November 2025 arXiv papers — page 73

Junhao Cheng, Liang Hou, Xin Tao, Jing Liao

While language models have become impactful in many real-world applications, video generation remains largely confined to entertainment. Motivated by video's inherent capacity to demonstrate physical-world information that is difficult to convey through language alone (e.g., imagine teaching someone to tie a tie using only text), we identify an underutilized

Yang Luo, Xuanlei Zhao, Baijiong Lin, Lingting Zhu

Recent progress in generative video models, such as Veo-3, has shown surprising zero-shot reasoning abilities, creating a growing need for systematic and reliable evaluation. We introduce V-ReasonBench, a benchmark designed to assess video reasoning across four key dimensions: structured problem-solving, spatial cognition, pattern-based inference, and physic

Christopher A. Leong, Bitan Roy

Starting from the nearest-neighbor tight-binding model on {10,3} and {14,3} hyperbolic lattices that, for a uniform hopping amplitude, gives rise to emergent Dirac fermions on a curved space with a constant negative curvature, displaying a vanishing density of states, we propose spatially modulated hopping pattern therein that preserve the underlying 5- and

Zhenyuan Qin, Xincheng Shuai, Henghui Ding

Controllable image generation has attracted increasing attention in recent years, enabling users to manipulate visual content such as identity and style. However, achieving simultaneous control over the 9D poses (location, size, and orientation) of multiple objects remains an open challenge. Despite recent progress, existing methods often suffer from limited

Qinghao Hu, Shang Yang, Junxian Guo, Xiaozhe Yao

The emergence of Large Language Models (LLMs) with strong reasoning capabilities marks a significant milestone, unlocking new frontiers in complex problem-solving. However, training these reasoning models, typically using Reinforcement Learning (RL), encounters critical efficiency bottlenecks: response generation during RL training exhibits a persistent long

Ali Taghibakhshi, Sharath Turuvekere Sreenivas, Saurav Muralidharan, Ruisi Cai

Training a family of large language models targeting multiple scales and deployment objectives is prohibitively expensive, requiring separate training runs for each different size. Recent work on model compression through pruning and knowledge distillation has reduced this cost; however, this process still incurs hundreds of billions of tokens worth of train

Changha Choi, Leon A. Takhtajan

We show that two elementary worldline path integrals-the thermal partition function of the harmonic oscillator and the one-loop effective action of scalar QED in a constant field strength-exhibit a natural form of supersymmetric localization. The mechanism relies on hidden fermionic symmetries of the worldline BRST formulation, rather than on standard BRST s

Eddie Pokming Sheung, Qihao Liu, Wufei Ma, Prakhar Kaushik

With the increasing demand for 3D animation, generating high-fidelity, controllable 4D avatars from textual descriptions remains a significant challenge. Despite notable efforts in 4D generative modeling, existing methods exhibit fundamental limitations that impede their broader applicability, including temporal and geometric inconsistencies, perceptual arti

Irmak Guzey, Haozhi Qi, Julen Urain, Changhao Wang

Learning multi-fingered robot policies from humans performing daily tasks in natural environments has long been a grand goal in the robotics community. Achieving this would mark significant progress toward generalizable robot manipulation in human environments, as it would reduce the reliance on labor-intensive robot data collection. Despite substantial effo

Priyanka Kargupta, Shuyue Stella Li, Haocheng Wang, Jinu Lee

Large language models (LLMs) solve complex problems yet fail on simpler variants, suggesting they achieve correct outputs through mechanisms fundamentally different from human reasoning. To understand this gap, we synthesize cognitive science research into a taxonomy of 28 cognitive elements spanning reasoning invariants, meta-cognitive controls, representat

Lorenza Cotugno, Mark de Rooij, Roberta Siciliano

In this paper, we introduce the Generalized Mixed Regularized Reduced Rank Regression model (GMR4), an extension of the GMR3 model designed to improve performance in high-dimensional settings. GMR3 is a regression method for a mix of numeric, binary and ordinal response variables, while also allowing for mixed-type predictors through optimal scaling. GMR4 ex

Zhaoning Wang, Xinyue Wei, Ruoxi Shi, Xiaoshuai Zhang

UV unwrapping flattens 3D surfaces to 2D with minimal distortion, often requiring the complex surface to be decomposed into multiple charts. Although extensively studied, existing UV unwrapping methods frequently struggle with AI-generated meshes, which are typically noisy, bumpy, and poorly conditioned. These methods often produce highly fragmented charts a

Juan C. King, Jose M. Amigo

This paper presents the implementation of an advanced artificial intelligence-based algorithmic trading system specifically designed for the EUR-USD pair within the high-frequency environment of the Forex market. The methodological approach centers on integrating a holistic set of input features: key fundamental macroeconomic variables (for example, Gross Do

Csongor Beke, Anqi Li, Julian Sahasrabudhe

In this paper, we determine the $r$-colour size Ramsey number of the path $P_k$, up to constants. In particular, for every fixed $r \geq 2$ and $k \geq 100\log r$, we have \[ \widehat{R}_r(P_k)=\Theta((r^2 \log r) \, k).\] Perhaps surprisingly, we do this by improving the lower bound on $\widehat{R}_r(P_k)$.

Vishaal Udandarao, Shyamgopal Karthik, Surabhi S. Nath, Andreas Hochlehnert

Cambrian-S aims to take the first steps towards improving video world models with spatial supersensing by introducing (i) two benchmarks, VSI-Super-Recall (VSR) and VSI-Super-Counting (VSC), and (ii) bespoke predictive sensing inference strategies tailored to each benchmark. In this work, we conduct a critical analysis of Cambrian-S across both these fronts.

Elias Lumer, Alex Cardenas, Matt Melich, Myles Mason

Recent advancements in Retrieval-Augmented Generation (RAG) have enabled Large Language Models (LLMs) to access multimodal knowledge bases containing both text and visual information such as charts, diagrams, and tables in financial documents. However, existing multimodal RAG systems rely on LLM-based summarization to convert images into text during preproce

Asya Y. Akkus, Bradley T. Wolfe, Pinghan Chu, Chengkun Huang

Neutron imaging is essential for diagnosing and optimizing inertial confinement fusion implosions at the National Ignition Facility. Due to the required 10-micrometer resolution, however, neutron image require image reconstruction using iterative algorithms. For low-yield sources, the images may be degraded by various types of noise. Gaussian and Poisson noi

Md. Samiul Alim, Sharjil Khan, Amrijit Biswas, Fuad Rahman

Unstructured pruning remains a powerful strategy for compressing deep neural networks, yet it often demands iterative train-prune-retrain cycles, resulting in significant computational overhead. To address this challenge, we introduce a novel teacher-guided pruning framework that tightly integrates Knowledge Distillation (KD) with importance score estimation

Bidipta Sarkar, Mattie Fellows, Juan Agustin Duque, Alistair Letcher

Evolution Strategies (ES) is a class of powerful black-box optimisation methods that are highly parallelisable and can handle non-differentiable and noisy objectives. However, na\"ive ES becomes prohibitively expensive at scale on GPUs due to the low arithmetic intensity of batched matrix multiplications with unstructured random perturbations. We introduce E

Yang Tian, Yuyin Yang, Yiman Xie, Zetao Cai

Recent works explore how real and synthetic data contribute to Vision-Language-Action (VLA) models' generalization. While current VLA models have shown the strong effectiveness of large-scale real-robot pre-training, synthetic data has not previously demonstrated comparable capability at scale. This paper provides the first evidence that synthetic data alone

Juan C. King, Jose M. Amigo

The aim of this paper is the analysis and selection of stock trading systems that combine different models with data of different nature, such as financial and microeconomic information. Specifically, based on previous work by the authors and applying advanced techniques of Machine Learning and Deep Learning, our objective is to formulate trading algorithms

Shuyu Cao, Chongshou Li, Jie Xu, Tianrui Li

3D hierarchical semantic segmentation (3DHS) is crucial for embodied intelligence applications that demand a multi-grained and multi-hierarchy understanding of 3D scenes. Despite the progress, previous 3DHS methods have overlooked following two challenges: I) multi-label learning with a parameter-sharing model can lead to multi-hierarchy conflicts in cross-h

Simone Blasi, Alberto Mariotti, Aäron Rase, Miguel Vanvlasselaer

Domain walls are topological defects that may have formed in the early Universe through the spontaneous breakdown of discrete symmetries, and can be a strong source of gravitational waves (GWs). We perform 3D lattice field theory simulations with CosmoLattice, considering grid sizes $N = 1250$, $2048$ and $4096$, to study the dynamics of the domain wall netw

Ignacio Magaña Hernandez, Katelyn Breivik

The asymmetric nature of GW190814, particularly its mass ratio ($q \approx 1/10$), has made its astrophysical origin elusive. We explore isolated binary evolution as a potential explanation for GW190814's formation. Using the binary population synthesis code COSMIC, and the backpop sampling technique to map the observed parameters of GW190814 to the initial

Jewel Mahajan

Let $(a_n)_{n \geq 1}$ be a sequence of distinct positive integers. The metric theory of minimal gaps for the sequence $\{\alpha a_n \text{ mod }1, 1\leq n \leq N\}$ as $N \to \infty$ was initiated by Rudnick, who established that the minimal gap admits an asymptotic upper bound expressible in terms of the additive energy of $\{a_1,\ldots,a_N\}$ for almost e

Mayur Singh, Lokanath Patra, Chengyang Zhang, Greg MacDougall

Metal-semiconductor interfaces play a central role in micro and nano-electronic devices as heat dissipation or temperature drop across these interfaces can significantly affect device performance. Prediction of accurate thermal boundary resistance (TBR) across these interfaces, considering realistic structures and their correlation with underlying thermal tr

Pazit Haim-Kislev

The symplectic Zoll property of smooth convex domains in the classical phase space has been extensively studied in recent years and, in particular, has been shown to detect local maximizers of the systolic ratio. We propose a dynamical extension of this property to the non-smooth setting, related to the behavior of the Ekeland-Hofer-Zehnder capacity with res

George Shaker, Beñat Martinez de Aguirre Jokisch, Pengning Chao, Steven G. Johnson

We demonstrate a new method that yields orders-of-magnitude acceleration in inverse design (e.g. topology optimization) of high-$Q$ resonant cavities to maximize the local density of states (LDOS), and which is also applicable to other resonant-response metrics. The key idea is that, once conventional LDOS optimization has identified a strong resonance, subs

Zeyuan Yin, Xiaoming Liu

Recent advances in 3D Gaussian diffusion models suffer from time-intensive denoising and post-denoising processing due to the massive number of Gaussian primitives, resulting in slow generation and limited scalability along sampling trajectories. To improve the efficiency of 3D diffusion models, we propose $\textbf{TRIM}$ ($\textbf{T}$rajectory $\textbf{R}$e

Run Hou, Andriy H. Nevidomskyy

The observation of fraction quantum Hall states in twisted MoTe2 has sparked a lof of interest in this phenomenon. Most theoretical works to date rely on the brute-force exact diagonalization which is limited to the one partially occupied band. In this work, we present strong evidence that the effect of higher lying bands cannot be ignored due to strong inte

Wei-Cheng Tseng, David Harwath

Recent advancements in neural audio codecs have not only enabled superior audio compression but also enhanced speech synthesis techniques. Researchers are now exploring their potential as universal acoustic feature extractors for a broader range of speech processing tasks. Building on this trend, we introduce Codec2Vec, the first speech representation learni

Jordan Bounds, Ellis Edinkrah

Given a ring $R$ with center $Z(R)$, we say a linear map $f:R\rightarrow R$ is commuting if $[f(x),x]=0$ for all $x\in R$. Such a map has a standard form if there exists $\lambda\in R$ and additive $\mu:R\rightarrow Z(R)$ such that $f(x)=\lambda x+\mu(x)$ for all $x\in R$. We characterize the linear commuting maps over the $n\times n$ Heisenberg algebra, sho

Markus Anders, Bart Bogaerts, Benjamin Bogø, Arthur Gontier

Symmetry breaking is a crucial technique in modern combinatorial solving, but it is difficult to be sure it is implemented correctly. The most successful approach to deal with bugs is to make solvers certifying, so that they output not just a solution, but also a mathematical proof of correctness in a standard format, which can then be checked by a formally

Benjamin Bedert

The lonely runner conjecture of Wills and Cusick asserts that if $n$ runners with distinct constant speeds run around a a circular unit length track, starting at a common time and place, then each runner will at some time be separated by a distance of at least $\frac{1}{n}$ from all other runners. A weaker lower bound of $\frac{1}{2n-2}$ follows from the so-

Guolin Huang, Wenting Chen, Jiaqi Yang, Xinheng Lyu

Survival analysis is critical for cancer prognosis and treatment planning, yet existing methods lack the transparency essential for clinical adoption. While recent pathology agents have demonstrated explainability in diagnostic tasks, they face three limitations for survival prediction: inability to integrate multimodal data, ineffective region-of-interest e

Mina Tarakemeh, Shenglong Xu

We show that operator dynamics in U(1) symmetric systems are constrained by two independent conserved charges and construct a non-Abelian operator size basis that respects both, enabling a symmetry-resolved characterization of operator growth. The non-Abelian operator size depends on the operator's nonlocal structure and is organized by an SU(2) algebra. Ope

Ashkan Javaherian

This study presents the first experimental validation of a Hessian-free ray-Born inversion technique for quantitative reconstruction of sound speed from transmission ultrasound data. The method combines single-scattering theory with high-frequency approximations, yielding an inversion framework well suited to the frequency ranges used in clinical ultrasound

Nolan J. Coble, Min Ye, Nicolas Delfosse

Neutral atom quantum computers and to a lesser extent trapped ions may suffer from atom loss. In this work, we investigate the impact of atom loss in long chains of trapped ions. Even though this is a relatively rare event, ion loss in long chains must be addressed because it destabilizes the entire chain resulting in the loss of all the qubits of the chain.

Haihao Shi, Zhenyang Huang, Junda Zhou, Guoliang Lü

We introduce a novel parameterization of supernova neutrino energy spectra with a clear physical motivation. Its central parameter, $\tau(t)$, quantifies the characteristic thermal-diffusion area during the explosion. When applied to the historic SN1987A data, this parameterization yields statistically significant fits and provides robust constraints on the

J. R. Fuentes, Ankit Barik, Jim Fuller

Thermal convection in rotating stars and planets drives anisotropic turbulence and differential rotation, both capable of feeding energy into global oscillations. Using 3D simulations of rotating convection in spherical shells, we show that inertial modes--oscillations restored by the Coriolis force--emerge naturally in rotationally constrained turbulence, w

Shihab Ahmed, El Houcine Bergou, Aritra Dutta, Yue Wang

Policy gradient methods, which have been extensively studied in the last decade, offer an effective and efficient framework for reinforcement learning problems. However, their performances can often be unsatisfactory, suffering from unreliable reward improvements and slow convergence, due to high variance in gradient estimations. In this paper, we propose a

Maurizio Atzori, Eleonora Calò, Loredana Caruccio, Stefano Cirillo

Although passwords remain the primary defense against unauthorized access, users often tend to use passwords that are easy to remember. This behavior significantly increases security risks, also due to the fact that traditional password strength evaluation methods are often inadequate. In this discussion paper, we present SODA ADVANCE, a data reconstruction

Tammam Bakeer, Max Herbers, Steffen Marx

In Structural Health Monitoring (SHM), the recovery of distributed mechanical parameters from sparse data is often ill-posed, raising critical questions about identifiability and the reliability of inferred states. While deterministic regularization methods such as Tikhonov stabilise the inversion, they provide little insight into the spatial limits of resol

Olivier Grasdijk, Jakob Kastelic, Jianhui Li, Oskari Timgren

We present a hyperfine-resolved state preparation scheme for thallium fluoride (TlF) molecules based on microwave-driven adiabatic passage (AP) in a spatially varying electric field. This method enables efficient and robust population transfer between selected $\left|J,m_J=0\right\rangle$ hyperfine sublevels of the $X\,^1\Sigma^+_0$ ground state in a cryogen

Jurgen Mezinaj

We develop a computational framework for classifying Galois groups of irreducible degree-7 polynomials over~$\mathbb{Q}$, combining explicit resolvent methods with machine learning techniques. A database of over one million normalized projective septics is constructed, each annotated with algebraic invariants~$J_0, \dots, J_4$ derived from binary transvectio

Tahere Fallahi Serish, Seyed Ali Hosseini Mansoori, Fereshteh Felegary, Özgür Akarsu

We investigate a non-separable subset of $k$-essence in which the kinetic and potential sectors interact through an $X^{\rho}V(\phi)$ coupling, implemented via a potential-dependent prefactor $f(\phi)=1+2\mathcal{K}V$. In slow roll, this structure preserves a constant sound speed $c_s^2=1/(2\rho-1)$ while modifying the Hubble flow in a controlled way, thereb

Imran Ansari, Pawanesh Pawanesh

Meso-scale structures, such as core-periphery (CP) and community structure, have attracted significant attention in modern network science. While communities are characterized by dense intra-group and sparse inter-group connections, CP structures consist of a densely interconnected core and a loosely connected periphery, where peripheral nodes are typically

Zenan Sun, Rashmie Abeysinghe, Xiaojin Li, Xinyue Hu

Objective This study introduces the Alzheimer's Disease Common Data Element Ontology for Clinical Trials (AD-CDO), a lightweight, semantically enriched ontology designed to represent and standardize key eligibility criteria concepts in Alzheimer's disease (AD) clinical trials. Materials and Methods We extracted high-frequency concepts from more than 1,500 AD

Reza Gheissari, Will Perkins, Corrine Yap

We study the fixed-magnetization ferromagnetic Ising model on random $d$-regular graphs for $d\ge 3$ and inverse temperature below the tree reconstruction threshold. Our main result is that for each magnetization $\eta$, the free energy density of the fixed-magnetization Ising model converges to the annealed free energy density, itself the Bethe free energy

Karl Kunisch, Sérgio S. Rodrigues

The stabilization of nonautonomous parabolic equations is achieved by feedback inputs tuning a finite number of actuators, where it is assumed that the input is subject to a time delay. To overcome destabilizing effects of the time delay, the input is based on a prediction of the state at a future time. This prediction is computed depending on a state-estima

Vincenzo Petrecca, Iossif E. Papadakis, Maurizio Paolillo, Demetra De Cicco

Variability is a striking features of quasars, observed at all timescales wavelengths. Studying its properties and the correlations with the physical parameters (e.g. black hole mass and accretion rate) provides significant insights into accretion physics. However, the detailed picture and the exact interplay between different emitting regions are not yet cl

A. Albert, S. Alves, M. André, M. Ardid

We present the $N$-fit algorithm designed to improve the reconstruction of neutrino events detected by a single line of the ANTARES underwater telescope, usually associated with low energy neutrino events ($\sim$ 100 GeV). $N$-Fit is a neural network model that relies on deep learning and combines several advanced techniques in machine learning --deep convol

Jingqiu Ding, Yiding Hua, Kasper Lindberg, David Steurer

We study community detection in the \emph{symmetric $k$-stochastic block model}, where $n$ nodes are evenly partitioned into $k$ clusters with intra- and inter-cluster connection probabilities $p$ and $q$, respectively. Our main result is a polynomial-time algorithm that achieves the minimax-optimal misclassification rate \begin{equation*} \exp \Bigl(-\bigl(

Alan Stapledon

In a companion paper, a canonical bijection was established between strong formal subdivisions of lower Eulerian posets and triples consisting of a lower Eulerian poset, a corresponding rank function, and a non-minimal element such that the join with any other element exists. The main goal of this paper is to relate the local $h$-polynomials of a strong form

Riccardo Venturi

This paper investigates the conditions under which a given circular (synchronizing) DFA is \emph{simple} (sometimes referred to as \emph{primitive}) and when it is \emph{irreducible}. Our notion of irreducibility slightly differs from the classical one, since we are considering our monoid representations to be over $\mathbb{C}$ instead of $\mathbb{Q}$; never

Flavia Lo Sardo, Marina Esposito, Tommaso Confalone, Christophe Tremblay

Artificially engineered twisted van der Waals (vdW) heterostructures have unlocked new pathways for exploring emergent quantum phenomena and strongly correlated electronic states. Many of these phenomena are highly sensitive to the twist angle, which can be deliberately tuned to tailor the interlayer interactions. This makes the twist angle a critical tunabl

J. -F. Donati, P. I. Cristofari, C. Moutou, A. Lavail

We outline in this paper observations of the young pre-main-sequence low-mass star TWA 7, hosting a debris disk and a distant planet. Using data collected with the near-infrared SPIRou spectropolarimeter / precision velocimeter at the Canada-France-Hawaii Telescope from early 2019 to mid 2021, we detected the magnetic field of TWA 7 from the circularly polar

Alan Stapledon

There is a natural notion of a subdivision of a lower Eulerian poset called a strong formal subdivision, which abstracts the notion of a polyhedral subdivision of a polytope, or a proper, surjective morphism of fans. We show that there is a canonical bijection between strong formal subdivisions and triples consisting of a lower Eulerian poset, a correspondin

J. Lange, B. M. Zawora

This article introduces two reduction schemes for Hamiltonian systems on an exact symplectic manifold admitting Lie group symmetries. It is demonstrated that these reduction procedures are equivalent by employing a modified Marsden-Meyer-Weinstein reduction theorem for exact symplectic manifolds and contact manifolds given by energy hypersurfaces. Each appro

Hao Wang, Yun-Song Piao

It has been noted that with the pre-recombination early dark energy (EDE) resolution of Hubble tension, the preference of recent datasets for the evolving dark energy (DE) can be suppressed significantly. In this work, we clarify and reconfirm this result with DESI DR2 and the latest ACT DR6 and SPT-3G D1, the tightest small-scale CMB constraints up to date.

Pratixan Sarmah, Andrzej Siódmok, James Whitehead

The combination of NLO matrix elements with parton showers is indispensable for LHC physics. Differences between matching methods introduce matching uncertainties, corresponding to formally higher-order terms. We recently presented the process-independent generalisation of the KrkNLO method for NLO matching, which employs a modified PDF factorisation scheme

Kabbo Jit Deb, Md. Azizul Hakim, Md Shamse Tabrej

In the era of digital communication, steganography allows covert embedding of data within media files. Adaptive Pixel Value Differencing (APVD) is a steganographic method valued for its high embedding capacity and invisibility, posing challenges for traditional steganalysis. This paper proposes a deep learning-based approach for detecting APVD steganography

Yi Zhang, Che Liu, Xiancong Ren, Hanchu Ni

Developing a universal and versatile embodied intelligence system presents two primary challenges: the critical embodied data bottleneck, where real-world data is scarce and expensive, and the algorithmic inefficiency of existing methods, which are resource-prohibitive. To address these limitations, we introduce Deliberate Practice Policy Optimization (DPPO)

Amandip Sangha

We develop an intrinsic information-theoretic approach for recovering Riemannian curvature from the small-time behaviour of heat diffusion. Given a point and a two-plane in the tangent space, we compare the heat mass transported along that plane with its Euclidean counterpart using the relative entropy of finite measures. We show that the leading small-time

Tianlong Zhang, Hongwei Xue, Shilin Yan, Di Wu

Multimodal large language models (MLLMs) show strong potential as judges. However, existing approaches face a fundamental trade-off: adapting MLLMs to output a single score misaligns with the generative nature of MLLMs and limits fine-grained requirement understanding, whereas autoregressively generating judging analyses is prohibitively slow in high-through

Lukas Billera, Hedwig Nora Nordlinder, Ben Murrell

This brief note clarifies that, in Generator Matching (which subsumes a large family of flow matching and diffusion models over continuous, manifold, and discrete spaces), both the Bregman divergence loss and the linear parameterization of the generator can depend on both the current state $X_t$ and the time $t$, and we show that the expectation over time in

Hyunuk Kim

Networks are useful representations for complex systems. Especially, heterogeneous and asymmetrical relations commonly found in complex systems can be converted to weighted directed edges between nodes. The disparity filter (Serrano et al., 2009) has successfully extracted backbones, sets of important edges, from empirical networks but is not designed to inc

Ivana Nikoloska, Osvaldo Simeone

The integration of sensing and communication functionalities within a common system is one of the main innovation drivers for next-generation networks. In this paper, we introduce a quantum integrated sensing and communication (QISAC) protocol that leverages entanglement in quantum carriers of information to enable both superdense coding and quantum sensing.

Zohar Rimon, Elisei Shafer, Tal Tepper, Efrat Shimron

Palpation, the use of touch in medical examination, is almost exclusively performed by humans. We investigate a proof of concept for an artificial palpation method based on self-supervised learning. Our key idea is that an encoder-decoder framework can learn a $\textit{representation}$ from a sequence of tactile measurements that contains all the relevant in

Boshen Xu, Zihan Xiao, Jiaze Li, Jianzhong Ju

We introduce TimeViper, a hybrid vision-language model designed to tackle challenges of long video understanding. Processing long videos demands both an efficient model architecture and an effective mechanism for handling extended temporal contexts. To this end, TimeViper adopts a hybrid Mamba-Transformer backbone that combines the efficiency of state-space

Bradley E. Schaefer, Andrew Pearce, Tom Love, Michael M. Shara

FQ Cir was an ordinary fast He/N classical nova, peaking at $V$=10.9. The pre-eruption and post-eruption counterpart was at $V$=14.0, making the smallest known classical nova amplitude of 3.1 mag. The nova light and the counterpart coincide to 0.034 arc-seconds, and the counterpart is a rare hot/blue emission-line star with flickering, so the identification

Diaeddin Rimawi

Cyber-physical systems (CPS) combine computational and physical components. Online Collaborative AI System (OL-CAIS) is a type of CPS that learn online in collaboration with humans to achieve a common goal, which makes it vulnerable to disruptive events that degrade performance. Decision-makers must therefore restore performance while limiting energy impact,

Daniil Tiapkin, Artem Agarkov, Nikita Morozov, Ian Maksimov

In this paper, we present gfnx, a fast and scalable package for training and evaluating Generative Flow Networks (GFlowNets) written in JAX. gfnx provides an extensive set of environments and metrics for benchmarking, accompanied with single-file implementations of core objectives for training GFlowNets. We include synthetic hypergrids, multiple sequence gen

Sen Chen, Tong Zhao, Yi Bin, Fei Ma

Developing intelligent agents capable of operating a wide range of Graphical User Interfaces (GUIs) with human-level proficiency is a key milestone on the path toward Artificial General Intelligence. While most existing datasets and benchmarks for training and evaluating GUI agents are static and idealized, failing to reflect the complexity and unpredictabil

Divan A. Burger, Sean van der Merwe, Emmanuel Lesaffre

We introduce a Bayesian quantile mixed-effects model for censored longitudinal outcomes based on the skew exponential power (SEP) error distribution. The SEP family separates tail behavior and skewness from the targeted quantile and includes the skew Laplace (SL) distribution as a special case. We derive analytic likelihood contributions for left, right, and

KM Khalid Saifullah, Daniel Palmer

Multi-agent coordination is critical for next-generation autonomous vehicle (AV) systems, yet naive implementations of communication-based rerouting can lead to catastrophic performance degradation. This study investigates a fundamental problem in decentralized multi-agent navigation: routing loops, where vehicles without persistent obstacle memory become tr

Jules Soria, Zakaria Chihani, Julien Girard-Satabin, Alban Grastien

Case-based reasoning networks are machine-learning models that make predictions based on similarity between the input and prototypical parts of training samples, called prototypes. Such models are able to explain each decision by pointing to the prototypes that contributed the most to the final outcome. As the explanation is a core part of the prediction, th

Amartya Mukherjee, Jun Liu

Differentially private stochastic gradient descent (DP-SGD) has become the standard algorithm for training machine learning models with rigorous privacy guarantees. Despite its widespread use, the theoretical understanding of its long-run behavior remains limited: existing analyses typically establish convergence in expectation or with high probability, but

Houwen Wu, Shuxuan Ying

We propose a new action for entanglement entropy in the framework of the AdS$_{3}$/CFT$_{2}$ correspondence. This action is constructed directly from the entanglement entropy of the CFT$_{2}$, and we show that the Einstein equations of AdS$_{3}$ gravity can be derived from it. In the near-coincidence limit, using Riemann normal coordinates, the action reduce

Ed Younis

High-performance numerical quantum compilers rely on classical optimization, but are limited by slow numerical evaluations and a design that makes extending them with new instructions a difficult, error-prone task for domain experts. This paper introduces OpenQudit, a compilation framework that solves these problems by allowing users to define quantum operat

Xinle Dai

Symmetric products of Riemann surfaces play a crucial role in symplectic geometry and low-dimensional topology. Symmetric products of punctured surfaces are Liouville manifolds of interest e.g. for Heegaard Floer theory. We study the symplectic topology of these spaces using Liouville sectorial techniques, along with examples and applications of these decomp

Jessica Anzanello, Pablo Spiga

Let $T$ be a finite non-abelian simple group. Giudici, Morgan and Praeger have shown that the order of $T$ is bounded above by a function depending on the maximum number of $\mathrm{Aut}(T)$-classes of elements of $T$ of prime-power order. In this note, we strengthen this result by showing, in particular, that prime-power can be replaced by prime.

Andres Campero, Derek Shiller, Jaan Aru, Jonathan Simon

We develop a taxonomical framework for classifying challenges to the possibility of consciousness in digital artificial intelligence systems. This framework allows us to identify the level of granularity at which a given challenge is intended (the levels we propose correspond to Marr's levels) and to disambiguate its degree of force: is it a challenge to com

L. Costa, I. Macías Tarrío, L. Roa-Leguizamón

Let $X$ be a smooth, irreducible, projective algebraic surface, and let $\alpha \in \mathbb{Q}[m]_{>0}$ be a polynomial. In this paper, we determine topological and geometric properties of the moduli space of $\alpha$-stable coherent systems of type $(n; c_{1}, c_{2}, k)$ with $k < n$ on $X$, for sufficiently large values of $\alpha$. We prove that, for $\al

El-Mehdi Mehiri

We introduce and study block-separated overpartitions, a constrained family of overpartitions in which no two consecutive distinct part-blocks are both overlined. This local restriction produces a new sequence that naturally interpolates between classical partitions and unrestricted overpartitions. We show that the internal decoration of distinct part-blocks

Gaspard Ohlmann, Edwin Hamel-De le Court, Francesco Belardinelli

Ensuring that agents satisfy safety specifications can be crucial in safety-critical environments. While methods exist for controller synthesis with safe temporal specifications, most existing methods restrict safe temporal specifications to probabilistic-avoidance constraints. Formal methods typically offer more expressive ways to express safety in probabil

Manish Kumar, Derek Waleffe, Anna Okounkova, Raveel Tejani

Magnetic fields typically suppress superconductivity once the Zeeman energy exceeds the pairing gap, unless mechanisms such as unconventional pairing, strong spin-orbit coupling, or intrinsic magnetism intervene. Several graphene platforms realize such mitigating routes, exhibiting superconductivity resilient to magnetic fields. Here we report superconductiv

Kexin Zhao, Ken Forbus

Word sense disambiguation is a fundamental challenge in natural language understanding. Current methods are primarily aimed at coarse-grained representations (e.g. WordNet synsets or FrameNet frames) and require hand-annotated training data to construct. This makes it difficult to automatically disambiguate richer representations (e.g. built on OpenCyc) that

Alima Subedi, Sankalpa Pokharel, Satish Puri

Similarity searches are a critical task in data mining. As data sets grow larger, exact nearest neighbor searches quickly become unfeasible, leading to the adoption of approximate nearest neighbor (ANN) searches. ANN has been studied for text data, images, and trajectories. However, there has been little effort to develop ANN systems for polygons in spatial

Fares Fourati, Mohamed-Slim Alouini, Vaneet Aggarwal

We propose ECPv2, a scalable and theoretically grounded algorithm for global optimization of Lipschitz-continuous functions with unknown Lipschitz constants. Building on the Every Call is Precious (ECP) framework, which ensures that each accepted function evaluation is potentially informative, ECPv2 addresses key limitations of ECP, including high computatio

Nirjhor Datta, Md. Golam Rabiul Alam

The ability to selectively remove knowledge from medical segmentation networks is increasingly important for privacy compliance, ethical deployment, and continual dataset revision. We introduce Erase to Retain, a controllable unlearning framework for medical image segmentation that achieves targeted forgetting without full retraining. Our method uses a teach

Huanshuo Dong, Hong Wang, Hao Wu, Zhiwei Zhuang

Neural operators have demonstrated considerable effectiveness in accelerating the solution of time-dependent partial differential equations (PDEs) by directly learning governing physical laws from data. However, for PDEs governed by conservation laws(e.g., conservation of mass, energy, or matter), existing neural operators fail to satisfy conservation proper

Herbert M. C. Maquera, Tiago Pereira, Matteo Tanzi

We investigate the dynamics of large heterogeneous network dynamical systems composed of nonlocally coupled chaotic maps. We show that the mean-field limit of such systems is governed by a suitably defined Self-Consistent Transfer Operator (STO) acting on graphons, describing the infinite-size limits of dense graphs, thereby allowing for a rigorous analysis

Md. Tawfique Ihsan, Md. Rakibul Hasan Rafi, Ahmed Shoyeb Raihan, Imtiaz Ahmed

Severe class imbalance is common in real-world tabular learning, where rare but important minority classes are essential for reliable prediction. Existing generative oversampling methods such as GANs, VAEs, and diffusion models can improve minority-class performance, but they often struggle with tabular heterogeneity, training stability, and privacy concerns

Angelo Farfan, Mehrdad Ghadiri, Junzhao Yang

We present an algorithm that given any invertible symmetric diagonally dominant M-matrix (SDDM), i.e., a principal submatrix of a graph Laplacian, $\boldsymbol{\mathit{L}}$ and a nonnegative vector $\boldsymbol{\mathit{b}}$, computes an entrywise approximation to the solution of $\boldsymbol{\mathit{L}} \boldsymbol{\mathit{x}} = \boldsymbol{\mathit{b}}$ in $

Natascia L. Fragapane, Volker L. Deringer

With the growing availability of machine-learned interatomic potential (MLIP) models for materials simulations, there is an increasing demand for robust, automated, and chemically insightful benchmarking methodologies. In response, we here introduce LiPS-25, a curated benchmark dataset for a canonical series of solid-state electrolyte materials from the Li2S

Lai Tian, Johannes O. Royset

We provide counterexamples showing that uniform laws of large numbers do not hold for subdifferentials under natural assumptions. Our constructions are univariate random Lipschitz functions and bivariate random convex functions with two smooth pieces. Consequently, they resolve the questions posed by Shapiro and Xu [J. Math. Anal. Appl., 325 (2007), 1390-139

Ye Mao, Weixun Luo, Ranran Huang, Junpeng Jing

In this paper, we introduce POMA-3D, the first self-supervised 3D representation model learned from point maps. Point maps encode explicit 3D coordinates on a structured 2D grid, preserving global 3D geometry while remaining compatible with the input format of 2D foundation models. To transfer rich 2D priors into POMA-3D, a view-to-scene alignment strategy i

Md. Mohaiminul Islam, Md. Mofazzal Hossen, Maher Ali Rusho, Nahiyan Nazah Ridita

This study deliberates on the application of advanced AI techniques for brain tumor classification through MRI, wherein the training includes the present best deep learning models to enhance diagnosis accuracy and the potential of usability in clinical practice. By combining custom convolutional models with pre-trained neural network architectures, our appro

Misaal Khan, Mayank Vatsa, Kuldeep Singh, Richa Singh

Child malnutrition remains a global crisis, yet existing screening methods are laborious and poorly scalable, hindering early intervention. In this work, we present NutriScreener, a retrieval-augmented, multi-pose graph attention network that combines CLIP-based visual embeddings, class-boosted knowledge retrieval, and context awareness to enable robust maln