October 2025 arXiv papers — page 48

Daniel Cui, Aaswath P. Raman

We show that purely lossy defects in one- and two-dimensional non-Hermitian photonic crystals can induce transmission matrix singularities not accessible with lossless defects. These singularities in turn can enable dramatic enhancement in overall system gain not accessible through conventional means. We further show that the underlying mechanism behind the

Xiao Lin, Yuhao Huang, Taimeng Fu, Xiaobin Xiong

Humanoid robots, designed to operate in human-centric environments, serve as a fundamental platform for a broad range of tasks. Although humanoid robots have been extensively studied for decades, a majority of existing humanoid robots still heavily rely on complex modular frameworks, leading to inflexibility and potential compounded errors from independent s

Masanori Asakura, Kei Hagihara

Recently, Kedlaya proves certain formula describing explicitly the Frobenius structure on a hypergeometric equation. In this paper, we give a generalization of it. In our case, the Frobenius matrix is no longer described by p-adic gamma function, and then we describe it by the p-adic polygamma functions. Since the p-adic polygamma values are linear combinati

Ranran Haoran Zhang, Soumik Dey, Ashirbad Mishra, Hansi Wu

Inference optimizations are routinely evaluated by throughput alone, without verifying output correctness. We conduct a forensic analysis of batch speculative decoding and find that several widely-used implementations silently produce corrupted outputs (repetitive tokens, <unk> symbols) while reporting competitive speed; failures invisible to metrics like RO

Miguel A. Alarcón, Karl Hauser, Nikolay V. Golubev

This work theoretically investigates possibilities of using the Stimulated Raman Adiabatic Passage (STIRAP) and its variants to control a coherent superposition of quantum states. We present a generalization of the so-called fractional STIRAP (f-STIRAP), demonstrating precise control over the mixing ratio of quantum states in the wave packet. In contrast to

S. J. Dilworth, S. R. Mane

We treat success runs of independent identically distributed Bernoulli trials (with success parameter $p$) distributed according to the Type II binomial distribution of order $k$. However, the success runs are separated by a gap $g\ge1$ (a failure followed by $g-1$ arbitrary outcomes). Most of the literature treats the case $g=1$ only. Our main results are e

Muhammad Adeel Ajaib

We show that spin-flip probabilities emerge in the relativistic regime for scalar potentials, absent in the standard Dirac representation. We examine 1D scattering for the Dirac equation employing an alternate matrix representation introduced by the Author in an earlier study. We demonstrate that the transmission (T) and reflection (R) coefficients can depen

Shuang Gao, Peter E. Caines

Transmission Neural Networks (TransNNs) proposed by Gao and Caines (2022) serve as both virus spread models over networks and neural network models with tuneable activation functions. This paper establishes that TransNNs provide upper bounds on the infection probability generated from the associated Markovian stochastic Susceptible-Infected-Susceptible (SIS)

Rohan Kadekodi, Haoran Peng, Gilbert Bernstein, Michael D. Ernst

A heterogeneous memory has a single address space with fast access to some addresses (a fast tier of DRAM) and slow access to other addresses (a capacity tier of CXL-attached memory or NVM). A tiered memory system aims to maximize the number of accesses to the fast tier via page migrations between the fast and capacity tiers. Unfortunately, previous tiered m

Jinmin Yi, Kangle Li, Chuan Liu, Zixuan Li

For spin and fermionic systems in any spatial dimension, we establish that the superpolynomial decay behavior of mutual information and conditional mutual information is a universal property of gapped pure- and mixed-state phases; i.e., all systems in such a phase possess this property if one system in this phase possesses this property. We further demonstra

Tiasa Singha Roy, Ayush Rajesh Jhaveri, Ilias Triantafyllopoulos

Despite their impressive capabilities, Large Language Models (LLMs) exhibit unwanted uncertainty, a phenomenon where a model changes a previously correct answer into an incorrect one when re-prompted. This behavior undermines trust and poses serious risks in high-stakes domains. In this work, we investigate the mechanisms that drive this phenomenon. We adapt

James Meese, Kyle Herbertson

Research in news recommendation systems (NRS) continues to explore the best ways to integrate normative goals such as editorial objectives and public service values into existing systems. Prior efforts have incorporated expert input or audience feedback to quantify these values, laying the groundwork for more civic-minded recommender systems. This paper cont

Lauri Laatu, Chang Sun, Arianna Cox, Abhijith Gandrakota

We present the first sub-microsecond transformer implementation on an FPGA achieving competitive performance for state-of-the-art high-energy physics benchmarks. Transformers have shown exceptional performance on multiple tasks in modern machine learning applications, including jet tagging at the CERN Large Hadron Collider (LHC). However, their computational

Zhexiao Lin, Peng Ding

Time-series experiments, also called switchback experiments or N-of-1 trials, play increasingly important roles in modern applications in medical and industrial areas. Under the potential outcomes framework, recent research has studied time-series experiments from the design-based perspective, relying solely on the randomness in the design to drive the stati

David Bom, Florian Unger, Birgit Vogtenhuber

We investigate a combinatorial reconfiguration problem on oriented graphs, where a reconfiguration step (edge-flip) is the inversion of the orientation of a single edge. A recently published conjecture that is relevant to the correctness of a Markov Chain Monte Carlo sampler for directed flag complexes states that any simple oriented graph admits a flip sequ

Linus Balicki, Serkan Gugercin

Many algorithms for approximating data with rational functions are built on interpolation or least-squares approximation. Inspired by the adaptive Antoulas-Anderson (AAA) algorithm for the univariate case, the parametric adaptive Antoulas-Anderson (p-AAA) algorithm extends this idea to the multivariate setting, combining least-squares and interpolation formu

Kathleen Kennedy, Anuj Gupta

This chapter introduces the AI & Data Acumen Learning Outcomes Framework, a comprehensive tool designed to guide the integration of AI literacy across higher education. Developed through a collaborative process, the framework defines key AI and data-related competencies across four proficiency levels and seven knowledge dimensions. It provides a structured a

Linyang He, Tianjun Zhong, Richard Antonello, Gavin Mischler

Understanding how the human brain progresses from processing simple linguistic inputs to performing high-level reasoning is a fundamental challenge in neuroscience. While modern large language models (LLMs) are increasingly used to model neural responses to language, their internal representations are highly "entangled," mixing information about lexicon, syn

Marmik Chaudhari, Jeremi Nuer, Rome Thorstenson

Mixture of Experts (MoE) models have become central to scaling large language models, yet their mechanistic differences from dense networks remain poorly understood. Previous work has explored how dense models use \textit{superposition} to represent more features than dimensions, and how superposition is a function of feature sparsity and feature importance.

Kaitong Cai, Jusheng Zhang, Jing Yang, Keze Wang

Hybrid offline--online reinforcement learning (O2O RL) promises both sample efficiency and robust exploration, but suffers from instability due to distribution shift between offline and online data. We introduce RLPD-GX, a framework that decouples policy optimization from safety enforcement: a reward-seeking learner explores freely, while a projection-based

Johann Verwee

We prove explicit Erd\H{o}s--Wintner bounds for Cantor numeration systems via a simple trailing-window decomposition. We temporarily discard the last block of digits (the ``window'') and analyze the remaining prefix. The resulting bound has three contributions: (i) a bridge loss from discarding the window; (ii) a variance-type tail for the prefix; and (iii)

Naisha Agarwal, Judith Amores, Andrew D. Wilson

While Cave Automatic Virtual Environment (CAVE) systems have long enabled room-scale virtual reality and various kinds of interactivity, their content has largely remained predetermined. We present \textit{Storycaster}, a generative AI CAVE system that transforms physical rooms into responsive storytelling environments. Unlike headset-based VR, \textit{Story

Patrick B. Warren, Andrew J. Masters

We address the problem of the vapor-liquid phase transition in the restricted primitive model (RPM) using Wertheim's statistical associating fluid theory to capture the effects of ion pairing which dominate the low-temperature vapor phase. For this we employ a reference system in which ion-pairing is suppressed by a judicious modification of the interaction

Yugong Zeng, Jiayuan Wang, Jonathan Wu

Precipitation prediction has undergone a profound transformation. A notable limitation of traditional NWP is the need for extensive statistical post-processing. To address this challenge, neural network-based approaches were developed. These approaches offer a framework that directly learns the mapping from atmospheric predictors to precipitation targets. Ba

Christian T. Covington, Jeffrey W. Miller

Goodness-of-fit (GoF) tests are a fundamental component of statistical practice, essential for checking model assumptions and testing scientific hypotheses. Despite their widespread use, popular GoF tests exhibit surprisingly low statistical power against substantial departures from the null hypothesis. To address this, we introduce PITOS, a novel GoF test b

Francesco Porro, Fabio Rapallo, Sara Sommariva

In this paper, we consider the problem of testing for a common principal subspace in compositional datasets with structural zeros. In particular, we address the problem in the general setting in which two groups are compared, each characterized by its own pattern of structural zeros. This situation prevents the direct use of standard logratio-based principal

Marianne Arriola, Yair Schiff, Hao Phung, Aaron Gokaslan

Discrete diffusion models enable parallel token sampling for faster inference than autoregressive approaches. However, prior diffusion models use a decoder-only architecture, which requires sampling algorithms that invoke the full network at every denoising step and incur high computational cost. Our key insight is that discrete diffusion models perform two

Lexiang Xiong, Chengyu Liu, Jingwen Ye, Yan Liu

Concept erasure in text-to-image diffusion models is crucial for mitigating harmful content, yet existing methods often compromise generative quality. We introduce Semantic Surgery, a novel training-free, zero-shot framework for concept erasure that operates directly on text embeddings before the diffusion process. It dynamically estimates the presence of ta

Nikolaos Georgiadis, Eleftherios Tiakas, Apostolos N. Papadopoulos

Community search in attributed networks poses a dual challenge: balancing structural connectivity -- the network's topological properties -- and attribute similarity -- the shared characteristics of nodes. This paper introduces a novel algorithm that integrates hop-based and random-walk-based methods to identify high-quality communities, effectively addressi

Adam Stein, Neelay Velingker, Mayur Naik, Eric Wong

Large language models (LLMs) excel at zero-shot inference but continue to struggle with complex, multi-step reasoning. Recent methods that augment LLMs with intermediate reasoning steps such as Chain of Thought (CoT) and Program of Thought (PoT) improve performance but often produce undesirable solutions, especially in algorithmic domains. We introduce Per-I

Divyesh Savaliya, Marius E. Yamakou

Self-induced stochastic resonance (SISR) is the emergence of coherent oscillations in slow-fast excitable systems driven solely by noise, without external periodic forcing or proximity to a bifurcation. This work presents a physics-informed machine learning framework for modeling and predicting SISR in the stochastic FitzHugh-Nagumo neuron. We embed the gove

Congzao Dong, Iryna Feshchenko, Alexander Iksanov

The decoupled standard random walk is a sequence of independent random variables $(\hat S_n)_{n\geq 1}$, in which $\hat S_n$ has the same distribution as the position at time $n$ of a standard random walk with nonnegative jumps. Denote by $\hat N(t)$ the number of elements of the decoupled standard random walk which do not exceed $t$. The random process $(\h

Dmytro Vovchuk, Oleg Torgovitsky, Mykola Khobzei, Vladyslav Tkach

Remote monitoring of drones has become a global objective due to emerging applications in national security and managing aerial delivery traffic. Despite their relatively small size, drones can carry significant payloads, which require monitoring, especially in cases of unauthorized transportation of dangerous goods. A drone's flight dynamics heavily depend

Harish Chandramouleeswaran, Ilan Newman, Tomer Pelleg, Nithin Varma

We study testing $\pi$-freeness of functions $f:[n]^d\to\mathbb{R}$, where $f$ is $\pi$-free if there there are no $k$ indices $x_1\prec\cdots\prec x_k\in [n]^d$ such that $f(x_i)<f(x_j)$ and $\pi(i) < \pi(j)$ for all $i,j \in [k]$, where $\prec$ is the natural partial order over $[n]^d$. Given $\epsilon\in(0,1)$, $\epsilon$-testing $\pi$-freeness asks to di

Audace A. V. Dossou-Olory, Eric O. Andriantiana

A $k$-nearly independent vertex subset of a graph $G$ is a set of vertices that induces a subgraph containing exactly $k$ edges. For $k = 0$, this coincides with the classical notion of independent subsets. This paper investigates the average size, $av_1(G)$ of the $1$-nearly independent vertex subsets of both graphs and trees of a given order $n$. Let $E_n$

Prerna Ravi, Dong Won Lee, Beatriz Flamia, Jasmine David

Understanding how ideas develop and flow in small-group conversations is critical for analyzing collaborative learning. A key structural feature of these interactions is threading, the way discourse talk naturally organizes into interwoven topical strands that evolve over time. While threading has been widely studied in asynchronous text settings, detecting

Omri Hirsch, Ron Shapira Weber, Shira Ifergane, Oren Freifeld

Joint Alignment (JA) of images aims to align a collection of images into a unified coordinate frame, such that semantically-similar features appear at corresponding spatial locations. Most existing approaches often require long training times, large-capacity models, and extensive hyperparameter tuning. We introduce FastJAM, a rapid, graph-based method that d

Antonio Raiola, Nazarii Salish

Empirical researchers often use slope-homogeneity tests to assess whether slopes can be treated as common across units. A key difficulty is that heterogeneity may be concentrated in a small number of units, so that a failure to reject homogeneity may reflect limited power rather than true homogeneity. We quantify this issue by analyzing the power of standard

Jesse Geneson

In threshold-linear networks (TLNs), a fixed point is called minimal if no proper subset of its support is also a fixed point. Curto et al (Advances in Applied Mathematics, 2024) conjectured that every stable fixed point of any TLN must be a minimal fixed point. We provide a counterexample to this conjecture: an explicit competitive TLN on 3 neurons that exh

Yifei Li, Erik-Jan van Kampen

A cascaded online learning flight control system has been developed and enhanced with respect to action smoothness. In this paper, we investigate the convergence performance of the control system, characterized by the increment of a Lyapunov function candidate. The derivation of this metric accounts for discretization errors and state prediction errors intro

Sagnik Mukherjee, Indrajit Barua

The optimization of structural parameters, such as mass(m), stiffness(k), and damping coefficient(c), is critical for designing efficient, resilient, and stable structures. Conventional numerical approaches, including Finite Element Method (FEM) and Computational Fluid Dynamics (CFD) simulations, provide high-fidelity results but are computationally expensiv

Aya Nakayama, Brian Wong, Yuji Nishimura, Kaito Tanaka

The "style trap" poses a significant challenge for Large Vision-Language Models (LVLMs), hindering robust semantic understanding across diverse visual styles, especially in in-context learning (ICL). Existing methods often fail to effectively decouple style from content, hindering generalization. To address this, we propose the Semantic-Preserving Cross-Styl

Bob Dong

We introduce the concept of a k-spine of a tree. A k-spine is essentially a path in the tree whose removal leaves only "less-bushy" components of a smaller pathwidth. Using a k-spine as a central guide, we introduce an O(klog dist) exponential search algorithm on a tree by searching mainly along the spine to narrow down the target's vicinity and then recursi

Guanyu Yao, Qiucheng Wu, Yang Zhang, Zhaowen Wang

Multimodal large language models (MLLMs) have demonstrated strong capabilities on vision-and-language tasks. However, recent findings reveal an imbalance in their reasoning capabilities across visual and textual modalities. Specifically, current MLLMs often over-rely on textual cues while under-attending to visual content, resulting in suboptimal performance

Dominik Meier, Shixing Yu, Sagnik Nandy, Promit Ghosal

Single-cell RNA sequencing (scRNA-seq) enables the study of cellular heterogeneity. Yet, clustering accuracy, and with it downstream analyses based on cell labels, remain challenging due to measurement noise and biological variability. In standard latent spaces (e.g., obtained through PCA), data from different cell types can be projected close together, maki

Daniele Maria Di Nosse, Federico Gatta, Fabrizio Lillo, Sebastian Jaimungal

Decentralized Exchanges (DEXs) are now a significant component of the financial world where billions of dollars are traded daily. Differently from traditional markets, which are typically based on Limit Order Books, DEXs typically work as Automated Market Makers, and, since the implementation of Uniswap v3, feature concentrated liquidity. By investigating th

Adrian Orenstein, Jessica Chen, Gwyneth Anne Delos Santos, Bayley Sapara

While reinforcement learning agents can achieve superhuman performance in many complex tasks, they typically do not become more computationally efficient as they improve. In contrast, humans gradually require less cognitive effort as they become more proficient at a task. If agents could reason about their compute as they learn, could they similarly reduce t

Long H Dang, David Rawlinson

The Hierarchical Reasoning Model (HRM) has impressive reasoning abilities given its small size, but has only been applied to supervised, static, fully-observable problems. One of HRM's strengths is its ability to adapt its computational effort to the difficulty of the problem. However, in its current form it cannot integrate and reuse computation from previo

Yitong Zhang, Siddhartha Sarkar, Xiaohan Wan, Daniel E. Parker

Enhancing the many-body gap of a fractional state is crucial for realizing robust fractional excitations. For fractional Chern insulators, existing studies suggest that making flat Chern bands closely resemble the lowest Landau level (LLL) seems to maximize the excitation gap, providing an apparently optimal platform. In this work, we demonstrate that deform

Haowei Hua, Hong Jiao, Xinyi Wang

BERT and its variants are extensively explored for automated scoring. However, a limit of 512 tokens for these encoder-based models showed the deficiency in automated scoring of long essays. Thus, this research explores generative language models for automated scoring of long essays via summarization and prompting. The results revealed great improvement of s

Nathan Rodkey, Jyotish Patidar, Kerstin Thorwarth, Sebastian Siol

The quality of high-power impulse magnetron sputtering (HiPIMS) deposited films can often improve through the effective use of metal-ion acceleration, requiring precise measurements of time-of-flight (ToF). These measurements are commonly done using time- and energy-resolved mass spectrometry but require careful consideration of the transit time of ions insi

Samaksh Bhargav, Zining Zhu

Large Language Model (LLM) deployment requires guiding the LLM to recognize and not answer unsafe prompts while complying with safe prompts. Previous methods for achieving this require adjusting model weights along with other expensive procedures. While recent advances in Sparse Autoencoders (SAEs) have enabled interpretable feature extraction from LLMs, exi

Aleksandar Pramov

This paper addresses the prediction of commercial (brand) memorability as part of "Subtask 2: Commercial/Ad Memorability" within the "Memorability: Predicting movie and commercial memorability" task at the MediaEval 2025 workshop competition. We propose a multimodal fusion system with a Gemma-3 LLM backbone that integrates pre-computed visual (ViT) and textu

Ye Shen, Rui Song, Alberto Abadie

The Synthetic Control method (SC) has become a valuable tool for estimating causal effects. Originally designed for single-treated unit scenarios, it has recently found applications in high-dimensional disaggregated settings with multiple treated units. However, challenges in practical implementation and computational efficiency arise in such scenarios. To t

Hanyang Chen, Yanchao Yang

Representation learning methods utilizing the InfoNCE loss have demonstrated considerable capacity in reducing human annotation effort by training invariant neural feature extractors. Although different variants of the training objective adhere to the information maximization principle between the data and learned features, data selection and augmentation st

Alexander S. Kuznetsov, Meysam Saeedi, Zixuan Wang, Kevin L. Silverman

Scalable solid-state single-photon sources with triggered single-photon emission rates exceeding a few GHz would aid in the wide technological adoption of photonic quantum technologies. We demonstrate triggering of a quantum dot (QD) single photon emission using dynamic Purcell effect induced at a frequency of several GHz by acoustic strain. To this end, InA

Nan Zhang

To enlarge the translational workspace of cable-driven robots, one common approach is to increase the number of cables. However, this introduces two challenges: (1) cable interference significantly reduces the rotational workspace, and (2) the solution of tensions in cables becomes non-unique, resulting in difficulties for kinematic control of the robot. In

Mirco A. Mannucci

GANs promise indistinguishability, logic explains it. We put the two on a budget: a discriminator that can only ``see'' up to a logical depth $k$, and a generator that must look correct to that bounded observer. \textbf{LOGAN} (LOGical GANs) casts the discriminator as a depth-$k$ Ehrenfeucht--Fra\"iss\'e (EF) \emph{Opponent} that searches for small, legible

Poli Nemkova, Amrit Adhikari, Matthew Pearson, Vamsi Krishna Sadu

Humanitarian organizations face a critical choice: invest in costly commercial APIs or rely on free open-weight models for multilingual human rights monitoring. While commercial systems offer reliability, open-weight alternatives lack empirical validation -- especially for low-resource languages common in conflict zones. This paper presents the first systema

Olena Podladchikova

Solar Orbiter EUV observations reveal ubiquitous small-scale brightenings in the quiet-Sun low corona. We analyze the spatial and temporal dissipation scales of these events with a focus on the formation, evolution, and dissipation of associated current sheets. The brightenings are observed at heights of 1-5 Mm and span energies of 10^20 - 10^24 erg, well be

Ricardo Vega, Connor Mattson, Kevin Zhu, Daniel S. Brown

Swarm robotics has potential for a wide variety of applications, but real-world deployments remain rare due to the difficulty of predicting emergent behaviors arising from simple local interactions. Traditional engineering approaches design controllers to achieve desired macroscopic outcomes under idealized conditions, while agent-based and artificial life s

Alexander Chernov

An induced additive action on a projective variety $X\subseteq\mathbb{P}^n$ is a regular action of the group $\mathbb{G}_a^n$ on $X$ with an open orbit that can be extended to a regular action on $\mathbb{P}^n$. Such actions are known to correspond to pairs $(A, U)$, where $A$ is a local algebra and $U$ is a generating subspace lying in the maximal ideal. Th

Jingxin Zhan, Yuze Han, Zhihua Zhang

The convergence analysis of online learning algorithms is central to machine learning theory, where the last-iterate convergence is particularly important, as it captures the learner&#39;s actual decisions and describes the evolution of the learning process over time. However, in multi-armed bandits, most existing algorithmic analyses mainly focus on the ord

Soham Pahari, Sandeep Chand Kumain

Air pollution remains a critical environmental and public health concern in Indian megacities such as Delhi, Kolkata, and Mumbai, where sudden spikes in pollutant levels challenge timely intervention. Accurate Air Quality Index (AQI) forecasting is difficult due to the coexistence of linear trends, seasonal variations, and volatile nonlinear patterns. This p

Yibo Sun

This study uses the Synthetic Control Method (SCM) to estimate the causal impact of a January 2025 wildfire on housing prices in Altadena, California. We construct a 'synthetic' Altadena from a weighted average of peer cities to serve as a counterfactual; this approach assumes no spillover effects on the donor pool. The results reveal a substantial negative

Honghao Lin, Zhao Song, David P. Woodruff, Shenghao Xie

In the distributed monitoring model, a data stream over a universe of size $n$ is distributed over $k$ servers, who must continuously provide certain statistics of the overall dataset, while minimizing communication with a central coordinator. In such settings, the ability to efficiently collect a random sample from the global stream is a powerful primitive,

Vasudev Vikram, Yuvraj Agarwal, Rohan Padhye

Library dependencies in software ecosystems play a crucial role in the development of software. As newer releases of these libraries are published, developers may opt to pin their dependencies to a particular version. While pinning may have benefits in ensuring reproducible builds and avoiding breaking changes, it bears larger risks in using outdated depende

Mohamed El Louadi, Emna Ben Romdhane

This article analyzes the existential risks artificial intelligence (AI) poses to humanity, tracing the trajectory from current AI to ultraintelligence. Drawing on Irving J. Good and Nick Bostrom's theoretical work, plus recent publications (AI 2027; If Anyone Builds It, Everyone Dies), it explores AGI and superintelligence. Considering machines' exponential

Feng Guo, Luis D. Couto, Khiem Trad, Guangdi Hu

This paper addresses state of charge (SOC) estimation for lithium iron phosphate (LFP) batteries, where the relatively flat open-circuit voltage (OCV-SOC) characteristic reduces observability. A residual bias compensation dual extended Kalman filter (RBC-DEKF) is developed. Unlike conventional bias compensation methods that treat the bias as an augmented sta

Shashank N. Sridhara, Birendra Kathariya, Fangjun Pu, Peng Yin

We introduce Region-Adaptive Learned Hierarchical Encoding (RALHE) for 3D Gaussian Splatting (3DGS) data. While 3DGS has recently become popular for novel view synthesis, the size of trained models limits its deployment in bandwidth-constrained applications such as volumetric media streaming. To address this, we propose a learned hierarchical latent represen

Jingyuan Liu, Hao Qiu, Lin Yang, Mengfan Xu

We study the distributed multi-agent multi-armed bandit problem with heterogeneous rewards over random communication graphs. Uniquely, at each time step $t$ agents communicate over a time-varying random graph $G_t$ generated by applying the Erd\H{o}s-R\'enyi model to a fixed connected base graph $G$ (for classical Erd\H{o}s-R\'enyi graphs, $G$ is a complete

Fatemeh Nazarieh, Zhenhua Feng, Diptesh Kanojia, Muhammad Awais

Audio-driven talking face generation has gained significant attention for applications in digital media and virtual avatars. While recent methods improve audio-lip synchronization, they often struggle with temporal consistency, identity preservation, and customization, especially in long video generation. To address these issues, we propose MAGIC-Talk, a one

Christopher J. Hazard, Michael Resnick, Jacob Beel, Jack Xia

Traditional machine learning relies on explicit models and domain assumptions, limiting flexibility and interpretability. We introduce a model-free framework using surprisal (information theoretic uncertainty) to directly analyze and perform inferences from raw data, eliminating distribution modeling, reducing bias, and enabling efficient updates including d

Denis Denisov, Will FitzGerald, Kaiyuan Zhang

We study a $d$-dimensional random walk with zero mean and finite variance in the Weyl chambers of type C and D. Under optimal moment assumptions we construct positive harmonic functions for random walks killed on exiting Weyl chambers. We also find the tail asymptotics for the exit time of the random walk from Weyl chambers.

Carlos Martínez-Ranero, Lucas Polymeris

Let $T^*$ be an almost Suslin tree, that is, an Aronszajn tree with no stationary antichains. Krueger introduced a forcing axiom, $\mathrm{PFA}(T^*)$, for the class of proper forcings that preserve that $T^*$ is almost Suslin. He showed that $\mathrm{PFA}(T^*)$ implies several well-known consequences of the Proper Forcing Axiom ($\mathrm{PFA}$), including Su

Seungman Choi, Peter Menart, Andrew Schramka, Shubhankar Jape

Ground-based imaging systems struggle to achieve diffraction-limited resolution when atmospheric turbulence and photon scarcity act simultaneously. In this regime, conventional adaptive optics, speckle imaging, and blind deconvolution lack sufficient information diversity to reliably estimate either the scene or the turbulence. We present Turbulence Aware Po

Jiro Adachi

The Cartan $(2,3,5)$-distribution is a tangent distribution of rank~$2$ on a $5$-dimensional manifold satisfying certain generic conditions. The necessary and sufficient condition for a manifold to admit such a structure is established in this paper. The condition obtained is purely topological. In addition, the classification of such structures, up to homot

Anton Shakov

We introduce the $2$-regular integer sequence A383066 $= (s(n))_{n \geq 1}$, which begins $0, 1, 1, 2, 3, 3, 2, \ldots$. We prove that the number of occurrences of an integer $m \geq 0$ in this sequence is equal to $\tau(m^2+1)$, the number of divisors of $m^2 + 1$. Using this fact, we give a generating function for $\tau(m^2+1)$. We also discuss other inter

Oleg Yu. Tsupko, Fabio Aratore, Volker Perlick

In this paper, we investigate the overlapping of photon rings - higher-order images of a black hole's luminous environment, concentrated near the shadow boundary and expected to be resolved in future observations. We consider a broad class of static spherically symmetric spacetimes and geometrically thin equatorial accretion disk with a prescribed inner radi

Hai-Dang Nguyen, Minh-Anh Dang, Minh-Tan Le, Minh-Tuan Le

Explainability is critical for the clinical adoption of medical visual question answering (VQA) systems, as physicians require transparent reasoning to trust AI-generated diagnoses. We present MedXplain-VQA, a comprehensive framework integrating five explainable AI components to deliver interpretable medical image analysis. The framework leverages a fine-tun

Ivan Hristov, Radoslava Hristova, Kiyotaka Tanikawa

A special 2D initial conditions' domain of the equal-mass zero angular momentum planar three-body problem, which has been formerly studied, is analyzed to deepen the knowledge of the stability regions in it. The decay times in the domain are carefully computed. Four stability regions are established. 971 verified initial conditions for linearly stable period

Griffin T. Goodwin, Jayant Biradar, Alison J. March, Christoph Schirninger

Understanding solar flares is critical for predicting space weather, as their activity shapes how the Sun influences Earth and its environment. The development of reliable forecasting methodologies of these events depends on robust flare catalogs, but current methods are limited to flare classification using integrated soft X-ray emission that are available

Peeyush Agarwal, Harsh Agarwal, Akshat Rana

Purpose: The rapid integration of artificial intelligence (AI) systems like ChatGPT, Claude AI, etc., has a deep impact on how work is done. Predicting how AI will reshape work requires understanding not just its capabilities, but how it is actually being adopted. This study investigates which intrinsic task characteristics drive users' decisions to delegate

Jing Xu

Probing the computational underpinnings of subjective experience, or qualia, remains a central challenge in cognitive neuroscience. This project tackles this question by performing a rigorous comparison of the representational geometry of color qualia between state-of-the-art AI models and the human brain. Using a unique fMRI dataset with a "no-report" parad

Florian Grötschla, Elia Trachsel, Luca A. Lanzendörfer, Roger Wattenhofer

Graph-based recommender systems are commonly trained in transductive settings, which limits their applicability to new users, items, or datasets. We propose NBF-Rec, a graph-based recommendation model that supports inductive transfer learning across datasets with disjoint user and item sets. Unlike conventional embedding-based methods that require retraining

Thu Phuong Nguyen, Duc M. Nguyen, Hyotaek Jeon, Hyunwook Lee

Automatically assessing handwritten mathematical solutions is an important problem in educational technology with practical applications, but it remains a significant challenge due to the diverse formats, unstructured layouts, and symbolic complexity of student work. To address this challenge, we introduce VEHME-a Vision-Language Model for Evaluating Handwri

Hemansh Shah, Sanved Kolekar

We investigate quantum field excitations in a rigid cavity that undergoes a transition from inertial motion to uniform acceleration while maintaining constant proper length. By constructing exact Bogoliubov transformations between inertial and accelerated mode bases, we analyze the induced excitations and identify a universal power-law decay in the excitatio

Bingying Zhao, Yin Song, Quanling Deng, Xin Li

The Generalized Finite Element Method (GFEM) is an effective unfitted numerical method for handling interface problems. By augmenting the standard FEM space with an appropriate enrichment space, GFEM can accurately capture C^0 solutions across the interfaces. While numerous GFEMs for interface problems have been studied, establishing a stable high-order GFEM

Michael Ungersböck, Florian Grötschla, Luca A. Lanzendörfer, June Young Yi

Generative models have made significant progress in synthesizing high-fidelity audio from short textual descriptions. However, editing existing audio using natural language has remained largely underexplored. Current approaches either require the complete description of the edited audio or are constrained to predefined edit instructions that lack flexibility

Mason Alexander Ross, Zi-Wei Lin

It has been challenging to quantitatively understand the stopping of incoming nucleons in nuclear collisions, and recently it has been proposed that comparing the baryon stopping with electric charge stopping can help address the question. Here we focus on the $B/Q\times Z/A$ ratio, which can strongly depend on rapidity although its value is one for the full

Florian Brück, Veronika Reimoser, Fabian Baier

This paper formally derives the asymptotic distribution of a goodness-of-fit test based on the Kernel Stein Discrepancy introduced in (Oscar Key et al., "Composite Goodness-of-fit Tests with Kernels", Journal of Machine Learning Research 26.51 (2025), pp. 1-60). The test enables the simultaneous estimation of the optimal parameter within a parametric family

Ibrahim Mohammed, Chris Robertson, M. Gabriela M. Gomes

Susceptible-Exposed-Infectious-Recovered (SEIR) models with inter-individual variation in susceptibility or exposure to infection were proposed early in the COVID-19 pandemic as a potential element of the mathematical/statistical toolset available to policy development. In comparison with other models employed at the time, those designed to fully estimate th

Abhijeet M. Kulkarni, Ioannis Poulakakis, Guoquan Huang

Accurate full-body motion prediction is essential for the safe, autonomous navigation of legged robots, enabling critical capabilities like limb-level collision checking in cluttered environments. Simplified kinematic models often fail to capture the complex, closed-loop dynamics of the robot and its low-level controller, limiting their predictions to simple

Ron Nissim

We establish a mass gap, prove the existence of a unique infinite volume limit, and give a new proof of the large $N$ limit for $\mathrm{U}(N)$ lattice Yang-Mills theory in the 't Hooft regime. These results were previously obtained for $\mathrm{SU}(N)$ and $\mathrm{SO}(N)$ lattice Yang-Mills theories as applications of the mixing of the associated Langevin

Kamil Szczepanik, Jarosław A. Chudziak

Software architecture design is a fundamental part of creating every software system. Despite its importance, producing a C4 software architecture model, the preferred notation for such architecture, remains manual and time-consuming. We introduce an LLM-based multi-agent system that automates this task by simulating a dialogue between role-specific experts

Benson Farb, Jesse Wolfson

We prove for various finite groups $G$ and integers $n\geq 1$ that there are families of equations with Galois group $G$ that cannot be simplified to a one-parameter family even after adjoining a root of a polynomial of degree at most $n$. In more geometric language, there are $G$-varieties $X$ with the following property: for any $G$-equivariant branched co

Jiaxiang Liu, Jiawei Du, Xiao Liu, Prayag Tiwari

Pre-trained vision-language models (VLMs) such as CLIP have demonstrated strong zero-shot capabilities across diverse domains, yet remain highly vulnerable to adversarial perturbations that disrupt image-text alignment and compromise reliability. Existing defenses typically rely on adversarial fine-tuning with labeled data, limiting their applicability in ze

Guangyao Shi, Yuwei Wu, Vijay Kumar, Gaurav S. Sukhatme

Enabling robot teams to execute natural language commands requires translating high-level instructions into feasible, efficient multi-robot plans. While Large Language Models (LLMs) combined with Planning Domain Description Language (PDDL) offer promise for single-robot scenarios, existing approaches struggle with multi-robot coordination due to brittle task

Mark Sellke, Jialu Shi, Jiamin Wang

A Gilbert-Shannon-Reeds (GSR) shuffle is performed on a deck of $N$ cards by cutting the top $n\sim Bin(N,1/2)$ cards and interleaving the two resulting piles uniformly at random. The celebrated "Seven shuffles suffice" theorem of [Bayer-Diaconis '92] established cutoff for this Markov chain: to leading order, total variation mixing occurs after precisely $\

Beyza Cinar, Maria Maleshkova

Individualized therapy is driven forward by medical data analysis, which provides insight into the patient's context. In particular, for Type 1 Diabetes (T1D), which is an autoimmune disease, relationships between demographics, sensor data, and context can be analyzed. However, outliers, noisy data, and small data volumes cannot provide a reliable analysis.