October 2025 arXiv papers — page 57

Vincent Flynn, Benedetta Flebus

The identification of platforms with independently tunable nonlinearity and non-Hermiticity promises a quantitative route to far-from-equilibrium universality across many-body systems. Here we show that a conventional ferromagnetic multilayer realizes this paradigm: balancing a dc drive against Gilbert damping stabilizes a self-organized, current-carrying no

Yunsen Lei, Kexin Bai, Quan Li, H. Howie Huang

Export controls have become one of America's most prominent tools of economic statecraft. They aim to block rival countries' access to sensitive technologies, safeguard U.S. supply chains, protect national security, and shape geopolitical competition. Among various instruments, the U.S. Entity List has emerged as the most salient, yet its dynamics remain und

Iskander Azangulov, Teodora Pandeva, Niranjani Prasad, Javier Zazo

Masked diffusion models (MDMs) offer a compelling alternative to autoregressive models (ARMs) for discrete text generation because they enable parallel token sampling, rather than sequential, left-to-right generation. This means potentially much faster inference. However, effective parallel sampling faces two competing requirements: (i) simultaneously update

Eduard Brüll, Samuel Mäurer, Davud Rostam-Afschar

We provide experimental evidence on how employers adjust expectations to automation risk in high-skill, white-collar work. Using a randomized information intervention among tax advisors in Germany, we show that firms systematically underestimate automatability. Information provision raises risk perceptions, especially for routine-intensive roles. Yet, it lea

Harrison F. Stropkay, Jiayi Chen, Mohammad J. Latifi, Daniel N. Rockmore

We show that large language models (LLMs) can be used to distinguish the writings of different authors. Specifically, an individual GPT-2 model, trained from scratch on the works of one author, will predict held-out text from that author more accurately than held-out text from other authors. We suggest that, in this way, a model trained on one author's works

Zhixin Pan, Ziyu Shu, Amberbir Alemayoh

Ransomware has become a critical threat to cybersecurity due to its rapid evolution, the necessity for early detection, and growing diversity, posing significant challenges to traditional detection methods. While AI-based approaches had been proposed by prior works to assist ransomware detection, existing methods suffer from three major limitations, ad-hoc f

Armin Gerami, Ramani Duraiswami

The original softmax-based attention mechanism (regular attention) in the extremely successful Transformer architecture computes attention between $N$ tokens, each embedded in a $D$-dimensional head, with a time complexity of $O(N^2D)$. Given the success of Transformers, improving their runtime during both training and inference is a popular research area. O

Mykola Haltiuk, Aleksander Smywinski-Pohl

Large Language Models (LLMs) are trained to support an increasing number of languages, yet their predefined tokenizers remain a bottleneck for adapting models to lower-resource or distinct-script languages. Existing tokenizer transfer methods typically rely on semantic heuristics to initialize new embeddings, ignoring higher-layer model dynamics and limiting

Adrián P. Bustamante, Alessandra Celletti, Christoph Lhotka

This work investigates different models of rotational dynamics of two rigid bodies with the shape of an ellipsoid, moving under their gravitational influence. The focus of this study is on their behavior, their linear stability, and numerical investigation of the main resonances. We assume that the spin axes of the two bodies are perpendicular to the orbital

J. Jon Ryu, Samuel Zhou, Gregory W. Wornell

Spectral decomposition of linear operators plays a central role in many areas of machine learning and scientific computing. Recent work has explored training neural networks to approximate eigenfunctions of such operators, enabling scalable approaches to representation learning, dynamical systems, and partial differential equations (PDEs). In this paper, we

Yijin Wang, Subhonmesh Bose

We propose a generalization of the Bass diffusion model in discrete-time that explicitly models the effect of price in adoption. Our model is different from earlier price-incorporated models and fits well to adoption data for various products. We then utilize this model to study two decision-making problems. First, we provide a series of structural results o

Nnamdi Daniel Aghanya, Romain Leemans

We study Heaven-Hell dynamics, a model for network consensus. A known result establishes an exact one-step convergence threshold for systems with a single uniform hub: the per-node inbound hub weight W suffices if and only if W >= maxrest, the maximum non-hub inbound mass. We develop scale laws and operational refinements that make this threshold robust to t

Mustafa Amin, Mason Daub, Mark A. Walton

Solutions of the time-dependent Schr\"odinger equation are mapped to other solutions for a (possibly) different potential by so-called form-preserving transformations. These time-dependent transformations of the space and time coordinates can produce remarkable solutions with surprising properties. A classic example is the force-free accelerating Airy beam f

Faizuddin Ahmed, Edilberto O. Silva

This study investigates a Schwarzschild-anti de Sitter black hole coupled to a cloud of strings featuring only the electric-like component of the string bivector, embedded in a perfect fluid dark matter and a quintessence field. We examine the dynamics of photons and massive particles, focusing on trajectories, photon spheres, BH shadows, their topological c

Kieu Dang, Phung Lai, NhatHai Phan, Yelong Shen

Large language models (LLMs) demonstrate remarkable capabilities across various tasks. However, their deployment introduces significant risks related to intellectual property. In this context, we focus on model stealing attacks, where adversaries replicate the behaviors of these models to steal services. These attacks are highly relevant to proprietary LLMs

Antoine Ledent, Rodrigo Alves, Yunwen Lei

It has been recently observed in much of the literature that neural networks exhibit a bottleneck rank property: for larger depths, the activation and weights of neural networks trained with gradient-based methods tend to be of approximately low rank. In fact, the rank of the activations of each layer converges to a fixed value referred to as the ``bottlenec

Tushar Sial, Abhishek Halder

We formulate and solve the fixed horizon linear quadratic covariance steering problem in continuous time with a terminal cost measured in Hilbert-Schmidt (i.e., Frobenius) norm error between the desired and the controlled terminal covariances. For this problem, the necessary conditions of optimality become a coupled matrix ODE two-point boundary value proble

Ruaridh Macdonald, Filippo Pecci, Luca Bonaldo, Jun Wen Law

MacroEnergy.jl (aka Macro) is an open-source framework for multi-sector capacity expansion modeling and analysis of macro-energy systems. It is written in Julia and uses the JuMP package to interface with a wide range of mathematical solvers. It enables researchers and practitioners to design and analyze energy and industrial systems that span electricity, f

Carlo Branchina, Angela Conaci, Stefania De Curtis, Luigi Delle Rose

We study how out-of-equilibrium effects modify the steady-state propagation of bubble walls during a cosmological first-order electroweak phase transition. Going beyond the local thermal equilibrium approximation, we numerically solve the coupled system of scalar field, hydrodynamic and Boltzmann equations using a spectral algorithm that allows a first-princ

Grzegorz Szamel

We propose and analyze a process that extracts useful work from a single active particle maintained at constant temperature in a harmonic potential by measuring the relative sign of the self-propulsion and the confining force and then adjusting the stiffness of the potential. First, we show analytically that useful work can be extracted by stepwise changes o

J. A. Gracey

We renormalize generalized 3-quark operators that relate to baryon states using the method devised by Kr\"{a}nkl and Manashov at four loops in the MSbar scheme. The anomalous dimensions of the four core operators used to compute nucleon matrix elements are determined and their associated critical exponents are studied in the conformal window using the Banks-

Mumnuna Aziz Qureshi

This paper presents a novel formalism for the out of equilibrium dynamics of the density matrix, capable of describing highly entangled many-body interactions. The evolution of quantum states is evaluated via eigenvalue dynamics of a general Hamiltonian system, perturbed by a parametrically evolving variable $\lambda(t)$ that carries the time-dependence. Thi

Pakapim Eua-anant, Papangkorn Apinyanon, Thunyatorn Jirachaisri, Nantapong Ruangsuksriwong

All or Nothing, Water Walk, and Remembered Length are pencil puzzles that involve constructing a continuous loop on a rectangular grid under specific constraints. In this paper, we analyze their computational complexity using the T-metacell framework developed by Tang and MIT Hardness Group. We establish that the puzzles are NP-complete by providing reductio

Alessandra Celletti, Christoph Lhotka, Giuseppe Pucacco

We study the dynamics of the collinear points in the planar, restricted three-body problem, assuming that the primaries move on an elliptic orbit around a common barycenter. The equations of motion can be conveniently written in a rotating pulsating barycentric frame, taking the true anomaly as independent variable. We consider the Hamiltonian modeling this

Ronan Kerr, Facundo Peŕez Paolino, Jonathan C. Tan, Joshua S. Speagle

Recent Gaia-based young stellar association surveys have revealed dozens of low-mass populations that have, until recently, been too small or sparse to detect. These populations represent a largely unstudied demographic with unknown origins, and their relative isolation may minimize gravitational disruptions that impact traceback, making them compelling targ

Samuel Bright-Thonney, Christina Reissel, Gaia Grosso, Nathaniel Woodward

Novelty detection in large scientific datasets faces two key challenges: the noisy and high-dimensional nature of experimental data, and the necessity of making statistically robust statements about any observed outliers. While there is a wealth of literature on anomaly detection via dimensionality reduction, most methods do not produce outputs compatible wi

Fardin Ganjkhanloo, Emmett Springer, Erik H. Hoyer, Daniel L. Young

Risk assessment tools in healthcare commonly employ point-based scoring systems that map patients to ordinal risk categories via thresholds. While electronic health record (EHR) data presents opportunities for data-driven optimization of these tools, two fundamental challenges impede standard supervised learning: (1) labels are often available only for extre

Joao Correia, Daniel Coutinho, Marco Castelluccio, Caio Barbosa

The use of Large Language Models (LLMs) to support tasks in software development has steadily increased over recent years. From assisting developers in coding activities to providing conversational agents that answer newcomers' questions. In collaboration with the Mozilla Foundation, this study evaluates the effectiveness of Retrieval-Augmented Generation (R

Yuansheng Ni, Songcheng Cai, Xiangchao Chen, Jiarong Liang

Large language models (LLMs) have recently enabled coding agents capable of generating, executing, and revising visualization code. However, existing models often fail in practical workflows due to limited language coverage, unreliable execution, and lack of iterative correction mechanisms. Progress has been constrained by narrow datasets and benchmarks that

Sophie Y. Lee, Philipp W. A. Schönhöfer, Sharon C. Glotzer

Collections of simple, self-propelled colloidal particles exhibit complex, emergent dynamical behavior, with promising applications in microrobotics. When confined within a deformable vesicle, self-propelled rods cluster and align, propelling the vesicle and inducing changes in the vesicle shape. We explore potential microrobotic capabilities of such vesicle

Adam Smith, Emre Özülker, Eleonora Di Valentino, Carsten van de Bruck

We investigate the interplay between varying electron mass ($m_e$) and dynamical dark energy by analysing the Chevallier-Polarski-Linder (CPL) parametrization and its non-crossing variants, both with and without a varying-$m_e$ component. Our aim is to assess whether the preference for late-time dynamics and phantom divide line (PDL) crossing persists when e

Harsh Kumar, Jasmine Chahal, Yinuo Zhao, Zeling Zhang

Seeking advice is a core human behavior that the internet has reinvented twice: first through forums and Q&A communities that crowdsource public guidance, and now through large language models (LLMs). Yet the quality of this LLM advice for everyday well-being scenarios remains unclear. How does it compare, not only against human comments, but against the wis

Amazigh Ouzriat, Viola Sordini, Francesco Di Renzo

Low-latency analyses of gravitational-wave (GW) data from LIGO, Virgo, and KAGRA enable rapid detection of compact binary coalescences (CBC) and prompt sky localization, essential for electromagnetic follow-up in multi-messenger astronomy. We evaluate the performance and limitations of low-latency sky localization using BAYESTAR algorithm, and investigate th

Carlos A. Argüelles, Toni Bertólez-Martínez, Alba Burgos-Mondéjar, Anne-Katherine Burns

In this article, we introduce the concept of \textit{topographic enhancement} in the context of ultra-high-energy neutrino detection by underwater neutrino telescopes. We demonstrate that the local topography around KM3NeT/ARCA can increase the detection efficiency in scenarios involving long-lived particles by up to a factor of $\sim 3$ due to the presence

Tianxing Shi, Chuhang Zhang, Liang Jin, Linhu Li

Controlling topological phases is a central goal in quantum materials and related fields, enabling applications such as robust transport and programmable edge states. Here we uncover a mechanism in which local on-site impurities act as knobs to decompose global topological properties in discrete steps. In non-Hermitian lattices with spectral winding topology

He-Ran Wang, Ilya Vilkoviskiy, Dmitry A. Abanin

The influence matrix (IM) provides a powerful framework for characterizing nonequilibrium quantum many-body dynamics by encoding multitime correlations into tensor-network states. Understanding how its computational complexity relates to underlying dynamics is crucial for both theoretical insight and practical utility, yet remains largely unexplored despite

Nasmi S Anand, Swarna Chatterjee, Ramij Raja, Majidul Rahaman

Recent advances in high-sensitivity radio observations have uncovered a population of faint, ultra-steep-spectrum sources in galaxy clusters, commonly known as radio phoenixes. However, their observational classification remains poorly constrained due to the limited number of confirmed detections. This study presents a detailed multi-frequency, high-sensitiv

Rongzhou Chen, Haitao Nie, Shuo Zhu, Yaping Zhao

Inverse design of metasurfaces for the joint optimization of optical modulation and algorithmic decoding in computational optics presents significant challenges, especially in applications such as hyperspectral imaging. We introduce a physics-data co-driven framework for designing reconfigurable metasurfaces fabricated from the phase-change material Ge2Sb2Se

Anurag Kaushal, Naveen S. Prabhakar, Spenta R. Wadia

We solve Einstein's equations with negative cosmological constant in $2+1$ dimensions in the Hamiltonian formulation. The spacetime has the topology of $\Sigma \times \mathbf{R}$ where $\mathbf{R}$ corresponds to the time direction and $\Sigma$ is a cylinder $\mathbf{R} \times \mathbf{S}^1$ and the spacetime metric satisfies asymptotically AdS (AAdS) boundar

Benjamin Elder, Kinga Gawrych, Arttu Rajantie

Instantons, localised saddle points of the action, play an important role in describing non-perturbative aspects of quantum field theories, for example vacuum decay or violation of conservation laws associated with anomalous symmetries. However, there are theories in which no saddle point exists. In this paper, we revisit the idea of constrained instantons,

Bingqian Ma, XiaoYing Pang, Sambaran Banerjee, Pengfei Ren

We analyze the velocity dispersion profiles of nine open clusters in the solar neighborhood using kinematic data from Gaia DR 3, aiming to identify potential dynamical signatures of stellar-mass black holes through a comparison of theoretical and observed dispersion profiles. The selected clusters include LP2373 gp4, NGC 1980, NGC 2451A, NGC 2516, NGC 3532,

Anurag Kaushal, Naveen S. Prabhakar, Spenta R. Wadia

We study the quantum dynamics of a probe scalar field in the background of a black hole in AAdS spacetime in the Hamiltonian formulation of general relativity in the maximal slicing gauge. The black hole solution in this gauge is expressed in terms of wormhole coordinates, a smooth coordinate system with constant time slices that cut across the horizon, and

Fabiola Cocchiararo, Alessia Franchini, Alessandro Lupi, Alberto Sesana

We investigate the impact of radiation pressure on electromagnetic signatures of accreting massive black hole binaries (MBHBs) at milli-parsec separations, using 3D hyper-Lagrangian resolution hydrodynamical simulations. We model binaries embedded in a self-gravitating circumbinary disc that evolves following an adiabatic equation of state, including viscous

D. Eckert, M. Markevitch, J. A. ZuHone, M. Regamey

The velocity field of intracluster gas in galaxy clusters contains key information on the virialization of infalling material, the dissipation of AGN energy into the surrounding medium, and the validity of the hydrostatic hypothesis. The statistical properties of the velocity field are characterized by its fluctuation power spectrum, which is usually expecte

Christopher Gerlach, Wolfram Ratzinger, Pedro Schwaller

In the presence of primordial isocurvature perturbations, for example in a separate dark radiation sector, the superhorizon evolution of curvature perturbations becomes nontrivial. If the dark sector is radiation-like and constitutes a significant fraction of the energy density, its isocurvature can imply isocurvature in the inflaton sector even without dire

Aidan Mai, Kyle Kremer, Fulya Kıroğlu

The LIGO/Virgo/Kagra (LVK) Collaboration has detected numerous binary black hole mergers with properties that challenge standard binary evolution scenarios, such as component masses above the pair-instability gap and high spin magnitudes. Dense stellar environments such as globular clusters provide a natural channel for producing such systems through hierarc

Giulio Barni

We present a comprehensive, self-contained pedagogical computation of the baryon asymmetry of the Universe within electroweak baryogenesis (EWBG), from the derivation of the semiclassical, CP-dependent force to the formulation and solution of the transport equations obtained from the Boltzmann equations$-$all implemented in the open-source code BARYONET. Our

Xuepeng Wang, Johannes S. Hofmann, Debanjan Chowdhury

Skyrmions are emergent many-body excitations that lie at the heart of both multi-component quantum Hall-like systems and deconfined quantum criticality. In a companion article (X. Wang et al., arXiv:2507.22971), we studied a microscopic time-reversal symmetric model of tunable interacting Chern bands using numerically exact determinant quantum Monte Carlo ca

Marcus Högås, Edvard Mörtsell

In many data analyses, each measurement may come with a simple yes/no correction; for example, belonging to one of two populations or being contaminated or not. Ignoring such binary effects may bias the results, while accounting for them explicitly quickly becomes infeasible as each of the $N$ data points introduces an additional parameter, resulting in an e

David S. Robertson, Thomas Jaki

Replication studies for scientific research are an important part of ensuring the reliability and integrity of experimental findings. In the context of clinical trials, the concept of replication has been formalised by the 'two-trials' rule, where two pivotal studies are required to show positive results before a drug can be approved. In experiments testing

Tobias Schmidt, Steffen Schneider, Matthias Bethge

We propose Equivariance by Contrast (EbC) to learn equivariant embeddings from observation pairs $(\mathbf{y}, g \cdot \mathbf{y})$, where $g$ is drawn from a finite group acting on the data. Our method jointly learns a latent space and a group representation in which group actions correspond to invertible linear maps -- without relying on group-specific ind

Christy Li, Josep Lopez Camuñas, Jake Thomas Touchet, Jacob Andreas

When a vision model performs image recognition, which visual attributes drive its predictions? Detecting unintended reliance on specific visual features is critical for ensuring model robustness, preventing overfitting, and avoiding spurious correlations. We introduce an automated framework for detecting such dependencies in trained vision models. At the cor

Kingsley E. Ehrich, Jason A. Dittmann, Samuel P. Halverson, Alejandro Camazón-Pinilla

Transmission spectroscopy allows us to detect molecules in planetary atmospheres, but is subject to contamination from inhomogeneities on the stellar surface. Quantifying the extent of this contamination is essential for accurate measurements of atmospheric composition, as stellar activity can manifest as false atmospheric signals in planetary transmission s

Hanqi Shi, Wenyuan Shi, Ian Whitehead, Ham Williams-Tracy

Haag, Kertzer, Rickards, and Stange disprove the Local-Global Conjecture for Apollonian circle packings. We extend their disproof to four more types of integral circle packing: the octahedral, cubic, square, and triangular packings. In each case, we find quadratic invariants which imply quadratic reciprocity obstructions to the conjecture in certain packings

Ayngaran Thavanesan, Aron C. Wall

Recent work on cosmological amplitudes has established reality conditions (derived from unitarity) for general particle-creation processes in flat FLRW cosmologies, in the Bunch-Davies wavefunction. In light of these results, we propose a new class of large-$N$ holographic gauge theories with $d$ spatial dimensions, which we call Kosmic Field Theories (KFTs)

Manami Roy, Kung-Yi Su, Stephanie Tonnesen, Yue Samuel Lu

Does cosmic ray (CR) pressure matter for the circumgalactic medium (CGM)? Despite growing interest, this remains a debated question, complicated by limited observational constraints and differing implementations of CR physics in simulations. While prior studies suggest that CRs influence the thermal and dynamical state of the CGM, their role in shaping cold

Daniel Bienstock, Matias Villagra

We propose a disciplined, numerically stable, and scalable approach to SDP relaxations of the ACOPF problem based on linear cutting-planes. Our method can be warm-started and, owing to its linear nature, enables the computation of tight and accurate bounds for large-scale multi-period relaxations -- well beyond what nonlinear convex solvers can achieve. Prel

Nir Goren, Shai Yehezkel, Omer Dahary, Andrey Voynov

In this paper we show that visual diffusion models can serve as effective geometric solvers: they can directly reason about geometric problems by working in pixel space. We first demonstrate this on the Inscribed Square Problem, a long-standing problem in geometry that asks whether every Jordan curve contains four points forming a square. We then extend the

Han Yan, Xibin Song, Yifu Wang, Hongdong Li

Diffusion Transformers (DiTs) have recently driven significant progress in text-to-video (T2V) generation. However, generating multiple videos with consistent characters and backgrounds remains a significant challenge. Existing methods typically rely on reference images or extensive training, and often only address character consistency, leaving background c

Edward Holmberg, Elias Ioup, Mahdi Abdelguerfi

The coordination of autonomous agents in dynamic environments is hampered by the semantic gap between high-level mission objectives and low-level planner inputs. To address this, we introduce a framework centered on a Knowledge Graph (KG) that functions as an intelligent translation layer. The KG's two-plane architecture compiles declarative facts into per-a

Stefan Taubenberger, Ana Acebron, Raoul Cañameras, Ting-Wan Chen

We present imaging and spectroscopic observations of supernova SN 2025wny, associated with the lens candidate PS1 J0716+3821. Photometric monitoring from the Lulin and Maidanak observatories confirms multiple point-like images, consistent with SN 2025wny being strongly lensed by two foreground galaxies. Optical spectroscopy of the brightest image with the No

Reuben Narad, Leonard Boussioux, Michael Wagner

Neural networks have advanced combinatorial optimization, with Transformer-based solvers achieving near-optimal solutions on the Traveling Salesman Problem (TSP) in milliseconds. However, these models operate as black boxes, providing no insight into the geometric patterns they learn or the heuristics they employ during tour construction. We address this opa

Edward Berman, Jacob Ginesin, Marco Pacini, Robin Walters

Data-sparse settings such as robotic manipulation, molecular physics, and galaxy morphology classification are some of the hardest domains for deep learning. For these problems, equivariant networks can help improve modeling across undersampled parts of the input space, and uncertainty estimation can guard against overconfidence. However, until now, the rela

P. M. Shober, J. Vaubaillon, S. Anghel, H. A. R. Devillepoix

Methods. We processed a dataset of 584 Geminid fireballs observed by FRIPON between 2016 and 2023. The single-station astrometric data is converted into the Global Fireball Exchange (GFE) standard format for uniform processing. We assess variations in trajectory, velocity, radiant, and orbital element calculations across the pipelines and compare them to pre

Jiayi Zhou, Günel Aghakishiyeva, Saagar Arya, Julian Dale

Computer vision can accelerate ecological research and conservation monitoring, yet adoption in ecology lags in part because of a lack of trust in black-box neural-network-based models. We seek to address this challenge by applying post-hoc explanations to provide evidence for predictions and document limitations that are important to field deployment. Using

Bogdan-Vasile Matioc, Christoph Walker

The inverse problem of reconstructing the initial state in quasilinear parabolic equations from time averages is investigated. Under suitable regularity assumptions on the quasilinear structure and a superlinear growth condition near zero for the semilinear part, it is shown that the initial state can be uniquely recovered from small time averages taken over

Raheem Karim Hashmani, Garrett W. Merz, Helen Qu, Mariel Pettee

We introduce a framework for generating highly multimodal datasets with explicitly calculable mutual information (MI) between modalities. This enables the construction of benchmark datasets that provide a novel testbed for systematic studies of mutual information estimators and multimodal self-supervised learning (SSL) techniques. Our framework constructs re

Teona Bagashvili, Tarikul Islam Papon, Subhadeep Sarkar, Manos Athanassoulis

Zoned Namespace (ZNS) SSDs offer a new storage model that allows for high throughput and low-latency storage by eliminating device-side garbage collection. The ZNS interface exposes storage as append-only zones, thus enforcing host applications (e.g., database systems) to append, read, and garbage collect their pages. However, the storage abstraction of ZNS

Jingyue Huang, Qihui Yang, Fei Yueh Chen, Julian McAuley

Existing pitch curve generators face two main challenges: they often neglect singer-specific expressiveness, reducing their ability to capture individual singing styles. And they are typically developed as auxiliary modules for specific tasks such as pitch correction, singing voice synthesis, or voice conversion, which restricts their generalization capabili

Albert Cheu, Artem Lagzdin, Brett McLarnon, Daniel Ramage

Large-scale systems that compute analytics over a fleet of devices must achieve high privacy and security standards while also meeting data quality, usability, and resource efficiency expectations. We present a next-generation federated analytics system that uses Trusted Execution Environments (TEEs) based on technologies like AMD SEV-SNP and Intel TDX to pr

Simon Butson, Mathew Cleveland, Alex Long, Todd Palmer

This work demonstrates algorithms to accurately compute solutions to thermal radiation transport problems using a reduced floating-point precision implementation of the Implicit Monte Carlo method. Several techniques falling into the categories of arithmetic manipulations and scaling methods are evaluated for their ability to improve the accuracy of reduced-

Sikuang Li, Chen Yang, Jiemin Fang, Taoran Yi

We tackle the challenge of generating the infinitely extendable 3D world -- large, continuous environments with coherent geometry and realistic appearance. Existing methods face key challenges: 2D-lifting approaches suffer from geometric and appearance inconsistencies across views, 3D implicit representations are hard to scale up, and current 3D foundation m

Mariano Devoto, Pablo Ariel Cipriotti

After major disasters, formal inquiries become arenas where responsibility is publicly contested. While extensive research has examined blame attribution through qualitative and actor-centred approaches, the relational structure of blame within formal accountability processes remains poorly understood. Using evidence from the Grenfell Tower Inquiry, this stu

Quentin Nicolas, Geoffrey K. Vallis

Superrotation is a common feature of quickly rotating gas giants, slowly rotating planetary bodies, and tidally-locked planets. In this paper we compare and contrast the mechanisms of superrotation in slow rotators and tidally-locked planets. We cover a wide range of planetary properties, varying in particular the thermal Rossby number Ro_T (controlled by pl

Marko Orešković, Ivana Kuzmanović Ivičić, Juraj Benić, Mario Essert

This paper presents a unified algebraic, topological, and logical framework for electrical one-port networks based on \v{S}are's $m$-theory. Within this formalism, networks are represented by $m$-words (jorbs) over an ordered alphabet, where series and parallel composition induce an $m$-topology on $m$-graphs with a theta mapping $\vartheta$ that preserves o

Mahavir Dabas, Tran Huynh, Nikhil Reddy Billa, Jiachen T. Wang

Large language models remain vulnerable to jailbreak attacks that bypass safety guardrails to elicit harmful outputs. Defending against novel jailbreaks represents a critical challenge in AI safety. Adversarial training -- designed to make models robust against worst-case perturbations -- has been the dominant paradigm for adversarial robustness. However, du

Catherine Arnett, Tyler A. Chang, Stella Biderman, Benjamin K. Bergen

The number of tokens it takes to encode parallel text in different languages is known to vary. These disparities are called token premiums. Having high token premiums leads to less throughput during training and increases costs at inference. In this paper, we show that even after controlling for dataset size, vocabulary size, and data content, monolingual to

Gaku Morio, Harri Rowlands, Dominik Stammbach, Christopher D. Manning

Companies spend large amounts of money on public relations campaigns to project a positive brand image. However, sometimes there is a mismatch between what they say and what they do. Oil & gas companies, for example, are accused of "greenwashing" with imagery of climate-friendly initiatives. Understanding the framing, and changes in framing, at scale can hel

Nabamita Banerjee, Vedant Bhutra, Suvankar Dutta, Soumava Kundu

We extend the Kac-Moody (KM) boundary conditions of AdS$_3$ gravity by incorporating fermionic fields. For $\mathcal{N}=(1,1)$ AdS$_3$ supergravity, we show that there are two possible ways to implement the fermionic extension. In the first, the extended KM boundary conditions are related to the standard super-Virasoro (VS) boundary conditions through a larg

Ying Wang

We solve a non-Archimedean Monge-Amp\`ere equation on the Berkovich analytification of a complex log Calabi-Yau pair whose dual complex is a standard simplex, answering a question of Collins-Li and offering a non-Archimedean analog of Ricci-flat metric potentials on complex affine varieties. This work builds on the solution to a complex Monge-Amp\`ere equati

Akash Shukla, Ammar Hakim, James Juno, Gregory Hammett

Structures in tokamak plasmas are elongated along the direction of the magnetic field and short in the directions perpendicular to the magnetic field. Many tokamak simulation codes take advantage of this by using a field aligned coordinate system. However, field aligned coordinate systems have a coordinate singularity at magnetic X-points where the poloidal

Michael C. Sitarz

Energetic astrophysical phenomena, such as $\gamma$-ray bursts and supernova explosion-driven shocks in collisionless plasmas, involve various plasma kinetic instabilities, such as the Weibel instability. These systems support various types of particle acceleration and radiation through a variety of mechanisms. In this paper, we explore the energy transforma

Z. Zeng, M. Först, M. Fechner, X. Deng

Chirality is a pervasive property of matter that underpins many important phenomena across physics, chemistry and biology. Given its broad significance, the development of protocols for rational control of chirality in solid state systems is highly desirable, especially if this effect can be tuned continuously and in two directions. Yet, this goal has remain

C. A. Downing, M. S. Ukhtary

In the quantum world, the process of energy storage can be enhanced thanks to various nonclassical phenomena. This inspiring fact suggests quantum batteries as plausible sources of power for future quantum devices, at least in principle. However, thermodynamically not all of the energy stored in a quantum battery is useful for doing work. By considering a cl

Yabo Yin, Yang Xi, Jialong Wang, Shanqi Wang

Multilingual e-commerce search suffers from severe data imbalance across languages, label noise, and limited supervision for low-resource languages--challenges that impede the cross-lingual generalization of relevance models despite the strong capabilities of large language models (LLMs). In this work, we present a practical, architecture-agnostic, data-cent

Zihao Jing, Yan Sun, Yan Yi Li, Sugitha Janarthanan

Multimodal molecular models often suffer from 3D conformer unreliability and modality collapse, limiting their robustness and generalization. We propose MuMo, a structured multimodal fusion framework that addresses these challenges in molecular representation through two key strategies. To reduce the instability of conformer-dependent fusion, we design a Str

Julien Brémont

The first-passage time (FPT) of a stochastic signal to a threshold is a fundamental observable across physics, biology, and finance. While renewal shot noise is a canonical model for such signals, analytical results for its FPT have remained confined to the Poisson (Markovian) case, despite the prevalence of non-Poisson arrival statistics in applications fro

Siddharth Chaudhary

Large language models display in-context learning as an emergent effect of scale, but they rely on static weights during inference. In contrast, biological systems continually adapt via synaptic plasticity. We investigate whether explicit, biologically inspired plasticity can endow Transformers with faster in-sequence adaptation. To this end, we augment deco

Maximilien Dreveton, Elaine Siyu Liu, Matthias Grossglauser, Patrick Thiran

This paper establishes the theoretical limits of graph clustering under the Popularity-Adjusted Block Model (PABM), addressing limitations of existing models. In contrast to the Stochastic Block Model (SBM), which assumes uniform vertex degrees, and to the Degree-Corrected Block Model (DCBM), which applies uniform degree corrections across clusters, PABM int

Qihui Yang, Randal Leistikow, Yongyi Zang

Virtual instrument generation requires maintaining consistent timbre across different pitches and velocities, a challenge that existing note-level models struggle to address. We present FlowSynth, which combines distributional flow matching (DFM) with test-time optimization for high-quality instrument synthesis. Unlike standard flow matching that learns dete

Michael C. Sitarz

Energetic astrophysical phenomena, such as $\gamma$-ray bursts, supernova explosions, and magnetar flares occur in collisionless plasmas and involve various plasma kinetic and magnetohydrodynamic instabilities. In this paper, we explore the spectral trends of the Weibel instability using spectral analysis of particle-in-cell simulations. Power dependence on

Chinmoy Bhattacharjee, Anna Gusakova

We consider Gaussian approximation in three particular models of Poisson-Laguerre tessellations, namely, the $\beta$-, $\beta'$- and Gaussian-Voronoi tessellations. The tessellations are constructed based on inhomogeneous Poisson point processes in space-time $\mathbb{R}^d \times \mathbb{R}$, where some of the points of the process give rise to a cell in $\m

Debdeep Sanyal, Manya Pandey, Dhruv Kumar, Saurabh Deshpande

Large language models (LLMs) often exhibit a puzzling disconnect between their asserted confidence and actual problem-solving competence. We offer a mechanistic account of this decoupling by analyzing the geometry of internal states across two phases - pre-generative assessment and solution execution. A simple linear probe decodes the internal "solvability b

Riya Gupta, Yiwei Zong, Dennis H. Murphree

The rapid generation of whole-slide images (WSIs) in dermatopathology necessitates automated methods for efficient processing and accurate classification. This study evaluates the performance of two foundation models, UNI and Virchow2, as feature extractors for classifying WSIs into three diagnostic categories: melanocytic, basaloid, and squamous lesions. Pa

Aarav Shetty, Gary B Huang

Separating synapses into different classes based on their appearance in EM images has many applications in biology. Examples may include assigning a neurotransmitter to a particular class, or separating synapses whose strength can be modulated from those whose strength is fixed. Traditionally, this has been done in a supervised manner, giving the classificat

Deepika Garg, Maxim Olshanskii

This paper addresses the analysis and numerical assessment of a computational method for solving the Cahn--Hilliard equation defined on a surface. The proposed approach combines the stabilized trace finite element method for spatial discretization with an implicit--explicit scheme for temporal discretization. The method belongs to a class of unfitted finite

Heyang Ji, Ufuk Beyaztas, Nicolas Escobar-Velasquez, Yuanyuan Luan

Functional data analysis (FDA) deals with high-resolution data recorded over a continuum, such as time, space or frequency. Device-based assessments of physical activity or sleep are objective yet still prone to measurement error. We present MECfda, an R package that (i) fits scalar-on-function, generalized scalar-on-function, and functional quantile regress

Leander Claes, Michael Winkler

In bounded $n$-dimensonal domains with $n\ge 1$, this manuscript examines an initial-boundary value problem for the system \[ \left\{ \begin{array}{l} u_{tt} = \nabla \cdot (\gamma(\Theta) \nabla u_t) + a \nabla \cdot (\gamma(\Theta) \nabla u) + \nabla\cdot f(\Theta), \Theta_t = D\Delta\Theta + \Gamma(\Theta) |\nabla u_t|^2 + F(\Theta)\cdot \nabla u_t, \end{

Yongyi Zang, Chris Manchester, David Young, Ivan Ivanov

Vocal recordings on consumer devices commonly suffer from multiple concurrent degradations: noise, reverberation, band-limiting, and clipping. We present Smule Renaissance Small (SRS), a compact single-stage model that performs end-to-end vocal restoration directly in the complex STFT domain. By incorporating phase-aware losses, SRS enables large analysis wi

Alexander Lai De Oliveira

We use Lazard's universal $\pi$-ring to construct a variation of the $\pi$-typical ramified Witt vector functors, which we call the Lazardian Witt vector functor. We then use the Lazardian Witt vector functor to construct the universal residual perfection of a Lazardian algebra.