November 2025 arXiv papers — page 182

Matheus Farias, Wanghley Martins, H. T. Kung

Manhattan Distance Mapping (MDM) is a post-training deep neural network (DNN) weight mapping technique for memristive bit-sliced compute-in-memory (CIM) crossbars that reduces parasitic resistance (PR) nonidealities. PR limits crossbar efficiency by mapping DNN matrices into small crossbar tiles, reducing CIM-based speedup. Each crossbar executes one tile, r

Jim James, Ben Wilson, Simon Lucey, James Hays

In this work, we introduce the 3D Gaussian Point Encoder, an explicit per-point embedding built on mixtures of learned 3D Gaussians. This explicit geometric representation for 3D recognition tasks is a departure from widely used implicit representations such as PointNet. However, it is difficult to learn 3D Gaussian encoders in end-to-end fashion with standa

Ting-Kuo Chen, Cheng-Wei Chiang, Sven Heinemeyer, Georg Weiglein

We analyze the excesses at 95 GeV in the light Higgs-boson searches in the di-photon decay channel reported by CMS and ATLAS, which combined are at the level of three standard deviations and are compatible with the excess in the $b\bar{b}$ final state observed at LEP, together with an excess in the di-photon channel at around 152 GeV reported based on a side

Ajendra Singh, Shubham Saurabh, Abhinav Gupta, Rajib Chowdhury

Designing metamaterials for extreme mechanical behavior involves the optimal selection of design parameters. However, identifying these optimal parameters through topology optimization (TO) across a large parametric space requires extensive computational resources. To address this challenge, we propose a novel deep learning framework for metamaterial topolog

Keith Moore, Jun W. Kim, David Lyu, Jeffrey Heo

We present Ask WhAI, a systems-level framework for inspecting and perturbing belief states in multi-agent interactions. The framework records and replays agent interactions, supports out-of-band queries into each agent's beliefs and rationale, and enables counterfactual evidence injection to test how belief structures respond to new information. We apply the

Aldo Morelli

It is well known that a continuous phase transition in Bernoulli bond percolation on the integer lattice is equivalent to a vanishing probability a vertex is invaded in invasion percolation. We provide a coupling between invasion percolation and first passage percolation with log-uniform passage times. This yields a new equivalent condition for a continuous

Aatiya Ali, Viacheslav Sadykov

Solar proton events (SPEs) pose radiation hazards, disrupt technology, and impact operations on Earth and in space, making continuous monitoring essential. We compare 10-50 MeV proton flux measurements from SOHO/EPHIN at Lagrange Point 1 (L1) with those from NOAA/GOES in geostationary orbit (GEO) during Solar Cycle 23 and most of Cycle 24. We identify 83 >=1

Gabriel Missael Barco, Ronan Legin, Connor Stone, Yashar Hezaveh

Score-based models can serve as expressive, data-driven priors for scientific inverse problems. In strong gravitational lensing, they enable posterior inference of a background galaxy from its distorted, multiply-imaged observation. Previous work, however, assumes that the lens mass distribution (and thus the forward operator) is known. We relax this assumpt

Caroline Uhler, Jiaqi Zhang

Massive data collection holds the promise of a better understanding of complex phenomena and, ultimately, better decisions. Representation learning has become a key driver of deep learning applications, as it allows learning latent spaces that capture important properties of the data without requiring any supervised annotations. Although representation learn

Xiaoda Wang, Yuji Zhao, Kaiqiao Han, Xiao Luo

Parkinson's disease (PD) shows heterogeneous, evolving brain-morphometry patterns. Modeling these longitudinal trajectories enables mechanistic insight, treatment development, and individualized 'digital-twin' forecasting. However, existing methods usually adopt recurrent neural networks and transformer architectures, which rely on discrete, regularly sample

Alexis Roger, Gwen Legate, Kashif Rasul, Yuriy Nevmyvaka

Tokenization and transfer learning are two critical components in building state of the art time series foundation models for forecasting. In this work, we systematically study the effect of tokenizer design, specifically scaling and quantization strategies, on model performance, alongside the impact of pretraining versus random initialization. We show that

Bradley E. Schaefer

For Galactic novae, I calculate and collect a comprehensive catalog of 208 measures of white dwarf (WD) masses ($M_{\rm WD}$) and 232 measures of average $V$ magnitudes in quiescence ($V_q$). These are collected into a comprehensive catalog of most fundamental properties of all 402 known Galactic novae. The nova light curve and spectral classes are determine

Madison VanWyngarden, Maya Fishbach, Aditya Vijaykumar, Alexandra G. Guerrero

Gravitational waves (GWs) serve as standard sirens by directly encoding the luminosity distance to their source. When the host galaxy redshift is known, for example, through observation of an electromagnetic (EM) counterpart, GW detections can provide an independent measurement of the Hubble constant, $H_0$. However, even in the absence of an EM counterpart,

Luca Danese, Riccardo Corradin, Andrea Ongaro

We introduce BayesChange, a computationally efficient R package, built on C++, for Bayesian change point detection and clustering of observations sharing common change points. While many R packages exist for change point analysis, BayesChange offers methods not currently available elsewhere. The core functions are implemented in C++ to ensures computational

Hamidreza Maleki Almani

Heavy-tailed phenomena appear across diverse domains --from wealth and firm sizes in economics to network traffic, biological systems, and physical processes-- characterized by the disproportionate influence of extreme values. These distributions challenge classical statistical models, as their tails decay too slowly for conventional approximations to hold.

Alexei Ilyin, Varga Kalantarov, Sergey Zelik

We study the dimensions of the attractors for the fractional Navier--Stokes--Voigt equations. These equations, which include a fractional order of the Stokes operator applied to the time derivative, serve as natural extensions and regularizations of the classical Navier--Stokes equations. We give a comprehensive analysis of the upper bounds for the fractal d

Haochun Ma, Jordan Roulleau-Pasdeloup

Depending on the persistence of a one-off shock bringing the economy to the Effective Lower Bound (ELB), the standard New Keynesian model predicts starkly different conclusions. We offer a potential solution to this morass by assuming that the one-off shock is such that the economy necessarily leaves the ELB in finite time. Under these assumptions, we prove

Yan Wang, Simon C. Warder, Ellyess F. Benmoufok, Andrew Wynn

Reanalysis datasets have become indispensable tools for wind resource assessment and wind power simulation, offering long-term and spatially continuous wind fields across large regions. However, they inherently contain systematic wind speed biases arising from various factors, including simplified physical parameterizations, observational uncertainties, and

Alejandro Guzmán-Ortega, Gustavo Bruzual, Vicente Rodriguez-Gomez, Lars Hernquist

We use the TNG50 cosmological simulation and three-dimensional radiative transfer post-processing to generate dust-aware synthetic observations of galaxies at $ 3 \leq z \leq 6 $ and $ \log_{10}(M_\ast/\mathrm{M}_\odot) \geq 8.5 $, tailored to match the depth and resolution of current deep JWST surveys (NGDEEP and JADES). We analyse the performance of spectr

Andrea Aspesi, Andrea Simpsi, Aaron Tognoli, Simone Mentasti

Event-based cameras are becoming a popular solution for efficient, low-power eye tracking. Due to the sparse and asynchronous nature of event data, they require less processing power and offer latencies in the microsecond range. However, many existing solutions are limited to validation on powerful GPUs, with no deployment on real embedded devices. In this p

Simran Joharle, Francisco Nogueras-Lara, Karl Fiteni

Determining the structure of the Milky Way is essential for understanding its morphology, dynamics, and evolution. However, studying its innermost regions is challenging due to high extinction and crowding. The detection of a double red clump (RC; core-helium-burning stars) feature at very low Galactic latitudes suggests the presence of a spiral arm beyond t

Dawn Virginillo, Asja Derviškadić, Mario Paolone

Due to recent blackout and system split incidents in power grids worldwide, as well as increased system complexity in view of the energy transition, there has been increasing interest in re-evaluating existing Power System Restoration (PSR) plans. In restoration scenarios, due to low island inertia, it is necessary to ensure not only the static, but also the

Ce Jin, Yael Kirkpatrick, Michał Stawarz, Virginia Vassilevska Williams

The All-Pairs Shortest Paths (APSP) is a foundational problem in theoretical computer science. Approximating APSP in undirected unweighted graphs has been studied for many years, beginning with the work of Dor, Halperin and Zwick [SICOMP'01]. Many recent works have attempted to improve these original algorithms using the algebraic tools of fast matrix multip

Zerui Bao, Di Zhu, Liu Jiang, Shiqi Sheng

Large-scale networked services rely on deep soft-ware stacks and microservice orchestration, which increase instruction footprints and create frontend stalls that inflate tail latency and energy. We revisit instruction prefetching for these cloud workloads and present a design that aligns with SLO driven and self optimizing systems. Building on the Entanglin

Shirin Ermis, Cesar Aybar, Lilli Freischem, Stella Girtsou

Accurate forecasting of tropical cyclones (TCs) remains challenging due to limited satellite observations probing TC structure and difficulties in resolving cloud properties involved in TC intensification. Recent research has demonstrated the capabilities of machine learning methods for 3D cloud reconstruction from satellite observations. However, existing a

Masatoshi Imanishi, Bernd Vollmer, Yoshiaki Hagiwara, Kouichiro Nakanishi

We present the results of our ALMA observations of the dense molecular HCN J=4-3 and HCO$^{+}$ J=4-3 lines at $\lesssim$1 pc ($\lesssim$14 mas) resolution in the nuclear region of the nearby ($\sim$14 Mpc) well-studied AGN NGC 1068. Both emission lines are clearly detected around the AGN along an almost east-west direction, which we ascribe to the dusty mole

Farshid Farhadi Khouzani, Abinash Kumar Shaw, Paul La Plante, Bryar Mustafa Shareef

Upcoming measurements of the kinetic Sunyaev-Zel'dovich (kSZ) effect, which results from Cosmic Microwave Background (CMB) photons scattering off moving electrons, offer a powerful probe of the Epoch of Reionization (EoR). The kSZ signal contains key information about the timing, duration, and spatial structure of the EoR. A precise measurement of the CMB op

Phat Nguyen, Tsun-Hsuan Wang, Zhang-Wei Hong, Erfan Aasi

Simulation plays a key role in scaling robot learning and validating policies, but constructing simulations remains a labor-intensive process. This paper introduces ReGen, a generative simulation framework that automates simulation design via inverse design. Given a robot's behavior -- such as a motion trajectory or an objective function -- and its textual d

Rubens Lacouture, Nathan Zhang, Ritvik Sharma, Marco Siracusa

As deep learning models scale, sparse computation and specialized dataflow hardware have emerged as powerful solutions to address efficiency. We propose FuseFlow, a compiler that converts sparse machine learning models written in PyTorch to fused sparse dataflow graphs for reconfigurable dataflow architectures (RDAs). FuseFlow is the first compiler to suppor

Kyle DeBry, Agustin Valdes-Martinez, David Reens, Colin D. Bruzewicz

We demonstrate a trapped-ion protocol in which a nearby, dedicated "monitor" qubit tracks magnetic-field drifts in real time without interrupting data-qubit operations. Using two $^{40}\mathrm{Ca}^+$ ions and the optical--metastable--ground architecture, we encode the data qubit in the ground-state manifold and the monitor qubit in a metastable-state manifol

Dhenenjay Yadav, Rohan Sawai

Foundation models (FMs) offer powerful representations for geospatial analysis, but adapting them effectively remains challenging. Standard adaptation methods, whether full fine-tuning or efficient frozen-backbone approaches, typically employ decoders with fixed regularization strategies, failing to account for the significant heterogeneity in satellite imag

Jovan Markovic, Mikhail M. Ivanov

Applications of effective field theory (EFT) and scattering amplitudes to gravitational problems have recently produced many unique results that advanced our understanding of the dynamics of compact binaries. Many of these results were made possible by comparing gravitational scattering amplitudes in EFT with exact expressions from general relativity. Howeve

Madhurjya Changmai, Jack M. Jenkins, Rony Keppens

Quiescent solar prominences show distinct small-scale dynamics in observations. Their internal density contrasts with the surrounding corona make them susceptible to Rayleigh-Taylor (RT) instabilities, leading to vertically structured prominence morphologies when observed at the solar limb. As a result, prominences develop bubbles and plumes, along with seco

Chris Kouvaris, Ian M. Shoemaker

Low mass particles with small electric charges can be produced abundantly in large electric fields via the Schwinger effect. We study the production rate of such particles inside the polar gap of nearby pulsars. After production they are accelerated above MeV energies by the local electric fields. These pulsar-produced millicharged particles can be detected

Pouya Shiri, Amirali Baniasadi

A Capsule Network (CapsNet) is a relatively new classifier and one of the possible successors of Convolutional Neural Networks (CNNs). CapsNet maintains the spatial hierarchies between the features and outperforms CNNs at classifying images including overlapping categories. Even though CapsNet works well on small-scale datasets such as MNIST, it fails to ach

Maksim Stebliy, Alex Jenkins, Luana Benetti, Ricardo Ferreira

Spin-torque nano-oscillators (STNOs) are promising nanoscale microwave sources for spintronic applications, serving as signal generators or elements in neuromorphic computing systems. In this paper, we investigate the experimental realization of an oscillator based on a magnetic tunnel junction (MTJ) comprising two magnetic layers: a reference layer (RL) and

Vaibhav Singh, Eugene Belilovsky, Rahaf Aljundi

In this paper, we investigate the phenomenon of grokking, where models exhibit delayed generalization following overfitting on training data. We focus on data-scarce regimes where the number of training samples falls below the critical threshold, making grokking unobservable, and on practical scenarios involving distribution shift. We first show that Knowled

Hala Lamdouar, Sairam Sundaresan, Anna Jungbluth, Sudeshna Boro Saikia

We present a hybridized unsupervised clustering algorithm Hisaxy as a novel way to identify frequently occurring magnetic structures embedded in the interplanetary magnetic field (IMF) carried by the solar wind. The Hisaxy algorithm utilizes a combination of indexable Symbolic Aggregate approXimation (iSAX) and Hierarchical Density-Based Spatial Clustering o

J. Gamboa

We investigate the infrared structure of QCD within the adiabatic approximation, where soft gluon configurations evolve slowly compared to the fermionic modes. In this formulation, the functional space of gauge connections replaces spacetime as the natural arena for the theory, and the long-distance behavior is encoded in quantized Berry phases associated wi

CMS Collaboration

A measurement of the substructure of bottom quark jets (b jets) in proton-proton (pp) collisions is presented. The measurement uses data collected in pp collisions at $\sqrt{s}$ = 5.02 TeV, with a low number of simultaneous interactions per bunch crossing, recorded by the CMS experiment in 2017, corresponding to an integrated luminosity of 301 pb$^{-1}$. An

Vibhu Pandya, Radhika Vinze, Anuradha Misra

Precision calculations in hadronic processes at high energy colliders are crucial for improving the understanding of the standard phenomena as well as for the discovery of new physics. Spinor-helicity formalism serves as one of the most efficient ways to simplify the calculations of $S$ matrix elements. In this article, we compute the $S$ matrix elements for

Ana Čolović

In this paper we offer alternate upper bound for the operator $\Pi_b^*\Pi_d$ to the ones present in literature, thus extending the known upper bounds from the $L^2(\mathbb{R})$ setting to $L^p(w)$, for $1<p<\infty,$ and a Muckenhoupt weight $w$. In the $L^2(w)$ setting, we fully characterize the boundedness of the operator.

Daeun Hwang, Saebyul Park

In this study, we introduce Emo100DB: a dataset consisting of improvised songs that were recorded and transcribed with emotion data based on Russell's circumplex model of emotion. The dataset was developed by collecting improvised songs that consist of melody, lyrics, and an instrumental accompaniment played, sung, and recorded by 20 young adults. Before rec

Nikolai Ilinykh, Simon Dobnik

We quantify linguistic diversity in image captioning with surprisal variance - the spread of token-level negative log-probabilities within a caption set. On the MSCOCO test set, we compare five state-of-the-art vision-and-language LLMs, decoded with greedy and nucleus sampling, to human captions. Measured with a caption-trained n-gram LM, humans display roug

Zonglin Lyu, Ming Li, Xinxin Liu, Chen Chen

To enhance controllability in text-to-image generation, ControlNet introduces image-based control signals, while ControlNet++ improves pixel-level cycle consistency between generated images and the input control signal. To avoid the prohibitive cost of back-propagating through the sampling process, ControlNet++ optimizes only low-noise timesteps (e.g., $t <

Riccardo Ciccone, Fabiana De Cesare, Lorenzo Di Pietro, Marco Serone

We study QCD on AdS space with scalars or fermions in the fundamental representation, extending earlier results on pure Yang-Mills theory. In the latter, the Dirichlet boundary condition is conjectured to disappear via merger and annihilation, as signaled by the lightest scalar singlet operator approaching marginality as the coupling increases. With matter,

Matteo Cercola, Michele Lomuscio, Dario Piga, Simone Formentin

Human-in-the-loop calibration is often addressed via preference-based optimization, where algorithms learn from pairwise comparisons rather than explicit cost evaluations. While effective, methods such as Preferential Bayesian Optimization or Global optimization based on active preference learning with radial basis functions (GLISp) treat the system as a bla

A. P. Jones, N. Ysard

Hydrogenated amorphous carbon materials, a-C(:H), are heterogeneous structures consisting of carbon atoms in different hybridisation states and bonding configurations and are thought to constitute a significant and observationally important fraction of the interstellar dust material. This work aims to characterise semi-conducting a-C(:H) nanoparticle structu

Ana Elisa D. Barioni, Arthur N. Montanari, Adilson E. Motter

Coupled lasers offer a promising approach to scaling the power output of photonic devices for applications demanding high frequency precision and beam coherence. However, maintaining coherence among lasers remains a fundamental challenge due to desynchronizing instabilities arising from time delay in the optical coupling. Here, we depart from the conventiona

Derek Perera, Daniel Gilman, Liliya L. R. Williams, Liang Dai

Astrometric perturbations of lensed arcs behind galaxy clusters have been recently suggested as promising probes of small-scale ($\lesssim10^9 M_{\odot}$) dark matter substructure. Populations of cold dark matter (CDM) subhalos, predicted in hierarchical structure formation theory, can break the symmetry of arcs near the critical curve, leading to positional

M. Escudero, G. Jackson, M. Laine, S. Sandner

Cosmological determinations of the number of relativistic neutrino species, $N^{ }_{\rm eff}$, are becoming increasingly accurate, and further improvements are expected both from CMB and BBN data. Given this context, we update the evaluation of $N^{ }_{\rm eff}$ and the current entropy density via the momentum-averaged approach. This allows for a numerically

Jan Vysoky

Lie theory is, beyond any doubt, an absolutely essential part of differential geometry. It is therefore necessary to seek its generalization to $\mathbb{Z}$-graded geometry. In particular, it is vital to construct non-trivial and explicit examples of graded Lie groups and their corresponding graded Lie algebras. Three fundamental families of graded Lie group

A. Camps-Fariña, M. Chamorro-Cazorla, S. F. Sánchez

We aim to measure the evolution of individual galaxies around the Star Formation Main Sequence (SFMS) during the last Gyr as a function of their stellar mass to quantify how much of its scatter is due to short-term variability.We derived star formation histories using full spectral fitting for a sample of 8,960 galaxies from the MaNGA survey to track the pos

Nicolás Abate, Horacio Casini, Marina Huerta, Leandro Martinek

We compute, for any R\'enyi index $n$, the exact difference between the mutual R\'enyi informations of a pair of free massless scalars and that of a Maxwell field in $d=4$ dimensions. Using the standard dimensional reduction method in polar coordinates, the problem is mapped to that of a single scalar field in $d=2$ with Dirichlet boundary conditions, which

Daniel A. Yahalomi, Matthew T. Scoggins, Nasiah Anderson, Mark Driker

NASA's Artemis Mission aims to return astronauts to the Moon and establish a base at the lunar south pole. A key challenge is understanding the threat from micrometeoroid impacts, which are too small to monitor directly. Using NASA's Meteoroid Engineering Model 3 (\texttt{MEM~3}), we estimate micrometeoroid impact rates on a base comparable in size to the In

V. A. Cúneo, A. D. Schwope, J. Kurpas, A. Avakyan

SRGt 062340.2-265751, a cataclysmic variable identified by SRG/eROSITA thanks to its significant X-ray variability, remains poorly characterised despite the multi-wavelength follow-up. We present spectral and timing analyses from the first dedicated X-ray and ultraviolet observations with XMM-Newton, complemented by SRG/eROSITA data from four all-sky surveys

J. M. Alarcón, E. Lope-Oter, Y. Cano

There is an increasing interest in the community for the Neutron Stars and what we can learn from them. In this review we show how chiral effective field theory, combined with many-body methods, can provide important results that connect Neutron Star properties at zero temperature to nuclear physics and allows to use these compact objects as laboratories of

Matthew T. Scoggins, Zoltan Haiman, Fabio Pacucci

The supermassive black holes (SMBHs) with mass $M_\bullet > 10^9 \, \rm M_\odot$ hosted by high-redshift galaxies have challenged our understanding of black hole formation and growth, as several pathways have emerged attempting to explain their existence. The "heavy-seed" pathway eases the problem with the progenitors of these SMBHs having birth masses up to

Peter B. Denton, Stephen J. Parke

Algorithms for computing neutrino oscillation probabilities in sharply varying matter potentials such as the Earth are becoming increasingly important. As the next generation of experiments, DUNE and HyperK as well as the IceCube upgrade and KM3NeT, come online, the computational cost for atmospheric and solar neutrinos will continue to increase. To address

Georgios K. Karananas, Mikhail Shaposhnikov

In this short note we analyze the inflationary dynamics in Weyl-invariant Einstein-Cartan gravity coupled to the Standard Model of particle physics. We take the axion-like particle of gravitational origin to be approximately massless in the early Universe and show how inflation with the Higgs field materializes.

Greta Heine, Fabio Mayer, Marc Neu, Jürgen Becker

We present a hardware-accelerated hit filtering system employing Graph Neural Networks (GNNs) on Field-Programmable Gate Arrays (FPGAs) for the Belle II Level-1 Trigger. The GNN exploits spatial and temporal relationships among sense wire hits and is optimized for high-throughput hardware operation via quantization, pruning, and static graph-building. Sector

A. Thomsen, J. Bucko, T. Kacprzak, V. Ajani

Data-driven approaches using deep learning are emerging as powerful techniques to extract non-Gaussian information from cosmological large-scale structure. This work presents the first simulation-based inference (SBI) pipeline that combines weak lensing and galaxy clustering maps in a realistic Dark Energy Survey Year 3 (DES Y3) configuration and serves as p

Connor Dunlop, Matthew Zheng, Kavana Venkatesh, Pinar Yanardag

Text-to-image (T2I) diffusion models have made remarkable strides in generating and editing high-fidelity images from text. Yet, these models remain fundamentally generic, failing to adapt to the nuanced aesthetic preferences of individual users. In this work, we present the first framework for personalized image editing in diffusion models, introducing Coll

Rafe Loya, Andrew Hamara, Benjamin Estell, Benjamin Kilpatrick

Automatic image cropping is a method for maximizing the human-perceived quality of cropped regions in photographs. Although several works have proposed techniques for producing singular crops, little work has addressed the problem of producing multiple, distinct crops with aesthetic appeal. In this paper, we motivate the problem with a discussion on modern s

Qingzhou Lu, Yao Feng, Baiyu Shi, Michael Piseno

Humanoid robots are expected to operate in human-centered environments where safe and natural physical interaction is essential. However, most recent reinforcement learning (RL) policies emphasize rigid tracking and suppress external forces. Existing impedance-augmented approaches are typically restricted to base or end-effector control and focus on resistin

Yihong Sun, Xinyu Yang, Jennifer J. Sun, Bharath Hariharan

Real-world objects frequently undergo state transformations. From an apple being cut into pieces to a butterfly emerging from its cocoon, tracking through these changes is important for understanding real-world objects and dynamics. However, existing methods often lose track of the target object after transformation, due to significant changes in object appe

Joaquin Tarraga-Moreno, Daniel Barley, Francisco J. Andujar Munoz, Jesus Escudero-Sahuquillo

The rapid growth of data-intensive applications such as generative AI, scientific simulations, and large-scale analytics is driving modern supercomputers and data centers toward increasingly heterogeneous and tightly integrated architectures. These systems combine powerful CPUs and accelerators with emerging high-bandwidth memory and storage technologies to

Ahmad Nouri-Zonoz, Farbod Hassani, Emilio Bellini, Martin Kunz

We present KGB-evolution, a relativistic $N$-body simulation code that extends the $k$-evolution code by incorporating an effective field theory parameterization of kinetic gravity braiding, while also including the $k$-essence model as a limiting case. As a first step, we implement the linearized dark energy stress-energy tensor and scalar field equations,

Jinlai Liu, Jian Han, Bin Yan, Hui Wu

We introduce InfinityStar, a unified spacetime autoregressive framework for high-resolution image and dynamic video synthesis. Building on the recent success of autoregressive modeling in both vision and language, our purely discrete approach jointly captures spatial and temporal dependencies within a single architecture. This unified design naturally suppor

Paul Fendley, Sascha Gehrmann, Eric Vernier, Frank Verstraete

Sutherland showed that the XYZ quantum spin-chain Hamiltonian commutes with the eight-vertex model transfer matrix, so that Baxter's subsequent tour de force proves the integrability of both. The proof requires parametrising the Boltzmann weights using elliptic theta functions and showing they satisfy the Yang-Baxter equation. We here give a simpler derivati

Caleb Lammers, Joshua N. Winn

We re-examine the expected yield of Gaia astrometric planet detections using updated models for giant-planet occurrence, the local stellar population, and Gaia's demonstrated astrometric precision. Our analysis combines a semi-analytic model that clarifies key scaling relations with more realistic Monte Carlo simulations. We predict $7{,}500 \pm 2{,}100$ pla

Lia Bronsard, Andrew Colinet, Dominik Stantejsky, Lee van Brussel

We consider minimizers $u_\varepsilon$ of the Ginzburg-Landau energy with quadratic divergence or curl penalization on a simply-connected two-dimensional domain $\Omega$. On the boundary, strong tangential anchoring is imposed. We prove a priori estimates for $u_\varepsilon$ in $L^\infty$ uniform in $\varepsilon$ and that the Lipschitz constant of $u_\vareps

Maximus A. Pace, Prithwish Dan, Chuanruo Ning, Atiksh Bhardwaj

Human videos are a scalable source of training data for robot learning. However, humans and robots significantly differ in embodiment, making many human actions infeasible for direct execution on a robot. Still, these demonstrations convey rich object-interaction cues and task intent. Our goal is to learn from this coarse guidance without transferring embodi

Shusheng Yang, Jihan Yang, Pinzhi Huang, Ellis Brown

We argue that progress in true multimodal intelligence calls for a shift from reactive, task-driven systems and brute-force long context towards a broader paradigm of supersensing. We frame spatial supersensing as four stages beyond linguistic-only understanding: semantic perception (naming what is seen), streaming event cognition (maintaining memory across

Ellis Brown, Arijit Ray, Ranjay Krishna, Ross Girshick

Despite impressive high-level video comprehension, multimodal language models struggle with spatial reasoning across time and space. While current spatial training approaches rely on real-world video data, obtaining diverse footage with precise spatial annotations remains a bottleneck. To alleviate this bottleneck, we present SIMS-V -- a systematic data-gene

Julian D. Allagan, Dasia A. Singleton, Shanae N. Perry, Gabrielle C. Morgan

This study evaluates a 40-item mathematics placement examination administered to 198 students using a multi-method framework combining Classical Test Theory, machine learning, and unsupervised clustering. Classical Test Theory analysis reveals that 55\% of items achieve excellent discrimination ($D \geq 0.40$) while 30\% demonstrate poor discrimination ($D <

Kaifeng Zhang, Shuo Sha, Hanxiao Jiang, Matthew Loper

Robotic manipulation policies are advancing rapidly, but their direct evaluation in the real world remains costly, time-consuming, and difficult to reproduce, particularly for tasks involving deformable objects. Simulation provides a scalable and systematic alternative, yet existing simulators often fail to capture the coupled visual and physical complexity

Phat Nguyen, Erfan Aasi, Shiva Sreeram, Guy Rosman

Autonomous driving systems remain brittle in rare, ambiguous, and out-of-distribution scenarios, where human driver succeed through contextual reasoning. Shared autonomy has emerged as a promising approach to mitigate such failures by incorporating human input when autonomy is uncertain. However, most existing methods restrict arbitration to low-level trajec

Ryan Golant, Luca Comisso, Philipp Kempski, Lorenzo Sironi

Large-amplitude turbulence -- characterized by a fluctuating magnetic field component, $\delta B$, that is stronger than the mean component, $B_0$ -- is generically intermittent, populated with intense localized structures such as sharp field-line bends and rapid field reversals. Recent MHD simulations suggest that these structures play an important role in

Yu Feng, Nathaniel Weir, Kaj Bostrom, Sam Bayless

LLMs can perform multi-step reasoning through Chain-of-Thought (CoT), but they cannot reliably verify their own logic. Even when they reach correct answers, the underlying reasoning may be flawed, undermining trust in high-stakes scenarios. To mitigate this issue, we introduce VeriCoT, a neuro-symbolic method that extracts and verifies formal logical argumen

Wanwan Zhang

In this paper, we propose and study a multi-dimensional nonlocal active scalar equation of the form \begin{eqnarray*} \partial_t\rho+g\mathcal{R}_a\rho\cdot \nabla\rho= 0,~\rho(\cdot,0)=\rho_{0}, \end{eqnarray*} where the transform $\mathcal{R}_a$ is defined by \begin{eqnarray*} \mathcal{R}_af(x)=\frac{\Gamma(\frac{n+1}{2})}{\pi^{\frac{n+1}{2}}}P.V.\int\limi

Huaguan Chen, Wei Han, Haofei Sun, Ning Lin

Reliable nowcasting of extreme precipitation remains difficult because convective systems are strongly nonlinear, multiscale, and nonstationary in 3D. Radar is the backbone of nowcasting, yet existing methods struggle to predict extremes: physics-based extrapolation cannot capture growth and decay, deterministic learning tends to oversmooth and underestimate

Adam Bouyamourn

How should researchers select experimental sites when the deployment population differs from observed data? I formulate the problem of experimental site selection as an optimal transport problem, developing methods to minimize downstream estimation error by choosing sites that minimize the Wasserstein distance between population and sample covariate distribu

Jasper Kranias, Christian Drago, Colin Vendromin, J. E. Sipe

The Whittaker-Shannon decomposition provides a temporally localized description of squeezed light, making it applicable in the CW limit and leading to a definition of squeezing strength based on the number of photon pairs at a time. We show examples of its usefulness by calculating quadrature variance in a homodyne detection scheme, coincidence detection pro

Jocelyn Finbar Russell

In this paper we build a geometric model for the renormalisation of irrationally indifferent fixed points of holomorphic maps with two critical points. The model incorporates arithmetic properties of the rotation number at the fixed point, as well as the "angle" between the two critical points. Using this model for the renormalisation, we build a topological

Ellis Brown, Jihan Yang, Shusheng Yang, Rob Fergus

Robust benchmarks are crucial for evaluating Multimodal Large Language Models (MLLMs). Yet we find that models can ace many multimodal benchmarks without strong visual understanding, instead exploiting biases, linguistic priors, and superficial patterns. This is especially problematic for vision-centric benchmarks that are meant to require visual inputs. We

Mohammad Atif Quamar, Mohammad Areeb

Chain-of-Thought (CoT) prompting is a key technique for enabling complex reasoning in large language models. However, generating full, fixed-length rationales is computationally wasteful, inflating both token usage and latency. We introduce LEASH: Logit-Entropy Adaptive Stopping Heuristic, a training-free decoding algorithm that adaptively halts rationale ge

Xinlu Zhang, Yansha Deng, Toktam Mahmoodi

Federated learning (FL) offers new opportunities in machine learning, particularly in addressing data privacy concerns. In contrast to conventional event-based federated learning, time-triggered federated learning (TT-Fed), as a general form of both asynchronous and synchronous FL, clusters users into different tiers based on fixed time intervals. However, t

Mantas Žurauskas, Tom Bu, Sanaz Alali, Beyza Kalkanli

Polarization-resolved near-infrared imaging adds a useful optical contrast mechanism to eye tracking by measuring the polarization state of light reflected by ocular tissues in addition to its intensity. In this paper we demonstrate how this contrast can be used to enable eye tracking. Specifically, we demonstrate that a polarization-enabled eye tracking (PE

Dražen Glavan, Alexander Vikman, Tom Zlosnik

Scalar fields with a global U(1) symmetry often appear in cosmology and astrophysics. We study the spherically-symmetric, stationary accretion of such a classical field onto a Schwarzschild black hole in the test-field approximation. Thus, we consider the relativistic Bondi accretion beyond a simplified perfect-fluid setup. We focus on the complex scalar fie

Krinio Marouda, Daniela Cors, Hannes R. Rüter, Alex Vaño-Viñuales

We investigate the threshold of collapse of a massless complex scalar field in axisymmetric spacetimes under the ansatz of Choptuik et al. 2004, in which a symmetry depending on the azimuthal parameter $m$ is imposed on the scalar field. This allows for both non-vanishing twist and angular momentum. We extend earlier work to include higher angular modes. Usi

Sitan Chen, Kevin Cong, Jerry Li

A major bottleneck of standard auto-regressive large language models is that their inference process is inherently sequential, resulting in very long and costly inference times. To circumvent this, practitioners proposed a class of language models called diffusion language models, of which the masked diffusion model (MDM) is the most successful. The MDM is a

Narjes Nourzad, Hanqing Yang, Shiyu Chen, Carlee Joe-Wong

Cooperative multi-agent planning requires agents to make joint decisions with partial information and limited communication. Coordination at the trajectory level often fails, as small deviations in timing or movement cascade into conflicts. Symbolic planning mitigates this challenge by raising the level of abstraction and providing a minimal vocabulary of ac

Miguel Sánchez

Finslerian extensions of Special and General Relativity -- commonly referred to as Very Special and Very General Relativity -- necessitate the development of a unified Lorentz-Finsler geometry. However, the scope of this geometric framework extends well beyond relativistic physics. Indeed, it offers powerful tools for modeling wave propagation in classical m

Stephen Ampleman, Himanshu Sharma, Sayak Mukherjee, Sonja Glavaski

Hybrid power plants (HPPs) combine multiple power generators (conventional/variable) and energy storage capabilities to support generation inadequacy and grid demands. This paper introduces a modeling and control design framework for hybrid power plants (HPPs) consisting of a wind farm, solar plant, and battery storage. Specifically, this work adapts establi

Alamgir Munir Qazi, John P. McCrae, Jamal Abdul Nasir

The proliferation of misinformation necessitates robust yet computationally efficient fact verification systems. While current state-of-the-art approaches leverage Large Language Models (LLMs) for generating explanatory rationales, these methods face significant computational barriers and hallucination risks in real-world deployments. We present DeReC (Dense

Jazmin Ordonez-Toro, Sergio A. Dzib, Laurent Loinard

Very Long Baseline Interferometry (VLBI) provides high angular resolution images and has been used for stellar astrometry for decades. The DYNAMO-VLBA project utilizes the Very Long Baseline Array (VLBA) to study tight binary and multiple pre-main sequence stars, whose components have detectable radio emission and typical separations on the order of milli-ar

Hans Harder, Abhijeet Vishwasrao, Luca Guastoni, Ricardo Vinuesa

This paper is concerned with probabilistic techniques for forecasting dynamical systems described by partial differential equations (such as, for example, the Navier-Stokes equations). In particular, it is investigating and comparing various extensions to the flow matching paradigm that reduce the number of sampling steps. In this regard, it compares direct

Pedro Gatón-Pérez, Enrique Rodriguez-Fernandez, Rodolfo Cuerno

The occurrence of strong coupling or nonlinear scaling behavior for kinetically rough interfaces whose dynamics are conserved, but not necessarily variational, remains to be fully understood. Here we formulate and study a family of conserved stochastic evolution equations for one-dimensional interfaces, whose nonlinearity depends on a parameter n, thus gener

Alberto Merino, Jesus Escudero-Sahuquillo, Pedro Javier Garcia, Francisco J. Quiles

The rise of distributed AI and large-scale applications has impacted the communication operations of data-center and Supercomputer interconnection networks, leading to dramatic incast or in-network congestion scenarios and challenging existing congestion control mechanisms, such as injection throttling (e.g., DCQCN) or congestion isolation (CI). While DCQCN