October 2025 arXiv papers — page 168

Hongyi Guan, Ananya Renuka Balakrishna

We introduce CuPyMag, an open-source, Python-based framework for large-scale micromagnetic simulations with magnetostriction. CuPyMag solves micromagnetics with finite elements in a GPU-resident workflow in which key operations, such as right-hand-side assembly, spatial derivatives, and volume averages, are tensorized using CuPy's BLAS-accelerated backend. B

Pavel Pyatov, Oleg Ogievetsky

For the family of the orthogonal quantum matrix algebras we investigate the structure of their characteristic subalgebras -- special commutative subalgebras, which for the subfamily of the reflection equation algebras appear to be central. In [OP1] we described three generating sets of the characteristic subalgebras of the symplectic and orthogonal quantum m

Romain Allart, Louis-Philippe Coulombe, Yann Carteret, Jared Splinter

Atmospheric escape of planets on short orbital periods, driven by the host star's irradiation, influences their evolution, composition, and atmospheric dynamics. Our main avenue to probe atmospheric escape is through the near-infrared metastable helium triplet, which has enabled mass loss rate measurements for tens of exoplanets. Among them, only a few studi

Hongzheng Shi, Yuhang Wang, Xiao Liu

As wildfires become increasingly destructive and expensive to control, effective management of active wildfires requires accurate, real-time fire spread predictions. To enhance the forecasting accuracy of active fires, data assimilation plays a vital role by integrating observations (such as remote-sensing data) and fire predictions generated from numerical

Marko Lela

This paper develops a compact, size-aware blueprint for transferring structure through gadget lifts. Two low-order invariants -- cumulative mod-$q$ Fourier mass up to degree $k$ and noise stability $\mathrm{Stab}_\rho$ -- are treated as a reusable "profile" tied to the gadget's affine interface. Under coordinate permutations ($\Delta=1$) the profile is prese

R. Thapa, E. Mustermann, H. Jain, V. Dierolf

Rotation of crystal seed during the early stages of growth in a glass matrix has been observed due to some torque, contradicting the expectations from the isotropic, uniform structure of the surrounding amorphous matrix. We establish an atomistic origin of this new phenomenon from molecular dynamics simulations using LiNbO3 and LiNbO3-SiO2 glasses as model s

Louis Desdoigts, Benjamin Pope, Max Charles, Peter Tuthill

The James Webb Space Telescope (JWST) hosts a non-redundant Aperture Masking Interferometer (AMI) in its Near Infrared Imager and Slitless Spectrograph (NIRISS) instrument, providing the only dedicated interferometric facility aboard - magnitudes more precise than any interferometric experiment previously flown. However, the performance of AMI (and other hig

Jeffrey Camlin

We present a latent-space formulation of adaptive temporal lifting for continuous-time dynamical systems. The method introduces a smooth monotone mapping $t \mapsto \tau(t)$ that regularizes near-singular behavior of the underlying flow while preserving its conservation laws. In the lifted coordinate, trajectories such as those of the incompressible Navier-S

Jesse G. Meyer

Building custom data analysis platforms traditionally requires extensive software engineering expertise, limiting accessibility for many researchers. Here, I demonstrate that modern large language models (LLMs) and autonomous coding agents can dramatically lower this barrier through a process called 'vibe coding', an iterative, conversational style of softwa

Md Tusher Ahmed, Farid Ahmed, Jianzhi Li

Tungsten (W) is widely valued for its exceptional thermal stability, mechanical strength, and corrosion resistance, making it an ideal candidate for high-performance military and aerospace applications. However, its high melting point and limited room-temperature plasticity pose significant challenges for processing W using additive manufacturing (AM). Cold

Gracyn Jewett, Mukremin Kilic, Adam Moss, Alejandro H. Córsico

We present time-series photometry of 31 massive DA white dwarfs with $M\gtrsim 0.9~M_\odot$ within the ZZ Ceti instability strip from the Montreal White Dwarf Database 100 pc sample. The majority of the targets had no previous time-series photometry available, though several were classified as non-variable or potential pulsators in the literature. Out of the

Lucas Wang

Let $g(k)$ be the maximum size of a planar set that determines at most $k$ distances. We prove $$\fracπ{3\,C(Λ_{hex})}\ k\sqrt{\log k} (1+o(1)) \le g(k) \le C k\log k,$$ so $g(k) \asymp k\sqrt{\log k}$ with an explicit constant from the hexagonal lattice. For any arithmetic lattice $Λ$ we show $$g_Λ(k)\ge (π/4) S^*(Λ) k\sqrt{\log k} (1+o(1)).$$ We also give

Alessio Maritan, Luca Schenato

We introduce and address a novel distributed clustering problem where each participant has a private dataset containing only a subset of all available features, and some features are included in multiple datasets. This scenario occurs in many real-world applications, such as in healthcare, where different institutions have complementary data on similar patie

Jordan T. McCourt, John Chiles, Chun-Chia Chen, Kenji Watanabe

The interfaces of quantum Hall insulators with superconductors have emerged as a promising platform to realise interesting physics that may be relevant for topologically protected quantum computing. However, these interfaces can host other effects which obscure the detection of the desired excitations. Here we present measurements of the thermoelectric effec

Elettra L. Piacentino, Aurelia Balkanski, Jenny Calahan, Anna Fitzsimmons

Aromaticity is a common chemical functionalities in bioactive molecules. In interstellar and circumstellar environments benzene and other small aromatics are considered the precursor for more complex prebiotic molecules and they have shown to potentially have rich ice-phase photochemistry. The availability of small organic molecules in prebiotic networks dep

Xiaoyu Wang, Alexandra Valavanis, Azhir Mahmood, Andreas Mang

The training of deep neural networks predominantly relies on a combination of gradient-based optimisation and back-propagation for the computation of the gradient. While incredibly successful, this approach faces challenges such as vanishing or exploding gradients, difficulties with non-smooth activations, and an inherently sequential structure that limits p

Fanfan Meng, Emma Van Meirvenne, Federica Luciano, Xiangyu Wu

Strain-mediated magnetoelectric (ME) heterostructures enable electric-field control of magnetism and are promising for ultra-low-power spintronic logic. Yet achieving spatially selective, low-voltage control in thin films and quantifying ME coupling across the full ferroelastic loop remains challenging. Here, we investigate sub-micrometer Ni/BPZT thin-film d

Lianghuan Huang, Yingshan Chang

Mechanistic interpretability seeks to uncover how internal components of neural networks give rise to predictions. A persistent challenge, however, is disentangling two often conflated notions: decodability--the recoverability of information from hidden states--and causality--the extent to which those states functionally influence outputs. In this work, we i

Kyle A. Hamer, Heman Gharibnejad, Luca Argenti, Nicolas Douguet

We present a time-dependent framework that combines a hybrid Gaussian-FEDVR basis with a multicenter grid to simulate strong-field and attosecond dynamics in atoms and molecules. The method incorporates the construction of the orthonormal hybrid basis, the evaluation of electronic integrals, a unitary time-propagation scheme, and the extraction of optical an

Vahidreza Jahanmard, Ali Ramezani-Kebrya, Robinson Hordoir

Neural operators are becoming the default tools to learn solutions to governing partial differential equations (PDEs) in weather and ocean forecasting applications. Despite early promising achievements, significant challenges remain, including long-term prediction stability and adherence to physical laws, particularly for high-frequency processes. In this pa

Sneha Gathani, Kevin Li, Raghav Thind, Sirui Zeng

What-if analysis is widely used to explore hypothetical scenarios and evaluate alternative pathways to desired results. However, current approaches are fragmented: systems implement what-if capabilities under diverse terminologies with different analytic techniques. Such fragmentation limits expressiveness, impedes flexible composition and reuse of workflows

Michael Freenor, Lauren Alvarez

Understanding how language and embedding models encode semantic relationships is fundamental to model interpretability. While early word embeddings exhibited intuitive vector arithmetic (''king'' - ''man'' + ''woman'' = ''queen''), modern high-dimensional text representations lack straightforward interpretable geometric properties. We introduce Rotor-Invaria

João Victor Lemos Vale, Lucas Cesena, Bruno H. S. Mendonça, Elizane E. de Moraes

Efficient water transport through nanostructure membranes is essential for advancing filtration and desalination technologies. In this study, we investigate the flow of water through molybdenum disulfide (MoS$_{2}$) nanopores of varying diameters using molecular dynamics simulations. The results demonstrate that both pore size and atomic edge composition pla

Kristaps John Balodis

In this article, we prove the $p$-adic Kazhdan-Lusztig hypothesis for $\mathrm{GL}_n(F)$. While the approach via graded affine Hecke algebras due to recent work of Solleveld leads to more general results, this article serves to completes and clarifies the approach via affine Hecke algebras of Chriss and Ginzburg. In particular, this article serves as an oppo

Yuang Lu, Song Wang, Xiao Han, Xuri Zhang

Temporal sequential tasks challenge humanoid robots, as existing Diffusion Policy (DP) and Action Chunking with Transformers (ACT) methods often lack temporal context, resulting in local optima traps and excessive repetitive actions. To address these issues, this paper introduces a Classifier-Free Guidance-Based Diffusion Policy (CFG-DP), a novel framework t

Vladimír Holý

We address the challenges of modeling high-frequency integer price changes in financial markets using continuous distributions, particularly the Student's t-distribution. We demonstrate that traditional GARCH models, which rely on continuous distributions, are ill-suited for high-frequency data due to the discreteness of price changes. We propose a modificat

Richard John, Yunrui Qiu, Lukas Herron, Pratyush Tiwary

Generative modeling becomes increasingly data-intensive in high-dimensional spaces. In molecular science, where data collection is expensive and important events are rare, compression to lower-dimensional manifolds is especially important for various downstream tasks, including generation. We combine a time-lagged information bottleneck designed to character

Yufa Zhou, Yixiao Wang, Xunjian Yin, Shuyan Zhou

We study how large language models (LLMs) ``think'' through their representation space. We propose a novel geometric framework that models an LLM's reasoning as flows -- embedding trajectories evolving where logic goes. We disentangle logical structure from semantics by employing the same natural deduction propositions with varied semantic carriers, allowing

Simone Carnemolla, Matteo Pennisi, Chiara Russo, Simone Palazzo

We introduce SeeingSounds, a lightweight and modular framework for audio-to-image generation that leverages the interplay between audio, language, and vision-without requiring any paired audio-visual data or training on visual generative models. Rather than treating audio as a substitute for text or relying solely on audio-to-text mappings, our method perfor

Yue Huang, Hang Hua, Yujun Zhou, Pengcheng Jing

While LLM agents can plan multi-step tasks, intervening at the planning stage-before any action is executed-is often the safest way to prevent harm, since certain risks can lead to severe consequences once carried out. However, existing guardrails mostly operate post-execution, which is difficult to scale and leaves little room for controllable supervision a

ChengAo Shen, Ziming Zhao, Hanghang Tong, Dongjin Song

Time series AI is crucial for analyzing dynamic web content, driving a surge of pre-trained large models known for their strong knowledge encoding and transfer capabilities across diverse tasks. However, given their energy-intensive training, inference, and hardware demands, using large models as a one-fits-all solution raises serious concerns about carbon f

Thomas Gschwind, Shramona Chakraborty, Nitin Gupta, Sameep Mehta

ETL (Extract, Transform, Load) tools such as IBM DataStage allow users to visually assemble complex data workflows, but configuring stages and their properties remains time consuming and requires deep tool knowledge. We propose a system that translates natural language descriptions into executable workflows, automatically predicting both the structure and de

Cinzia Bisi, Antonio Carbone

The purpose of this paper is to introduce the notion of Nash functions in the context of slice regular functions of one quaternionic or octonionic variable. We begin with a detailed analysis of the possible definitions of Nash slice regular functions which leads us to the definition of \textit{slice-Nash} function proposed in this paper (and which we strongl

Ilya Kosolapov, Tatiana Sheloput, Sergey Matveev

In this work we investigate efficient data compression for spatiotemporal Black, Azov and Marmara Seas temperature tensors that contain significant number of missing values. These tensors have a complex structure influenced by the coastlines and bathymetry, as well as temporal temperature changes. While such missing data typically provokes utilization of ten

Michael J. Desrochers, Dominic Marchand, P. C. E. Stamp

At low temperature T we expect vacuum tunneling processes to occur in superfluid $^{4}$ He films. We distinguish between extrinsic processes, in which single vortices nucleate by tunneling off boundaries in the system, and intrinsic processes, in which vortex/anti-vortex pairs nucleate far from boundaries. It is crucial to incorporate the varying effective m

Tycho J. Blom, Matthijs Rog, Marieke Altena, Andrea Capa Salinas

Materials with a Kagome lattice are intensely studied because they host exotic states that combine strong correlations and topology. Recently, critical current oscillations were observed in an unstructured flake of CsV3Sb5 . In this work, we show that the origin of these oscillations is a network of Josephson junctions intrinsic to the flake that emerges bel

Yufa Zhou, Yixiao Wang, Surbhi Goel, Anru R. Zhang

Time series forecasting (TSF) remains a challenging and largely unsolved problem in machine learning, despite significant recent efforts leveraging Large Language Models (LLMs), which predominantly rely on Transformer architectures. Empirical evidence consistently shows that even powerful Transformers often fail to outperform much simpler models, e.g., linea

Alex Hiles, Bashar I. Ahmad

Fingerprinting radio frequency (RF) emitters typically involves finding unique characteristics that are featured in their received signal. These fingerprints are nuanced, but sufficiently detailed, motivating the pursuit of methods that can successfully extract them. The downstream task that requires the most meticulous RF fingerprinting (RFF) is known as sp

Raed Diab, Alvaro Herrera, Chance Jackson, Paul Fulda

Precise laser alignment in optical cavities is essential for high-precision laser interferometry. We report on a table-top optical experiment featuring two alignment sensing schemes: the conventional Wavefront Sensing (WFS) scheme which uses quadrant photodetectors (QPDs) to recover optical alignment, and the newly developed Radio Frequency Jitter Alignment

Nora Basha, Bechir Hamdaoui, Attila A. Yavuz, Thang Hoang

Secret-key generation and agreement based on wireless channel reciprocity offers a promising avenue for securing IoT networks. However, existing approaches predominantly rely on the similarity of instantaneous channel measurement samples between communicating devices. This narrow view of reciprocity is often impractical, as it is highly susceptible to noise,

J. Bätz, M. Mugrauer, K. -U. Michel, J. Reichert

We present new radial velocity measurements of 13 selected intermediate mass stars (2 - 6 M$_\odot$). The measurements were performed between 29 April and 6 September 2024 at the University Observatory Jena using the \'echelle spectrograph FLECHAS. The radial velocity of eight stars was found to be constant during our spectroscopic monitoring, namely: 17 Dra

Rakib Hossan, Shubhashis Roy Dipta

The BLP-2025 Task 1A requires Bengali hate speech classification into six categories. Traditional supervised approaches need extensive labeled datasets that are expensive for low-resource languages. We developed PromptGuard, a few-shot framework combining chi-square statistical analysis for keyword extraction with adaptive majority voting for decision-making

Adam Byerly, Daniel Khashabi

Large language models (LLMs) exhibit pronounced position bias in long-context needle-in-haystack problems, systematically prioritizing the location of information over its relevance. While current mitigations rely on white-box access, this is effectively impossible for many state-of-the-art models. We introduce GOLD PANNING, a black-box Bayesian framework th

Gabriel Currier

We present some new sharp constructions for the Szemer\'{e}di-Trotter theorem. These constructions generalize previous work of Erd\H{o}s, Elekes, Sheffer and Silier, Guth and Silier, and the author. In the past, arguments showing the optimality of many of these constructions have required some elementary number theory and have been rather technical, thus lim

Khang Ngo, Siamak Ravanbakhsh

We present an empirical study in the geometric task of learning interatomic potentials, which shows equivariance matters even more at larger scales; we show a clear power-law scaling behaviour with respect to data, parameters and compute with ``architecture-dependent exponents''. In particular, we observe that equivariant architectures, which leverage task s

Sisir Doppalapudi

Multi-modal threat detection faces a fundamental challenge that involves security tools operating in isolation, and this creates streams of network, email, and system data with no natural alignment or correlation. We present Hierarchical Multi-Modal Threat Intelligence Fusion (HM-TIF), a framework explicitly designed for this realistic scenario where natural

Yifan Lu, Ziyun Zou, Belal Alsinglawi, Islam Al-Qudah

Graph Transformers have recently achieved remarkable progress in graph representation learning by capturing long-range dependencies through self-attention. However, their quadratic computational complexity and inability to effectively model heterogeneous semantics severely limit their scalability and generalization on real-world heterogeneous graphs. To addr

Mehdi Rezaei, Supriya Pan, Weiqiang Yang, David F. Mota

This study explores the possibility of a time-varying dark energy (DE) equation of state (EoS) deviating from -1. We employ a comprehensive dataset of usual astronomical probes (Type Ia supernovae, baryon acoustic oscillations, Big Bang nucleosynthesis, Hubble data, and Planck 2018 CMB) alongside future mock gravitational wave (GW) distance measurements from

Bhanu Pratap Yadav, Mahdi Bayanifar, Olav Tirkkonen

We consider a global phase-invariant metric in the projective unitary group PUn, relevant for universal quantum computing. We obtain the volume and measure of small metric ball in PUn and derive the Gilbert-Varshamov and Hamming bounds in PUn. In addition, we provide upper and lower bounds for the kissing radius of the codebooks in PUn as a function of the m

Donghan Hu, Rameen Mahmood, Annabelle David, Danny Yuxing Huang

AI-driven applications have become woven into students' academic and creative workflows, influencing how they learn, write, and produce ideas. Gaining a nuanced understanding of these usage patterns is essential, yet conventional survey and interview methods remain limited by recall bias, self-presentation effects, and the underreporting of habitual behavior

Ruo Yang, Sai Krishna Reddy Mudhiganti, Manali Sharma

Patent drafting is complex due to its need for detailed technical descriptions, legal compliance, and visual elements. Although Large Vision Language Models (LVLMs) show promise across various tasks, their application in automating patent writing remains underexplored. In this paper, we present PatentVision, a multimodal framework that integrates textual and

Mariia Marinichenko, Marcel P. van Daalen, Elena Sellentin, Jeger C. Broxterman

The scattering transform is a wavelet-based statistic capable of capturing non-Gaussian features in weak lensing (WL) convergence maps and has been proven to tighten cosmological parameter constraints by accessing information beyond two-point functions. However, its application in cosmological inference requires a clear understanding of its sensitivity to as

Anna Ordog, Rebecca A. Booth, T. L. Landecker, Ettore Carretti

Polarized synchrotron emission at meter to centimeter wavelengths provides an effective tracer of the Galactic magnetic field. Calculating Faraday depth, the most useful parameter for mapping the line-of-sight magnetic field, requires observations covering wide frequency bands with many channels. As part of the Global Magneto-Ionic Medium Survey (GMIMS), we

Hsuan-Kai Huang, Joseph Kuo, Seonyeong Park, Umberto Villa

Transcranial photoacoustic computed tomography (PACT) is an emerging neuroimaging modality, but skull-induced aberrations can result in severe image artifacts if not compensated for during image reconstruction. The development of advanced image reconstruction methods for transcranial PACT is hindered by the lack of well-characterized, clinically relevant eva

Brian Aevermann, Andrea Califano, Chi-Li Chiu, Nathan Clack

Biology is at the precipice of a new era where AI accelerates and amplifies the ability to study how cells operate, organize, and work as systems, revealing why disease happens and how to correct it. Organizations globally are prioritizing AI to accelerate basic research, drug discovery, personalized medicine, and synthetic biology. However, despite these op

Myung Ho Kim

Large language models exhibit intelligence without genuine epistemic understanding, exposing a key gap: the absence of epistemic architecture. This paper introduces the Structured Cognitive Loop (SCL) as an executable epistemological framework for emergent intelligence. Unlike traditional AI research asking "what is intelligence?" (ontological), SCL asks "un

Philip F. Hopkins, Elias R. Most

Many astrophysical simulations involve extreme dynamic range of timescales around 'special points' in the domain (e.g. black holes, stars, planets, disks, galaxies, shocks, mixing interfaces), where processes on small scales couple strongly to those on large scales. Adaptive resolution, multi-physics, and hybrid numerical methods have enabled tremendous prog

Kaitlyn Zhou, Kristina Gligorić, Myra Cheng, Michelle S. Lam

Although language model-based chat systems are increasingly used in daily life, most Americans remain non-adopters of chat-based LLMs -- as of June 2025, 66% had never used ChatGPT. At the same time, LLM development and evaluation rely mainly on data from adopters (e.g., logs, preference data), focusing on the needs and tasks for a limited demographic group

I. Flores, M. Mezcua, V. Rodríguez Morales

Low-mass black holes hosted by dwarf galaxies offer valuable insights into galaxy formation and the growth of the massive black holes found in massive galaxies. Their detection as AGN is challenging due to their low luminosity and compact size. This can be circumvented employing multi-wavelength observational strategies, such as combining optical and radio o

Dominika Ďurovčíková, Anna-Christina Eilers, Yuzo Ishikawa, Minghao Yue

Measurements of quasar lifetimes at high redshift indicate that the earliest billion-solar-mass supermassive black holes (SMBHs) have only been active as luminous quasars for less than a million years. Recently, extended Ly$\alpha$ nebulae around $z\sim6$ quasars have revealed that these short observed lifetimes are unlikely a sightline-dependent effect. How

Sai Krishna Reddy Mudhiganti, Juanyan Wang, Ruo Yang, Manali Sharma

Patent drafting presents significant challenges due to its reliance on the extensive experience and specialized expertise of patent attorneys, who must possess both legal acumen and technical understanding of an invention to craft patent applications in a formal legal writing style. This paper presents a demonstration of Patentformer, an AI-powered automated

Jason A. S. Hunt, Michael S. Petersen, Martin D. Weinberg, Kathryn V. Johnston

The Milky Way is known to contain a stellar bar, as are a significant fraction of disc galaxies across the universe. Our understanding of bar evolution, both theoretically and through analysis of simulations indicates that bars both grow in amplitude and slow down over time through interaction and angular momentum exchange with the galaxy's dark matter halo.

Michael M. Wyatt, Steven R. Furlanetto, Mary H. Minasyan

Recent measurements of the mean free path (MFP) of ionizing photons at $z=6$ find that it is significantly shorter than extrapolations from lower $z$. This has a substantial impact on the topology of reionization and thus the prospects of tomography of the 21-cm signal from upcoming radio interferometers. In this work we develop the first analytic model of r

Jeffrey Z. Song, Gilad Kishony, Erez Berg, Mark S. Rudner

We introduce a variational approach for preparing low energy states of arbitrary target Hamiltonians. The protocol is defined in terms of a repeated cycle consisting of p layers of unitary gates applied to the system and ancilla "bath" qubits, followed by reset of the bath qubits. The gate parameters within each cycle are optimized such that the steady state

Michelle Ntampaka, A. Ciprijanovic, Ana Maria Delgado, John Soltis

The application of deep machine learning methods in astronomy has exploded in the last decade, with new models showing remarkably improved performance on benchmark tasks. Not nearly enough attention is given to understanding the models' robustness, especially when the test data are systematically different from the training data, or "out of domain." Domain s

Pravita Hallur, Lia Medeiros, Pierre Christian, George N. Wong

We characterize the spatial power spectrum of density fluctuations in magnetohydrodynamic flows in a suite of high-resolution, long-time-span general relativistic magnetohydrodynamic (GRMHD) simulations. Extracting the local spatial power spectrum in curved spacetime directly from GRMHD simulations can be challenging for several conceptual and mechanical rea

Genevieve Schroeder, Ben Margalit, Brian D. Metzger, Wen-fai Fong

In addition to a $\gamma$-ray burst (GRB), the merger of two neutron stars may produce a temporarily or indefinitely stable neutron star remnant with a strong magnetic field (a "magnetar"). As this magnetar remnant spins down, it can deposit its rotational energy into the surrounding kilonova ejecta, producing synchrotron emission that peaks in the radio ban

Lazaros Souvaitzis, Antti Rantala, Thorsten Naab

The most massive galaxies in the Universe also host the largest supermassive black holes (SMBHs), with masses of $10^9 \: \mathrm{M_{\odot}}$ and above. During their hierarchical assembly, these galaxies have experienced only a few major mergers at low redshift, but have accreted many low-mass galaxies across cosmic time, possibly hosting intermediate mass b

Jamie Bamber, Antonios Tsokaros, Milton Ruiz, Stuart L. Shapiro

The gravitational wave signal produced by the merger of two compact objects includes both an oscillatory transient and a non-oscillatory part, the so-called memory effect. This produces a permanent displacement of test masses and has not yet been measured. We use general relativistic magnetohydrodynamic simulations, including neutrinos, with several represen

Shaoqi Dong, Chaoyou Fu, Haihan Gao, Yi-Fan Zhang

Vision-Language Action (VLA) models significantly advance robotic manipulation by leveraging the strong perception capabilities of pretrained vision-language models (VLMs). By integrating action modules into these pretrained models, VLA methods exhibit improved generalization. However, training them from scratch is costly. In this work, we propose a simple y

Adam Coscia, Alex Endert

Intelligence analysts perform sensemaking over collections of documents using various visual and analytic techniques to gain insights from large amounts of text. As data scales grow, our work explores how to leverage two AI technologies, large language models (LLMs) and knowledge graphs (KGs), in a visual text analysis tool, enhancing sensemaking and helping

Colin Holm-Hansen, Yingtian Chen, Oleg Y. Gnedin

Dynamically cold stellar streams from tidally dissolved globular clusters (GCs) serve as excellent tools to measure the Galactic mass distribution and show promise to probe the nature of dark matter. For successful application of these tools to observations, it is essential to have models of stellar stream properties on the Galactic scale. To this end we pro

Christopher Daw, Martin Orr, Georgios Papas

We prove the Zilber-Pink conjecture for curves in $Y(1)^3$ that intersect a modular curve in the boundary. We also give an unconditional result for unlikely intersection points having few places of supersingular reduction where they are close to a fixed base point. Both results are proved using the G-functions method for unlikely intersections.

Shibendu Gupta Choudhury, Purba Mukherjee, Anjan Ananda Sen

We study the robustness and physical implications of a set of characteristic redshifts that capture key features of the late-time Universe. Using both model-independent reconstructions as well as different dark energy (DE) parameterizations, we show that these redshifts remain stable across cosmological models and reconstruction algorithm, making them reliab

Min-Hsiu Hsieh, Xingjian Li, Ting-Chun Lin

Quantum weight reduction is the task of transforming a quantum code with large check weight into one with small check weight. Low-weight codes are essential for implementing quantum error correction on physical hardware, since high-weight measurements cannot be executed reliably. Weight reduction also serves as a critical theoretical tool, which may be relev

T. Mera, C. Ashall, P. Hoeflich, K. Medler

We present James Webb Space Telescope (JWST) NIRSpec observations of SN~2024ggi, spanning wavelengths of 1.7--5.5 micron at +285.51 and +385.27 days post-explosion. These nebular spectra are dominated by asymmetric emission lines from atomic species including H, Ca, Ar, C, Mg, Ni, Co, and Fe, indicative of an aspherical explosion. The other strong features a

Dwip Dalal, Gautam Vashishtha, Utkarsh Mishra, Jeonghwan Kim

Multimodal large language models (MLLMs) often miss small details and spatial relations in cluttered scenes, leading to errors in fine-grained perceptual grounding. We introduce AttWarp, a lightweight method that allocates more resolution to query-relevant content while compressing less informative areas, all while preserving global context. At test time, th

Sondos Mahmoud Bsharat, Zhiqiang Shen

Large language models (LLMs) have demonstrated impressive reasoning capabilities when provided with chain-of-thought exemplars, but curating large reasoning datasets remains laborious and resource-intensive. In this work, we introduce Prompting Test-Time Scaling (P-TTS), a simple yet effective inference-time data augmentation strategy for enhancing LLM reaso

Antonio R. Linero

Some applied researchers hesitate to use nonparametric methods, worrying that they will lose power in small samples or overfit the data when simpler models are sufficient. We argue that at least some of these concerns are unfounded when nonparametric models are strongly shrunk toward parametric submodels. We consider expanding a parametric model with a nonpa

Gabriel Andres Piovano

We present a family of analytic solutions for the nearly-equatorial motion of a test particle with precessing spin in Kerr spacetime. We solve the equations of motion up to linear order in the small body's spin for periodic and homoclinic orbits. At zero order, the particle moves along equatorial geodesics. The spin-curvature force introduces post-geodesic c

Sangyun Lee, Brandon Amos, Giulia Fanti

Today's generative models thrive with large amounts of supervised data and informative reward functions characterizing the quality of the generation. They work under the assumptions that the supervised data provides knowledge to pre-train the model, and the reward function provides dense information about how to further improve the generation quality and

Nathan Quiblier, Roy Friedman, Matthew Ricci

Dynamical systems in the life sciences are often composed of complex mixtures of overlapping behavioral regimes. Cellular subpopulations may shift from cycling to equilibrium dynamics or branch towards different developmental fates. The transitions between these regimes can appear noisy and irregular, posing a serious challenge to traditional, flow-based mod

Disharee Bhowmick, Ranjith Ramanathan, Sathyanarayanan N. Aakur

Time series data often contain latent temporal structure, transitions between locally stationary regimes, repeated motifs, and bursts of variability, that are rarely leveraged in standard representation learning pipelines. Existing models typically operate on raw or fixed-window sequences, treating all time steps as equally informative, which leads to ineffi

Donghang Wu, Haoyang Zhang, Jun Chen, Xiangyu

Real-time Spoken Language Models (SLMs) struggle to leverage Chain-of-Thought (CoT) reasoning due to the prohibitive latency of generating the entire thought process sequentially. Enabling SLMs to think while speaking, similar to humans, is attracting increasing attention. We present, for the first time, Mind-Paced Speaking (MPS), a brain-inspired framework

Atharv Goel, Sharat Agarwal, Saket Anand, Chetan Arora

Active Learning (AL) promises to reduce annotation cost by prioritizing informative samples, yet its reliability is undermined when labels are noisy or when the data distribution shifts. In practice, annotators make mistakes, rare categories are ambiguous, and conventional AL heuristics (uncertainty, diversity) often amplify such errors by repeatedly selecti

Saad Ahmed Bazaz, Mirza Omer Beg

All widely used and useful programming languages have a common problem. They restrict entry on the basis of knowledge of the English language. The lack of knowledge of English poses a major hurdle to many newcomers who do not have the resources, in terms of time and money, to learn the English language. Studies show that people learn better in their own lang

Martyna Kobus, Radosław Kurek, Thomas Parker

Strong empirical evidence from laboratory experiments, and more recently from population surveys, shows that individuals, when evaluating their situations, pay attention to whether they experience gains or losses, with losses weighing more heavily than gains. The electorate's loss aversion, in turn, influences politicians' choices. We propose a new framework

Jaehong Oh

We prove that ONN achieves order-optimal performance on convergence rate ($\mu \propto \lambda_2$), edge efficiency ($E = N$ for minimal connectivity $k = 2$), and computational complexity ($O(N d^2)$). Empirical validation on 3M-node semantic networks demonstrates 99.75\% improvement over baseline methods, confirming exponential convergence ($\mu = 3.2 \tim

Carlos Rito, Xavier Roulleau

Starting from computer experiments with the fundamental group of the Cartwright--Steger surface, we construct an infinite tower $(X_n)_{n\ge 1}$ of normal projective surfaces obtained by successive $\mathbb Z/3$-Galois covers $X_{n}\to X_{n-1}$. For $n>1$, their minimal resolutions $\widetilde{X}_n$ lie on the line $K^2 = 9\chi - 18$ (equivalently $c_1^2 = 3

Shivprasad S. Shastri, Antonio Cammarata, Tomas Polcar

Semiconductor photocatalysis offers a sustainable route for converting solar energy into chemical energy, enabling the production of clean fuels and valuable chemical products. To this aim, we explore van der Waals heterostructures made up of Janus PtSSe and WXY (X, Y $=$ S, Se, Te and X $\neq$Y), in the context of photocatalytic applications. The redox capa

Fengming Lin

We present a transparent, reproducible measurement of research trends across 26,104 accepted papers from CVPR, ICLR, and NeurIPS spanning 2023-2025. Titles and abstracts are normalized, phrase-protected, and matched against a hand-crafted lexicon to assign up to 35 topical labels and mine fine-grained cues about tasks, architectures, training regimes, object

Arianna Francesconi, Donato Cappetta, Fabio Rebecchi, Paolo Soda

Parkinson's disease (PD) presents a growing global challenge, affecting over 10 million individuals, with prevalence expected to double by 2040. Early diagnosis remains difficult due to the late emergence of motor symptoms and limitations of traditional clinical assessments. In this study, we propose a novel pipeline that leverages keystroke dynamics as a no

Julien Boulanger, Rodolfo Gutiérrez-Romo, Erwan Lanneau

Fix $g \geq 2$. Let $\mathsf{t}(g)$ be the maximal order of the translation group among all genus-$g$ abelian differentials. By work of Schlage-Puchta and Weitze-Schmith\"usen, $\mathsf{t}(g) \leq 4(g - 1)$. They also classify the $g$ attaining this bound. We assume $g$ is outside this class. We first prove that either $\mathsf{t}(g) = (2(m + 1) / m) (g - 1)

Shubham Trehan, Udhav Ramachandran, Akash Rao, Ruth Scimeca

Object detection in biomedical settings is fundamentally constrained by the scarcity of labeled data and the frequent emergence of novel or rare categories. We present FSP-DETR, a unified detection framework that enables robust few-shot detection, open-set recognition, and generalization to unseen biomedical tasks within a single model. Built upon a class-ag

Dhilan Lahoti, Deven Manam

We show that the natural map from the syntomification of a ring $R$ to the stack of $R$-algebra stacks is fully faithful, answering a question of Drinfeld, and we describe its essential image in terms of underlying monoid stacks. We also give similar statements in the characteristic 0 filtered de Rham, $\ell = p$ \'etale, and Betti settings.

Jessie E. An, Chi-Huan Tung, Changwoo Do, Wei-Ren Chen

Small-Angle Neutron Scattering (SANS) data analysis often relies on fixed-width binning schemes that overlook variations in signal strength and structural complexity. We introduce a statistically grounded approach based on the Freedman-Diaconis (FD) rule, which minimizes the mean integrated squared error between the histogram estimate and the true intensity

Margarita Belova, Jiaxin Xiao, Shikhar Tuli, Niraj K. Jha

Researchers have pursued neurosymbolic artificial intelligence (AI) applications for nearly three decades. A marriage of the neural and symbolic components can lead to rapid advancements in AI. Yet, the field has not realized this promise since most neurosymbolic AI frameworks fail to scale. In addition, the implicit representations and approximate reasoning

Konrad Eder, Wladimir Zholobenko, Andreas Stegmeir, Kaiyu Zhang

Numerical modeling of the edge and scrape-off layer (SOL) must account for atomic processes such as hydrogenic ionization and recombination, charge-exchange, and line radiation. Their reaction rates depend non-linearly on density and temperature and are thus sensitive to turbulent fluctuations, whose inclusion/omission may significantly affect model outcomes

Yuqian Huo, David Quiroga, Anastasios Kyrillidis, Tirthak Patel

Variational quantum algorithms (VQAs) have the potential to demonstrate quantum utility on near-term quantum computers. However, these algorithms often get executed on the highest-fidelity qubits and computers to achieve the best performance, causing low system throughput. Recent efforts have shown that VQAs can be run on low-fidelity qubits initially and hi

Xiao Yu, Baolin Peng, Michel Galley, Hao Cheng

Reasoning models have recently shown remarkable progress in domains such as math and coding. However, their expert-level abilities in math and coding contrast sharply with their performance in long-horizon, interactive tasks such as web navigation and computer/phone-use. Inspired by literature on human cognition, we argue that current AI agents need ''vicari