May 2025 arXiv papers — page 7

Juho Kim

The independent chip model (ICM) forms a cornerstone of all modern poker tournament strategy. However, despite its prominence, the ICM's performance in the real world has not been sufficiently scrutinized, especially at a large scale. In this paper, we introduce our new dataset of poker tournaments, consisting of results of over ten thousand events. Then, us

G. L. Klimchitskaya, A. S. Korotkov, V. V. Loboda, V. M. Mostepanenko

The stability of nanoelectromechanical pressure sensors working in the Casimir regime is considered with account of surface roughness on both the sensor membrane and the ground plate. The equilibrium positions of the sensor membrane are found from the balance between the external measured, elastic, electric pressures, and the Casimir pressure computed by mea

Anirudh Nair, Adi Banerjee, Laurent Mombaerts, Matthew Hagen

Prompt engineering represents a critical bottleneck to harness the full potential of Large Language Models (LLMs) for solving complex tasks, as it requires specialized expertise, significant trial-and-error, and manual intervention. This challenge is particularly pronounced for tasks involving subjective quality assessment, where defining explicit optimizati

Jasmine Kalia, Jared Rivera, Rubaiya R Emran, William J Solorio Hernandez

We report the realization of a degenerate mixture of $^{166}$Er and $^{7}$Li atoms in their energetically lowest spin states. The two species are sequentially laser-cooled and loaded into an optical dipole trap, then transported to a glass cell and simultaneously evaporated to degeneracy. Er serves as the coolant for Li, and we observe efficient sympathetic

Muhammad Syafrudin, Ganjar Alfian, Norma Latif Fitriyani

With growing consumer health awareness, ensuring food safety and quality throughout the supply chain is crucial, particularly for perishable goods. Contamination can occur during production, processing, or distribution, making real-time monitoring essential. This study proposes an affordable Smartphone-based food traceability system (FTS) that utilizes RFID

Malena Mendez Isla, Agustin Mauro, Diego Kozlowski

Interdisciplinary scientific research is increasingly important in knowledge production, funding policies, and academic discussions on scholarly communication. While many studies focus on interdisciplinary corpora defined a priori -- usually through keyword-based searches within assumed interdisciplinary domains -- few explore interdisciplinarity as an emerg

J. M. Gibson, Robe Elliman, T. Susi, C. Mangler

Self-ion implantation amorphization is an established approach to study the structure and properties of amorphous silicon (a-Si). Fluctuation Electron Microscopy (FEM) has consistently observed Medium-Range Order (MRO) in this system that is not consistent with the Continuous Random Network (CRN) model. Using this technique we find that the degree of MRO fir

Aziida Nanyonga, Joiner Keith, Turhan Ugur, Wild Graham

This study compares the effectiveness of BERTopic and Probabilistic Latent Semantic Analysis (PLSA) in extracting meaningful topics from aviation safety reports aiming to enhance the understanding of patterns in aviation incident data. Using a dataset of over 36,000 National Transportation Safety Board (NTSB) reports from 2000 to 2020, BERTopic employed tran

Yezhuo Li, Qiong Zhang, Madhura Limaye, Gang Li

Bayesian Optimization, leveraging Gaussian process models, has proven to be a powerful tool for minimizing expensive-to-evaluate objective functions by efficiently exploring the search space. Extensions such as constrained Bayesian Optimization have further enhanced Bayesian Optimization's utility in practical scenarios by focusing the search within feasible

Mingyi Shi, Wei Liu, Jidong Mei, Wangpok Tse

We present MotionPersona, a novel real-time character controller that allows users to characterize a character by specifying attributes such as physical traits, mental states, and demographics, and projects these properties into the generated motions for animating the character. In contrast to existing deep learning-based controllers, which typically produce

Kaivalya Hariharan, Uzay Girit, Atticus Wang, Jacob Andreas

Benchmarks for large language models (LLMs) have predominantly assessed short-horizon, localized reasoning. Existing long-horizon suites (e.g. SWE-bench) rely on manually curated issues, so expanding or tuning difficulty demands expensive human effort and evaluations quickly saturate. However, many real-world tasks, such as software engineering or scientific

Nicolás García Trillos, Chenghui Li, Raghavendra Venkatraman

We study the problem of estimating eigenpairs of elliptic differential operators from samples of a distribution $\rho$ supported on a manifold $M$. The operators discussed in the paper are relevant in unsupervised learning and in particular are obtained by taking suitable scaling limits of widely used graph Laplacians over data clouds. We study the minimax r

Julian Bushelli

The classical theory of free analysis generalizes the noncommutative (nc) polynomials and rational functions, easily providing such results as an nc analogue of the Jacobian conjecture. However, the classical theory misses out on important functions, such as the Schur complement. This paper presents a generalization of free functions, viewing them as a natur

Aziida Nanyonga, Graham Wild

The volume of textual data available in aviation safety reports presents a challenge for timely and accurate analysis. This paper examines how Artificial Intelligence (AI) and, specifically, Natural Language Processing (NLP) can automate the process of extracting valuable insights from this data, ultimately enhancing aviation safety. The paper reviews ongoin

Edward H. Rohr, John T. Nardini

Agent-based models (ABMs) are widely used in biology to understand how individual actions scale into emergent population behavior. Modelers employ sensitivity analysis (SA) algorithms to quantify input parameters' impact on model outputs, however, it is hard to perform SA for ABMs due to their computational and complex nature. In this work, we develop the Si

Ehsan Ghoreishi, Bahman Abolhassani, Yan Huang, Shiva Acharya

Puncturing is a promising technique in 3GPP to multiplex Enhanced Mobile Broadband (eMBB) and Ultra-Reliable Low Latency Communications (URLLC) traffic on the same 5G New Radio (NR) air interface. The essence of puncturing is to transmit URLLC packets on demand upon their arrival, by preempting the radio resources (or subcarriers) that are already allocated

Kundan Krishna, Joseph Y Cheng, Charles Maalouf, Leon A Gatys

Existing paradigms for ensuring AI safety, such as guardrail models and alignment training, often compromise either inference efficiency or development flexibility. We introduce Disentangled Safety Adapters (DSA), a novel framework addressing these challenges by decoupling safety-specific computations from a task-optimized base model. DSA utilizes lightweigh

Jie Gao, Rajesh Jayaram, Benedikt Kolbe, Shay Sapir

Randomized dimensionality reduction is a widely-used algorithmic technique for speeding up large-scale Euclidean optimization problems. In this paper, we study dimension reduction for a variety of maximization problems, including max-matching, max-spanning tree, max TSP, as well as various measures for dataset diversity. For these problems, we show that the

Agustín Roca, Gabriel Torre, Juan I. Giribet, Gastón Castro

This paper examines the use of Unmanned Aerial Vehicles (UAVs) and deep learning for detecting endangered deer species in their natural habitats. As traditional identification processes require trained manual labor that can be costly in resources and time, there is a need for more efficient solutions. Leveraging high-resolution aerial imagery, advanced compu

Fabian Achammer, Stefan Hetzl, Renate A. Schmidt

Second-order quantifier-elimination is the problem of finding, given a formula with second-order quantifiers, a logically equivalent first-order formula. While such formulas are not computable in general, there are practical algorithms and subclasses with applications throughout computational logic. One of the most prominent algorithms for second-order quant

Saheli Mukherjee, Bivas Mallick, Sahil Gopalkrishna Naik, Ananda G. Maity

Genuine multipartite entanglement (GME) represents the strongest form of entanglement in multipartite systems, providing significant advantages in various quantum information processing tasks. In this work, we propose an experimentally feasible scheme for detecting GME, based on the truncated moments of positive maps. Our method avoids the need for full stat

Sebastian Hedwig, Gregor Zinke, Jürgen Braun, Benito Arnoldi

Layered 2D van der Waals materials, such as transition metal dichalcogenides, are promising for nanoscale spintronic and optoelectronic applications. Harnessing their full potential requires understanding both intrinsic transport and the dynamics of optically excited spin and charge carriers -- particularly the transition between excited spin polarization an

Qihui Fan, Wenbo Li, Enfu Nan, Yixiao Chen

The growing popularity of social deduction games has created an increasing need for intelligent frameworks where humans can collaborate with AI agents, particularly in post-pandemic contexts with heightened psychological and social pressures. Social deduction games like Werewolf, traditionally played through verbal communication, present an ideal application

Peter R. Young

HelioIndex is a directory of authors who are active in solar and heliospheric physics (SHP). It is available at the webpage HelioIndex.org, and it includes several derived products such as publication lists, country and institute data, journal data, and lists of the most cited articles in the field. HelioIndex is built from ORCID identifiers and publication

Rachael K. Ross, Ivan Diaz, Amy J. Pitts, Elizabeth A. Stuart

Randomized clinical trials are considered the gold standard for informing treatment guidelines, but results may not generalize to real-world populations. Generalizability is hindered by distributional differences in baseline covariates and treatment-outcome mediators. Approaches to address differences in covariates are well established, but approaches to add

Moritz Laber, Samantha Dies, Joseph Ehlert, Brennan Klein

The spread of information through socio-technical systems determines which individuals are the first to gain access to opportunities and insights. Yet, the pathways through which information flows can be skewed, leading to systematic differences in access across social groups. These inequalities remain poorly characterized in settings involving nonlinear soc

Ni Zhan, William A. Wheeler, Gil Goldshlager, Elif Ertekin

Neural network wave functions have shown promise as a way to achieve high accuracy on the many-body quantum problem. These wave functions most commonly use a determinant or sum of determinants to antisymmetrize many-body orbitals which are described by a neural network. In many cases, the wave function is projected onto a fixed-spin state. Such a treatment i

Agustín Roca, Gastón Castro, Gabriel Torre, Leonardo J. Colombo

This study compares the performance of state-of-the-art neural networks including variants of the YOLOv11 and RT-DETR models for detecting marsh deer in UAV imagery, in scenarios where specimens occupy a very small portion of the image and are occluded by vegetation. We extend previous analysis adding precise segmentation masks for our datasets enabling a fi

Łukasz Baran, Cosmin A. Dicu-Gohoreanu, Luis G. MacDowell

Efficient large-scale computer simulations of aqueous solutions require the use of accurate but simple empirical force fields for water. However, the complexity of these systems evidences the difficulties in describing solution properties without due account of polarization. Different strategies to remedy this problem are parametrizing water force fields to

Erfan Loghmani

Large language models are being widely used across industries to generate content that contributes directly to key performance metrics, such as conversion rates. Pretrained models, however, often fall short when it comes to aligning with human preferences or optimizing for business objectives. As a result, fine-tuning with good-quality labeled data is essent

M. Ryleigh Davis, Samantha K. Trumbo, Michael E. Brown, Matthew Belyakov

We present global, spatially resolved ultraviolet-visible spectra of Callisto obtained with HST/STIS and explore possible compositions of Callisto's surface material. We map the strength of a widespread downturn toward the near-UV and the NIR spectral slope from 700 to 1000 nm, which varies from slightly blue (reflectance decreasing from 700 to 1000 nm) to r

Rafael Capilla, J. Andrés Díaz-Pace, Yamid Ramírez, Jennifer Pérez

Architecture evaluation methods have long been used to evaluate software designs. Several evaluation methods have been proposed and used to analyze tradeoffs between different quality attributes. Having competing qualities leads to conflicts for selecting which quality-attribute scenarios are the most suitable ones that an architecture should tackle and for

B. A. McCullian, M. Chilcote, H. Yusuf, E. Johnston-Halperin

We demonstrate quantitative measurement of the ferromagnetic resonance (FMR) precession cone angle of a micro-scale sample of vanadium tetracyanoethylene (V[TCNE]$_{x\sim 2}$) using diamond spins. V[TCNE]$_{x\sim 2}$ is a low-damping, low-magnetization ferrimagnet with potential for scalable spintronics applications. Our study is motivated by the persistent

Enming Xing, Junjie Zhang, Shen Wang, Xiaolin Cheng

Deep learning-based prediction of protein-ligand complexes has advanced significantly with the development of architectures such as AlphaFold3, Boltz-1, Chai-1, Protenix, and NeuralPlexer. Multiple sequence alignment (MSA) has been a key input, providing coevolutionary information critical for structural inference. However, recent benchmarks reveal a major l

Keng-Jung Lee, Dharanya Sampath, Konstantinos Mavrommatis

Cell-free DNA (cfDNA) analysis is a powerful, minimally invasive tool for monitoring disease progression, treatment response, and early detection. A major challenge, however, is accurately determining the tissue of origin, especially in complex or heterogeneous disease contexts. To address this, we developed a machine learning framework that leverages tissue

Sujeet Kumar, Pretam Ray, Abhinay Beerukuri, Shrey Kamoji

Sanskrit, an ancient language with a rich linguistic heritage, presents unique challenges for automatic speech recognition (ASR) due to its phonemic complexity and the phonetic transformations that occur at word junctures, similar to the connected speech found in natural conversations. Due to these complexities, there has been limited exploration of ASR in S

Nóra Takács, Csaba Kiss, Róbert Szakáts, Emese Plachy

Hilda asteroids, which orbit in a 3:2 resonance with Jupiter, serve as key indicators of dynamical processes in the early solar system. Their spin rates, an important probe of these mechanisms, can constrain their density and collisional evolution, offering valuable insights into their origin. In this paper, we report on the identification of three fast-rota

Biqi Rebekah Zhao, Alexander Chou, Robert Peltekov, Elad Alon

Magnetic resonance imaging (MRI) exhibits rich and clinically useful endogenous contrast mechanisms, which can differentiate soft tissues and are sensitive to flow, diffusion, magnetic susceptibility, blood oxygenation level, and more. However, MRI sensitivity is ultimately constrained by Nuclear Magnetic Resonance (NMR) physics, and its spatiotemporal resol

Reydne Santos, Rafa Prado, Ana Paula de Holanda Silva, Kiev Gama

Context: Diversity can impact team communication, productivity, cohesiveness, and creativity. Analyzing the existing knowledge about diversity in open source software (OSS) projects can provide directions for future research and raise awareness about barriers and biases against underrepresented groups in OSS. Objective: This study aims to analyze the knowled

Philip A. Isenberg, Bernard J. Vasquez, Benjamin D. G. Chandran, Peera Pongkitiwanichakul

We investigate the quasilinear effects of the resonant wave-particle interaction under conditions of imbalanced turbulent heating in the collisionless coronal hole. We find that velocity-space transport of protons from the heated part of the distribution leads to strong wave growth in the minority (sunward) direction. In the present quasilinear analysis, the

Jesse Thibodeau, Hadi Nekoei, Afaf Taïk, Janarthanan Rajendran

Dynamic, risk-based pricing can systematically exclude vulnerable consumer groups from essential resources such as health insurance and consumer credit. We show that a regulator can realign private incentives with social objectives through a learned, interpretable tax schedule. First, we provide a formal proposition that bounding each firm's \emph{local} dem

Mirco Guerrini, Giuseppe Pagliara, Andrea Lavagno, Alessandro Drago

We present a framework that aims to investigate the role of thermal fluctuations of the matter composition and color-superconductivity in the nucleation of three-flavor deconfined quark matter in the typical conditions of high-energy astrophysical systems related to compact stars. It is usually assumed that the flavor composition is locally fixed during the

Reece Keller, Alyn Kirsch, Felix Pei, Xaq Pitkow

Autonomy is a hallmark of animal intelligence, enabling adaptive and intelligent behavior in complex environments without relying on external reward or task structure. Existing reinforcement learning approaches to exploration in reward-free environments, including a class of methods known as model-based intrinsic motivation, exhibit inconsistent exploration

Alireza Salemi, Hamed Zamani

Personalization is essential for question answering systems that are user-centric. Despite its importance, personalization in answer generation has been relatively underexplored. This is mainly due to lack of resources for training and evaluating personalized question answering systems. We address this gap by introducing LaMP-QA -- a benchmark designed for e

Magdalena Proszewska, Nikolay Malkin, N. Siddharth

Diffusion autoencoders (DAs) are variants of diffusion generative models that use an input-dependent latent variable to capture representations alongside the diffusion process. These representations, to varying extents, can be used for tasks such as downstream classification, controllable generation, and interpolation. However, the generative performance of

Idan Attias, Steve Hanneke, Arvind Ramaswami

We study online and transductive online learning when the learner interacts with the concept class only via Empirical Risk Minimization (ERM) or weak consistency oracles on arbitrary instance subsets. This contrasts with standard online models, where the learner knows the entire class. The ERM oracle returns a hypothesis minimizing loss on a given subset, wh

Fardin Ahsan Sakib, Ziwei Zhu, Karen Trister Grace, Meliha Yetisgen

Social determinants of health (SDOH) extraction from clinical text is critical for downstream healthcare analytics. Although large language models (LLMs) have shown promise, they may rely on superficial cues leading to spurious predictions. Using the MIMIC portion of the SHAC (Social History Annotation Corpus) dataset and focusing on drug status extraction a

Mohammadhossein Homaei, Mehran Tarif, Agustin Di Bartolo, Victor Gonzalez Morales

The Internet of Underwater Things (IoUT) has a lot of problems, like low bandwidth, high latency, mobility, and not enough energy. Routing protocols that were made for land-based networks, like RPL, don't work well in these underwater settings. This paper talks about RL-RPL-UA, a new routing protocol that uses reinforcement learning to make things work bette

William J. Huggins, Tanuj Khattar, Nathan Wiebe

For many practical applications of quantum computing, the most costly steps involve coherently accessing classical data. We help address this challenge by applying mass production techniques, which can reduce the cost of applying an operation multiple times in parallel. We combine these techniques with modern approaches for classical data loading based on "q

Simon Sinong Zhan, Qingyuan Wu, Philip Wang, Frank Yang

Offline-to-online deployment of reinforcement-learning (RL) agents must bridge two gaps: (1) the sim-to-real gap, where real systems add latency and other imperfections not present in simulation, and (2) the interaction gap, where policies trained purely offline face out-of-distribution states during online execution because gathering new interaction data is

Catherine Leroux, Joseph K. Iverson

We introduce and analyze a family of Clifford-deformed bivariate bicycle codes that are tailored for biased noise. Our qLDPC codes are defined on a bipartite hexagonal lattice with limited-range gates and low-weight stabilizers. The code is non-CSS, featuring stabilizer generators that are each half X and half Z. We find small examples with high encoding rat

Edward Fish, Richard Bowden

Recent progress in Sign Language Translation (SLT) has focussed primarily on improving the representational capacity of large language models to incorporate Sign Language features. This work explores an alternative direction: enhancing the geometric properties of skeletal representations themselves. We propose Geo-Sign, a method that leverages the properties

Gabriel Aracena, Kyle Luster, Fabio Santos, Igor Steinmacher

Effective prioritization of issue reports in software engineering helps to optimize resource allocation and information recovery. However, manual issue classification is laborious and lacks scalability. As an alternative, many open source software (OSS) projects employ automated processes for this task, yet this method often relies on large datasets for adeq

Seyed Danial Hashemi, Sunil Mittal

The emergence of dissipative Kerr solitons (DKS) in nonlinear resonators has revolutionized the generation of on-chip coherent optical frequency combs. The formation of DKS in conventional single resonators hinges on balancing the resonator dissipation against the parametric gain and balancing the resonator dispersion against the resonance frequency shifts i

M. Vieweg, V. Kott, L. Lenke, A. Schellenberger

We investigate the $U(1)$ checkerboard toric code which corresponds to the $U(1)$-symmetry enriched toric code with two distinct star sublattices. One can therefore tune from the limit of isolated stars to the uniform system. The uniform system has been conjectured to possess non-Abelian topological order based on quantum Monte Carlo simulations suggesting a

Phillip Bridgham, Alex Delacroix, Laura Domine, Andriy Fedorenko

Scientific investigation of Unidentified Anomalous Phenomena (UAP) is limited by poor data quality and a lack of transparency. Existing data are often fragmented, uncalibrated, and missing critical metadata. To address these limitations, the authors present the Observatory Class Integrated Computing Platform (OCICP), a system designed for the systematic and

Iyán Méndez Veiga, Esther Hänggi

Quantum cryptographic protocols do not rely only on quantum-physical resources, they also require reliable classical communication and computation. In particular, the secrecy of any quantum key distribution protocol critically depends on the correct execution of the privacy amplification step. This is a classical post-processing procedure transforming a part

Gen Luo, Ganlin Yang, Ziyang Gong, Guanzhou Chen

The remarkable progress of Multimodal Large Language Models (MLLMs) has attracted increasing attention to extend them to physical entities like legged robot. This typically requires MLLMs to not only grasp multimodal understanding abilities, but also integrate visual-spatial reasoning and physical interaction capabilities. Nevertheless,existing methods strug

Dajun Liu, Hanpeng Gao, Yu-Zhe Liu

We constructed some tensor functors that send each exceptional sequence in a module category to another exceptional sequence in another module category by using split extensions and recollements.

Adam Boesky, V. Ashley Villar, Alexander Gagliano, Brian Hsu

The upcoming Legacy Survey of Space and Time (LSST) conducted by the Vera C. Rubin Observatory will detect millions of supernovae (SNe) and generate millions of nightly alerts, far outpacing available spectroscopic resources. Rapid, scalable photometric classification methods are therefore essential for identifying young SNe for follow-up and enabling large-

Raji Ashenafi Mamade, Barton Zwiebach

We study the theory of massless fields of type II strings arising from the string field theory that uses two string fields, a physical one and an extra one that allows the writing of an action, but whose degrees of freedom ultimately decouple. The mechanism allowing the description of the self-dual five-form of type IIB, anticipated by Sen, is used by the SF

Chi Lung Cheng, Ranit Das, Runze Li, Radha Mastandrea

Statistical inference in physics is often based on samples from a generator (sometimes referred to as a ``forward model") that emulate experimental data and depend on parameters of the underlying theory. Modern machine learning has supercharged this workflow to enable high-dimensional and unbinned analyses to utilize much more information than ever before. W

Antoine Lemelin, Christophe Pere, Olivier Landon-Cardinal, Camille Coti

We explore the usefulness of mid-circuit measurements to enhance quantum algorithmics. Specifically, we assess how quantum phase estimation (QPE) and mid-circuit measurements can improve the performance of variational quantum algorithms. Our focus is on the single-qubit version of QPE namely, the Hadamard test applied to the Quantum Approximate Optimization

J. M. Bestenlehner, Paul A. Crowther, V. A. Bronner, S. Simon-Diaz

We aim to determine the physical properties of OB stars from the multi-epoch VLT/FLAMES BLOeM spectroscopic survey of the Small Magellanic Cloud. We apply a pipeline designed to analyse large spectroscopic samples of OB stars to the co-added, initial 9 epochs of the BLOeM survey, utilising grids of synthetic model spectra computed with the stellar atmosphere

Piotr Sierant, Paolo Stornati, Xhek Turkeshi

Understanding the computational complexity of quantum states is a central challenge in quantum many-body physics. In qubit systems, fermionic Gaussian states can be efficiently simulated on classical computers and hence can be employed as a natural baseline for evaluating quantum complexity. In this work, we develop a framework for quantifying fermionic magi

Jiaxin Qiao, Yoshito Watanabe, Simon Trebst

Motivated by the recent introduction of a $U(1)$-symmetric toric code model, we investigate symmetry-based deformations of topological order by systematically deconstructing the Gauss-law-enforcing star terms of the toric code (TC) Hamiltonian. This "term-dropping" protocol introduces global symmetries that go beyond the alternative framework of "ungauging"

Sean Benevedes, Jesse Thaler

We introduce wifi ensembles as a novel framework to obtain asymptotic frequentist uncertainties on density ratios, with a particular focus on neural ratio estimation in the context of high-energy physics. When the density ratio of interest is a likelihood ratio conditioned on parameters, wifi ensembles can be used to perform simulation-based inference on tho

David Schmid, Yìlè Yīng, Matthew Leifer

We define a class of Copenhagenish interpretations encompassing modern interpretations that follow the Copenhagen spirit. These interpretations are characterized by four postulates: Observers Observe, Universality, Anti-$\psi$-ontology, and Completeness. We explain why such interpretations are not equivalent to the textbook (or orthodox) interpretation, nor

Lori E. Porter, Matthew Abruzzo, Greg L. Bryan, Mary Putman

High-velocity clouds (HVCs) may fuel future star formation in the Milky Way, but they must first survive their passage through the hot halo. While recent work has improved our understanding of the survival criterion for cloud-wind interactions, few observational comparisons exist that test this criterion. We therefore present an initial comparison of simulat

Ya-Hui Zhang

Recent experiments have observed superconductivity proximate to the nu = -2/3 fractional quantum anomalous Hall (FQAH) insulator in twisted MoTe2. A critical open question is whether the underlying normal state is a Fermi liquid with a large Fermi surface or a strongly correlated metal with low carrier density. In this work, we develop a theory of the phases

Maria Renee Meza, Xiaoshan Huang, Shane W. Davis, Yan-Fei Jiang

We perform 3D radiation-magnetohydrodynamic simulations of the evolution of the fallback debris after a tidal disruption event. We focus on studying the effects of magnetic fields on the formation and early evolution of the accretion flow. We find that large magnetic fields can increase the debris stream thickness, moderately reducing the efficiency of the r

Giorgi Tukhashvili

Higher curvature corrections to the Einstein-Hilbert term may play an important role in probing the strong-field regime of gravity. In this letter, we demonstrate that the local effective action reproducing the trace anomaly can resemble the Einstein-Gauss-Bonnet theory in four dimensions on specific backgrounds. The two key observations support this claim:

Yaxin Luo, Zhaoyi Li, Jiacheng Liu, Jiacheng Cui

CAPTCHAs have been a critical bottleneck for deploying web agents in real-world applications, often blocking them from completing end-to-end automation tasks. While modern multimodal LLM agents have demonstrated impressive performance in static perception tasks, their ability to handle interactive, multi-step reasoning challenges like CAPTCHAs is largely unt

Yangyi Huang, Ye Yuan, Xueting Li, Jan Kautz

Existing methods for image-to-3D avatar generation struggle to produce highly detailed, animation-ready avatars suitable for real-world applications. We introduce AdaHuman, a novel framework that generates high-fidelity animatable 3D avatars from a single in-the-wild image. AdaHuman incorporates two key innovations: (1) A pose-conditioned 3D joint diffusion

Yu Zhang, Yunqi Li, Yifan Yang, Rui Wang

Although chain-of-thought reasoning and reinforcement learning (RL) have driven breakthroughs in NLP, their integration into generative vision models remains underexplored. We introduce ReasonGen-R1, a two-stage framework that first imbues an autoregressive image generator with explicit text-based "thinking" skills via supervised fine-tuning on a newly gener

Adam Stein, Aaditya Naik, Neelay Velingker, Mayur Naik

Neuro-symbolic learning was proposed to address challenges with training neural networks for complex reasoning tasks with the added benefits of interpretability, reliability, and efficiency. Neuro-symbolic learning methods traditionally train neural models in conjunction with symbolic programs, but they face significant challenges that limit them to simplist

Bojia Zi, Weixuan Peng, Xianbiao Qi, Jianan Wang

Recent advances in video diffusion models have driven rapid progress in video editing techniques. However, video object removal, a critical subtask of video editing, remains challenging due to issues such as hallucinated objects and visual artifacts. Furthermore, existing methods often rely on computationally expensive sampling procedures and classifier-free

Zilin Xiao, Jaywon Koo, Siru Ouyang, Jefferson Hernandez

Recent advancements in reinforcement learning with verifiable rewards have pushed the boundaries of the visual reasoning capabilities in large vision-language models (LVLMs). However, training LVLMs with reinforcement fine-tuning (RFT) is computationally expensive, posing a significant challenge to scaling model size. In this work, we propose ProxyThinker, a

Yiqing Liang, Jielin Qiu, Wenhao Ding, Zuxin Liu

Reinforcement Learning with Verifiable Rewards (RLVR) has recently emerged as a powerful paradigm for post-training large language models (LLMs), achieving state-of-the-art performance on tasks with structured, verifiable answers. Applying RLVR to Multimodal LLMs (MLLMs) presents significant opportunities but is complicated by the broader, heterogeneous natu

Zehan Wang, Jiayang Xu, Ziang Zhang, Tianyu Pang

Humans can intuitively compose and arrange scenes in the 3D space for photography. However, can advanced AI image generators plan scenes with similar 3D spatial awareness when creating images from text or image prompts? We present GenSpace, a novel benchmark and evaluation pipeline to comprehensively assess the spatial awareness of current image generation m

Ce Zhang, Yan-Bo Lin, Ziyang Wang, Mohit Bansal

Recent advances in test-time optimization have led to remarkable reasoning capabilities in Large Language Models (LLMs), enabling them to solve highly complex problems in math and coding. However, the reasoning capabilities of multimodal LLMs (MLLMs) still significantly lag, especially for complex video-language tasks. To address this issue, we present SILVR

Kalle Alaluusua, Konstantin Avrachenkov, B. R. Vinay Kumar, Lasse Leskelä

Subspace clustering becomes inherently difficult near intersections, where points from different subspaces are barely separated. Most existing theoretical results address this issue by imposing separation or sampling assumptions that limit the statistical effect of points near the intersection. We study a minimal setting of two intersecting lines in which th

Ujjwal Upadhyay, Mukul Ranjan, Zhiqiang Shen, Mohamed Elhoseiny

Recent advances in vision-language models (VLMs) have made impressive strides in understanding spatio-temporal relationships in videos. However, when spatial information is obscured, these models struggle to capture purely temporal patterns. We introduce $\textbf{SpookyBench}$, a benchmark where information is encoded solely in temporal sequences of noise-li

Xinqi Xiong, Prakrut Patel, Qingyuan Fan, Amisha Wadhwa

The rapid advancement of talking-head deepfake generation fueled by advanced generative models has elevated the realism of synthetic videos to a level that poses substantial risks in domains such as media, politics, and finance. However, current benchmarks for deepfake talking-head detection fail to reflect this progress, relying on outdated generators and o

Marco Scigliuzzo, Léo Peyruchat, Riccardo Maria Marabini, Carla Becker

Precise control of mechanical modes in the quantum regime is a key resource for quantum technologies, offering promising pathways for quantum sensing with macroscopic systems and scalable architectures for quantum simulation. In this work, we realise a multimode mechanical cavity coupled to a superconducting Kerr resonator, which induces nonlinearity in the

Mingjie Liu, Shizhe Diao, Ximing Lu, Jian Hu

Recent advances in reasoning-centric language models have highlighted reinforcement learning (RL) as a promising method for aligning models with verifiable rewards. However, it remains contentious whether RL truly expands a model's reasoning capabilities or merely amplifies high-reward outputs already latent in the base model's distribution, and whether cont

Le Jiang, Liyan Ma, Guang Yang

Federated learning (FL) has emerged as a transformative framework for privacy-preserving distributed training, allowing clients to collaboratively train a global model without sharing their local data. This is especially crucial in sensitive fields like healthcare, where protecting patient data is paramount. However, privacy leakage remains a critical challe

Junyu Zhang, Runpei Dong, Han Wang, Xuying Ning

This paper presents AlphaOne ($\alpha$1), a universal framework for modulating reasoning progress in large reasoning models (LRMs) at test time. $\alpha$1 first introduces $\alpha$ moment, which represents the scaled thinking phase with a universal parameter $\alpha$. Within this scaled pre-$\alpha$ moment phase, it dynamically schedules slow thinking transi

Cailin Zhuang, Ailin Huang, Yaoqi Hu, Jingwei Wu

Story visualization aims to generate coherent image sequences that faithfully represent a narrative and match given character references. Despite progress in generative models, existing benchmarks remain narrow in scope, often limited to short prompts, lacking character references, or single-image cases, failing to reflect real-world narrative complexity and

Avery S. Williamson, Michael J. Bennington, Ravesh Sukhnandan, Mrinali Nakhre

Many bioinspired robots mimic the rigid articulated joint structure of the human hand for grasping tasks, but experience high-frequency mechanical perturbations that can destabilize the system and negatively affect precision without a high-frequency controller. Despite having bandwidth-limited controllers that experience time delays between sensing and actua

Joschka Braun, Carsten Eickhoff, Seyed Ali Bahrainian

Steering vectors are a lightweight method for controlling text properties by adding a learned bias to language model activations at inference time. While predominantly studied for multiple-choice and toy tasks, their effectiveness in free-form generation remains largely unexplored. Moving "Beyond Multiple Choice," we evaluate steering vectors for controlling

Siru Ouyang, Xinyu Zhu, Zilin Xiao, Minhao Jiang

Reinforcement learning (RL) has become a powerful approach for improving the reasoning capabilities of large language models (LLMs), as evidenced by recent successes such as OpenAI's o1 and Deepseek-R1. However, applying RL at scale remains intimidatingly resource-intensive, requiring multiple model copies and extensive GPU workloads. On the other hand, whil

Gabrielle Kaili-May Liu, Gal Yona, Avi Caciularu, Idan Szpektor

A critical component in the trustworthiness of LLMs is reliable uncertainty communication, yet LLMs often use assertive language when conveying false claims, leading to over-reliance and eroded trust. We present the first systematic study of $\textit{faithful confidence calibration}$ of LLMs, benchmarking models' ability to use linguistic expressions of unce

Heli Ben-Hamu, Itai Gat, Daniel Severo, Niklas Nolte

Recent masked diffusion models (MDMs) have shown competitive performance compared to autoregressive models (ARMs) for language modeling. While most literature has focused on performance enhancing sampling procedures, efficient sampling from MDMs has been scarcely explored. We make the observation that often a given sequence of partially masked tokens determi

Brendan P. Crill, Yoonsoo P. Bach, Sean A. Bryan, Jean Choppin de Janvry

We describe the SPHEREx Sky Simulator, a software tool designed to model science data for NASA's SPHEREx mission that will carry out a series of all-sky spectrophotometric surveys at $\sim$6'' spatial resolution in 102 spectral channels spanning 0.75 to 5 $\mu$m. The Simulator software implements models for astrophysical emission, instrument characteristics,

Giulio Costantini, Lorenzo Caprini, Umberto Marini Bettolo Marconi, Fabio Cecconi

Understanding the link between structure and function in proteins is fundamental in molecular biology and proteomics. A central question in this context is whether allostery - where the binding of a molecule at one site affects the activity of a distant site - emerges as a further manifestation of the intricate interplay between structure, function, and intr

Nicola Agnew, Veronika Vohníková, Erling Riis, Graham Machin

Doppler-broadening thermometry (DBT) can be used as a calibration-free primary reference suitable for practical applications, e.g. reliably measuring temperatures over long periods of time in environments where sensor retrieval is impractical. We report on our proof-of-concept investigations into DBT with alkali metal vapour cells, with a particular focus on

Zhao Mandi, Yifan Hou, Dieter Fox, Yashraj Narang

We study the problem of functional retargeting: learning dexterous manipulation policies to track object states from human hand-object demonstrations. We focus on long-horizon, bimanual tasks with articulated objects, which is challenging due to large action space, spatiotemporal discontinuities, and embodiment gap between human and robot hands. We propose D

Michelle Chalupnik, Brian Doolittle, Suparna Seshadri, Eric G. Brown

Quantum key distribution (QKD) can provide secure key material between two parties without relying on assumptions about the computational power of an eavesdropper. QKD is performed over quantum links and quantum networks, systems which are resource-intensive to deploy and maintain. To evaluate and optimize performance prior to, during, and after deployment,

Shuyao Xu, Cheng Peng, Jiangxuan Long, Weidi Xu

Recent advances in model distillation show that data from advanced reasoning models can effectively train smaller student models. However, standard practices discard incorrect reasoning traces -- valuable, yet underutilized data. This paper addresses the critical question: How can both positive and negative distilled reasoning traces be effectively leveraged