March 2025 arXiv papers — page 72

Andrei Neguţ

The q-characters of quantum loop algebras are very important objects in representation theory. In [20], we showed that q-characters factor as a power series of the form studied in [9] times a character, an important phenomenon which had already been known in finite types. In the present paper, we prove a conjectural formula for the aforementioned character f

Andrew McNutt, Maggie K McCracken, Ishrat Jahan Eliza, Daniel Hajas

Data visualizations are typically not accessible to blind and low-vision (BLV) users. Automatically generating text descriptions offers an enticing mechanism for democratizing access to the information held in complex scientific charts, yet appropriate procedures for generating those texts remain elusive. Pursuing this issue, we study a single complex chart

Arkady Kitover, Mehmet Orhon

We obtain a complete description of semi-Fredholm spectra of operators of the form $(Tf)(z) = w(z)f(B(z)$ acting on the disc algebra in the case when $B$ is either elliptic or double parabolic finite Blaschke product of degree $d \geq 2$ and $w$ has no zeros on the unit circle. In the case when $B$ has zeros on the unit circle we provide only some partial re

Ítalo Romani de Oliveira, Samet Ayhan, Glaucia Balvedi, Michael Biglin

Predicting air traffic congestion and flow management is essential for airlines and Air Navigation Service Providers (ANSP) to enhance operational efficiency. Accurate estimates of future airport capacity and airspace density are vital for better airspace management, reducing air traffic controller workload and fuel consumption, ultimately promoting sustaina

Ken Ziyu Liu, Christopher A. Choquette-Choo, Matthew Jagielski, Peter Kairouz

An important question today is whether a given text was used to train a large language model (LLM). A \emph{completion} test is often employed: check if the LLM completes a sufficiently complex text. This, however, requires a ground-truth definition of membership; most commonly, it is defined as a member based on the $n$-gram overlap between the target text

Giuseppe Franco, Pablo Monteagudo-Lago, Ian Colbert, Nicholas Fraser

The size of a model has been a strong predictor of its quality, as well as its cost. As such, the trade-off between model cost and quality has been well-studied. Post-training optimizations like quantization and pruning have typically focused on reducing the overall volume of pre-trained models to reduce inference costs while maintaining model quality. Howev

Alessandro Bellina, Vito D. P. Servedio

Statistical regularities in human language have fascinated researchers for decades, suggesting deep underlying principles governing its evolution and information structuring for efficient communication. While Zipf's Law describes the frequency-rank distribution of words, deviations from this pattern-particularly for less frequent words-challenge the notion o

Ayberk Acar, Jumanh Atoum, Peter S. Connor, Clifford Pierre

Ureteroscopy is the standard of care for diagnosing and treating kidney stones and tumors. However, current ureteroscopes have a limited field of view, requiring significant experience to adequately navigate the renal collecting system. This is evidenced by the fact that inexperienced surgeons have higher rates of missed stones. One-third of patients with re

Jeremiah Gbadegoye, Mustafa C. Camur, Xueping Li

With the steady increase in global logistics and freight transport demand, the need for efficient and sustainable intermodal transport systems becomes increasingly important. This study addresses the optimization of container movement by intermodal transport with fixed train schedules. We emphasize the integration of road-rail intermodal transport amid uncer

Joseph Gatto, Parker Seegmiller, Timothy Burdick, Inas S. Khayal

Follow-up question generation is an essential feature of dialogue systems as it can reduce conversational ambiguity and enhance modeling complex interactions. Conversational contexts often pose core NLP challenges such as (i) extracting relevant information buried in fragmented data sources, and (ii) modeling parallel thought processes. These two challenges

Drossos Gintides, Leonidas Mindrinos

We present two applications of the integro-differential volume equation for the eigenstrain, building on Eshelby's inclusion method [15,16], in the contexts of both static and dynamic linear elasticity. The primary objective is to address the inverse problem of recovering the elastic moduli of the inhomogeneity using a limited number of incident fields. In t

Ahmed H. Salamah, Pierre McWhannel, Nicole Yan

Information retrieval systems have traditionally relied on exact term match methods such as BM25 for first-stage retrieval. However, recent advancements in neural network-based techniques have introduced a new method called dense retrieval. This approach uses a dual-encoder to create contextual embeddings that can be indexed and clustered efficiently at run-

Xinwei Liu, Vladimir Dvorkin

When solving decision-making problems with mathematical optimization, some constraints or objectives may lack analytic expressions but can be approximated from data. When an approximation is made by neural networks, the underlying problem becomes optimization over trained neural networks. Despite recent improvements with cutting planes, relaxations, and heur

Navaneeth N, Souvik Chakraborty

Modeling cardiovascular anatomies poses a significant challenge due to their complex, irregular structures and inherent pathological conditions. Numerical simulations, while accurate, are often computationally expensive, limiting their practicality in clinical settings. Traditional machine learning methods, on the other hand, often struggle with some major h

Buse Carik, Victoria Izaac, Xiaohan Ding, Angela Scarpa

Autistic individuals often experience negative self-talk (NST), leading to increased anxiety and depression. While therapy is recommended, it presents challenges for many autistic individuals. Meanwhile, a growing number are turning to large language models (LLMs) for mental health support. To understand how autistic individuals perceive AI's role in coping

Anran Xu, Lindsey J. Heagy

In this work, we employ neural fields, which use neural networks to map a coordinate to the corresponding physical property value at that coordinate, in a test-time learning manner. For a test-time learning method, the weights are learned during the inversion, as compared to traditional approaches which require a network to be trained using a training datase

Emanuela Barzi, Masaki Takeuchi, Daniele Turrioni, Akihiro Kikuchi

The joint expertise of ANL and FNAL has led to the production of Nb3Sn undulator magnets in operation in the ANL Advanced Photon Source (APS). These magnets showed performance reproducibility close to the short sample limit, and a design field increase of 20% at 820A. However, the long training did not allow obtaining the expected 50% increase of the on-axis

Hamed Jelodar, Mohammad Meymani, Roozbeh Razavi-Far

Large language models (LLMs) and transformer-based architectures are increasingly utilized for source code analysis. As software systems grow in complexity, integrating LLMs into code analysis workflows becomes essential for enhancing efficiency, accuracy, and automation. This paper explores the role of LLMs for different code analysis tasks, focusing on thr

Christopher J. Ford, Haoran Li, Manuel G. Catalano, Matteo Bianchi

This paper presents a shear-based control scheme for grasping and manipulating delicate objects with a Pisa/IIT anthropomorphic SoftHand equipped with soft biomimetic tactile sensors on all five fingertips. These `microTac' tactile sensors are miniature versions of the TacTip vision-based tactile sensor, and can extract precise contact geometry and force inf

Louis Owen, Abhay Kumar, Nilabhra Roy Chowdhury, Fabian Güra

The outcome of Large Language Model (LLM) pre-training strongly depends on weight initialization and variance control strategies. Although the importance of initial variance control has been well documented in neural networks in general, the literature on initialization and management of its growth during LLM pre-training, specifically, is somewhat sparse. I

Joey Mulé, Dhandeep Challagundla, Rachit Saini, Riadul Islam

Event-based vision revolutionizes traditional image sensing by capturing asynchronous intensity variations rather than static frames, enabling ultrafast temporal resolution, sparse data encoding, and enhanced motion perception. While this paradigm offers significant advantages, conventional event-based datasets impose a fixed thresholding constraint to deter

Catherine Cerny, Guillaume Mahler, Keren Sharon, Mathilde Jauzac

We leverage JWST's superb resolution to derive strong lensing mass maps of 14 clusters, spanning a redshift range of $z\sim0.25 - 1.06$ and a mass range of $M_{500}\sim2-12 \times 10^{14}M_\odot$, from the Strong LensIng and Cluster Evolution (SLICE) JWST program. These clusters represent a small subsample of the first clusters observed in the SLICE program

Daniel Kuhse, Harun Teper, Sebastian Buschjäger, Chien-Yao Wang

We introduce AnytimeYOLO, a family of variants of the YOLO architecture that enables anytime object detection. Our AnytimeYOLO networks allow for interruptible inference, i.e., they provide a prediction at any point in time, a property desirable for safety-critical real-time applications. We present structured explorations to modify the YOLO architecture, en

Cade Ballew, Thomas Trogdon, Heather Wilber

Two inverse-free iterative methods are developed for solving Sylvester matrix equations when the spectra of the coefficient matrices are on, or near, known disjoint subintervals of the real axis. Both methods use the recently-introduced Akhiezer iteration: one to address an equivalent problem of approximating the matrix sign function applied to a block matri

Collin Nolte

In 2025, we identified a methodological issue in the bootstrapped differences of times series (BDOTS) first introduced in 2017 resulting in a significant inflation of the family-wise error rate. The goal of the present manuscript is threefold: to identify the problem in the original methodology, to present two alternative solutions, and to compare estimates

Shivam Gupta, Sushrut Karmalkar

Knowledge distillation is a technique used to train a small student network using the output generated by a large teacher network, and has many empirical advantages~\citep{Hinton2015DistillingTK}. While the standard one-shot approach to distillation only uses the output of the final teacher network, recent work~\citep{panigrahi2024progressive} has shown that

Aidos Konyspay, Pakizar Shamoi, Malika Ziyada, Zhusup Smambayev

Internet memes are a central element of online culture, blending images and text. While substantial research has focused on either the visual or textual components of memes, little attention has been given to their interplay. This gap raises a key question: What methodology can effectively compare memes and the emotions they elicit? Our study employs a multi

Cas D. van Erp, Ana Monreal-Ibero, Jelmer C. Stroo, Peter M. Weilbacher

[ABRIDGED] DIBs are faint absorption features of mainly unknown origin. Observational constraints on their carriers have been provided in the vast majority of the cases via observations in our Galaxy. Detections in other galaxies are scarce. However, they can further constrain the nature of the carriers by sampling different environments and they can put int

Emanuele Giacomini, Luca Di Giammarino, Lorenzo De Rebotti, Giorgio Grisetti

LiDARs provide accurate geometric measurements, making them valuable for ego-motion estimation and reconstruction tasks. Although its success, managing an accurate and lightweight representation of the environment still poses challenges. Both classic and NeRF-based solutions have to trade off accuracy over memory and processing times. In this work, we build

Mael Cavan-Piton, Diego Guadagnoli, Axel Iohner, Pablo Fernandez-Menendez

We consider an axion flux on Earth consistent with emission from the Supernova explosion SN 1987A. Using Chiral Perturbation Theory augmented with an axion, we calculate the energy spectrum of $a + N \to N + \gamma$ as well as $a + N \to N + \pi^0$, where $N$ denotes a nucleon in a water tank, such as the one planned for the Hyper-Kamiokande neutrino detecti

Shu Pu, Yaochen Wang, Dongping Chen, Yuhang Chen

Evaluating generative foundation models on open-ended multimodal understanding (MMU) and generation (MMG) tasks across diverse modalities (e.g., images, audio, video) poses significant challenges due to the complexity of cross-modal interactions. To this end, the idea of utilizing Multimodal LLMs (MLLMs) as automated judges has emerged, with encouraging resu

Tianwen Zhou, Jing Wang, Songtao Wu, Kuanhong Xu

Recent approaches using large-scale pretrained diffusion models for image dehazing improve perceptual quality but often suffer from hallucination issues, producing unfaithful dehazed image to the original one. To mitigate this, we propose ProDehaze, a framework that employs internal image priors to direct external priors encoded in pretrained models. We intr

Zhengqing Gao, Dongting Hu, Jia-Wang Bian, Huan Fu

3D Gaussian Splatting (3DGS) has made significant strides in novel view synthesis but is limited by the substantial number of Gaussian primitives required, posing challenges for deployment on lightweight devices. Recent methods address this issue by compressing the storage size of densified Gaussians, yet fail to preserve rendering quality and efficiency. To

Lama Ayash, Hassan Alhuzali, Ashwag Alasmari, Sultan Aloufi

Large Language Models (LLMs) have demonstrated remarkable capabilities in natural language processing; however, they often struggle to accurately capture and reflect cultural nuances. This research addresses this challenge by focusing on Saudi Arabia, a country characterized by diverse dialects and rich cultural traditions. We introduce SaudiCulture, a novel

Chengyin Wu, Yanzhou Ji

Phase-field models have proven indispensable for deciphering the microstructure complexities inherent in multicomponent systems. The confluence of varying phase molar volumes presents unique challenges. Understanding the impact of molar volume differences on multiphase systems is of crucial significance, as it directly influences the system's thermodynamic a

Jonah J. Glunt, Joshua A. Robbins, Jacob A. Siefert, Daniel Silvestre

Mixed integer set representations, and specifically hybrid zonotopes, have enabled new techniques for reachability and verification of nonlinear and hybrid systems. Mixed-integer sets which have the property that their convex relaxation is equal to their convex hull are said to be sharp. This property allows the convex hull to be computed with minimal overhe

Keyon Vafa, Sarah Bentley, Jon Kleinberg, Sendhil Mullainathan

How should we evaluate the quality of generative models? Many existing metrics focus on a model's producibility, i.e. the quality and breadth of outputs it can generate. However, the actual value from using a generative model stems not just from what it can produce but whether a user with a specific goal can produce an output that satisfies that goal. We ref

Emily Townsend, Joshua Pomeroy, Garnett W. Bryant

We show that resonant driving between the spin up and spin down states of an electron spin-qubit in a quantum dot reduces the occupancy of the dot through leakage to an appropriately tuned lead. A nearby charge sensor measuring the occupancy of the dot should be able to detect the narrow resonant condition upon sweeping the driving frequency. The presence of

Jaromír Fiurášek

We investigate fundamental bounds on the ability to determine photon number distribution and other related quantities from tomographically incomplete measurements with an array of M detectors that can only distinguish the absence or presence of photons. We show that the lower and upper bounds on photon number probabilities can be determined by solving a line

Yiwen Xu, Monideep Chakraborti, Tianyi Zhang, Katelyn Eng

In this paper, we present Speak Ease: an augmentative and alternative communication (AAC) system to support users' expressivity by integrating multimodal input, including text, voice, and contextual cues (conversational partner and emotional tone), with large language models (LLMs). Speak Ease combines automatic speech recognition (ASR), context-aware LLM-ba

Miguel López-Pérez, Marco Miani, Valery Naranjo, Søren Hauberg

Generative models are popular for medical imaging tasks such as anomaly detection, feature extraction, data visualization, or image generation. Since they are parameterized by deep learning models, they are often sensitive to distribution shifts and unreliable when applied to out-of-distribution data, creating a risk of, e.g. underrepresentation bias. This b

P G Kubendran Amos

Efforts to promote equity in higher education often rely on shared intent among instructors and students. Yet, as demonstrated in this study, when students form their own teams for Team-Based Learning (TBL) tasks, they unintentionally cluster with peers of similar socio-economic backgrounds, ultimately undermining equity. This study introduces a simple strat

Gary Y. Li, Li Chen, Bryson Hicks, Nikolai Schnittke

Computer-aided pathology detection algorithms for video-based imaging modalities must accurately interpret complex spatiotemporal information by integrating findings across multiple frames. Current state-of-the-art methods operate by classifying on video sub-volumes (tubelets), but they often lose global spatial context by focusing only on local regions with

B. Hamil, B. C. Lütfüoğlu

We study the properties of a nonlinear magnetic-charged black hole in the presence of a phantom global monopole. By incorporating nonlinear electrodynamics (NLE) and exotic scalar fields, we derive an exact black hole solution and analyze its geometric structure, causal properties, and thermodynamic behavior. We examine how the presence of a phantom global m

Cathy Mengying Fang, Auren R. Liu, Valdemar Danry, Eunhae Lee

As people increasingly seek emotional support and companionship from AI chatbots, understanding how such interactions impact mental well-being becomes critical. We conducted a four-week randomized controlled experiment (n=981, >300k messages) to investigate how interaction modes (text, neutral voice, and engaging voice) and conversation types (open-ended, no

H. X. Ren, Q. Remy, S. Ravikularaman, M. Bouyahiaoui

Our goal is to estimate the total gas mass in the direction of the Central Molecular Zone (CMZ), quantify the various uncertainties associated, and discuss the implications for the estimates of CR energy densities and dust opacities. The $H_{\rm{I}}$ 21 cm line and the carbon monoxide isotopes ($^{12}\rm{CO}$, $^{13}\rm{CO}$ and $\rm{C}^{18}\rm{O}$) line emi

Batoul Banihashemi, Edgar Shaghoulian, Sanjit Shashi

Recent work has studied the thermodynamics of gravity subjected to conformal boundary conditions, where the trace of the extrinsic curvature $K$ and the conformal class of metrics are kept fixed. The case $K < 0$ seems tied to solutions with cosmic horizons. To study this situation further, we analyze Einstein-Maxwell theory, where we find solutions with cos

Le Wang, Krishna Prasad Koirala, Shuhang Wu, Jueli Shi

High-entropy oxides (HEOs) offer exceptional compositional flexibility and structural stability, making them promising materials for energy and catalytic applications. Here, we investigate Sr doping effects on B-site cation oxidation states, local composition, and structure in epitaxial La1-xSrx(Cr0.2Mn0.2Fe0.2Co0.2Ni0.2)O3 thin films. X-ray spectroscopies r

Sahil Tyagi, Prateek Sharma

Deep learning systems are optimized for clusters with homogeneous resources. However, heterogeneity is prevalent in computing infrastructure across edge, cloud and HPC. When training neural networks using stochastic gradient descent techniques on heterogeneous resources, performance degrades due to stragglers and stale updates. In this work, we develop an ad

Mbaimou Auxence Ngremmadji, Freddy Odille, Charline Bertholdt, Marine Beaumont

Intravoxel incoherent motion (IVIM) is a diffusion-weighted magnetic resonance imaging (MRI) method that may be applied to the placenta to help diagnose abnormal pregnancies. IVIM requires prolonged scan times, followed by a model-based estimation procedure. Maternal or fetal motion during the scan affects the accuracy of this estimation. In this work, we pr

Yuxuan Wei, Zehan Wang, Tian Guo, Hao Liu

Point clouds, which directly record the geometry and attributes of scenes or objects by a large number of points, are widely used in various applications such as virtual reality and immersive communication. However, due to the huge data volume and unstructured geometry, efficient compression of point clouds is very crucial. The Moving Picture Expert Group is

André Pedroso Kowacs, Alexandre Kirilov

This paper provides a complete characterization of global hypoellipticity and solvability with loss of derivatives for Fourier multiplier operators on the $n$-dimensional torus. We establish necessary and sufficient conditions for these properties and examine their connections with classical notions of global hypoellipticity and solvability, particularly in

Florent Fessler, Pierre Muller, Antonio Stocco

Transport of microscopic objects across biological membranes usually involves membrane deformation to enclose the object followed by detachment of the engulfed particle. However, in artificial membranes, this last topological remodelling step is in many cases not spontaneous due to the elastic stability of the neck structure formed upon complete particle wra

Nicolás Allo-Gómez

It is a well-known and easily established fact that every Euclidean domain is also a principal ideal domain. However, the converse statement is not true, and this is usually shown by exhibiting as a counterexample the ring of algebraic integers in a certain, very specific quadratic field, and the proof that this works is quite unnatural and technical. In thi

Victor Zucatti, Matthew J. Zahr

Reduced-order models (ROMs) remain generally unreliable for convection-dominated problems, such as those encountered in hypersonic flows, due to the slowly decaying Kolmogorov $n$-width of linear subspace approximations, known as the Kolmogorov barrier. This limitation hinders the accuracy of traditional ROMs and necessitates impractical amounts of training

Hebert Pérez-Rosés

We make an asymptotic analysis via singularity analysis of generating functions of a number sequence that involves the Fibonacci numbers and generalizes the binomial coefficients.

Reem Gody, Mahmoud Goudy, Ahmed Y. Tawfik

In this paper, we present ConvoGen: an innovative framework for generating synthetic conversational data using multi-agent systems. Our method leverages few-shot learning and introduces iterative sampling from a dynamically updated few-shot hub to create diverse and realistic conversational scenarios. The generated data has numerous applications, including t

Quan Wang, Yuanyuan Hua, Chenxu Li, Zhizheng Yuan

Diffuse correlation spectroscopy (DCS) is a noninvasive optical technique that probes microvascular blood flow in deep tissues. Here, we present and validate a new on-chip hardware correlator for high-speed DCS measurements. The correlator is embedded in a custom-built 512 x 512 single-photon avalanche diode (SPAD) array named ATLAS, which computes intensity

Taylor Lundy, Narun Raman, Scott Duke Kominers, Kevin Leyton-Brown

Conspicuous consumption occurs when a consumer derives value from a good based on its social meaning as a signal of wealth, taste, and/or community affiliation. Common conspicuous goods include designer footwear, country club memberships, and artwork; conspicuous goods also exist in the digital sphere, with non-fungible tokens (NFTs) as a prominent example.

Soumen Kumar Mondal, Sayambhu Sen, Abhishek Singhania, Preethi Jyothi

Multilingual large language models (LLMs) aim towards robust natural language understanding across diverse languages, yet their performance significantly degrades on low-resource languages. This work explores whether existing techniques to identify language-specific neurons can be leveraged to enhance cross-lingual task performance of lowresource languages.

Euclid Collaboration, C. Bellhouse, J. B. Golden-Marx, S. P. Bamford

The intracluster light (ICL) permeating galaxy clusters is a tracer of the cluster's assembly history, and potentially a tracer of their dark matter structure. In this work we explore the capability of the Euclid Wide Survey to detect ICL using H-band mock images. We simulate clusters across a range of redshifts (0.3-1.8) and halo masses ($10^{13.9}$-$10^{15

Ali Beikmohammadi, Sarit Khirirat, Peter Richtárik, Sindri Magnússon

Federated reinforcement learning (FedRL) enables collaborative learning while preserving data privacy by preventing direct data exchange between agents. However, many existing FedRL algorithms assume that all agents operate in identical environments, which is often unrealistic. In real-world applications, such as multi-robot teams, crowdsourced systems, and

Ran Liu, Fengyu Zhang, Cong Yu, Longjiang Yang

This article presents our results for the eighth Affective Behavior Analysis in-the-wild (ABAW) competition.Multimodal emotion recognition (ER) has important applications in affective computing and human-computer interaction. However, in the real world, compound emotion recognition faces greater issues of uncertainty and modal conflicts. For the Compound Exp

Gideon Stein, Maha Shadaydeh, Jan Blunk, Niklas Penzel

Causal discovery, or identifying causal relationships from observational data, is a notoriously challenging task, with numerous methods proposed to tackle it. Despite this, in-the-wild evaluation of these methods is still lacking, as works frequently rely on synthetic data evaluation and sparse real-world examples under critical theoretical assumptions. Real

John O. Dabiri, Anthony Leonard

Recent work demonstrated that alternative models to the "no-slip" boundary condition for incipient flow perturbations can produce linear instabilities that do not arise in the classical formulation. The present study introduces a Robin-type boundary condition with explicit Reynolds-number dependence, which leads to a more physically realistic transition to i

Nileema Sharma, Matthew Toole, James McKenzie, Fangjun Cheng

Superconducting vortices can reveal electron pairing details and nucleate topologically protected states. Yet, vortices of bulk spin-triplet superconductors have never been visualized. Recently, UTe$_2$ has emerged as a nominative spin-triplet superconductor, but its superconducting order parameter is elusive, and whether time-reversal symmetry is broken rem

Joshua Davies, Kay Schönwald, Matthias Steinhauser

We compute the three-loop form factors for $gg\to HH$ in the limit of vanishing transverse momentum of the Higgs boson which provides a reasonable approximation of the cross section. In our calculations we adopt the large-$N_c$ limit, which already includes non-trivial non-planar Feynman diagrams. We discuss the results for top quark masses in the pole and $

Tanay Kibe, Pratik Roy

The quantum null energy condition (QNEC) is a lower bound on the expectation value of the null-null component of the energy-momentum tensor in terms of null variations of the entanglement entropy. A stronger version of the QNEC (the primary QNEC) is expected to hold in 1+1 dimensional conformal field theories (CFT). QNEC has been shown to impose non-trivial

Andrea Barontini, Mark N. Costantini, Giovanni De Crescenzo, Stefano Forte

We critically assess the robustness of uncertainties on parton distribution functions (PDFs) determined using neural networks from global sets of experimental data collected from multiple experiments. We view the determination of PDFs as an inverse problem, and we study the way the neural network model tackles it when inconsistencies between input datasets a

Ignacio Ruiz Cejudo, Ignacio Trujillo, Giulia Golini, Nafise Sedighi

Ultra-deep optical surveys have reached unprecedented depths, facilitating the study of faint galactic structures. However, the ultraviolet bands, crucial for stellar population studies, remain essentially unexplored at these depths. We present a detailed surface brightness and color analysis of 20 nearby galaxies in the LIGHTS fields observed by GALEX in th

Maximilian Jacobi, Fabio Magistrelli, Eleonora Loffredo, Giacomo Ricigliano

We investigate the nucleosynthesis and kilonova emission based on numerical-relativity binary neutron star merger simulations that incorporate a two-moment neutrino-transport scheme. Unlike in previous works with simpler neutrino treatments, a massive, fast (up to $v=0.3c$), proton-rich neutrino-driven wind develops in the post-merger phase of the simulation

Mitchell Revalski, D. Michael Crenshaw, Garrett E. Polack, Marc Rafelski

We present spatially-resolved measurements of the ionized gas masses and mass outflow rates for six low-redshift ($z \leq$ 0.02) active galaxies. In this study, we expand our sample to galaxies with more complex gas kinematics modeled as outflows along a galactic disk that is ionized by the active galactic nucleus (AGN) bicone. We use Hubble Space Telescope

Hossein Hosseinabadi, Oksana Chelpanova, Jamir Marino

We put forward a user-friendly framework of the truncated Wigner approximation (TWA) for dissipative quantum many-body systems. Our approach is computationally affordable and it features a straightforward implementation. The leverage of the method can be ultimately traced to an intimate connection between the TWA and the semi-classical limit of the quantum L

Shuai Wang, Koichi Hattori, Xu-Guang Huang, Andrey V. Sadofyev

We revisit the collective modes of chiral matter described by the second-order chiral hydrodynamics, noticing that chiral shear waves (CSWs) may become unstable for momenta above a characteristic scale. In the absence of sufficient dissipation, this instability emerges within the hydrodynamic regime, depending on the interplay between shear viscosity and the

Thomas G. Williams, Francesco Belfiore, Martin Bureau, Ashley T. Barnes

Understanding how and why star formation varies between galaxies is fundamental to our comprehension of galaxy evolution. In particular, the star-formation efficiency (SFE; star-formation rate or SFR per unit cold gas mass) has been shown to vary substantially both across and within galaxies. Early-type galaxies (ETGs) constitute an extreme case, as about a

James G. Rogers

The detection of young transiting exoplanets represents a new frontier in our understanding of planet formation and evolution. For the population of observed close-in sub-Neptunes, two proposed formation pathways can reproduce their observed masses and radii at $\sim$Gyr ages: the "gas dwarf" hypothesis and the "water world" hypothesis. We show that a sub-Ne

Yansi Li, Jiahao Xu, Tian Liang, Xingyu Chen

Enhancing the reasoning capabilities of large language models (LLMs), particularly for complex tasks requiring multi-step logical deductions, remains a significant challenge. Traditional inference time scaling methods utilize scalar reward signals from process reward models to evaluate candidate reasoning steps, but these scalar rewards lack the nuanced qual

Sophia Tang, Yinuo Zhang, Alexander Tong, Pranam Chatterjee

Flow matching in the continuous simplex has emerged as a promising strategy for DNA sequence design, but struggles to scale to higher simplex dimensions required for peptide and protein generation. We introduce Gumbel-Softmax Flow and Score Matching, a generative framework on the simplex based on a novel Gumbel-Softmax interpolant with a time-dependent tempe

C. Serafim, S. Calatroni, F. Djurabekova, M. C. Giordano

Vacuum breakdown in accelerator structures is a critical challenge that occurs under high electric fields. In environments subjected to hydrogen ion irradiation or high beam losses, such as in Radio-Frequency Quadrupoles (RFQ), residual hydrocarbons from the vacuum may result in carbon contamination of the metal surfaces from charged particle induced crackin

Jiwen Yu, Yiran Qin, Haoxuan Che, Quande Liu

Modern game development faces significant challenges in creativity and cost due to predetermined content in traditional game engines. Recent breakthroughs in video generation models, capable of synthesizing realistic and interactive virtual environments, present an opportunity to revolutionize game creation. In this position paper, we propose Interactive Gen

Zhuoshi Pan, Yu Li, Honglin Lin, Qizhi Pei

Large language models (LLMs) have demonstrated remarkable reasoning capability in solving mathematical problems. However, existing approaches primarily focus on improving the quality of correct training data, e.g., distilling high-quality correct solutions from advanced models, neglecting the value contained in error data, potentially hindering the model's r

Jerred Chen, Ronald Clark

In many robotics and VR/AR applications, fast camera motions lead to a high level of motion blur, causing existing camera pose estimation methods to fail. In this work, we propose a novel framework that leverages motion blur as a rich cue for motion estimation rather than treating it as an unwanted artifact. Our approach works by predicting a dense motion fl

Ryan Abbott, Daniel C. Hackett, George T. Fleming, Dimitra A. Pefkou

Recent work has shown that the (block) Lanczos algorithm can be used to extract approximate energy spectra and matrix elements from (matrices of) correlation functions in quantum field theory, and identified exact coincidences between Lanczos analysis methods and others. In this work, we note another coincidence: the Lanczos algorithm is equivalent to the we

Brandon Augustino, Dylan Herman, Enrico Fontana, Junhyung Lyle Kim

We study quantum algorithms based on quantum (sub)gradient estimation using noisy function evaluation oracles, and demonstrate the first dimension-independent query complexities (up to poly-logarithmic factors) for zeroth-order convex optimization in both smooth and nonsmooth settings. Interestingly, only using noisy function evaluation oracles, we match the

Haoming Wang, Lek-Heng Lim

We extend the celebrated Glivenko-Cantelli theorem, sometimes called the fundamental theorem of statistics, from its standard setting of total variation distance to all $f$-divergences. A key obstacle in this endeavor is to define $f$-divergence on a subcollection of a $\sigma$-algebra that forms a $\pi$-system but not a $\sigma$-subalgebra. This is a side c

David Rein, Joel Becker, Amy Deng, Seraphina Nix

To understand and predict the societal impacts of highly autonomous AI systems, we need benchmarks with grounding, i.e., metrics that directly connect AI performance to real-world effects we care about. We present HCAST (Human-Calibrated Autonomy Software Tasks), a benchmark of 189 machine learning engineering, cybersecurity, software engineering, and genera

Alex Reneau, Jerry Yao-Chieh Hu, Zhongfang Zhuang, Ting-Chun Liu

In deep learning, processing multidimensional inputs (e.g., images, medical scans, and time series) is an important task that often requires flattening the inputs. We introduce $\mathit{NdLinear}$, a drop-in replacement for linear layers that operates directly on tensors, requiring no flattening. By applying transformations separately along each dimension, N

Yihe Deng, Hritik Bansal, Fan Yin, Nanyun Peng

We introduce OpenVLThinker, one of the first open-source large vision-language models (LVLMs) to exhibit sophisticated chain-of-thought reasoning, achieving notable performance gains on challenging visual reasoning tasks. While text-based reasoning models (e.g., Deepseek R1) show promising results in text-only tasks, distilling their reasoning into LVLMs via

Vineet R. Shenoy, Shaoju Wu, Armand Comas, Tim K. Marks

Remote estimation of vital signs enables health monitoring for situations in which contact-based devices are either not available, too intrusive, or too expensive. In this paper, we present a modular, interpretable pipeline for pulse signal estimation from video of the face that achieves state-of-the-art results on publicly available datasets.Our imaging pho

Qingyu Shi, Jianzong Wu, Jinbin Bai, Jiangning Zhang

The motion transfer task aims to transfer motion from a source video to newly generated videos, requiring the model to decouple motion from appearance. Previous diffusion-based methods primarily rely on separate spatial and temporal attention mechanisms within the 3D U-Net. In contrast, state-of-the-art video Diffusion Transformers (DiT) models use 3D full a

Jianing Qi, Jiawei Liu, Hao Tang, Zhigang Zhu

Vision Language Models (VLMs) excel at identifying and describing objects but often fail at spatial reasoning. We study why VLMs, such as LLaVA, underutilize spatial cues despite having positional encodings and spatially rich vision encoder features. Our analysis reveals a key imbalance: vision token embeddings have much larger norms than text tokens, suppre

Alice Contat, Nicolas Curien

We show that critical parking trees conditioned to be fully parked converge in the scaling limits towards the Brownian growth-fragmentation tree, a self-similar Markov tree different from Aldous' Brownian tree recently introduced and studied by Bertoin, Curien and Riera. As a by-product of our study, we prove that positive non-linear polynomial equations inv

Jichen Hu, Chen Yang, Zanwei Zhou, Jiemin Fang

Reflection removal of a single image remains a highly challenging task due to the complex entanglement between target scenes and unwanted reflections. Despite significant progress, existing methods are hindered by the scarcity of high-quality, diverse data and insufficient restoration priors, resulting in limited generalization across various real-world scen

T. Vrignaud, A. Lecavelier des Etangs, P. A. Strom, F. Kiefer

Beta Pic is a young, A5V star, known for harbouring a large number of exocomets, which frequently transit the star and produce absorption signatures. The physical and chemical properties of these exocomets can be probed by the recently introduced curve of growth approach, which enables column densities measurements in exocomets using observations in numerous

C. Hickman, K. K. H. Leung, A. H. Al-Tawhid, B. W. Filippone

Solders with superconducting properties around $4\,{\rm K}$ are useful in low magnetic field environments for AC current leads or in electrical and mechanical bonds. Accurate knowledge of these properties are needed in high precision experiments. We have measured the electrical resistance of five commercially-available solders: 50\%Sn-50\%Pb, 60\%Sn-40\%Pb,

Isabella Armstrong, Fraser A. Evans, Jo Bovy

Hypervelocity stars (HVSs) ejected from the Galactic Center (GC) at speeds faster than the Galactic escape velocity are useful tools to provide insight into the Milky Way's dark matter halo. However, most characterizations of HVS orbits assume static models of the Milky Way's gravitational potential. In this work, we assess the influence of the Galactic bar

Yi Ching Chou, Long Chen, Hengzhi Wang, Feng Wang

Low Earth Orbit (LEO) satellite networks, characterized by their high data throughput and low latency, have gained significant interest from both industry and academia. Routing data efficiently within these networks is essential for maintaining a high quality of service. However, current routing strategies, such as bent-pipe and inter-satellite link (ISL) ro

A. N. Ormondroyd, W. J. Handley, M. P. Hobson, A. N. Lasenby

We present an updated reconstruction of the dark energy equation of state, $w(a)$, using the newly released DESI DR2 Baryon Acoustic Oscillation (BAO) data in combination with Pantheon+ and DES5Y Type Ia supernovae measurements, respectively. Building on our previous analysis in arXiv:2503.08658, which employed a nonparametric flexknot reconstruction approac

G. B. Lima Neto, H. V. Capelato, F. Durret, R. E. G. Machado

Most galaxies in the Universe are found in groups, which have various morphologies and dynamical states. Studying how groups evolve is an important step for our understanding in both large-scale structure formation and galaxy evolution. We analysed the system composed by two groups at z = 0.037, NGC 5098, a group dominated by a pair of elliptical galaxies, a

Congyi Fan, Jian Guan, Xuanjia Zhao, Dongli Xu

Automatically generating natural, diverse and rhythmic human dance movements driven by music is vital for virtual reality and film industries. However, generating dance that naturally follows music remains a challenge, as existing methods lack proper beat alignment and exhibit unnatural motion dynamics. In this paper, we propose Danceba, a novel framework th