December 2024 arXiv papers — page 120

Ziyu Wang, Ivan C. Christov

We propose an analytical approach to solving nonlocal generalizations of the Euler--Bernoulli beam. Specifically, we consider a version of the governing equation recently derived under the theory of peridynamics. We focus on the clamped--clamped case, employing the natural eigenfunctions of the fourth derivative subject to these boundary conditions. Static s

Amanda Rios, Ibrahima Ndiour, Parual Datta, Jerry Sydir

AI deployed in many real-world use cases should be capable of adapting to novelties encountered after deployment. Here, we consider a challenging, under-explored and realistic continual adaptation problem: a deployed AI agent is continuously provided with unlabeled data that may contain not only unseen samples of known classes but also samples from novel (un

T. Gastine, B. Favier

A better understanding of the ice-ocean couplings is required to better characterise the hydrosphere of the icy moons. Using global numerical simulations in spherical geometry, we have investigated here the interplay between rotating convection and a melting boundary. To do so, we have implemented and validated a phase field formulation in the open-source co

Irene Tamborra

With the advent of time-domain astronomy and the game-changing next generation of telescopes, we have unprecedented opportunities to explore the most energetic events in our Universe through electromagnetic radiation, gravitational waves, and neutrinos. These are elementary particles, which exist in three different flavors and change the latter as they propa

Matej Benko, Iwona Chlebicka, Jørgen Endal, Błażej Miasojedow

We present a significant advancement in the field of Langevin Monte Carlo (LMC) methods by introducing the Inexact Proximal Langevin Algorithm (IPLA). This novel algorithm broadens the scope of problems that LMC can effectively address while maintaining controlled computational costs. IPLA extends LMC's applicability to potentials that are convex, strongly c

Sangjin Lee, Kwonsang Lee

Tests for paired censored outcomes have been extensively studied, with some justified in the context of randomization-based inference. These tests are primarily designed to detect an overall treatment effect across the entire follow-up period, providing limited insight into when the effect manifests and how it changes over time. In this article, we introduce

Bitgoeul Kim, Samuel W. Blair, Talukder Z. Jubery, Soumik Sarkar

Plant breeding programs require assessments of days to maturity for accurate selection and placement of entries in appropriate tests. In the early stages of the breeding pipeline, soybean breeding programs assign relative maturity ratings to experimental varieties that indicate their suitable maturity zones. Traditionally, the estimation of maturity value fo

Tom Holberton

This article evaluates how creative uses of machine learning can address three adjacent terms: ambiguity, uncertainty and indeterminacy. Through the progression of these concepts it reflects on increasing ambitions for machine learning as a creative partner, illustrated with research from Unit 21 at the Bartlett School of Architecture, UCL. Through indetermi

Miguel Sales-Cabrera, Xaro Soler-Escrivà, Víctor Sotomayor

In this paper we obtain the Wedderburn-Artin decomposition of a semisimple group algebra associated to a direct product of finite groups. We also provide formulae for the number of all possible group codes, and their dimensions, that can be constructed in a group algebra. As particular cases, we present the complete algebraic description of the group algebra

Guocheng Qian, Kuan-Chieh Wang, Or Patashnik, Negin Heravi

We introduce Omni-ID, a novel facial representation designed specifically for generative tasks. Omni-ID encodes holistic information about an individual's appearance across diverse expressions and poses within a fixed-size representation. It consolidates information from a varied number of unstructured input images into a structured representation, where eac

Fabien Defrance, Andrew D. Beyer, Jordan Wheeler, Jack Sayers

At sub-Kelvin temperatures, two-level systems (TLS) present in amorphous dielectrics source a permittivity noise, degrading the performance of a wide range of devices using superconductive resonators such as qubits or kinetic inductance detectors. We report here on measurements of TLS noise in hydrogenated amorphous silicon (a-Si:H) films deposited by plasma

Yunna Lv, Long Tang, Dengpan Ye, Caiyun Xie

Deep hash-based retrieval techniques are widely used in facial retrieval systems to improve the efficiency of facial matching. However, it also carries the danger of exposing private information. Deep hash models are easily influenced by adversarial examples, which can be leveraged to protect private images from malicious retrieval. The existing adversarial

Graham Van Goffrier, Debasish Banerjee, Bipasha Chakraborty, Emilie Huffman

Quantum link models (QLMs) are generalizations of Wilson's lattice gauge theory formulated with finite-dimensional link Hilbert spaces. In certain cases, the non-Abelian Gauss Law constraint can be exactly solved, and the gauge invariant subspace embedded onto local spin Hamiltonians for efficient quantum simulation. In $(1+1)d$ previous studies of the $SO(3

Sebastián Rodríguez-Martínez, Giancarlo Troni

Despite their widespread use in determining system attitude, Micro-Electro-Mechanical Systems (MEMS) Attitude and Heading Reference Systems (AHRS) are limited by sensor measurement biases. This paper introduces a method called MAgnetometer and GYroscope Calibration (MAGYC), leveraging three-axis angular rate measurements from an angular rate gyroscope to est

Michael Neilan, Hongzhi Wan

We propose a $C^0$ interior penalty method for the fourth-order stream function formulation of the surface Stokes problem. The scheme utilizes continuous, piecewise polynomial spaces defined on an approximate surface. We show that the resulting discretization is positive definite and derive error estimates in various norms in terms of the polynomial degree o

Luisa Boateng, Matilde Marcolli

A construction that assigns a Boolean 1D TQFT with defects to a finite state automaton was recently developed by Gustafson, Im, Kaldawy, Khovanov, and Lihn. We show that the construction is functorial with respect to the category of finite state automata with transducers as morphisms. Certain classes of subregular languages correspond to additional cohomolog

Wenhao Hu, Paul Henderson, José Cano

Quantization of Deep Neural Network (DNN) activations is a commonly used technique to reduce compute and memory demands during DNN inference, which can be particularly beneficial on resource-constrained devices. To achieve high accuracy, existing methods for quantizing activations rely on complex mathematical computations or perform extensive searches for th

Rupkatha Hira, Dominik Kau, Jessica Sorrell

Active learning aims to reduce the number of labeled data points required by machine learning algorithms by selectively querying labels from initially unlabeled data. Ensuring replicability, where an algorithm produces consistent outcomes across different runs, is essential for the reliability of machine learning models but often increases sample complexity.

Attila Szabó

Based on the relationship between reduced and thermal density matrices in conformal field theory (CFT), we show that the entanglement spectrum of a conformal critical chain with exponentially decaying terms consists of conformal towers of the associated chiral CFT, with only weak finite-size effects. Through free-fermion and interacting examples, we show tha

Alessandro Marcomini, Akihiro Mizutani, Fadri Grünenfelder, Marcos Curty

Current implementations of quantum key distribution (QKD) typically rely on prepare-and-measure (P&M) schemes. Unfortunately, these implementations are not completely secure, unless security proofs fully incorporate all imperfections of real devices. So far, existing proofs have primarily focused on imperfections of either the light source or the measurement

Tom Benhamou, Gabriel Goldberg

We characterize sums of normal ultrafilters after the Magidor iteration (product) of Prikry forcings over a discrete set of measurable cardinals. We apply this to show that the weak Ultrapower Axiom is not equivalent to the Ultrapower Axiom. We also construct a non-rigid ultrapower and two uniform ultrafilters on different cardinals that have the same ultrap

Alberto Guilabert, Pedro Bargueño

The connection between classical thermodynamics and black hole horizons is a fundamental topic in gravitational physics, offering a potential pathway to understanding quantum aspects of gravity. However, while black hole mechanics exhibits well-known thermodynamic parallels, a rigorous geometric interpretation of thermodynamic variables directly from the fie

Sagar Airen, Edward Broadberry, Gustavo Marques-Tavares, Lorenzo Ricci

We assess the sensitivity of future lepton colliders to weakly coupled vector dark portals (aka ``$ Z' $ bosons'') with masses ranging from tens of GeV to a few TeV. Our analysis focuses on dark photons and $ L_{\mu} - L_{\tau} $ gauge bosons. We consider both visible and invisible decay channels. We demonstrate that both high energy $\mu$ colliders and futu

Sean Wu, Shamik Basu, Tim Broedermann, Luc Van Gool

We tackle the ill-posed inverse rendering problem in 3D reconstruction with a Neural Radiance Field (NeRF) approach informed by Physics-Based Rendering (PBR) theory, named PBR-NeRF. Our method addresses a key limitation in most NeRF and 3D Gaussian Splatting approaches: they estimate view-dependent appearance without modeling scene materials and illumination

Julianne Goddard, Isaac Shlosman, Emilio Romano-Diaz

We use high-resolution cosmological zoom-in simulations to model feedback from Seyfert-type supermassive black hole (SMBH) jets onto galaxies with identical dark matter (DM) halos of log(M/M$_\odot$) ~ 11.8. The low mass, ~10$^6$ M$_\odot$, seed SMBHs, have been introduced when the parent DM halos have reached log(M/M$_\odot$) ~ 11, at z ~ 3.7. In a controll

Deepanshu Bisht, Sabyasachi Chakraborty, Atanu Samanta

We present a comprehensive study of axion-like particles (ALPs) through flavor changing neutral current processes, such as $B\to K a$ followed by $a\to\text{hadronic}, \gamma\gamma,\mu^+\mu^-$ channels. Our generic framework encompasses different ultraviolet scenarios similar to KSVZ, DFSZ and flavorful axions etc. Starting from the effective Lagrangian writ

Romain Daviet, Carl Philipp Zelle, Armin Asadollahi, Sebastian Diehl

We discuss the universal behavior linked to the Goldstone mode associated with the spontaneous breaking of time-translation symmetry in many-body systems, in which the order parameter traces out a limit cycle. We show that this universal behavior is closely tied to Kardar-Parisi-Zhang physics, which can strongly affect the scaling properties in all dimension

M. J. Mróz, R. Poleski, A. Udalski, T. Sumi

We present a comprehensive analysis of a planetary microlensing event OGLE-2015-BLG-1609. The planetary anomaly was detected by two survey telescopes, OGLE and MOA. Each of these surveys collected enough data over the planetary anomaly to allow for an unambiguous planet detection. Such survey detections of planetary anomalies are needed to build a robust sam

Federico Pozzoli, Jonathan Gair, Riccardo Buscicchio, Lorenzo Speri

Separating a stochastic gravitational wave background (SGWB) from noise is a challenging statistical task. One approach to establishing a detection criterion for the SGWB is using Bayesian evidence. If the evidence ratio (Bayes factor) between models with and without the signal exceeds a certain threshold, the signal is considered detected. We present a form

Hooman Davoudiasl, Marvin Schnubel

We propose that the Higgs potential, a key element in our understanding of Nature, is partially generated by the instantons of new confining dynamics, perhaps from a hidden sector. In this picture, while the Higgs itself is a fundamental field, it controls the strength of the non-perturbative interactions that give rise to its potential. We examine a simple

Danielle Dickinson, Dan Milisavljevic, Braden Garretson, Luc Dessart

We present high resolution WIYN/NEID echelle spectroscopy (R $\approx70$,000) of the supernova (SN) 2023ixf in M101, obtained 1.51 to 18.51 days after explosion over nine epochs. Daily monitoring for the first four days after explosion shows narrow emission features ($\leq200$ km s$^{-1}$), exhibiting predominantly blueshifted velocities, that rapidly weaken

Riccardo Bartocci, Anke Biekötter, Tobias Hurth

Global analyses in the Standard Model Effective Field Theory (SMEFT) framework serve as a tool to probe potential directions of new physics. To break degeneracies between the Wilson coefficients of the SMEFT, it is essential to combine observables from various experiments. Since different observables entering global fits may be measured at different energy s

Rathindra N. Das, Saskia Demulder, Johanna Erdmenger, Christian Northe

Complexity is a fundamental characteristic of states within a quantum system. Its use is however mostly limited to bosonic systems, inhibiting its present applicability to supersymmetric theories. This is also relevant to its application to the AdS/CFT correspondence. To address this limitation, we extend the framework of spread complexity beyond bosonic sys

Jakub Czartowski, Karol Życzkowski

Designs, structures connected to averaging with respect to a given measure using finite sets of points, have proven themselves as invaluable tools across the field of quantum information, finding their uses in state and process tomography, key distribution and others. In this work, we introduce a new concept of pushforward designs, which allows us to obtain

Olaf Massen, Govert Nijs, Mike Sas, Wilke van der Schee

Thermal electromagnetic radiation is emitted by the quark-gluon plasma (QGP) throughout its space-time evolution, with production rates that depend characteristically on the temperature. We study this temperature using thermal photons and dileptons using the Trajectum heavy ion code, which is constrained by Bayesian analysis. In addition we present the ellip

Wenzhao Zheng, Zetian Xia, Yuanhui Huang, Sicheng Zuo

End-to-end autonomous driving has received increasing attention due to its potential to learn from large amounts of data. However, most existing methods are still open-loop and suffer from weak scalability, lack of high-order interactions, and inefficient decision-making. In this paper, we explore a closed-loop framework for autonomous driving and propose a

Haonan Qiu, Shiwei Zhang, Yujie Wei, Ruihang Chu

Visual diffusion models achieve remarkable progress, yet they are typically trained at limited resolutions due to the lack of high-resolution data and constrained computation resources, hampering their ability to generate high-fidelity images or videos at higher resolutions. Recent efforts have explored tuning-free strategies to exhibit the untapped potentia

Yue Feng, Vaibhav Sanjay, Spencer Lutz, Badour AlBahar

Automatically generating multiview illusions is a compelling challenge, where a single piece of visual content offers distinct interpretations from different viewing perspectives. Traditional methods, such as shadow art and wire art, create interesting 3D illusions but are limited to simple visual outputs (i.e., figure-ground or line drawing), restricting th

Taiming Lu, Tianmin Shu, Junfei Xiao, Luoxin Ye

Understanding, navigating, and exploring the 3D physical real world has long been a central challenge in the development of artificial intelligence. In this work, we take a step toward this goal by introducing GenEx, a system capable of planning complex embodied world exploration, guided by its generative imagination that forms priors (expectations) about th

Weiqi Li, Shijie Zhao, Chong Mou, Xuhan Sheng

As virtual reality gains popularity, the demand for controllable creation of immersive and dynamic omnidirectional videos (ODVs) is increasing. While previous text-to-ODV generation methods achieve impressive results, they struggle with content inaccuracies and inconsistencies due to reliance solely on textual inputs. Although recent motion control technique

Enis Simsar, Thomas Hofmann, Federico Tombari, Pinar Yanardag

Recent advances in text-to-image customization have enabled high-fidelity, context-rich generation of personalized images, allowing specific concepts to appear in a variety of scenarios. However, current methods struggle with combining multiple personalized models, often leading to attribute entanglement or requiring separate training to preserve concept dis

Linyi Jin, Richard Tucker, Zhengqi Li, David Fouhey

Learning to understand dynamic 3D scenes from imagery is crucial for applications ranging from robotics to scene reconstruction. Yet, unlike other problems where large-scale supervised training has enabled rapid progress, directly supervising methods for recovering 3D motion remains challenging due to the fundamental difficulty of obtaining ground truth anno

Yunzhong Hou, Liang Zheng, Philip Torr

This study seeks to automate camera movement control for filming existing subjects into attractive videos, contrasting with the creation of non-existent content by directly generating the pixels. We select drone videos as our test case due to their rich and challenging motion patterns, distinctive viewing angles, and precise controls. Existing AI videography

Dongting Hu, Jierun Chen, Xijie Huang, Huseyin Coskun

Existing text-to-image (T2I) diffusion models face several limitations, including large model sizes, slow runtime, and low-quality generation on mobile devices. This paper aims to address all of these challenges by developing an extremely small and fast T2I model that generates high-resolution and high-quality images on mobile platforms. We propose several t

Zhuofan Zong, Dongzhi Jiang, Bingqi Ma, Guanglu Song

Significant achievements in personalization of diffusion models have been witnessed. Conventional tuning-free methods mostly encode multiple reference images by averaging their image embeddings as the injection condition, but such an image-independent operation cannot perform interaction among images to capture consistent visual elements within multiple refe

Hung-Jui Huang, Michael Kaess, Wenzhen Yuan

Tactile sensing is crucial for robots aiming to achieve human-level dexterity. Among tactile-dependent skills, tactile-based object tracking serves as the cornerstone for many tasks, including manipulation, in-hand manipulation, and 3D reconstruction. In this work, we introduce NormalFlow, a fast, robust, and real-time tactile-based 6DoF tracking algorithm.

Junqi Ge, Ziyi Chen, Jintao Lin, Jinguo Zhu

Vision-Language Models (VLMs) have shown promising capabilities in handling various multimodal tasks, yet they struggle in long-context scenarios, particularly in tasks involving videos, high-resolution images, or lengthy image-text documents. In our work, we first conduct an empirical analysis of the long-context capabilities of VLMs using our augmented lon

Metehan Calis, Narasimha Boddeti, J. Scott Bunch

We measure the out-of-plane shear modulus of few-layer graphene (FLG) by a blister test. During the test, we employed a monolayer molybdenum disulfide (MoS2) membrane stacked onto FLG wells to facilitate the separation of FLG from the silicon oxide (SiOx) substrate. Using the deflection profile of the blister, we determine an average shear modulus G of 0.97

Kavana Venkatesh, Yusuf Dalva, Ismini Lourentzou, Pinar Yanardag

Despite impressive visual fidelity, current text-to-image (T2I) diffusion models struggle to depict rare, complex, or culturally nuanced concepts due to training data limitations. We introduce RAVEL, a training-free framework that significantly improves rare concept generation, context-driven image editing, and self-correction by integrating graph-based retr

Chenyu Yang, Xuan Dong, Xizhou Zhu, Weijie Su

Large Vision-Language Models (VLMs) have been extended to understand both images and videos. Visual token compression is leveraged to reduce the considerable token length of visual inputs. To meet the needs of different tasks, existing high-performance models usually process images and videos separately with different token compression strategies, limiting t

Yuanze Lin, Yunsheng Li, Dongdong Chen, Weijian Xu

We introduce Olympus, a new approach that transforms Multimodal Large Language Models (MLLMs) into a unified framework capable of handling a wide array of computer vision tasks. Utilizing a controller MLLM, Olympus delegates over 20 specialized tasks across images, videos, and 3D objects to dedicated modules. This instruction-based routing enables complex wo

Yusuf Dalva, Kavana Venkatesh, Pinar Yanardag

Rectified flow models have emerged as a dominant approach in image generation, showcasing impressive capabilities in high-quality image synthesis. However, despite their effectiveness in visual generation, rectified flow models often struggle with disentangled editing of images. This limitation prevents the ability to perform precise, attribute-specific modi

Mehmet Asim Gumus, Damien Leflot, Piotr Tourkine, Alexander Zhiboedov

In this work, we develop machine learning techniques to study nonperturbative scattering amplitudes. We focus on the two-to-two scattering amplitude of identical scalar particles, setting the double discontinuity to zero as a simplifying assumption. Neural networks provide an efficient parameterization for scattering amplitudes, offering a flexible toolkit t

Gabriel Malave, Rodrigo Soto-Garrido, Vladimir Juricic, Bitan Roy

We formulate a field-theoretic description for $d$-dimensional interacting nodal semimetals, featuring dispersion that scales with the linear and $n$th power of momentum along $d_L$ and $d_M$ mutually orthogonal directions around a few isolated points in the reciprocal space, respectively, with $d_L+d_M=d$, and residing at the brink of isotropic insulation,

Zhen Xu, Yinghao Xu, Zhiyuan Yu, Sida Peng

This paper aims to address the challenge of reconstructing long volumetric videos from multi-view RGB videos. Recent dynamic view synthesis methods leverage powerful 4D representations, like feature grids or point cloud sequences, to achieve high-quality rendering results. However, they are typically limited to short (1~2s) video clips and often suffer from

Carlos Esteves, Mohammed Suhail, Ameesh Makadia

Image tokenizers map images to sequences of discrete tokens, and are a crucial component of autoregressive transformer-based image generation. The tokens are typically associated with spatial locations in the input image, arranged in raster scan order, which is not ideal for autoregressive modeling. In this paper, we propose to tokenize the image spectrum in

Yue Chen, Xingyu Chen, Anpei Chen, Gerard Pons-Moll

Given that visual foundation models (VFMs) are trained on extensive datasets but often limited to 2D images, a natural question arises: how well do they understand the 3D world? With the differences in architecture and training protocols (i.e., objectives, proxy tasks), a unified framework to fairly and comprehensively probe their 3D awareness is urgently ne

Yiheng Xu, Dunjie Lu, Zhennan Shen, Junli Wang

Graphical User Interface (GUI) agents can automate complex tasks across digital environments, but their development is hindered by the scarcity of high-quality trajectory data for training. Existing approaches rely on expensive human annotation, making them unsustainable at scale. We propose AgentTrek, a scalable data synthesis pipeline that generates web ag

Hao Li, Changyao Tian, Jie Shao, Xizhou Zhu

The remarkable success of Large Language Models (LLMs) has extended to the multimodal domain, achieving outstanding performance in image understanding and generation. Recent efforts to develop unified Multimodal Large Language Models (MLLMs) that integrate these capabilities have shown promising results. However, existing approaches often involve complex des

Jiaying Lin, Shuquan Ye, Dan Xu, Wanli Ouyang

While Multimodal Large Language Models (MLLMs) excel at many vision tasks, it is unknown if they exhibit human-like perceptual behaviors. To evaluate this, we introduce HVSBench, the first large-scale benchmark with over 85,000 samples designed to test MLLM alignment with the human visual system (HVS). The benchmark covers 13 categories across 5 key fields:

Julian Zimmerlin, Jens Beißwenger, Bernhard Jaeger, Andreas Geiger

End-to-end driving systems have made rapid progress, but have so far not been applied to the challenging new CARLA Leaderboard 2.0. Further, while there is a large body of literature on end-to-end architectures and training strategies, the impact of the training dataset is often overlooked. In this work, we make a first attempt at end-to-end driving for Lead

Xizi Wang, Feng Cheng, Ziyang Wang, Huiyu Wang

Video temporal grounding aims to localize relevant temporal boundaries in a video given a textual prompt. Recent work has focused on enabling Video LLMs to perform video temporal grounding via next-token prediction of temporal timestamps. However, accurately localizing timestamps in videos remains challenging for Video LLMs when relying solely on temporal to

Yuanhui Huang, Wenzhao Zheng, Yuan Gao, Xin Tao

Video generation models (VGMs) have received extensive attention recently and serve as promising candidates for general-purpose large vision models. While they can only generate short videos each time, existing methods achieve long video generation by iteratively calling the VGMs, using the last-frame output as the condition for the next-round generation. Ho

Ayaka Higami, Karin Oshima, Tomoyo Isoguchi Shiramatsu, Hirokazu Takahashi

Analyzing rat behavior lies at the heart of many scientific studies. Past methods for automated rodent modeling have focused on 3D pose estimation from keypoints, e.g., face and appendages. The pose, however, does not capture the rich body surface movement encoding the subtle rat behaviors like curling and stretching. The body surface lacks features that can

Tibor Rakovszky, Benedikt Placke, Nikolas P. Breuckmann, Vedika Khemani

We prove an analogue of the "bottleneck theorem", well-known for classical Markov chains, for Markovian quantum channels. In particular, we show that if two regions (subspaces) of Hilbert space are separated by a region that has very low weight in the channel's steady state, then states initialized on one side of this barrier will take a long time to relax,

Yabo Chen, Chen Yang, Jiemin Fang, Xiaopeng Zhang

Single-image 3D reconstruction remains a fundamental challenge in computer vision due to inherent geometric ambiguities and limited viewpoint information. Recent advances in Latent Video Diffusion Models (LVDMs) offer promising 3D priors learned from large-scale video data. However, leveraging these priors effectively faces three key challenges: (1) degradat

Pan Zhang, Xiaoyi Dong, Yuhang Cao, Yuhang Zang

Creating AI systems that can interact with environments over long periods, similar to human cognition, has been a longstanding research goal. Recent advancements in multimodal large language models (MLLMs) have made significant strides in open-world understanding. However, the challenge of continuous and simultaneous streaming perception, memory, and reasoni

David Alonso-González, David Cerdeño, Marina Cermeño, Andres D. Perez

In this article, we claim that axion-like particles (ALPs) with MeV masses can be produced with semi-relativistic velocities in core-collapse supernovae (SNe), generating a diffuse galactic flux. We show that these ALPs can be detected in neutrino water Cherenkov detectors via $a \, p \rightarrow p \, \gamma$ interactions. Using Super-Kamiokande data, we der

Jingruo Sun, Wenzhi Gao, Ellen Vitercik, Yinyu Ye

Online linear programming (OLP) has found broad applications in revenue management and resource allocation. State-of-the-art OLP algorithms achieve low regret by repeatedly solving linear programming (LP) subproblems that incorporate updated resource information. However, LP-based methods are computationally expensive and often inefficient for large-scale ap

Zexin He, Tengfei Wang, Xin Huang, Xingang Pan

Recovering the geometry and materials of objects from a single image is challenging due to its under-constrained nature. In this paper, we present Neural LightRig, a novel framework that boosts intrinsic estimation by leveraging auxiliary multi-lighting conditions from 2D diffusion priors. Specifically, 1) we first leverage illumination priors from large-sca

M. L. Hamadouche, R. J. McLure, A. Carnall, D. J. McLeod

We present the results of a study investigating the galaxy stellar-mass function (GSMF), size-mass relations and morphological properties of star-forming and quiescent galaxies over the redshift range $0.25<z<2.25$, using the JWST PRIMER survey. The depth of the PRIMER near-IR imaging allows us to confirm the double Schechter function shape of the quiescent

Stephen W. Yan, Diego Barberena, Matthew P. A. Fisher, Sagar Vijay

We investigate a quantum dynamical phase transition induced by the competition between local unitary evolution and dissipation in a qubit chain with a strong, on-site $\mathbb{Z}_2$ symmetry. While the steady-state of this evolution is always maximally-mixed, we show that the dynamical behavior of certain non-local observables on the approach to this steady-

Abhishek Chaudhary, Ujjwal Koley, Emil Wiedemann

We introduce the concept of a dissipative measure-valued solution to the Euler alignment system. This approach incorporates a modified total energy balance, utilizing a binary tensor Young measure. The central finding is a weak (measure-valued)--strong uniqueness principle: if both a dissipative measure-valued solution and a classical smooth solution origina

Tuomas Lappi, Heikki Mäntysaari, Hannu Paukkunen, Mirja Tevio

We formulate and numerically solve the Dokshitzer-Gribov-Lipatov-Altarelli-Parisi~(DGLAP) evolution equations at next-to-leading order in perturbation theory directly for a basis of 6 physical, observable structure functions in deeply inelastic scattering. By expressing the evolution in the physical basis one evades the factorization scale and scheme depende

Yago Bea, Mauro Giliberti, David Mateos, Mikel Sanchez-Garitaonandia

A line of first-order phase transitions is conjectured in the phase diagram of Quantum Chromodynamics at non-zero baryon density. If this is the case, numerical simulations of neutron star mergers suggest that various regions of the stars may cross this line multiple times. This results in the nucleation of bubbles of the preferred phase, which subsequently

Chester Palen-Michel, Maxwell Pickering, Maya Kruse, Jonne Sälevä

We present OpenNER 1.0, a standardized collection of openly-available named entity recognition (NER) datasets. OpenNER contains 36 NER corpora that span 52 languages, human-annotated in varying named entity ontologies. We correct annotation format issues, standardize the original datasets into a uniform representation with consistent entity type names across

Fiona Ryan, Ajay Bati, Sangmin Lee, Daniel Bolya

We address the problem of gaze target estimation, which aims to predict where a person is looking in a scene. Predicting a person's gaze target requires reasoning both about the person's appearance and the contents of the scene. Prior works have developed increasingly complex, hand-crafted pipelines for gaze target estimation that carefully fuse features fro

Mytraya Gattu, J. K. Jain

We demonstrate that formulating the composite-fermion theory of the fractional quantum Hall (FQH) effect in terms of quaternions greatly expands its reach and opens the door into many interesting issues that were previously beyond the reach of quantitative theoretical investigation. As an illustration, we investigate the possibility of a nematic or a charge-

Jitesh Jain, Zhengyuan Yang, Humphrey Shi, Jianfeng Gao

In recent times, the standard practice for developing MLLMs is to feed features from vision encoder(s) into the LLM and train with natural language supervision. This approach often causes models to lean towards language comprehension and undermine the rich visual perception signals present in the data, which are critical for tasks involving spatial reasoning

Keyi Shen, Jiangwei Yu, Jose Barreiros, Huan Zhang

Neural-network-based dynamics models learned from observational data have shown strong predictive capabilities for scene dynamics in robotic manipulation tasks. However, their inherent non-linearity presents significant challenges for effective planning. Current planning methods, often dependent on extensive sampling or local gradient descent, struggle with

David Jobst

Nowadays, weather forecasts are commonly generated by ensemble forecasts based on multiple runs of numerical weather prediction models. However, such forecasts are usually miscalibrated and/or biased, thus require statistical postprocessing. Non-homogeneous regression models, such as the ensemble model output statistics are frequently applied to correct thes

Arsha Nagrani, Mingda Zhang, Ramin Mehran, Rachel Hornung

We introduce Neptune, a benchmark for long video understanding that requires reasoning over long time horizons and across different modalities. Many existing video datasets and models are focused on short clips (10s-30s). While some long video datasets do exist, they can often be solved by powerful image models applied per frame (and often to very few frames

Mohammad Taha Askari, Lutz Lampe

Optimizing the input probability distribution of a discrete-time channel is a standard step in the information-theoretic analysis of digital communication systems. Nevertheless, many practical communication systems transmit uniformly and independently distributed symbols drawn from regular constellation sets. The introduction of the probabilistic amplitude s

Daniel Drayss, Dengyang Fang, Alban Sherifaj, Huanfa Peng

The conventional way of generating optical waveforms relies on the in-phase and quadrature (IQ) modulation of a continuous wave (CW) laser tone. In this case, the bandwidth of the resulting optical waveform is limited by the underlying electronic components, in particular by the digital-to-analog converters (DACs) generating the drive signals for the IQ modu

Saptarshi Mandal, Xiaojun Lin, R. Srikant

Knowledge distillation, where a small student model learns from a pre-trained large teacher model, has achieved substantial empirical success since the seminal work of \citep{hinton2015distilling}. Despite prior theoretical studies exploring the benefits of knowledge distillation, an important question remains unanswered: why does soft-label training from th

Ovidiu Cristinel Stoica

The physical meaning of the operators is not reducible to the intrinsic relations of the quantum system, since unitary transformations can find other operators satisfying the exact same relations. The physical meaning is determined empirically. I propose that the assignment of physical meaning to operators spreads through observation, along with the values o

Messi H. J. Lee, Soyeon Jeon

Vision-Language Models (VLMs) extend Large Language Models' capabilities by integrating image processing, but concerns persist about their potential to reproduce and amplify human biases. While research has documented how these models perpetuate stereotypes across demographic groups, most work has focused on between-group biases rather than within-group diff

Caleb Stam, Emily Saldanha, Mahantesh Halappanavar, Anurag Acharya

The emergence of the COVID-19 pandemic resulted in a significant rise in the spread of misinformation on online platforms such as Twitter. Oftentimes this growth is blamed on the idea of the "echo chamber." However, the behavior said to characterize these echo chambers exists in two dimensions. The first is in a user's social interactions, where they are sai

Ilyoun Na, Jack Kemp, Sinéad M. Griffin, Yang Peng

We present a systematic framework for Floquet prethermalization under strong resonant driving, emphasizing the pivotal role of dynamical space-time symmetries. Our approach demonstrates how dynamical space-time symmetries map onto the projective static symmetry group of the prethermal Hamiltonian governing the prethermal regime. We introduce techniques for d

Irakli Giorgadze, Haixuan Huang, Jordan Gaines, Elio J. König

Fermionic Hamiltonians play a critical role in quantum chemistry, one of the most promising use cases for near-term quantum computers. However, since encoding nonlocal fermionic statistics using conventional qubits results in significant computational overhead, fermionic quantum hardware, such as fermion atom arrays, were proposed as a more efficient platfor

Sara Rufrano Aliberti, Gaetano Lambiase, Tanmay Kumar Poddar

Dark matter (DM) within the solar system induces deviations in the geodetic drift of gyroscope spin due to its gravitational interaction. Assuming a constant DM density as a minimal scenario, we constrain DM overdensity within the Gravity Probe B (GP-B) orbit and project limits for Earth's and Neptune's orbits around the Sun. The presence of electrons in gra

Róbert Németh, Vatsal K. Bandaru, Pedro Alves, Emma Brann

Conveyor-mode shuttling is a key approach for implementing intermediate-range coupling between electron-spin qubits in quantum dots. Initial implementations are encouraging; however, long shuttling trajectories are guaranteed to encounter regions of low conduction-band valley energy splittings, due to the presence of random-alloy disorder in Si/SiGe quantum

Jiale Xu, Shenghua Gao, Ying Shan

Sparse-view reconstruction models typically require precise camera poses, yet obtaining these parameters from sparse-view images remains challenging. We introduce FreeSplatter, a scalable feed-forward framework that generates high-quality 3D Gaussians from uncalibrated sparse-view images while estimating camera parameters within seconds. Our approach employs

Yu Feng, Phu Mon Htut, Zheng Qi, Wei Xiao

Quantifying uncertainty in black-box LLMs is vital for reliable responses and scalable oversight. Existing methods, which gauge a model's uncertainty through evaluating self-consistency in responses to the target query, can be misleading: an LLM may confidently provide an incorrect answer to a target query, yet give a confident and accurate answer to that sa

Yury Malyshkin

We study the asymptotic behavior of the maximal in-degree in the spatial preferential attachment model with a choice-based edge step. We prove different types of behavior of maximal in-degree based on the model's parameters.

Zhaohuan Zhu, Shangjia Zhang, Ted Johnson

Protoplanetary disks can exhibit asymmetric temperature variations due to phenomena such as shadows cast by the inner disk or localized heating by young planets. We investigate the disk features induced by these asymmetric temperature variations. We find that spirals are initially excited, then break into two and reconnect to form rings. By carrying out line

Ariel Gera, Odellia Boni, Yotam Perlitz, Roy Bar-Haim

Given the rapid progress of generative AI, there is a pressing need to systematically compare and choose between the numerous models and configurations available. The scale and versatility of such evaluations make the use of LLM-based judges a compelling solution for this challenge. Crucially, this approach requires first to validate the quality of the LLM j

Robin Y. Wen, Lukas T. Hergt, Niayesh Afshordi, Douglas Scott

One regime where we might see departures from general relativity is at the largest accessible scales, with a natural choice in cosmology being the cosmological horizon (or Hubble) scale. We investigate a single-parameter extension to the standard cosmological model with a different strength of gravity above and below this scale -- a "cosmic glitch" in gravit

Tom Davot, Jessica Enright, Jayakrishnan Madathil, Kitty Meeks

A graph G is c-closed if every two vertices with at least c common neighbors are adjacent to each other. Introduced by Fox, Roughgarden, Seshadhri, Wei and Wein [ICALP 2018, SICOMP 2020], this definition is an abstraction of the triadic closure property exhibited by many real-world social networks, namely, friends of friends tend to be friends themselves. So

Olivia Mostow, Paul Torrey, Jonah C. Rose, Alex M. Garcia

We present a novel method for systematically assessing the impact of central potential fluctuations associated with bursty outflows on the structure of dark matter halos for classical and ultra-faint dwarf galaxies. Specifically, we use dark-matter-only simulations augmented with a manually-added massive particle that modifies the central potential and appro