February 2025 arXiv papers — page 11

Jonathan Tonglet, Tinne Tuytelaars, Marie-Francine Moens, Iryna Gurevych

Visualizations play a pivotal role in daily communication in an increasingly data-driven world. Research on multimodal large language models (MLLMs) for automated chart understanding has accelerated massively, with steady improvements on standard benchmarks. However, for MLLMs to be reliable, they must be robust to misleading visualizations, i.e., charts tha

Lance Ying, Katherine M. Collins, Lionel Wong, Ilia Sucholutsky

Recent advances in Artificial Intelligence (AI) have yielded powerful computational models that, by learning from vast amounts of human-generated data, are increasingly posited as approximate models of human cognition. However, we argue that many current evaluation paradigms for AI are insufficient for assessing human-like cognitive capabilities in these mod

Johanna Teske, Natasha E. Batalha, Nicole L. Wallack, James Kirk

The atmospheres of planets between the size of Earth and Neptune at short orbital periods have been under intense scrutiny. Of the ~dozen planets in this regime with atmospheres studied so far, a few appear to have prominent molecular features while others appear relatively void of detectable atmospheres. Further work is therefore needed to understand the at

Beomyeol Yu, Taeyoung Lee

Improving sampling efficiency and generalization capability is critical for the successful data-driven control of quadrotor unmanned aerial vehicles (UAVs) that are inherently unstable. While various reinforcement learning (RL) approaches have been applied to autonomous quadrotor flight, they often require extensive training data, posing multiple challenges

Felipe del Rio, Alain Raymond-Saez, Daniel Florea, Rodrigo Toro Icarte

Deep neural networks (DNNs) struggle at systematic generalization (SG). Several studies have evaluated the possibility to promote SG through the proposal of novel architectures, loss functions or training methodologies. Few studies, however, have focused on the role of training data properties in promoting SG. In this work, we investigate the impact of certa

J. Kluson

In this short note we find covariant canonical formulation of modified Eddington gravity action coupled to scalar field. We also discuss limitation of this formulation and suggest its possible generalization.

Harrison Grodin, Runming Li, Robert Harper

Software development depends on the use of libraries whose public specifications inform client code and impose obligations on private implementations; it follows that verification at scale must also be modular, preserving such abstraction. Hoare's influential methodology uses abstraction functions to demonstrate the coherence between such concrete implementa

Yisheng Tu, Zhi-Yun Li, Zhaohuan Zhu, Chun-Yen Hsu

Jets and outflows are commonly observed in young stellar objects (YSOs), yet their origins remain debated. Using 3D non-ideal magnetohydrodynamic (MHD) simulations of a circumstellar disk threaded by a large-scale open poloidal magnetic field, we identify three components in the disk-driven outflow: (1) a fast, collimated jet, (2) a less collimated, slower l

John Moffat

An approach to cosmological modelling is presented that incorporates the inhomogeneous structure of the Cosmic Web, specifically focusing on the interplay between cosmic voids and density walls. We extend the standard homogeneous and isotropic cosmological model to account for the observed large-scale structure of the universe. By modifying the Friedmann equ

Vijay Srinivas Tida, Md Imran Hossen, Liqun Shan, Sai Venkatesh Chilukoti

The optimization of the transpose convolution layer for deep learning applications is achieved with the kernel segregation mechanism. However, kernel segregation has disadvantages, such as computing extra elements to obtain the output feature map with odd dimensions while launching a thread. To mitigate this problem, we introduce a unified kernel segregation

Aybars Yunusoglu, Dexter Le, Murat Isik, I. Can Dikmen

In the rapidly advancing field of neuromorphic computing, integrating biologically-inspired models like the Leaky Integrate-and-Fire Astrocyte (LIFA) into spiking neural networks (SNNs) enhances system robustness and performance. This paper introduces the LIFA model in SNNs, addressing energy efficiency, memory management, routing mechanisms, and fault toler

Jackie Chan, Fred Choi, Koustuv Saha, Eshwar Chandrasekharan

Platforms are increasingly relying on algorithms to curate the content within users' social media feeds. However, the growing prominence of proprietary, algorithmically curated feeds has concealed what factors influence the presentation of content on social media feeds and how that presentation affects user behavior. This lack of transparency can be detrimen

MohammadHossein Rezaei, Yicheng Fu, Phil Cuvin, Caleb Ziems

Human activity is moderated by norms; however, supervision for normative reasoning is sparse, particularly where norms are physically- or socially-grounded. We thus present EGONORMIA $\|\epsilon\|$, comprising 1,853 (200 for EGONORMIA-verified) multiple choice questions (MCQs) grounded within egocentric videos of human interactions, enabling the evaluation a

Linying Lv

This paper documents novel investment value in analyst report text. Using 1.2 million reports from 2000-2023, I embed narratives with large language models (LLMs) and fit machine learning (ML) forecasts of future long-term returns. Portfolios formed on the report narrative forecasts earn sizable and significant performance that is incremental to analysts' nu

Amiya Mishra

We consider the Chern-Simons theory coupled to massive fundamental matter in three spacetime dimensions, focusing on the Higgsed phase of the bosonic matter in the large N limit. To study the characteristics of the Z-boson in the Higgsed phase and address an apparent puzzle regarding its fermionic analogue, we employ the 3d Bose-Fermi duality to derive the e

Heshan Aravinda

Jakimiuk et al. (2024) have proved that, if $X$ is an ultra log-concave random variable with integral mean, then $$\max_n \mathbb{P}\{X=n\} \geq \max_n \mathbb{P} \{Z=n\}\,,$$ where $Z$ is a Poisson random variable with the parameter $\mathbb{E}[X]$. In this note, we show that this inequality does not always hold true when $X$ is ultra log-concave with $\mat

Jonathan Chan, Stephanie Weirich

In dependent type theory, being able to refer to a type universe as a term itself increases its expressive power, but requires mechanisms in place to prevent Girard's paradox from introducing logical inconsistency in the presence of type-in-type. The simplest mechanism is a hierarchy of universes indexed by a sequence of levels, typically the naturals. To im

Jianqiao Long, Lei Zhang, Miaowen Wen, Kezhi Wang

In current molecular communication (MC) systems, performing computational operations at the nanoscale remains challenging, restricting their applicability in complex scenarios such as adaptive biochemical control and advanced nanoscale sensing. To overcome this challenge, this paper proposes a novel framework that seamlessly integrates computation into the m

Miguel A. Porras, Spencer W. Jolly

Hermite-Gaussian (HG) beams are standard modes delivered by continuous or pulsed lasers systems, and pulse-front tilt is one of the most common, detrimental or beneficial, spatiotemporal couplings affecting ultrashort pulses. Combining them, we show that focusing a pulsed HG beam with a tilt generates an elliptical spatiotemporal optical vortex (STOV), or a

Yongchao Huang

A reward-guided, gradient-free ParVI method, \textit{R-ParVI}, is proposed for sampling partially known densities (e.g. up to a constant). R-ParVI formulates the sampling problem as particle flow driven by rewards: particles are drawn from a prior distribution, navigate through parameter space with movements determined by a reward mechanism blending assessme

Apratim Majumder, John A. Doughty, Tina H. Hayward Henry I. Smith, Rajesh Menon

Rigorous simulations challenge recent claims that metalenses outperform conventional diffractive lenses, such as Fresnel Zone Plates (FZPs), in focusing efficiency at high numerical apertures (NAs). Across various lens diameters, FZPs exhibit a pronounced asymmetry in the shadow effect, leading to significantly higher focusing efficiency when optimally orien

Chaoyu Li, Sid Padmanabhuni, Maryam Cheema, Hasti Seifi

Video descriptions are crucial for blind and low vision (BLV) users to access visual content. However, current artificial intelligence models for generating descriptions often fall short due to limitations in the quality of human annotations within training datasets, resulting in descriptions that do not fully meet BLV users' needs. To address this gap, we i

Daniel Moreno-Cartagena, Pavlos Protopapas, Guillermo Cabrera-Vives, Martina Cádiz-Leyton

This study investigates the potential of a pre-trained vision Transformer (VT) model, specifically the Swin Transformer V2 (SwinV2), to classify photometric light curves without the need for feature extraction or multi-band preprocessing. The goal is to assess whether this image-based approach can accurately differentiate astronomical phenomena and serve as

Jun Hou, Lucy Lu Wang

Explainable AI (XAI) techniques are necessary to help clinicians make sense of AI predictions and integrate predictions into their decision-making workflow. In this work, we conduct a survey study to understand clinician preference among different XAI techniques when they are used to interpret model predictions over text-based EHR data. We implement four XAI

Gabriel Fernández-Blanco, Pedro García-Cereijo, David Lema-Núñez, Diego Ramil-López

In the last years, especially since the COVID-19 pandemic, precision medicine platforms emerged as useful tools for supporting new tests like the ones that detect the presence of antibodies and antigens with better sensitivity and specificity than traditional methods. In addition, the pandemic has also influenced the way people interact (decentralization), b

Elizabeth Anne Watkins, Emanuel Moss, Ramesh Manuvinakurike, Meng Shi

In this short paper we address issues related to building multimodal AI systems for human performance support in manufacturing domains. We make two contributions: we first identify challenges of participatory design and training of such systems, and secondly, to address such challenges, we propose the ACE paradigm: "Action and Control via Explanations". Spec

Yankai Li, Mo Chen

Model Predictive Path Integral (MPPI) control, Reinforcement Learning (RL), and Diffusion Models have each demonstrated strong performance in trajectory optimization, decision-making, and motion planning. However, these approaches have traditionally been treated as distinct methodologies with separate optimization frameworks. In this work, we establish a uni

Tianyi Lorena Yan, Robin Jia

To answer one-to-many factual queries (e.g., listing cities of a country), a language model (LM) must simultaneously recall knowledge and avoid repeating previous answers. How are these two subtasks implemented and integrated internally? Across multiple datasets, models, and prompt templates, we identify a promote-then-suppress mechanism: the model first rec

Michael Hoefnagel, Zurab Janelidze

It is known that in (regular) unital and in subtractive categories, internal abelian groups are simply behaved; e.g., they are the same as internal algebras $(A,s)$ satisfying $s(x,0)=x$ and $s(x,x)=0$, i.e., \emph{subtraction algebras}. Moreover, in these categorical settings, such internal abelian group structures are unique, and every morphism between the

Mohammadreza Iranpour, Mohammad Rasoul Narimani

False Data Injection (FDI) attacks are a significant threat to modern power systems. Although numerous research studies have focused on FDI attacks on power systems, these studies have primarily concentrated on designing or detecting DC FDI attacks, with less attention given to the impact analysis of AC FDI attacks. AC FDI attacks are potentially more harmfu

Alexey Potapov, Hendrik Linz, Jeroen Bouwman, Will Rocha

While the number of detected molecules, particularly complex organic molecules, in the solid-state in astrophysical environments is still rather limited, laboratory experiments and astrochemical models predict many potential candidates. Detection of molecules in protoplanetary disks provides a bridge between the chemical evolution of the interstellar medium

Rashid Mushkani, Shravan Nayak, Hugo Berard, Allison Cohen

We introduce the Local Intersectional Visual Spaces (LIVS) dataset, a benchmark for multi-criteria alignment, developed through a two-year participatory process with 30 community organizations to support the pluralistic alignment of text-to-image (T2I) models in inclusive urban planning. The dataset encodes 37,710 pairwise comparisons across 13,462 images, s

Marcos A. G. Garcia, Wenqi Ke, Yann Mambrini, Keith A. Olive

One of the simplest possible candidates for dark matter is a stable scalar singlet beyond the Standard Model. If its mass is below the Hubble scale during inflation, long-wavelength modes of this scalar will be excited during inflation, and their subsequent evolution may lead to the correct relic density of dark matter. In this work, we provide a comprehensi

Fred B. Holt

Viewing Eratosthenes sieve as a discrete dynamic system, we show that every admissible instance of every admissible constellation of gaps arises and persists in Eratosthenes sieve. For an admissible constellation of length J, we show that its population across stages of the sieve is consistent with the Hardy and Littlewood estimates from 1923. This work stro

Jennifer Hu, Felix Sosa, Tomer Ullman

Some things are impossible, but some things may be even more impossible than impossible. Levitating a feather using one's mind is impossible in our world, but fits into our intuitive theories of possible worlds, whereas levitating a feather using the number five cannot be conceived in any possible world ("inconceivable"). While prior work has examined the di

Nancy Lynch

In this manuscript I overview my work on developing a Theory for Distributed Systems -- work that has involved many students and other collaborators. This effort started at Georgia Tech in the late 1970s, and has continued at MIT since 1981. This manuscript emphasizes the earlier contributions, and their impact on the directions of the field. These contribut

María González-Calderón, María Jaenada, Leandro Pardo

A one-shot device is a unit that operates only once, after which it is either destroyed or needs to be rebuilt. For this type of device, the operational status can only be assessed at a specific inspection time, determining whether failure occurred before or after it. Consequently, lifetimes are subject to left- or right-censoring. One-shot devices are usual

Mete Şeref Ahunbay, Martin Bichler

Projected gradient ascent is known to satisfy no-external regret as a learning algorithm. However, recent empirical work shows that projected gradient ascent often finds the Nash equilibrium in settings beyond two-player zero-sum interactions or potential games, including those where the set of coarse correlated equilibria is very large. We show that gradien

Yu-Ching Tseng, Chamika Goonetilleke, Xiaotian Lu, Niladri Sekhar Mandal

Through a combination of experiments and modeling, we have demonstrated a novel pattern formation phenomenon in an isothermal miscible fluid system involving simple protein and sugar solutions. We introduced dye-tagged protein solution into a petri dish with sugar solutions, which had higher density than the added protein solution. Initially, the protein spr

Mohammadreza Iranpour, Mohammad Rasoul Narimani

FDI (False Data Injection) attacks are critical to address as they can compromise the integrity and reliability of data in cyber-physical systems, leading to potentially severe consequences in sectors such as power systems. The feasibility of FDI attacks has been extensively studied from various perspectives, including access to measurements and sensors, kno

Tejasvi Chebrolu, Ponnurangam Kumaraguru, Ashwin Rajadesingan

Engaging in political discussions is crucial in democratic societies, yet many individuals remain politically disinclined due to various factors such as perceived knowledge gaps, conflict avoidance, or a sense of disconnection from the political system. In this paper, we explore the potential of personal narratives-short, first-person accounts emphasizing pe

Leopoldo Agorio, Sean Van Alen, Santiago Paternain, Miguel Calvo-Fullana

Constrained multi-agent reinforcement learning offers the framework to design scalable and almost surely feasible solutions for teams of agents operating in dynamic environments to carry out conflicting tasks. We address the challenges of multi-agent coordination through an unconventional formulation in which the dual variables are not driven to convergence

Tomáš Hons

A theorem by Ding, Oporowski, Oxley, and Vertigan states that every sufficiently large bipartite graph without twins contains a matching, co-matching, or half-graph of any given size as an induced subgraph. We prove that this Ramsey statement has polynomial dependency assuming bounded VC-dimension of the initial graph, using the recent verification of the Er

Arshia Sobhan, Philippe Pasquier, Gabriela Aceves Sepulveda

Text-to-image generative AI systems exhibit significant limitations when engaging with under-represented domains, including non-Western art forms, often perpetuating biases and misrepresentations. We present a focused case study on the generative AI system DALL-E 3, examining its inability to properly represent calligraphic Arabic script, a culturally signif

Edoardo Levati, Alejandro Cárdenas-Avendaño, Kyriakos Destounis, Paolo Pani

Orbital resonances in extreme-mass-ratio inspirals (EMRIs) have been proven to be a key feature for accurate gravitational-wave template modeling. Decades of research have led to schemes that can not only model the adiabatic inspiral of such a binary system, but also account for the effects of resonances on their evolution. In this work, we use an effective

Anna Grigoreva, Anton Andreev, Leonid Glazman

We study the electromagnetic (EM) response of the excitonic insulator phase of a time-reversal (TR) invariant Weyl semimetal (WSM). At low temperatures, the system develops two exciton condensates. The condensates are related to each other by TR symmetry and weakly coupled by a Josephson tunneling term. The latter leads to the formation of the Leggett mode [

Beatrice Magni, Alexios Christopoulos, Andrea De Luca, Xhek Turkeshi

Anticoncentration describes how an ensemble of quantum states spreads over the allowed Hilbert space, leading to statistically uniform output probability distributions. In this work, we investigate the anticoncentration of random Clifford circuits toward the overlap distribution of random stabilizer states. Using exact analytical techniques and extensive num

Luna Y. Liu, Steffi Y. Woo, Jinyuan Wu, Bowen Hou

Excitons -- elementary excitations formed by bound electron-hole pairs -- govern the optical properties and excited-state dynamics of materials. In two-dimensions (2D), excitons are theoretically predicted to have a linear energy-momentum relation with a non-analytic discontinuity in the long wavelength limit, mimicking the dispersion of a photon. This resul

Jaco ter Hoeve, Luca Mantani, Alejo N. Rossia, Juan Rojo

Global interpretations of particle physics data within the framework of the Standard Model Effective Field Theory (SMEFT), including their matching to UV-complete models, involve energy scales potentially spanning several orders of magnitude. Relating these measurements among them in terms of a common energy scale is enabled by the Renormalisation Group Equa

Munshi G. Mustafa, Aritra Bandyopadhyay, Chowdhury Aminul Islam

This review has explored the fundamental principles of thermal field theory in the context of a background magnetic field, highlighting its theoretical framework and some of its applications to the thermo-magnetic QCD plasma generated in heavy-ion collisions. Our discussion has been limited to equilibrium systems for clarity and conciseness. We analysed bulk

Experiment Galactic Science Group, Julian Borrill, Susan E. Clark, Jacques Delabrouille

Polarized foreground emission from the Galaxy is one of the biggest challenges facing current and upcoming cosmic microwave background (CMB) polarization experiments. We develop new models of polarized Galactic dust and synchrotron emission at CMB frequencies that draw on the latest observational constraints, that employ the ``polarization fraction tensor''

Mathias Garny, Roman Scoccimarro

We develop a new approach to Vlasov Perturbation Theory (VPT) that solves for the hierarchy of cumulants of the phase-space distribution function to arbitrarily high truncation order in the context of cosmological structure formation driven by collisionless dark matter. We investigate the impact of higher cumulants on density and velocity power spectra as we

Fei Huang, V. Knapp-Perez

Cosmological stasis is a new type of epoch in the cosmological timeline during which the cosmological abundances of different energy components -- such as vacuum energy, matter, and radiation -- remain constant despite the expansion of the universe. Previous studies have shown that stasis naturally arises in various scenarios beyond the Standard Model, eithe

Caterina Bracci, Francesco Belfiore, Michele Ginolfi, Anna Feltre

Emission-line regions are key to understanding the properties of galaxies, as they trace the exchange of matter and energy between stars and the interstellar medium (ISM). In nearby galaxies, individual nebulae can be identified as HII regions, planetary nebulae (PNe), supernova remnants (SNR), and diffuse ionised gas (DIG) with criteria on single or multipl

Samuel Gagnon-Hartman, James E. Davies, Andrei Mesinger

The cosmic 21-cm signal promises to revolutionize studies of the Epoch of Reionization (EoR). Radio interferometers are aiming for a preliminary, low signal-to-noise (S/N) detection of the 21-cm power spectrum. Cross-correlating 21-cm with galaxies will be especially useful in these efforts, providing both a sanity check for initial 21-cm detection claims an

Hernán Muriel, David Pérez-Millán, Martín de los Rios, Andrea Biviano

Clusters of galaxies have proven to be efficient systems in modifying various properties of galaxies, such as star formation or morphology. However, projection effects impose serious challenges in determining how, when, and to what extent galaxies are affected by the cluster environment. Using innovative techniques to classify galaxies based on their history

Vasco Gennari, Simone Mastrogiovanni, Nicola Tamanini, Sylvain Marsat

The population of the observed gravitational wave events encodes unique information on the formation and evolution of stellar-mass black holes, from the underlying astrophysical processes to the large-scale dynamics of the Universe. We use the ICAROGW analysis infrastructure to perform hierarchical Bayesian inference on the gravitational wave signals from th

Luca Cassia, Kiril Hristov

We study the equivariant generalization of topological strings on toric manifolds, focusing in particular on defining the contributions of constant maps in the genus expansion of the partition function. This approach regularizes the integration over non-compact Calabi-Yau spaces, producing finite results at each order in the expansion, as illustrated by a br

Tao Han, Matthew Low, Tong Arthur Wu, Keping Xie

A high-energy $\mu^+\mu^-$ collider provides a wide variety of mechanisms for the production of new heavy particles. While the reach for such particles via the direct annihilation of $\mu^+\mu^-$ will approach the center-of-mass energy of the collider, the partonic fusions from gauge bosons, quarks, and gluons, originating from the incoming muon beams will o

Wuji Wang, Carlos De Breuck, Dominika Wylezalek, Joël Vernet

Mergers play important roles in galaxy evolution at and beyond Cosmic Noon ($z\sim3$). They are found to be a trigger of active galactic nuclei (AGN) activity and a process for growing stellar mass and black hole mass. High-$z$ radio galaxies (HzRGs=type-2 radio-loud AGN) are among the most massive galaxies known, and reside in dense environments on scales o

M. L. L. Dantas, R. Smiljanic, R. S. de Souza, P. B. Tissera

We employ a Generalised Additive Model (GAM) to address the limitations inherent in radial metallicity gradients predicted by chemical evolution models, thereby facilitating the estimation of birth radii for the thin disc stars in our sample based on their ages and chemical composition. We then juxtapose the birth radius predictions derived from the GAM with

Lakshya Bhardwaj, Sakura Schafer-Nameki, Apoorv Tiwari, Alison Warman

We use the Symmetry Topological Field Theory (SymTFT) to systematically characterize gapped phases in 2+1 dimensions with categorical symmetries. The SymTFTs that we consider are (3+1)d Dijkgraaf-Witten (DW) theories for finite groups $G$, whose gapped boundaries realize all so-called ``All Bosonic type" fusion 2-category symmetries. In arXiv:2408.05266 we p

Yllari K. González-Koda, Tomás Ruiz-Lara, Carme Gallart, Edoardo Ceccarelli

Context. Gaia-Sausage-Enceladus is considered the last major merger that contributed to the formation of the Milky Way. Its remnants dominate the nearby accreted stellar halo of the Milky Way. Aim. We aim to characterise the star formation history of Gaia-Sausage-Enceladus through the age and metallicity of its stellar populations. Methods. From Gaia DR3 dat

Ivano Basile, Alessandro Borys, Joaquin Masias

We build a novel realization of dark bubble cosmology in non-supersymmetric string theory. Among the simplest models in ten dimensions, the type 0'B orientifold is the unique option which yields a scale-separated construction. The resulting setting produces a logarithmically varying dynamical dark energy, reflecting its holographic counterpart in terms of ru

Rahul Kannan, Ewald Puchwein, Aaron Smith, Josh Borrow

We introduce the THESAN-ZOOM project, a comprehensive suite of high-resolution zoom-in simulations of $14$ high-redshift ($z>3$) galaxies selected from the THESAN simulation volume. This sample encompasses a diverse range of halo masses, with $M_\mathrm{halo} \approx 10^8 - 10^{13}~\mathrm{M}_\odot$ at $z=3$. At the highest-resolution, the simulations achiev

Fernando Aguado-Agelet, Davide Massari, Matteo Monelli, Santi Cassisi

We present the age determination of 13 globular clusters dynamically associated with the Gaia-Sausage-Enceladus (GSE) merger event, as part of the CARMA project effort to trace the Milky Way assembly history. We used deep and homogeneous archival $Hubble$ $Space$ $Telescope$ data, and applied isochrone-fitting to derive homogeneous age estimates. We find tha

Weiguang Cao, Masahito Yamazaki, Linhao Li

Recent advancements in generalized symmetries have drawn significant attention to gapped phases of matter exhibiting novel symmetries, such as noninvertible symmetries. By leveraging the duality transformations, the classification and construction of gapped phases with noninvertible symmetry can be mapped to those involving conventional group symmetries. We

Matthew R. Buckley, Peizhi Du, Nicolas Fernandez, Mitchell J. Weikert

Current cosmological data are well-described by the Lambda-Cold Dark Matter ($\Lambda$CDM) model, which assumes adiabatic initial conditions for the primordial density perturbations. This agreement between data and theory enables strong constraints on new physics that generates isocurvature perturbations. Existing constraints typically assume a simple power

Kaley Brauer, Jennifer Mead, John H. Wise, Greg L. Bryan

We explore the effect of variations in the Population III (Pop III) initial mass function (IMF) and star-by-star feedback on early galaxy formation and evolution using the Aeos simulations. We compare simulations with two different Pop III IMFs: $M_\text{char} = 10 \, \mathrm{M}_\odot$ and $M_{\rm max} = 100 \, \mathrm{M}_\odot$ (Aeos10) and $M_\text{char} =

Elena Arjona-Galvez, Salvador Cardona-Barrero, Robert J. J. Grand, Arianna Di Cintio

Galaxy sizes are a key parameter to distinguishing between different galaxy types and morphologies, reflecting their formation and assembly histories. Several methods define galaxy boundaries, often relying on light concentration or isophotal densities. However, these approaches were often constrained by observational limitations and did not necessarily prov

Andrew Kuzovchikov

We study the connection between $\mathrm{SU}(n)$ spin chains and one-dimensional sigma models on flag manifolds. Using this connection, we calculate the spectrum of the Laplace-Beltrami operator and geodesics for a particular class of metrics on $\mathbb{CP}^1$ and $\mathcal{F}_3$, which is a manifold of complete flags in $\mathbb{C}^3$.

Toru Lin, Kartik Sachdev, Linxi Fan, Jitendra Malik

Learning generalizable robot manipulation policies, especially for complex multi-fingered humanoids, remains a significant challenge. Existing approaches primarily rely on extensive data collection and imitation learning, which are expensive, labor-intensive, and difficult to scale. Sim-to-real reinforcement learning (RL) offers a promising alternative, but

Zhongyang Li, Ziyue Li, Tianyi Zhou

In large multimodal models (LMMs), the perception of non-language modalities (e.g., visual representations) is usually not on par with the large language models (LLMs)' powerful reasoning capabilities, deterring LMMs' performance on challenging downstream tasks. This weakness has been recently mitigated by replacing the vision encoder with a mixture-of-exper

Sk Asrap Murshed, Sanjib Kumar Das, Bitan Roy

From a leading-order unbiased renormalization group analysis we here showcase the emergence of superconductivity (including the topological ones) from purely repulsive electron-electron interactions in two-dimensional doped Dirac insulators, featuring a Fermi surface. In the absence of chemical doping, such systems describe quantum anomalous or spin Hall and

Susmit Agrawal, Deepika Vemuri, Sri Siddarth Chakaravarthy P, Vineeth N. Balasubramanian

Concept-based methods have emerged as a promising direction to develop interpretable neural networks in standard supervised settings. However, most works that study them in incremental settings assume either a static concept set across all experiences or assume that each experience relies on a distinct set of concepts. In this work, we study concept-based mo

Matthew Tamayo-Rios, Alexander Schell, Rima Alaifari

The signature kernel is a recent state-of-the-art tool for analyzing high-dimensional sequential data, valued for its theoretical guarantees and strong empirical performance. In this paper, we present a novel method for efficiently computing the signature kernel of long, high-dimensional time series via adaptively truncated recursive local power series expan

Siddhant Haldar, Lerrel Pinto

Building robotic agents capable of operating across diverse environments and object types remains a significant challenge, often requiring extensive data collection. This is particularly restrictive in robotics, where each data point must be physically executed in the real world. Consequently, there is a critical need for alternative data sources for robotic

Sirui Xu, Hung Yu Ling, Yu-Xiong Wang, Liang-Yan Gui

Achieving realistic simulations of humans interacting with a wide range of objects has long been a fundamental goal. Extending physics-based motion imitation to complex human-object interactions (HOIs) is challenging due to intricate human-object coupling, variability in object geometries, and artifacts in motion capture data, such as inaccurate contacts and

Ang Cao, Sergio Arnaud, Oleksandr Maksymets, Jianing Yang

3D vision-language grounding faces a fundamental data bottleneck: while 2D models train on billions of images, 3D models have access to only thousands of labeled scenes--a six-order-of-magnitude gap that severely limits performance. We introduce $\textbf{LIFT-GS}$, a practical distillation technique that overcomes this limitation by using differentiable rend

Sucheng Ren, Qihang Yu, Ju He, Xiaohui Shen

Autoregressive (AR) modeling, known for its next-token prediction paradigm, underpins state-of-the-art language and visual generative models. Traditionally, a ``token'' is treated as the smallest prediction unit, often a discrete symbol in language or a quantized patch in vision. However, the optimal token definition for 2D image structures remains an open q

Jingru Jia, Zehua Yuan, Junhao Pan, Paul E. McNamara

Strategic decision-making involves interactive reasoning where agents adapt their choices in response to others, yet existing evaluations of large language models (LLMs) often emphasize Nash Equilibrium (NE) approximation, overlooking the mechanisms driving their strategic choices. To bridge this gap, we introduce an evaluation framework grounded in behavior

Jiahe Li, Jiawei Zhang, Xiao Bai, Jin Zheng

Despite exhibiting impressive performance in synthesizing lifelike personalized 3D talking heads, prevailing methods based on radiance fields suffer from high demands for training data and time for each new identity. This paper introduces InsTaG, a 3D talking head synthesis framework that allows a fast learning of realistic personalized 3D talking head from

Dexter Ong, Yuezhan Tao, Varun Murali, Igor Spasojevic

We address the challenge of task-oriented navigation in unstructured and unknown environments, where robots must incrementally build and reason on rich, metric-semantic maps in real time. Since tasks may require clarification or re-specification, it is necessary for the information in the map to be rich enough to enable generalization across a wide range of

David Bolin, Alexandre B. Simas

The R software package rSPDE contains methods for approximating Gaussian random fields based on fractional-order stochastic partial differential equations (SPDEs). A common example of such fields are Whittle-Mat\'ern fields on bounded domains in $\mathbb{R}^d$, manifolds, or metric graphs. The package also implements various other models which are briefly in

Özgür Akarsu, Arman Çam, Evangelos A. Paraskevas, Leandros Perivolaropoulos

We investigate linear matter density perturbations in the $\Lambda_{\rm s}$CDM, in which the $\Lambda$ is replaced by late-time ($z\sim2$) mirror AdS-dS transition, resulting in distinct growth dynamics. We use two complementary approaches: (i) determining the initial density contrast and its evolution rate for a given collapse scale factor, (ii) computing t

Jeffrey Yang Fan Chiang, Seungjae Lee, Jia-Bin Huang, Furong Huang

Recent advancements in Web AI agents have demonstrated remarkable capabilities in addressing complex web navigation tasks. However, emerging research shows that these agents exhibit greater vulnerability compared to standalone Large Language Models (LLMs), despite both being built upon the same safety-aligned models. This discrepancy is particularly concerni

Lujie Yang, H. J. Terry Suh, Tong Zhao, Bernhard Paus Graesdal

We present a low-cost data generation pipeline that integrates physics-based simulation, human demonstrations, and model-based planning to efficiently generate large-scale, high-quality datasets for contact-rich robotic manipulation tasks. Starting with a small number of embodiment-flexible human demonstrations collected in a virtual reality simulation envir

Dmitrii Zendrikov, Alessio Franci, Giacomo Indiveri

Neural systems use the same underlying computational substrate to carry out analog filtering and signal processing operations, as well as discrete symbol manipulation and digital computation. Inspired by the computational principles of canonical cortical microcircuits, we propose a framework for using recurrent spiking neural networks to seamlessly and robus

Arnav Kumar Jain, Gonzalo Gonzalez-Pumariega, Wayne Chen, Alexander M Rush

We address the problem of code generation from multi-turn execution feedback. Existing methods either generate code without feedback or use complex, hierarchical reinforcement learning to optimize multi-turn rewards. We propose a simple yet scalable approach, $\mu$Code, that solves multi-turn code generation using only single-step rewards. Our key insight is

Shalev Lifshitz, Sheila A. McIlraith, Yilun Du

By utilizing more computational resources at test-time, large language models (LLMs) can improve without additional training. One common strategy uses verifiers to evaluate candidate outputs. In this work, we propose a novel scaling dimension for test-time compute: scaling the number of verifiers. We introduce Multi-Agent Verification (MAV) as a test-time co

Hanyang Kong, Xingyi Yang, Xinchao Wang

Rendering dynamic scenes from monocular videos is a crucial yet challenging task. The recent deformable Gaussian Splatting has emerged as a robust solution to represent real-world dynamic scenes. However, it often leads to heavily redundant Gaussians, attempting to fit every training view at various time steps, leading to slower rendering speeds. Additionall

Zhuangwei Chen, Marco Calvi, John Durrell, Cristian Boffo

Considerable effort has been devoted to the development of superconducting undulators (SCUs) intended for particle accelerator-based light sources, including synchrotrons and free electron laser (FEL) facilities. Recently, a high-temperature superconducting (HTS) undulator prototype, consisting of staggered-array Re-Ba-Cu-O bulks, achieved an on-axis sinusoi

Albert Gong, Kamilė Stankevičiūtė, Chao Wan, Anmol Kabra

High-quality benchmarks are essential for evaluating reasoning and retrieval capabilities of large language models (LLMs). However, curating datasets for this purpose is not a permanent solution as they are prone to data leakage and inflated performance results. To address these challenges, we propose PhantomWiki: a pipeline to generate unique, factually con

Edo Kadosh, Nir Goren, Or Patashnik, Daniel Garibi

Text-to-image diffusion models offer powerful image editing capabilities. To edit real images, many methods rely on the inversion of the image into Gaussian noise. A common approach to invert an image is to gradually add noise to the image, where the noise is determined by reversing the sampling equation. This process has an inherent tradeoff between reconst

Aravind Gollakota, Parikshit Gopalan, Aayush Karan, Charlotte Peale

Given a predictor and a loss function, how well can we predict the loss that the predictor will incur on an input? This is the problem of loss prediction, a key computational task associated with uncertainty estimation for a predictor. In a classification setting, a predictor will typically predict a distribution over labels and hence have its own estimate o

James Purcell, Abhishek Rajput, Toby Cubitt

Dissipative processes have long been proposed as a means of performing computational tasks on quantum computers that may be intrinsically more robust to noise. In this work, we prove two main results concerning the error-resilience capabilities of two types of dissipative algorithms: dissipative ground state preparation in the form of the dissipative quantum

Lixing Zhang, Ze-Xun Lin, Prineha Narang, Di Luo

Hybrid quantum systems with different particle species are fundamental in quantum materials and quantum information science. In this work, we establish a rigorous theoretical framework proving that, given access to an unknown spin-boson type Hamiltonian, our algorithm achieves Heisenberg-limited estimation for all coupling parameters up to error $\epsilon$ w

Gareth Mansfield

The Veneziano amplitude describing the tree-level scattering of four open superstrings is expected to be consistent with unitarity in ten spacetime dimensions. While this follows indirectly from the no-ghost theorem, a direct proof at the level of the amplitude has only been found for $D\leq 6$. In this article, we close this gap by providing a complete proo

Saeid Naderiparizi, Xiaoxuan Liang, Berend Zwartsenberg, Frank Wood

In this paper we describe a novel framework for diffusion-based generative modeling on constrained spaces. In particular, we introduce manual bridges, a framework that expands the kinds of constraints that can be practically used to form so-called diffusion bridges. We develop a mechanism for combining multiple such constraints so that the resulting multiply

Zhi Cen, Huaijin Pi, Sida Peng, Qing Shuai

This paper addresses the task of generating two-character online interactions. Previously, two main settings existed for two-character interaction generation: (1) generating one's motions based on the counterpart's complete motion sequence, and (2) jointly generating two-character motions based on specific conditions. We argue that these settings fail to mod