May 2025 arXiv papers — page 40

Majda Smole, Miroslav Micic, Ana Mitrašinović

We investigate galaxy groups that reside in the field but have been previously processed by galaxy clusters. Observationally, they would appear to have the same properties as regular field groups at first glance. However, one would expect to find quantifiable differences in processed groups as dynamical interactions within clusters perturb them. We use Illus

Rodolfo Capdevilla, Roni Harnik, Taegyun Kim, Tom Krokotsch

The use of light axion dark matter experiments as high-frequency gravitational wave (HFGW) detectors has garnered increasing attention in recent years. We explore the capabilities of the Broadband Reflector Experiment for Axion Detection (BREAD) in probing the GW parameter space and study the directional dependence of its coverage. This detector can investig

Ange-Clement Akazan, Verlon Roel Mbingui, Gnankan Landry Regis N'guessan, Issa Karambal

Weather forecasting is crucial for managing risks and economic planning, particularly in tropical Africa, where extreme events severely impact livelihoods. Yet, existing forecasting methods often struggle with the region's complex, non-linear weather patterns. This study benchmarks deep recurrent neural networks such as $\texttt{LSTM, GRU, BiLSTM, BiGRU}$, a

Nicolas Guerra, Nicholas H. Nelsen, Yunan Yang

In scientific machine learning, models are routinely deployed with parameter values or boundary conditions far from those used in training. This paper studies the learning-where-to-learn problem of designing a training data distribution that minimizes average prediction error across a family of deployment regimes. A theoretical analysis shows how the trainin

H. J. Hoeijmakers, K. P. Jaworska, B. Prinoth

The young $\beta$ Pictoris system has been monitored with high-resolution optical spectrographs for decades. These observations have revealed strongly variable absorption in the Ca II H\&K lines attributed to in-falling cometary bodies. Since 2003, over 9000 HARPS observations of $\beta$ Pictoris have been taken and many of these have not yet been used for e

Elena Rasia, Roberta Tripodi, Stefano Borgani, Veronica Biffi

The baryon fraction of galaxy clusters is a powerful tool to inform on the cosmological parameters while the hot-gas fraction provides indications on the physics of the intracluster plasma and its interplay with the processes driving galaxy formation. Using cosmological hydrodynamical simulations from The Three Hundred collaboration of about 300 simulated ma

Kai-Peng Lu, H. Lu, Liang Ma

In this paper, we consider the recently proposed bosonic/heterotic duality that relates the heterotic superstrings to the noncritical bosonic string. Although the latter is nonsupersymmetric, it can be viewed as pseudo-supersymmetric in that the theory admits a consistent set of Killing spinor equations whose integrability conditions are satisfied by the equ

Aurélien Genin, Marko Shuntov, Gabe Brammer, Natalie Allen

To better understand how galaxies assemble their structure and evolve over cosmic time, we present a new catalog of morphological measurements for over 340,000 sources spanning $0 < z < 12$, derived from deep JWST NIRCam imaging across four major extragalactic fields (CEERS, PRIMER-UDS, PRIMER-COSMOS, GOODS) compiled in the DAWN JWST Archive (DJA). We perfor

Gefen Baranes, Iria W. Wang, Francisco Machado, Aziza Suleymanzade

Blind quantum computing (BQC) is a computational paradigm that allows a client with limited quantum capabilities to delegate quantum computations to a more powerful server while keeping both the algorithm and data hidden. However, in practice, existing BQC protocols face significant challenges when scaling to large-scale computations due to photon losses, lo

Zhengyuan Jiang, Moyang Guo, Kecen Li, Yuepeng Hu

The rapid development of video generative models has led to a surge in highly realistic synthetic videos, raising ethical concerns related to disinformation and copyright infringement. Recently, video watermarking has been proposed as a mitigation strategy by embedding invisible marks into AI-generated videos to enable subsequent detection. However, the robu

Nai Chao Hu, Rui-Zhen Huang, Nick Bultinck

We consider a one-dimensional multi-orbital Kondo lattice model and show that by tuning the kinetic energy of the itinerant electrons it is possible to stabilize Kondo insulators with non-trivial spin physics. In particular, depending on the size of the exchange coupling between the local moments, we find kinetic-energy-driven transitions between a featurele

Cameron Calcluth

(Abridged.) Quantum computers promise to solve some problems exponentially faster than traditional computers, but we still do not fully understand why this is the case. While the most studied model of quantum computation uses qubits, which are the quantum equivalent of a classical bit, an alternative method for building quantum computers is gaining traction.

Y. Lerner, N. C. Stone, D. D. Ofengeim

We present a parameter survey of fragmentation in collapsar disks, using a revised version of the Chen & Beloborodov (2007) model that determines the structure of steady state hyperaccretion disks in a general relativistic and neutrino cooled framework. We map out the range of disk conditions leading to gravitational instability alongside an exploration of t

Adarsh S. Patri, Marcel Franz

Recent experiments in rhombohedrally-stacked multilayer graphene heterostructures have reported signatures of chiral superconductivity, emerging from a spin and valley-polarized normal state with broken time-reversal symmetry and an associated anomalous Hall effect. These findings bring into focus the role of the electronic Bloch wavefunction and the quantum

Andrea Trost, Catarina M. J. Marques, Stefano Cristiani, Guido Cupani

The measurement of the temporal evolution in the redshift of distant objects, the redshift drift, is a probe of universal expansion and cosmology. We perform the first steps towards a measurement of such effect using the Lyman-$\alpha$ forest in the spectra of bright quasars as a tracer of cosmological expansion. Our goal is to determine to which precision a

Nicolas Dirnegger, Marie Wesson, Arpit Arora, Ioannis Petrides

Time-reversal symmetry breaking (TRSB) has been central to detecting exotic phases of matter. Here, we leverage the circuit electrodynamics capabilities of superconducting devices to propose a novel scheme based on a multimode superconducting ring resonator for sensitive probing of TRSB in quantum materials. A ring resonator enables nonlinear cross-interacti

Sóley Ó. Hyman, S. P. Willner, Belinda J. Wilkes

In the course of studying the 3C 220.3 lensing system, spectra were obtained with the Binospec instrument on the MMT for 511 additional objects in 3C 220.3's vicinity. These gave 146 good-quality galaxy redshifts and identified 126 Galactic stars. The galaxy redshift histogram shows a peak near 3C 220.3's redshift, but there is no evidence for or against a g

Rossella Gamba, Jacob Lange, Danilo Chiaramello, Jacopo Tissino

The first direct detection of gravitational waves by the LIGO collaboration, GW150914, marked the start of a new exciting era in astronomy, enabling the study of the Universe through a new messenger. Since then, the field has grown rapidly, with the development of increasingly more sophisticated techniques to detect, analyze and interpret the signals. In thi

Fabiana De Cesare, Slava Rychkov

The $O(N)$ Non-Linear Sigma Model (NLSM) in $d=2+\epsilon$ has long been conjectured to describe the same conformal field theory (CFT) as the Wilson-Fisher (WF) $O(N)$ fixed point obtained from the $\lambda(\phi^2)^2$ model in $d=4-\epsilon$. In this work, we put this conjecture into question, building on the recent observation [Jones (2024)] that the NLSM C

Luca Tortorelli, Silvan Fischbacher, Daniel Grün, Alexandre Refregier

Next generation photometric and spectroscopic surveys will enable unprecedented tests of the concordance cosmological model and of galaxy formation and evolution. Fully exploiting their potential requires a precise understanding of the selection effects on galaxies and biases on measurements of their properties, required, above all, for accurate estimates of

Shimao Zhang, Zhejian Lai, Xiang Liu, Shuaijie She

Multilingual Alignment is an effective and representative paradigm to enhance LLMs' multilingual capabilities, which transfers the capabilities from the high-resource languages to the low-resource languages. Meanwhile, some research on language-specific neurons provides a new perspective to analyze and understand LLMs' mechanisms. However, we find that there

Yihan Wang, Qiao Yan, Zhenghao Xing, Lihao Liu

Large language models (LLMs) have demonstrated strong potential in clinical question answering, with recent multi-agent frameworks further improving diagnostic accuracy via collaborative reasoning. However, we identify a recurring issue of Silent Agreement, where agents prematurely converge on diagnoses without sufficient critical analysis, particularly in c

Yipengjing Sun, Shengping Zhang, Chenyang Wang, Shunyuan Zheng

We propose GRGS, a generalizable and relightable 3D Gaussian framework for high-fidelity human novel view synthesis under diverse lighting conditions. Unlike existing methods that rely on per-character optimization or ignore physical constraints, GRGS adopts a feed-forward, fully supervised strategy projecting geometry, material, and illumination cues from m

Yinjie Chen, Zipeng Yan, Chong Zhou, Bo Dai

Vision Transformers (ViTs) have emerged as the dominant architecture for visual processing tasks, demonstrating excellent scalability with increased training data and model size. However, recent work has identified the emergence of artifact tokens in ViTs that are incongruous with local semantics. These anomalous tokens degrade ViT performance in tasks that

Dingming Li, Hongxing Li, Zixuan Wang, Yuchen Yan

Vision-language models (VLMs) have demonstrated remarkable capabilities in understanding and reasoning about visual content, but significant challenges persist in tasks requiring cross-viewpoint understanding and spatial reasoning. We identify a critical limitation: current VLMs excel primarily at egocentric spatial reasoning (from the camera's perspective)

Mathew J. Walter, Aaron Barrett, Kimberly Tam

Adversarial artificial intelligence (AI) attacks pose a significant threat to autonomous transportation, such as maritime vessels, that rely on AI components. Malicious actors can exploit these systems to deceive and manipulate AI-driven operations. This paper addresses three critical research challenges associated with adversarial AI: the limited scope of t

Haowei Wang, Junjie Wang, Xiaojun Jia, Rupeng Zhang

Vision-Language Model (VLM) based Web Agents represent a significant step towards automating complex tasks by simulating human-like interaction with websites. However, their deployment in uncontrolled web environments introduces significant security vulnerabilities. Existing research on adversarial environmental injection attacks often relies on unrealistic

Wei Pang, Kevin Qinghong Lin, Xiangru Jian, Xi He

Academic poster generation is a crucial yet challenging task in scientific communication, requiring the compression of long-context interleaved documents into a single, visually coherent page. To address this challenge, we introduce the first benchmark and metric suite for poster generation, which pairs recent conference papers with author-designed posters a

Han Xiao, Guozhi Wang, Yuxiang Chai, Zimu Lu

In this paper, we introduce UI-Genie, a self-improving framework addressing two key challenges in GUI agents: verification of trajectory outcome is challenging and high-quality training data are not scalable. These challenges are addressed by a reward model and a self-improving pipeline, respectively. The reward model, UI-Genie-RM, features an image-text int

Pranav N. Thakkar, Shubhangi Sinha, Karan Baijal, Yuhan

Robust robot manipulation in unstructured environments often requires understanding object properties that extend beyond geometry, such as material or compliance-properties that can be challenging to infer using vision alone. Multimodal haptic sensing provides a promising avenue for inferring such properties, yet progress has been constrained by the lack of

Miao Peng, Nuo Chen, Jianheng Tang, Jia Li

Large Language Models (LLMs) have shown remarkable capabilities in knowledge-intensive tasks, while they remain vulnerable when encountering misinformation. Existing studies have explored the role of LLMs in combating misinformation, but there is still a lack of fine-grained analysis on the specific aspects and extent to which LLMs are influenced by misinfor

Xiaojun Jia, Sensen Gao, Simeng Qin, Tianyu Pang

Multimodal large language models (MLLMs) remain vulnerable to transferable adversarial examples. While existing methods typically achieve targeted attacks by aligning global features-such as CLIP's [CLS] token-between adversarial and target samples, they often overlook the rich local information encoded in patch tokens. This leads to suboptimal alignment and

Di Che, Brian Stern, Kwangwoong Kim, Cagri Ozdilek

We propose a comb-based WDM transmitter capable of modulating independent signals to comb lines without demultiplexing them and prove its concept and potential scalability in a WDM transmitter consisting of a Kerr microcomb and a silicon I/Q modulator array.

Xiangxin Zhou, Zichen Liu, Anya Sims, Haonan Wang

The recent paradigm shift towards training large language models (LLMs) using DeepSeek-R1-Zero-style reinforcement learning (RL) on verifiable rewards has led to impressive advancements in code and mathematical reasoning. However, this methodology is limited to tasks where rule-based answer verification is possible and does not naturally extend to real-world

Aleksandra Leśniewska, Jens Hjorth, Christa Gall

Understanding the evolution of dust in galaxies is crucial because it affects the dynamics and cooling of gas, star formation, and chemical evolution. Recent work on dust removal in galaxies indicates timescales of gigayears, with old stellar populations and AGNs as the primary drivers of this process. However, most statistically significant studies are focu

Boyang Wang, Xuweiyi Chen, Matheus Gadelha, Zezhou Cheng

Controllability, temporal coherence, and detail synthesis remain the most critical challenges in video generation. In this paper, we focus on a commonly used yet underexplored cinematic technique known as Frame In and Frame Out. Specifically, starting from image-to-video generation, users can control the objects in the image to naturally leave the scene or p

Tsering Dolkar, Marco A. R. Ferreira, Hwasoo Shin, Allison N. Tegge

We propose novel Bayesian Dynamic Clustering Factor Models (BDCFM) for the analysis of multivariate longitudinal data. BDCFM combines factor models with hidden Markov models to concomitantly perform dimension reduction, clustering, and estimation of the dynamic transitions of subjects through clusters. We develop an efficient Gibbs sampler for exploration of

Samim Akhtar, Yilber Fabian Bautista, Cristoforo Iossa, Zihan Zhou

In this Letter, we study scalar wave perturbations of arbitrary frequency to the 5D Schwarzschild-Tangherlini black hole (STBH) within general relativity. For the first time, we derive a closed formula for the 5D partial wave gravitational Raman scattering amplitude applicable to a broad class of boundary conditions, expressed in terms of the Nekrasov-Shatas

Omer Dahary, Yehonathan Cohen, Or Patashnik, Kfir Aberman

Generating multiple distinct subjects remains a challenge for existing text-to-image diffusion models. Complex prompts often lead to subject leakage, causing inaccuracies in quantities, attributes, and visual features. Preventing leakage among subjects necessitates knowledge of each subject's spatial location. Recent methods provide these spatial locations v

Ted Zadouri, Hubert Strauss, Tri Dao

LLM decoding is bottlenecked for large batches and long contexts by loading the key-value (KV) cache from high-bandwidth memory, which inflates per-token latency, while the sequential nature of decoding limits parallelism. We analyze the interplay among arithmetic intensity, parallelization, and model quality and question whether current architectures fully

Yang Yang, Jiemin Wu, Yutao Yue

Inductive Logic Programming (ILP) is a principled approach for generalizing regularities from data and constructing hypotheses as interpretable logic programs. However, a key limitation is its reliance on expert-crafted language bias - the predicate inventory, types, and mode declarations that delimit the search space. We propose hypothesis generation via LL

Gianluca Calcagni, Giuseppe Nardelli

We construct representations of complex powers of the d'Alembertian operator $\Box$ in Lorentzian signature and pinpoint one which is self-adjoint and suitable for classical and quantum fractional field theory. This self-adjoint fractional d'Alembertian is associated with complex-conjugate poles, which are removed from the physical spectrum via the Anselmi--

Tom Hutchcroft, Nicolas Monod, Omer Tamuz

We establish a fixed-point theorem for the face maps that consist in deleting the $i$th entry of an ordered set. Furthermore, we show that there exists random finite sets of integers that are almost invariant under such deletions. Consequences for various monoids of order-preserving transformations of $\mathbf{N}$ are discussed in an appendix.

Kerui Ren, Jiayang Bai, Linning Xu, Lihan Jiang

Object compositing offers significant promise for augmented reality (AR) and embodied intelligence applications. Existing approaches predominantly focus on single-image scenarios or intrinsic decomposition techniques, facing challenges with multi-view consistency, complex scenes, and diverse lighting conditions. Recent inverse rendering advancements, such as

A. Gregory DiRienzo, Elie Massaad, Hutan Ashrafian

Multi-Cancer Early Detection (MCED) testing with tissue localization aims to detect and identify multiple cancer types from a single blood sample. Such tests have the potential to aid clinical decisions and significantly improve health outcomes. Despite this promise, MCED testing has not yet achieved regulatory approval, reimbursement or broad clinical adopt

Kyrylo Gerashchenko, Rémi Rousseau, Léo Balembois, Himanshu Patange

Long-lived mechanical resonators like drums oscillating at MHz frequencies and operating in the quantum regime are a powerful platform for quantum technologies and tests of fundamental physics. Yet, quantum control of such systems remains challenging, owing to their low energy scale and the difficulty of achieving efficient coupling to other well-controlled

Antonis Ballis

Global financial systems are undergoing strategic shifts as geopolitical tensions reshape international trade and payments. The United States (US)-China trade war, sanctions regimes, and rising concerns over the weaponization of financial infrastructures like SWIFT have led countries to seek alternative networks, including China's CIPS and emerging cross-bor

Keenan Samway, Max Kleiman-Weiner, David Guzman Piedrahita, Rada Mihalcea

As AI systems increasingly navigate applications in healthcare, law, and governance, understanding how they handle ethically complex scenarios becomes critical. Previous work has mainly examined the moral judgments in large language models (LLMs), rather than their underlying moral reasoning process. In contrast, we focus on a large-scale analysis of the mor

Uri Gadot, Rinon Gal, Yftah Ziser, Gal Chechik

Text-to-image generation has evolved beyond single monolithic models to complex multi-component pipelines. These combine fine-tuned generators, adapters, upscaling blocks and even editing steps, leading to significant improvements in image quality. However, their effective design requires substantial expertise. Recent approaches have shown promise in automat

Jean-Louis Colliot-Thélène, Alena Pirutka, Federico Scavia

Let $R$ be the field of real Puiseux series. It is a real closed field. We construct the first examples of smooth intersections of two quadrics in $\mathbb{P}_R^5$ and smooth cubic hypersurfaces in $\mathbb{P}_R^4$ which are not stably rational but for which the space $X(R)$ of $R$-points is semi-algebraically connected. The question of constructing such exa

R. Aliberti, T. Aoyama, E. Balzani, A. Bashir

We present the current Standard Model (SM) prediction for the muon anomalous magnetic moment, $a_\mu$, updating the first White Paper (WP20) [1]. The pure QED and electroweak contributions have been further consolidated, while hadronic contributions continue to be responsible for the bulk of the uncertainty of the SM prediction. Significant progress has been

Ilias Diakonikolas, Giannis Iakovidis, Daniel M. Kane, Lisheng Ren

We study the complexity of learning real-valued Multi-Index Models (MIMs) under the Gaussian distribution. A $K$-MIM is a function $f:\mathbb{R}^d\to \mathbb{R}$ that depends only on the projection of its input onto a $K$-dimensional subspace. We give a general algorithm for PAC learning a broad class of MIMs with respect to the square loss, even in the pres

Fengqing Jiang, Fengbo Ma, Zhangchen Xu, Yuetai Li

Large language models (LLMs) exhibit advancing capabilities in complex tasks, such as reasoning and graduate-level question answering, yet their resilience against misuse, particularly involving scientifically sophisticated risks, remains underexplored. Existing safety benchmarks typically focus either on instructions requiring minimal knowledge comprehensio

A. Kuzin, D. Malyshev, M. Chernyakova, B. van Soelen

PSR B1259-63/LS 2883 is a well-studied gamma-ray binary hosting a pulsar in a 3.4-year eccentric orbit around a Be-type star. Its non-thermal emission spans from radio to TeV energies, exhibiting a significant increase near the periastron passage. This paper is dedicated to the analysis of INTEGRAL observations of the system following its last periastron pas

Kristina Radivojevic, Caleb Reinking, Shaun Whitfield, Paul Brenner

Social media serves as a primary communication and information dissemination platform for major global events, entertainment, and niche or topically focused community discussions. Therefore, it represents a valuable resource for researchers who aim to understand numerous questions. However, obtaining data can be difficult, expensive, and often unreliable due

Yiheng Liu, Liao Qu, Huichao Zhang, Xu Wang

This paper presents DetailFlow, a coarse-to-fine 1D autoregressive (AR) image generation method that models images through a novel next-detail prediction strategy. By learning a resolution-aware token sequence supervised with progressively degraded images, DetailFlow enables the generation process to start from the global structure and incrementally refine d

Mehrdad Fazli, Bowen Wei, Ahmet Sari, Ziwei Zhu

Large vision-language models (LVLMs) achieve impressive performance on multimodal tasks but often suffer from hallucination, and confidently describe objects or attributes not present in the image. Current training-free interventions struggle to maintain accuracy in open-ended and long-form generation scenarios. We introduce the Confidence-Aware Attention Ca

Zijun Liu, Zhennan Wan, Peng Li, Ming Yan

With the rapid advancement of post-training techniques for reasoning and information seeking, large language models (LLMs) can incorporate a large quantity of retrieved knowledge to solve complex tasks. However, the limited context window of LLMs obstructs scaling the amount of external knowledge input, prohibiting further improvement. Existing context windo

Julia Lena Lienert, Bertram Bitsch, Thomas Henning

The chemical evolution of the inner regions of protoplanetary discs is a complex process. Several factors influence it, one being the inward drift and evaporation of volatile-rich pebbles. During the disc's evolution, its inner part is first enriched with evaporating water-ice, resulting in a low C/O ratio. Afterwards, C-rich gas from the outer disc is trans

Cheng Zeng, Jirui Jin, Connor Ambrose, George Karypis

Molecule generation is advancing rapidly in chemical discovery and drug design. Flow matching methods have recently set the state of the art (SOTA) in unconditional molecule generation, surpassing score-based diffusion models. However, diffusion models still lead in property-guided generation. In this work, we introduce PropMolFlow, an approach for property-

Simon Dirmeier, Antonietta Mira

We present Causal Posterior Estimation (CPE), a novel method for Bayesian inference in simulator models, i.e., models where the evaluation of the likelihood function is intractable or too computationally expensive, but where one can simulate model outputs given parameter values. CPE utilizes a normalizing flow-based (NF) approximation to the posterior distri

Zhanqiu Hu, Jian Meng, Yash Akhauri, Mohamed S. Abdelfattah

Diffusion language models offer parallel token generation and inherent bidirectionality, promising more efficient and powerful sequence modeling compared to autoregressive approaches. However, state-of-the-art diffusion models (e.g., Dream 7B, LLaDA 8B) suffer from slow inference. While they match the quality of similarly sized autoregressive (AR) models (e.

Mathew A. Johnson, Jeffrey Oregero, Wesley R. Perkins

We consider the nonlinear wave modulation of arbitrary amplitude periodic traveling wave solutions of the Ostrovsky equation, which arises as a model for the unidirectional propagation of small-amplitude, weakly nonlinear surface and internal gravity waves in a rotating fluid of finite depth. While the modulation of such waves with asymptotically small ampli

Bozhou Li, Wentao Zhang

Currently, a prevalent approach for enhancing Vision-Language Models (VLMs) performance is to encode both the high-resolution version and the thumbnail of an image simultaneously. While effective, this method generates a large number of image tokens. When combined with the widely used Rotary Position Embedding (RoPE), its long-term decay property hinders the

Edouard Daviaud

In this article, we establish optimality results regarding the dynamical Borel-Cantelli lemma and the the Hausdorff dimension of certain dynamical diophantine sets.

Jake Hassan, Rosalba Perna, Matteo Cantiello, Tyler Parsotan

Population III (Pop III) stars, the first generation of stars formed from primordial gas, played a fundamental role in shaping the early universe through their influence on cosmic reionization, early chemical enrichment, and the formation of the first galaxies. However, to date they have eluded direct detection due to their short lifetimes and high redshifts

Yali Yuan, Yu Huang, Xingjian Zeng, Hantao Mei

The growing complexity of encrypted network traffic presents dual challenges for modern network management: accurate multiclass classification of known applications and reliable detection of unknown traffic patterns. Although deep learning models show promise in controlled environments, their real-world deployment is hindered by data scarcity, concept drift,

Joan Gutierrez-Florensa, Alvaro Ortega, Lukas Sigrist, Federico Milano

Accurate frequency estimation is critical for the control, monitoring and protection of electrical power systems, in particular, of systems with a high penetration of power electronics. This paper introduces the novel concept of Quasi Steady-State (QSS) frequency as a quantity that fills the gap between stationary and instantaneous frequency. QSS frequency c

Maxwell Fishelson, Noah Golowich, Mehryar Mohri, Jon Schneider

We study online calibration of multi-dimensional forecasts over an arbitrary convex set $P \subset \mathbb{R}^d$ relative to an arbitrary norm $|\cdot|$. We connect this to external regret minimization for online linear optimization (OLO): if one can guarantee $O(\sqrt{ρT})$ worst-case regret after $T$ rounds when actions are drawn from $P$ and losses from t

Xiangru Jian, Wei Pang, Zhengyuan Dong, Chao Zhang

Current video analytics approaches face a fundamental trade-off between flexibility and efficiency. End-to-end Vision Language Models (VLMs) often struggle with long-context processing and incur high computational costs, while neural-symbolic methods depend heavily on manual labeling and rigid rule design. In this paper, we introduce LazyVLM, a neuro-symboli

Shintaro Ozaki, Tatsuya Hiraoka, Hiroto Otake, Hiroki Ouchi

Large Language Models (LLMs) are known to process information using a proficient internal language consistently, referred to as latent language, which may differ from the input or output languages. However, how the discrepancy between the latent language and the input and output language affects downstream task performance remains largely unexplored. While m

Gourab Ghatak

We present a novel analytical framework to characterize the distribution of the conditional receiver operating characteristic (ROC) in radar systems operating within a realization of a Poisson field of interferers and clutters. While conventional stochastic geometry based studies focus on the distribution of signal to interference and noise ratio (SINR), the

Yue Li Du, Ben Alexander, Mikhail Antonenka, Rohan Mahadev

Retrieving semantically similar but visually distinct contents has been a critical capability in visual search systems. In this work, we aim to tackle this problem with Visual Product Graph (VPG), leveraging high-performance infrastructure for storage and state-of-the-art computer vision models for image understanding. VPG is built to be an online real-time

Michael L. Shull, Ephraim M. Hanks, James C. Russell, Robert K. Murphy

With technological advancements, the quantity and quality of animal movement data have increased greatly. Currently, no movement model can be used to describe full-year data from migratory species by leveraging both individual movement and species distribution data. Herein we propose a full-year stochastic differential equation model for jointly modeling bot

Xiangxin Zhou, Mingyu Li, Yi Xiao, Jiahan Li

Cyclic peptides offer inherent advantages in pharmaceuticals. For example, cyclic peptides are more resistant to enzymatic hydrolysis compared to linear peptides and usually exhibit excellent stability and affinity. Although deep generative models have achieved great success in linear peptide design, several challenges prevent the development of computationa

Jocelyn Shen, Akhila Yerukola, Xuhui Zhou, Cynthia Breazeal

Conversational breakdowns in close relationships are deeply shaped by personal histories and emotional context, yet most NLP research treats conflict detection as a general task, overlooking the relational dynamics that influence how messages are perceived. In this work, we leverage nonviolent communication (NVC) theory to evaluate LLMs in detecting conversa

Harmender Gahlawat

\textsc{Cops and Robber} is a game played on graphs where a set of \textit{cops} aim to \textit{capture} the position of a single \textit{robber}. The main parameter of interest in this game is the \textit{cop number}, which is the minimum number of cops that are sufficient to guarantee the capture of the robber. In a directed graph $\overrightarrow{G}$, the

Ziqiao Peng, Jiwen Liu, Haoxian Zhang, Xiaoqiang Liu

Lip synchronization is the task of aligning a speaker's lip movements in video with corresponding speech audio, and it is essential for creating realistic, expressive video content. However, existing methods often rely on reference frames and masked-frame inpainting, which limit their robustness to identity consistency, pose variations, facial occlusions, an

Dane Isenberg, Michael O. Harhay, Andrew B. Forbes, Paul J. Young

In cluster-randomized crossover (CRXO) trials, groups of individuals are randomly assigned to two or more sequences of alternating treatments. Since clusters serve as their own control, the CRXO design is typically more statistically efficient than the usual parallel-arm design. CRXO trials are increasingly popular in many areas of health research where the

Alexis Poncet, Aurélien Grabsch, Olivier Bénichou

Tracer diffusion in single-file systems, where particles are restricted to move on a line without passing each other, has been a fertile ground to investigate anomalous diffusion and strong memory effects. While the long-time behavior of such a tracer has been well studied, with a known subdiffusive dynamics and a Gaussian description for the rescaled positi

Zhiqi Ai, Meixuan Bao, Zhiyong Chen, Zhi Yang

The performance of speaker verification systems is adversely affected by speaker aging. However, due to challenges in data collection, particularly the lack of sustained and large-scale longitudinal data for individuals, research on speaker aging remains difficult. In this paper, we present VoxAging, a large-scale longitudinal dataset collected from 293 spea

Andre Massahiro Shimaoka, Renato Cordeiro Ferreira, Alfredo Goldman

This study explores the integration of eXtreme Programming (XP) and the Cross-Industry Standard Process for Data Mining (CRISP-DM) in agile Data Science projects. We conducted a case study at the e-commerce company Elo7 to answer the research question: How can the agility of the XP method be integrated with CRISP-DM in Data Science projects? Data was collect

Sheikh Shafayat, Fahim Tajwar, Ruslan Salakhutdinov, Jeff Schneider

Recent successes of reinforcement learning (RL) in training large reasoning models motivate the question of whether self-training - the process where a model learns from its own judgments - can be sustained within RL. In this work, we study this question using majority voting as a simple self-feedback mechanism. On a comprehensive set of experiments on both

Pawan Khatiwada, Xiao-Feng Qian

We present a systematic framework to quantify the interplay between coherence and wave-particle duality in generic two-path interference systems. Our analysis reveals a closed-form duality ellipse (DE) equality, that rigorously unifies visibility (a traditional waveness measure) and predictability (a particleness measure) with degree of coherence, providing

Pouria Fallahpour, Alex B. Grilo, Garazi Muguruza, Mahshid Riahinia

One-way functions (OWFs) form the foundation of modern cryptography, yet their unconditional existence remains a major open question. In this work, we study this question by exploring its relation to lossy reductions, i.e., reductions $R$ for which it holds that $I(X;R(X)) \ll n$ for all distributions $X$ over inputs of size $n$. Our main result is that eith

Binh Duc Vu, Jan Kapar, Marvin Wright, David S. Watson

We propose a principled method for autoencoding with random forests. Our strategy builds on foundational results from nonparametric statistics and spectral graph theory to learn a low-dimensional embedding of the model that optimally represents relationships in the data. We provide exact and approximate solutions to the decoding problem via constrained optim

Tiziano Bertoli, Elena Stellino, Francesco Minati, Camilla Belloni

alpha-Sn has recently been attracting significant interest due to its unique electronic properties. However, this allotrope of Sn is stable only below 13 {\deg}C and alternative options to the conventional stabilization by epitaxial growth on InSb are still a challenge. In this work, nanoparticles with inner alpha-Sn nanocrystals were synthesized on a Silico

Yuchen Zhuang, Aaron Trinh, Rushi Qiang, Haotian Sun

Modern information retrieval (IR) models, trained exclusively on standard <query, passage> pairs, struggle to effectively interpret and follow explicit user instructions. We introduce InF-IR, a large-scale, high-quality training corpus tailored for enhancing retrieval models in Instruction-Following IR. InF-IR expands traditional training pairs into over 38,

Leiming Chen, Patrick Jentsch, Chiu Fan Lee, Ananyo Maitra

This is our response to the comment arXiv:2504.13683 posted by Chat\'e and Solon in reference to our preprint arXiv:2503.17064

Yiwen Tu, Ziqi Liu, Jiaqi W. Ma, Weijing Tang

Measuring task relatedness and mitigating negative transfer remain a critical open challenge in Multitask Learning (MTL). This work extends data attribution -- which quantifies the influence of individual training data points on model predictions -- to MTL setting for measuring task relatedness. We propose the MultiTask Influence Function (MTIF), a method th

Huaijin Pi, Zhi Cen, Zhiyang Dou, Taku Komura

Synthesizing whole-body manipulation of articulated objects, including body motion, hand motion, and object motion, is a critical yet challenging task with broad applications in virtual humans and robotics. The core challenges are twofold. First, achieving realistic whole-body motion requires tight coordination between the hands and the rest of the body, as

James William Bruce, Marco Antônio do Couto Fernandes, Farid Tari

Given the germ of a smooth plane curve $(\{f(x,y)=0\},0)\subset (\mathbb{K}^2,0), \mathbb{K}=\mathbb{R}, \mathbb{C}$, with an isolated singularity, we define two invariants $I_f$ and $V_f \in \mathbb{N} \cup\{\infty\}$, which count the number of inflections and vertices (suitably interpreted in the complex case) concentrated at the singular point. The first

Amnon Balanov, Wasim Huleihel, Tamir Bendory

We study the multi-reference alignment (MRA) problem of recovering a signal from noisy observations acted on by unknown random circular shifts. While the information-theoretic limits of MRA are well characterized in many settings, the algorithmic behavior at low signal-to-noise ratio (SNR), the regime of practical interest, remains poorly understood. In this

Neil Mulcahy, Syeda Ramin Jannat, Geri Topore, Lukas Worch

Understanding solid liquid interfaces at high spatial and chemical resolution is crucial for advancing electrochemical energy storage technologies, yet this remains a persistent challenge due to the lack of characterisation techniques that can capture dynamic processes and preserve fragile interfacial chemistries. In lithium ion batteries, interfacial phenom

Nadym Mallek, Kirill Simonov

We study the Requirement Cut problem, a generalization of numerous classical graph partitioning problems including Multicut, Multiway Cut, $k$-Cut, and Steiner Multicut among others. Given a graph with edge costs, terminal groups $(S_1, ..., S_g)$ and integer requirements $(r_1,... , r_g)$; the goal is to compute a minimum-cost edge cut that separates each g

Haoming Song, Delin Qu, Yuanqi Yao, Qizhi Chen

Humans practice slow thinking before performing actual actions when handling complex tasks in the physical world. This thinking paradigm, recently, has achieved remarkable advancement in boosting Large Language Models (LLMs) to solve complex tasks in digital domains. However, the potential of slow thinking remains largely unexplored for robotic foundation mo

Raúl Carballo-Rubio, Héloïse Delaporte, Astrid Eichhorn, Pedro G. S. Fernandes

Non-minimal couplings between the electromagnetic field strength and the spacetime curvature are part of the effective field theory of gravity and matter. They alter the local propagation of light in a significant way if the ratio of spacetime curvature to the non-minimal coupling is of order one. Spacetime curvature can become appreciable around black holes

Shiwei Zeng, Jie Shen

Attribute-efficient PAC learning of sparse halfspaces has been a fundamental problem in machine learning theory. In recent years, machine learning algorithms are faced with prevalent data corruptions or even malicious attacks. It is of central interest to design computationally and attribute-efficient algorithms that are robust to extreme corruptions. In thi

Yuta Nasuda, Koki Tokeshi, Yuki Watanabe

The stochastic dynamics of a scalar field in de Sitter spacetime can be regarded as a non-perturbative diffusion process, to which exact distribution and correlation functions are constructed by utilising the correspondence between diffusion and Schr\"{o}dinger equations. The Krein--Adler transformation of the quantum harmonic oscillator deletes several pair

Marius Juul Nielsen, Luuk H. E. Kempen, Julie de Neergaard Ravn, Raffaele Cheula

The conversion of $\mathrm{CO_2}$ to value-added compounds is an important part of the effort to store and reuse atmospheric $\mathrm{CO_2}$ emissions. Here we focus on $\mathrm{CO_2}$ hydrogenation over so-called inverse catalysts: transition metal oxide clusters supported on metal surfaces. The conventional approach for computational screening of such cand