles, and robotics. Drawing inspiration from biological systems where roles naturally emerge to enable coordination, role-based MARL methods have been proposed to enhance cooperation learning for complex tasks. However, existing methods

Xin Quan, Marco Valentino, Louise A. Dennis, André Freitas

Natural language explanations play a fundamental role in Natural Language Inference (NLI) by revealing how premises logically entail hypotheses. Recent work has shown that the interaction of large language models (LLMs) with theorem provers (TPs) can help verify and improve the validity of NLI explanations. However, TPs require translating natural language i

Naila Shafirni Hidayat, Muhammad Dehan Al Kautsar, Alfan Farizki Wicaksono, Fajri Koto

The performance of large language models (LLMs) continues to improve, as reflected in rising scores on standard benchmarks. However, the lack of transparency around training data raises concerns about potential overlap with evaluation sets and the fairness of reported results. Although prior work has proposed methods for detecting data leakage, these approac

Hiroki Naganuma, Kotaro Yoshida, Laura Gomezjurado Gonzalez, Takafumi Horie

Model editing techniques, particularly task arithmetic with task vectors, offer an efficient alternative to full fine-tuning by enabling direct parameter updates through simple arithmetic operations. While this approach promises substantial computational savings, its impact on fairness has remained largely unexplored -- despite growing concern over biased ou

Weiyi Wang, Junwei Deng, Yuzheng Hu, Shiyuan Zhang

Data attribution methods, which quantify the influence of individual training data points on a machine learning model, have gained increasing popularity in data-centric applications in modern AI. Despite a recent surge of new methods developed in this space, the impact of hyperparameter tuning in these methods remains under-explored. In this work, we present

Mingyi He, Yuebing Liang, Shenhao Wang, Yunhan Zheng

Urban design is a multifaceted process that demands careful consideration of site-specific constraints and collaboration among diverse professionals and stakeholders. The advent of generative artificial intelligence (GenAI) offers transformative potential by improving the efficiency of design generation and facilitating the communication of design ideas. How

Yang Sui, Qi Xu, Ting Li, Yang Bai

Alzheimer's Disease Neuroimaging Initiative (ADNI) diagnostic groups present strong heterogeneous associations among demographic, imaging, and cognitive data. We propose a novel PArtially-shared Imaging Regression (PAIR) model to represent imaging coefficients as weighted combinations of smooth spatial components. A Total Variation penalty is applied to enfo

Vishal Pallagani, Nitin Gupta, John Aydin, Biplav Srivastava

Understanding how data moves, transforms, and persists, known as data flow, is fundamental to reasoning in procedural tasks. Despite their fluency in natural and programming languages, large language models (LLMs), although increasingly being applied to decisions with procedural tasks, have not been systematically evaluated for their ability to perform data-

Sahithya Ravi, Gabriel Sarch, Vibhav Vineet, Andrew D. Wilson

An embodied AI assistant operating on egocentric video must integrate spatial cues across time - for instance, determining where an object A, glimpsed a few moments ago lies relative to an object B encountered later. We introduce Disjoint-3DQA , a generative QA benchmark that evaluates this ability of VLMs by posing questions about object pairs that are not

Alexander V. Smirnov, Boris I. Rozhnov, Vadim V. Voevodin

The Zippel algorithm performs a rational reconstruction of multivariate polynomials and aims specifically at the sparse case. It is applied in different fields of science, lately becoming an important step in Feynman integral reduction in elementary particle physics. In some cases with multiple variables it might become a bottleneck for the whole evaluation

Neemesh Yadav, Yihuai Lan, Shan Dong, Mai Hieu Hien

Large Language Models (LLMs) have shown potential in simulating human behaviors and performing theory-of-mind (ToM) reasoning, crucial for complex social interactions. We investigate ToM reasoning&#39;s role in aligning agentic behaviors with human norms in negotiation tasks, using the ultimatum game as our referenced task. We initialized LLM agents with dif

Zheng Wang, Wanhao Yu, Li Yang, Sen Lin

Continual Learning (CL) seeks to build an agent that can continuously learn a sequence of tasks, where a key challenge, namely Catastrophic Forgetting, persists due to the potential knowledge interference among different tasks. On the other hand, deep neural networks (DNNs) are shown to converge to a terminal state termed Neural Collapse during training, whe