December 2025 arXiv papers — page 41

Miles Q. Li, Benjamin C. M. Fung, Martin Weiss, Pulei Xiong

As autonomous AI agents are increasingly deployed in high-stakes environments, ensuring their safety and alignment with human values is becoming a practical deployment concern. Current benchmarks for AI agents primarily evaluate refusal of explicitly harmful instructions or completion of complex multi-step tasks. However, there is a lack of benchmarks design

Haizhou Yang, Jiyang Zhang, Brahmajee K. Nallamothu, Krishna Garikipati

Coronary microvascular dysfunction (CMD), characterized by impaired regulation of blood flow in the coronary microcirculation, plays a key role in the pathogenesis of ischemic heart disease and is increasingly recognized as a contributor to adverse cardiovascular outcomes. Despite its clinical importance, CMD remains underdiagnosed due to the reliance on inv

Zhengyang Shan, Aaron Mueller

We investigate how independent demographic bias mechanisms are from general demographic recognition in language models. Using a multi-task evaluation setup where demographics are associated with names, professions, and education levels, we measure whether models can be debiased while preserving demographic detection capabilities. We compare attribution-based

Aymen Alsaadi, Mason Hooten, Mariya Goliyad, Andre Merzky

Hybrid AI-HPC workflows combine large-scale simulation, training, high-throughput inference, and tightly coupled, agent-driven control within a single execution campaign. These workflows impose heterogeneous and often conflicting requirements on runtime systems, spanning MPI executables, persistent AI services, fine-grained tasks, and low-latency AI-HPC coup

Shariqah Hossain, Lalana Kagal

Machine unlearning aims to remove unwanted information from a model, but many methods are inefficient for LLMs with large numbers of parameters or fail to fully remove the intended information without degrading performance on knowledge that should be retained. Model editing algorithms solve a similar problem of changing information in models, but they focus

Ian Matthias Flór, Claudia Artiaco, Thomas Klein Kvorning, Jens H. Bardarson

We generalize the information lattice, originally defined for one-dimensional open-boundary chains, to characterize quantum many-body states in higher-dimensional geometries. In one dimension, the information lattice provides a position- and scale-resolved decomposition of von Neumann information. Its generalization is nontrivial because overlapping subsyste

Hongqin Liu

The potential energy landscape, PEL, theory stands as one of the most successful frameworks for understanding supercooled liquids and glassy systems. A central element of this theory is the configurational entropy, Sc, which is traditionally represented by a symmetric Gaussian distribution. However, the asymmetric nature of the potential energy of inherent s

Pavel Dvurechensky, Meggie Marschner, Shimrit Shtern, Mathias Staudigl

In the framework of a real Hilbert space we consider the problem of approaching solutions to a class of hierarchical variational inequality problems, subsuming several other problem classes including certain mathematical programs under equilibrium constraints, constrained min-max problems, hierarchical game problems, optimal control under VI constraints, and

Yihan, Wen, Xin Chen

The growing complexity of power system operations has created an urgent need for intelligent, automated tools to support reliable and efficient grid management. Conventional analysis tools often require significant domain expertise and manual effort, which limits their accessibility and adaptability. To address these challenges, this paper presents X-GridAge

Alejandro Borda, Julian Rincon, César Galindo

We show that single-qudit universality in Clifford-based gate sets follows a trichotomy determined by the prime factorization of the local dimension $d$. For prime $d$, any gate outside the Clifford group is universal. For prime-power dimensions $d=p^m$ with $m\ge 2$, not every non-Clifford gate is universal, but it can be achieved by suitable members of a f

F. Yusef-Zadeh, M. Wardle, R. G. Arendt, C. O. Heinke

Flux variability is a fundamental channel of information from Sgr A* because of its direct probe of processes occurring within an accretion disk under strong gravity. We present simultaneous IR, X-ray and radio observations of Sgr A* on 2024 Apr 05 using JWST, NuSTAR, and VLA. We report the detection of a strong X-ray flare with a luminosity of $\sim5.2x10^{

Mikhail Lazarev, Andrey Ustyuzhanin

Machine learning models have become firmly established across all scientific fields. Extracting features from data and making inferences based on them with neural network models often yields high accuracy; however, this approach has several drawbacks. Symbolic regression is a powerful technique for discovering analytical equations that describe data, providi

Raja Mallina, Bryar Shareef

Breast ultrasound (BUS) segmentation provides lesion boundaries essential for computer-aided diagnosis and treatment planning. While promptable methods can improve segmentation performance and tumor delineation when text or spatial prompts are available, many public BUS datasets lack reliable metadata or reports, constraining training to small multimodal sub

Facundo Tarasi, Esteban D. Gadea, Tchavdar Todorov, Damian A. Scherlis

Since the first evidence of luminescence of organic polymers in STM junctions, efforts have been invested in elucidating how to leverage the voltage, anchoring chemistry, and molecular structure to optimize emission power and efficiency. Understanding the fundamentals underlying current-driven molecular emission is important not only for OLED engineering, bu

Youjin Jung, Seongwoo Cho, Hyun-seok Min, Sungchul Choi

Composed Image Retrieval (CIR) aims to find a target image that aligns with user intent, expressed through a reference image and a modification text. While Zero-shot CIR (ZS-CIR) methods sidestep the need for labeled training data by leveraging pretrained vision-language models, they often rely on a single fused query that merges all descriptive cues of what

Moshe Rafaeli Shimron, Vadim Indelman

Multi-agent decision-making under uncertainty is fundamental for effective and safe autonomous operation. In many real-world scenarios, each agent maintains its own belief over the environment and must plan actions accordingly. However, most existing approaches assume that all agents have identical beliefs at planning time, implying these beliefs are conditi

Jorge Sastre, Daniel Faronbi, José Miguel Alonso, Peter Traver

The matrix exponential is a fundamental operator in scientific computing and system simulation, with applications ranging from control theory and quantum mechanics to modern generative machine learning. While Pad\'e approximants combined with scaling and squaring have long served as the standard, recent Taylor-based methods, which utilize polynomial evaluati

Chih-Yuan Lin, Pia Bhatia, Alexandra Sofia Uy-Tioco, Kyril Kavetsky

We demonstrate a dual nanopore platform (DNP) containing a top 2D MoS2 pore in series with a 3 to 5 nm thick SiN pore, vertically separated by 30 nm, with diameters of 1.0 and 3.0 nm, respectively. This platform enables independent probing of analytes by each pore, thereby providing complementary information. We measure translocations of single amino acids (

Joaquin Estevez-Delgado, Roberto De Arcia, Gabino Estevez-Delgado, Israel Quiros

In this work we investigate the asymptotic cosmological dynamics of a modified gravity model based on the $f(R,T^\phi)$ theory, where $R$ denotes the Ricci scalar and $T^\phi$ is the trace of the stress-energy tensor of a scalar field. Despite the extensive study of $f(R,T)$ gravity, the asymptotic implications of quadratic trace couplings in scalar field co

Alex Dantart

T\'ecnicas recientes de "Contextualized Chunking" inyectan res\'umenes para mejorar el contexto en RAG, pero introducen una "diluci\'on vectorial" que opaca el contenido local. Evaluando distintos ratios de inyecci\'on, demostramos una curva en "U invertida": una inyecci\'on moderada mejora el "Recall" (+18%), pero superar un umbral cr\'itico (CIR > 0.4) red

Ziyi Zhu, Olivier Tieleman, Caitlin A. Stamatis, Luka Smyth

Realistic user simulation is crucial for training and evaluating multi-turn dialogue systems, yet creating simulators that accurately replicate human behavior remains a significant challenge. An effective simulator must expose the failure modes of the systems under evaluation. This work introduces Direct Iterative Adversarial Learning (DIAL), an adversarial

Daniel Cortild, Meggie Marschner, Mathias Staudigl

We consider hierarchical variational inequality problems, or more generally, variational inequalities defined over the set of zeros of a monotone operator. This framework includes convex optimization over equilibrium constraints and equilibrium selection problems. In a real Hilbert space setting, we combine a Tikhonov regularization and a proximal penalizati

Tiago Pereira

These lectures are based on material which was presented in the 2025 Summer school at Funda\c{c}\~ao Getulio Vargas. The aim of this series is to introduce graduate students with a little background in the field of dynamical systems and network theory to epidemic models. Our goal is to give a succinct and self-contained description of the models

Markus Gross, Sai B. Matha, Aya Fahmy, Rui Song

Semantic Scene Completion (SSC) is essential for 3D perception in mobile robotics, as it enables holistic scene understanding by jointly estimating dense volumetric occupancy and per-voxel semantics. Although SSC has been widely studied in terrestrial domains such as autonomous driving, aerial settings like autonomous flying remain largely unexplored, thereb

Tanmay P. Patel, Erica L. Tevere, Erik H. Kramer, Rudranarayan M. Mukherjee

This paper presents a general purpose framework for autonomous, vision-based interception of dynamic, non-cooperative targets, validated across three distinct mobility platforms: an unmanned aerial vehicle (UAV), a four-wheeled ground rover, and an air-thruster spacecraft testbed. The approach relies solely on a monocular camera with fiducials for target tra

Waltraud Lederle

We construct, for the free group acting on its Cayley tree, boomerang subgroups whose critical exponent is arbitrarily close to the critical exponent of a given finitely generated subgroup.

Rose Ranson, Yifan Zhou, Michael Hesford, Jack Drouin

We demonstrate an injection-locked 399 nm laser system with up to 1 W output power and a locked power fraction of 0.57. The system consists of a high power, multimode diode laser that is seeded by 5 mW from a single-mode external cavity diode laser. The locked high-power laser inherits the frequency agility and linewidth of the seed laser with 3.9 kHz broade

Marcio Santetti

Energy efficiency gains in production and consumption are undisputed economic and environmental goals. However, potential energy savings derived from efficiency innovations may have short-lasting effects due to increased demand for more affordable energy services. Measuring the size of this rebound effect is a critical tool for better assessing the reliabili

Víctor Becerril, Marco A. Pérez

We define and study induced duality pairs under Foxby equivalences. Given a semidualizing $(S,R)$-bimodule ${}_S C_R$, if $(\mathcal{A}_C(R),\mathcal{B}_C(R^{\rm op}))$ and $(\mathcal{A}_C(S^{\rm op}),\mathcal{B}_C(S))$ denote the duality pairs formed by the corresponding classes of Auslander and Bass modules, and if $(\mathcal{M,N})$ is a duality pair over

Tim Brüers, Christoph Lehrenfeld, Max Wardetzky

This paper presents a streamfunction-vorticity formulation for the Navier--Stokes and Euler equations on general surfaces. Notably, this includes non-simply connected surfaces, on which the harmonic components of the velocity field play a fundamental role in the dynamics. By relying only on scalar and finite-dimensional quantities, our formulation ensures th

Zachary Izzo, Iain Melvin

We study the problem of subgroup discovery for survival analysis, where the goal is to find an interpretable subset of the data on which a Cox model is highly accurate. Our work is the first to study this particular subgroup problem, for which we make several contributions. Subgroup discovery methods generally require a "quality function" in order to sift th

Marcel Meyer, Sascha Kaltenpoth, Henrik Albers, Kevin Zalipski

Time Series Foundation Models (TSFMs) are transforming the field of forecasting. However, evaluating them on historical data is increasingly difficult due to the risks of train-test sample overlaps and temporal overlaps between correlated train and test time series. To address this, we introduce TS-Arena, a live forecasting platform that shifts evaluation fr

Brian Lu, Hongyu Zhao, Shuo Sun, Hao Peng

Reinforcement learning with verifiable rewards (RLVR) has emerged as a promising paradigm for post-training large language models (LLMs) on complex reasoning tasks. Yet, the conditions under which RLVR yields robust generalization remain underexplored. This paper provides an empirical study of RLVR generalization in the setting of probabilistic inference ove

Daniel A. López Aguilar, Antonio Rodríguez Sánchez, Pablo Roig, Hanchen Yu

In case there are tensor interactions beyond the SM in the low-energy effective Lagrangian, their interference with the SM V-A currents yields a contribution for tau decays into three or more mesons that cannot be included in the famous structure functions introduced by K\"uhn and Mirkes in their seminal 1992 paper. We present this generalization here, highl

Petar Tomić, Patrick Bütler, Yuze Wu, Bart Raes

Silicon spin qubits offer long coherence times, a compact footprint and compatibility with industrial CMOS manufacturing. Here, we investigate spin qubits hosted in quantum dots fabricated in a state-of-the-art 300 mm nanoelectronics foundry and demonstrate substantially enhanced coherence, achieving a Hahn-echo time of $T_2^{\text{Hahn}} = 4\,\mathrm{ms}$ f

M. E. Abishev, V. D. Ivashchuk, A. N. Malybayev, S. Toktarbay

Dilatonic black hole dyon-like solutions in the gravitational $4d$ model with a scalar field, two 2-forms, two dilatonic coupling constants $\lambda_i \neq 0$, $i =1,2$, obeying $\lambda_1 \neq - \lambda_2$ and the sign parameter $\varepsilon = \pm 1$ for scalar field kinetic term are overviewed. Here $\varepsilon = - 1$ corresponds to a phantom scalar field

Yizhak Yisrael Elboher, Avraham Raviv, Amihay Elboher, Zhouxing Shi

Ensuring the safety and efficiency of AI systems is a central goal of modern research. Formal verification provides guarantees of neural network robustness, while early exits improve inference efficiency by enabling intermediate predictions. Yet verifying networks with early exits introduces new challenges due to their conditional execution paths. In this wo

Daine L. Danielson, Gautam Satishchandran

It was recently shown that black holes decohere any quantum superpositions in their vicinity. This decoherence is mediated by soft radiation through the horizon, and can be understood as the result of the fact that quantum states in the exterior source distinguishable states of long-range fields in the interior. To study this phenomenon and others, we extend

Lisa Rohde, Gordei Anchutkin, Viktor Holubec, Frank Cichos

Self-propelled microparticles create flow fields that determine how they interact with surfaces, external flows, and each other. These flow fields fall into distinct classes--pushers, pullers, and neutral swimmers--each exhibiting fundamentally different collective behaviors. In all existing synthetic systems, this hydrodynamic character is permanently set d

Rakshita Goel, S Pranav Kumar, Anmol Agrawal, Divyan Poddar

Legal research in India involves navigating long and heterogeneous documents spanning statutes, constitutional provisions, penal codes, and judicial precedents, where purely keyword-based or embedding-only retrieval systems often fail to support structured legal reasoning. Recent retrieval augmented generation (RAG) approaches improve grounding but struggle

Renan J. S. Isneri, Eric B. Santiago, Severino H. da Silva

In this work we study the asymptotic behavior of a class of damped second-order gradient systems $$ \ddot{u}(t) + a\dot{u}(t) + \nabla W(u(t)) = 0, $$ under assumptions ensuring local convexity of the potential near equilibrium and coercivity at infinity. By introducing a Lyapunov functional adapted to the geometry of the system, we establish uniform asympto

V. N. Temlyakov

This paper is devoted to the theoretical study of the efficiency, namely, stability of some greedy algorithms. In the greedy approximation theory researchers are mostly interested in the following two important properties of an algorithm -- convergence and rate of convergence. In this paper we present some results on one more important property of an algorit

Diyar Altinses, Andreas Schwung

In recent years, the development of multimodal autoencoders has gained significant attention due to their potential to handle multimodal complex data types and improve model performance. Understanding the stability and robustness of these models is crucial for optimizing their training, architecture, and real-world applicability. This paper presents an analy

Hanzhi Yang, Nina Mahmoudian

Underwater gliders are increasingly deployed in challenging missions - such as hurricane-season observations and long-endurance environmental monitoring - where strong currents and turbulence pose significant risks to navigation safety. To address these practical challenges, this paper presents a fixed-time prescribed performance control scheme for the 3D pa

Alan Inglis, Fiona Doohan, Subramani Natarajan, Breige McNulty

Mycotoxin contamination poses a significant risk to cereal crop quality, food safety, and agricultural productivity. Accurate prediction of mycotoxin levels can support early intervention strategies and reduce economic losses. This study investigates the use of neural networks and transfer learning models to predict mycotoxin contamination in Irish oat crops

Subhamoy Chatterjee, Mausumi Dikpati

Solar active regions (ARs) do not appear randomly but cluster along longitudinally warped toroidal bands ('toroids') that encode information about magnetic structures in the tachocline, where global-scale organization likely originates. Global MagnetoHydroDynamic Shallow-Water Tachocline (MHD-SWT) models have shown potential to simulate such toroids, matchin

Tony Tran, Bin Hu

This paper addresses trash detection on the TACO dataset under strict TinyML constraints using an iterative hardware-aware neural architecture search framework targeting edge and IoT devices. The proposed method constructs a Once-for-All-style ResDets supernet and performs iterative evolutionary search that alternates between backbone and neck/head optimizat

Haipeng Luo, Huawen Feng, Qingfeng Sun, Can Xu

Large Reasoning Models (LRMs) like o3 and DeepSeek-R1 have achieved remarkable progress in reasoning tasks with long cot. However, they remain computationally inefficient and struggle with accuracy when solving problems requiring complex mathematical operations. In this work, we present AgentMath, an agent framework that seamlessly integrates language models

Shaurya Gaur, Michel Vitale, Alessa Hering, Johan Kwisthout

Lung cancer is the leading cause of cancer-related mortality in adults worldwide. Screening high-risk individuals with annual low-dose CT (LDCT) can support earlier detection and reduce deaths, but widespread implementation may strain the already limited radiology workforce. AI models have shown potential in estimating lung cancer risk from LDCT scans. Howev

Mark Ballard, Guanqun Song, Ting Zhu

The rapid proliferation of satellite constellations, particularly in Low Earth Orbit (LEO), has fundamentally altered the global space infrastructure, shifting the risk landscape from purely kinetic collisions to complex cyber-physical threats. While traditional safety frameworks focus on debris mitigation, ground-based adversaries increasingly exploit radio

Keegan J. Flood, Gabriele Lobbia, Giacomo Tendas

Given a category $\mathcal{E}$, we establish sufficient conditions on a faithful isofibration $\mathcal{E}\rightarrow\operatorname{Mon}(\mathcal{V})$ valued in the category of monoids internal to a monoidal additive category $\mathcal{V}$ such that $\mathcal{E}$ admits a canonical functor to the category of first order differential calculi in $\mathcal{V}$.

Valentina Visani, Marco Pinamonti, Valentina Sammassimo, Manuela Moretto

The Choroid Plexus (ChP) is a highly vascularized brain structure that plays a critical role in several physiological processes. ASCHOPLEX, a deep learning-based segmentation toolbox with an integrated fine-tuning stage, provides accurate ChP delineations on non-contrast-enhanced T1-weighted MRI scans; however, its performance is hindered by inter-dataset va

Stanislav Jabuka

In graph theory an interesting question is whether for a fixed choice of $p\in [0,\infty]$, all simple graphs appear as sphere-of-influence graphs in some Euclidean space with respect to the $\ell_p$ metric. The answer is affirmative for $p=\infty$, negative for any $p\in (0,\infty)$, and unknown for $p=1$. The result of this work shows that for the case of

Anshul Sharma, Shujaatali Badami, Biky Chouhan, Pushpanjali Pandey

The 6G wireless aims at the Tb/s peak data rates are expected, a sub-millisecond latency, massive Internet of Things/vehicle connectivity, which requires sustainable access to audio over the air and energy-saving functionality. Cognitive Radio Networks CCNs help in alleviating the problem of spectrum scarcity, but classical sensing and allocation are still e

José Rodal

In the Alcubierre warp-drive spacetime, we investigate the following scalar curvature invariants: the scalar $I$, derived from a quadratic contraction of the Weyl tensor, the trace $R$ of the Ricci tensor, and the quadratic $r1$ and cubic $r2$ invariants from the trace-adjusted Ricci tensor. In four-dimensional spacetime the trace-adjusted Einstein and Ricci

Shijing Wang, Chaoqun Cui, Yaping Huang, Hyung Jin Chang

Human gaze provides essential cues for interpreting attention, intention, and social interaction in visual scenes, yet gaze understanding remains largely unexplored in current vision-language models (VLMs). While recent VLMs achieve strong scene-level reasoning across a range of visual tasks, there exists no benchmark that systematically evaluates or trains

Enrique Rozas Garcia, Johannes Hofmann

We provide a field-theoretical description of thermal nucleation in a one-dimensional ferromagnetic superfluid, a quantum-gas analogue of false-vacuum decay. The rate at which ground-state domains nucleate follows an Arrhenius law, with an exponential factor determined by a saddle-point of the energy functional -- the critical droplet or instanton -- and a m

Xavi Masip-Bruin, Eva Rodríguez, Admela Jukan, Panos Trakadas

6G networks promise to be the proper technology to support a wide deployment of highly demanding services, satisfying key users-related aspects such as extremely high quality, and persistent communications. However, there is no service to support if the network is not reliable enough. In this direction, it is with no doubt that security guarantees become a m

Siegfried Janz, Shurui Wang, Rubin Ma, Jean Lapointe

Silicon photonic components based on freely propagating beams in 220 nm thick Si slab waveguides are described and characterized. Examples include optical relays, waveguide crossings, couplers, and resonators with wavelength independent Q factors. The unconfined beams are manipulated using reflecting mirrors etched into the Si layer, which operate in the tot

Tanmoy Modak

The $R^2$ and the single-field-like regime of $R^2$-Higgs inflation are disfavored by the observed high spectral index $n_s$ from the combined cosmic microwave background (CMB) and baryon acoustic oscillation (BAO) measurements at the $\sim2\sigma$ level. The addition of a dimension-six $R^3$ term in the action helps alleviate this tension. We show that the

Darren J. Edwards

Shifted partial derivative (SPD) methods are a central algebraic tool for circuit lower bounds, measuring the dimension of spaces of shifted derivatives of a polynomial. We develop the Shifted Partial Derivative Polynomial (SPDP) framework, packaging SPD into an explicit coefficient-matrix formalism. This turns shifted-derivative spans into concrete linear-a

Vilas J. Shinde

The classical Fourier analysis of a time signal, in the discrete sense, provides the frequency content of signal under the assumption of periodicity. Although the original signal can be exactly recovered using an inverse transform, the time dependence of the spectrum remains inaccessible. There exist various time-frequency analysis techniques, such as the sh

Davide Breoni, Emanuele Locatelli, Luca Tubiana

We simulate the formation of cyclical arms in star polymers, focusing on the effects of solvent quality on their resulting linking complexity and gyration radius. We find that polymers circularized in bad solvent present a higher degree of linking among arms with respect to those circularized in good solvent. When both are transported to good solvent, this r

Alexandros Christoforos, Chadbourne Davis

Diffusion based approaches to long form text generation suffer from prohibitive computational cost and memory overhead as sequence length increases. We introduce SA-DiffuSeq, a diffusion framework that integrates sparse attention to fundamentally improve scalability for long document modeling. By selectively allocating attention within the diffusion process,

Prajwal Ghimire, Keyoumars Ashkan

The term artificial implies an inherent dichotomy from the natural or organic. However, AI, as we know it, is a product of organic ingenuity: designed, implemented, and iteratively improved by human cognition. The very principles that underpin AI systems, from neural networks to decision-making algorithms, are inspired by the organic intelligence embedded in

Jie Chen, Xianbin Wang

Conventional mobile networks, including both localized cell-centric and cooperative cell-free networks (CCN/CFN), are built upon rigid network topologies. However, neither architecture is adequate to flexibly support distributed integrated sensing and communication (ISAC) services, due to the increasing difficulty of aligning spatiotemporally distributed het

Balma Duch, Pere Masjuan

In this letter, we discuss the determination of the $f_0(500)$ resonance by analytic continuation through Pad\'e approximants of the $\pi\pi$-scattering amplitude from the physical region to the pole in the complex energy plane. Using as input a class of admissible parametrizations of the scalar-isoscalar $\pi\pi$ partial wave and imposing now the correct th

Shibam Das, Debanjan Karan, Babli Khatun, Nilay Kundu

We derive discontinuity relations, also known as cutting rules, and explore the analytic properties of cosmological correlators, fundamental observables of the primordial universe. Our emphasis is on how these relations arise from unitarity and hermitian analyticity in interacting quantum field theories on de Sitter space-time. Instead of analyzing wave-func

Hritik Gopal Shah, Catherine Tajmajer, Elli Ntakou

Natural disasters often inflict severe damage on distribution grids. Rapid, reliable damage assessment (DA) is essential for storm restoration, yet most optimization work targets repair dispatch after faults are identified. This paper presents a production, rolling horizon DA crew allocation system deployed across multiple U.S. states in Eversource Energy's

Sébastien Breteaux, Jérémy Faupin, Viviana Grasselli

We consider the boson star equation with a general two-body interaction potential $w$ and initial data $\psi_0$ in a Sobolev space. Under general assumptions on $w$, namely that $w$ decomposes as a sum of a finite, signed measure and an essentially bounded function, we prove that the (local in time) solution cannot propagate faster than the speed of light, u

Mattia Arundine

The Anti-de Sitter/Conformal Field Theory correspondence (AdS/CFT) is one of the most significant findings in theoretical physics and forms the basis of this thesis. Although highly powerful, the main limitation of AdS/CFT is that AdS does not appear in the real world outside of very specific limits. This limitation justifies the attempt to generalize the ho

Sondre Vik Furuseth, Guillaume Aulanier

Context. Space weather and its potential negative consequences for life on Earth has received increasing attention in recent decades. Particularly predicting CME onset has become important from a security perspective. To predict CMEs, one must first understand the dynamics leading to pre-eruptive magnetic field configurations such as flux ropes. Aims. In thi

Yash Saraswat, Abhimanyu Nag

While Ethereum has successfully achieved dynamic availability together with safety, a fundamental delay remains between transaction execution and immutable finality. In Ethereum's current Gasper protocol, this latency is on the order of 15 minutes, exposing the network to ex ante reorganization attacks, enabling MEV extraction, and limiting the efficiency of

Iman Reihanian, Yunfei Hou, Qingquan Sun

Generative AI enables personalized computer science education at scale, yet questions remain about whether such personalization supports or undermines learning. This scoping review synthesizes 32 studies (2023-2025) purposively sampled from 259 records to map personalization mechanisms and effectiveness signals in higher-education computer science contexts.

Garrett Richards, Svetlana Barkanova

STEM (Science, Technology, Engineering, and Mathematics) outreach programs in Canada, especially those oriented towards youth, play a critical role in supporting the nation's future workforce, innovation capacity, and equity across social groups in STEM fields. They constitute a large, multi-layered ecosystem connecting universities, national laboratories, n

Omer Gazit, Yael Itzhakev, Yuval Elovici, Asaf Shabtai

Radio frequency (RF) based systems are increasingly used to detect drones by analyzing their RF signal patterns, converting them into spectrogram images which are processed by object detection models. Existing RF attacks against image based models alter digital features, making over-the-air (OTA) implementation difficult due to the challenge of converting di

Jan Mikula, Miroslav Kulich

Expected-time mobile search (ETS) is a fundamental robotics task where a mobile sensor navigates an environment to minimize the expected time required to locate a hidden object. Global route optimization for ETS in static 2D continuous environments remains largely underexplored due to the intractability of objective evaluation, stemming from the continuous n

Denis Karateev, Petr Kravchuk, Andrea Manenti, Marten Reehorst

We study four-dimensional conformal field theories (CFTs) with an abelian $U(1)$ global symmetry using the conformal bootstrap approach. We obtain numerical bounds on the scaling dimensions of low-lying operators, the stress-tensor central charge, and a particular combination of the 't Hooft anomaly and the current central charge. Our analysis provides t

Władysław Sokołowski, Huma Jamil, Karol Makuch

Diffusion of particles in complex fluids and gels is difficult to describe and often lies beyond the scope of the classical Stokes-Einstein relation. One of the main lines of research over the past few decades has sought to relate diffusivity to a fundamental dissipative property of the fluid: the wave-vector-dependent shear viscosity function. Here, we use

János Kollár, Wenhao Ou

We show that a compact, complex analytic space $X$ has a bimeromorphic orbifold modification that is an isomorphism over the locally trivial orbifold locus of $X$.

Yang-Ren Liu, Biao Huang

We introduce an analytical framework to calculate the values of key observables in a strongly disordered discrete time crystal (DTC) without fitting parameter. The perturbatively obtained closed-form formulae show quantitative agreement with numerical simulations of inverse participation ratios for eigenstate localization in Fock space, Edwards-Anderson para

Meng Wang, Philipp Görz, Joschua Schilling, Keno Hassler

Detecting business logic vulnerabilities is a critical challenge in software security. These flaws come from mistakes in an application's design or implementation and allow attackers to trigger unintended application behavior. Traditional fuzzing sanitizers for dynamic analysis excel at finding vulnerabilities related to memory safety violations but largely

Fang Xu

We study neutrino non-standard interactions (NSI) induced by the trilinear $LLE$ couplings in $R$-parity-violating supersymmetry in a split-family spectrum where third-generation sleptons can be comparatively light. We systematically classify the relevant coupling patterns, derive the corresponding effective interactions and NSI parameters, and confront the

Matthias Stierle, Karsten Kraume, Martin Matzner

Data-driven analysis of business processes has a long tradition in research. However, recently the term of process mining is mostly used when referring to data-driven process analysis. As a consequence, awareness for the many facets of process analysis is decreasing. In particular, while an increasing focus is put onto technical aspects of the analysis, huma

Jarred Gillette, Michael J. Koss, Darshan Kakkad, Federica Ricci

We present an analysis of near-infrared (NIR) emission-line properties, AGN diagnostics, and circumnuclear gas dynamics for 453 hard X-ray selected (14$-$195 keV) AGN from the BAT AGN Spectroscopic Survey (BASS) NIR Data Release 3 (DR3; $\langle z \rangle=0.036$, $z<1.0$). This dataset is the largest compilation of rest-frame NIR spectroscopic observations o

Ingmar Metzler

We establish a new converse theorem for Borcherds products. Moreover, the injectivity of the Kudla-Millson theta lift is demonstrated in the O$(n,2)$ case in greater generality than is currently available in the literature. Both results are derived under the assumption of a single hyperbolic split of the base lattice.

Filippo Santoliquido, Jacopo Tissino, Ulyana Dupletsa, Marica Branchesi

The coming decade will be crucial for determining the final design and configuration of a global network of next-generation (XG) gravitational-wave detectors, including the Einstein Telescope (ET) and Cosmic Explorer (CE). In this study, and for the first time, we assessed the performance of various network configurations using neural posterior estimation (N

Pouya Asadi, Graham D. Kribs, Markus A. Luty

We demonstrate that cosmological observations place strong bounds on the reheat temperature $T_\text{RH}$ of the Standard Model (SM) in minimal models of `quirks' -- heavy fermions transforming under the SM gauge group together with a new non-Abelian gauge interaction with a confinement scale far below the mass of the fermions. These models have unique colli

Ferruccio Feruglio, Gabriele Levati, Robert Ziegler

We analyze the effective couplings of a light, spinless, gauge-singlet particle $\phi$ to on-shell photons. Starting from the most general theory at the electroweak scale, which allows for CP-violating interactions suppressed by inverse powers of an ultraviolet scale $\Lambda$, we derive the corresponding low-energy effective theory valid below the GeV scale

Jianhong Bai, Xiaoshi Wu, Xintao Wang, Xiao Fu

State-of-the-art video generative models typically learn the distribution of video latents in the VAE space and map them to pixels using a VAE decoder. While this approach can generate high-quality videos, it suffers from slow convergence and is computationally expensive when generating long videos. In this paper, we introduce SemanticGen, a novel solution t

Runtao Liu, Ziyi Liu, Jiaqi Tang, Yue Ma

Recent advances in multimodal LLMs and systems that use tools for long-video QA point to the promise of reasoning over hour-long episodes. However, many methods still compress content into lossy summaries or rely on limited toolsets, weakening temporal grounding and missing fine-grained cues. We propose a multi-agent framework in which a master LLM coordinat

Yuxi Xiao, Longfei Li, Shen Yan, Xinhang Liu

Cognitive science suggests that spatial ability develops progressively-from perception to reasoning and interaction. Yet in multimodal LLMs (MLLMs), this hierarchy remains poorly understood, as most studies focus on a narrow set of tasks. We introduce SpatialTree, a cognitive-science-inspired hierarchy that organizes spatial abilities into four levels: low-l

Javier de Cruz Pérez, Adrià Gómez-Valent, Joan Solà Peracaula

In this paper, we study several models and parameterizations of dynamical dark energy (DE) that have been studied already in the past, in conjunction with the recently proposed model $w$XCDM, the running vacuum model (RVM) with and without a threshold at $z=1$ and two variants of it, the RRVM and the ``flipped RVM'', and compare them all with the concordance

Xuanhua He, Tianyu Yang, Ke Cao, Ruiqi Wu

Current video avatar generation methods excel at identity preservation and motion alignment but lack genuine agency, they cannot autonomously pursue long-term goals through adaptive environmental interaction. We address this by introducing L-IVA (Long-horizon Interactive Visual Avatar), a task and benchmark for evaluating goal-directed planning in stochastic

Debashish Dutta, Sayan Choudhury

We study a non-Hermitian extension of the Creutz ladder with generic non-reciprocal hopping. By mapping the ladder onto two decoupled non-Hermitian Su--Schrieffer--Heeger (SSH) chains, we uncover a rich structure in parameter space under different boundary conditions. Under periodic boundary conditions, the spectrum admits a fine-tuned line in parameter spac

Yibin Lei, Shwai He, Ang Li, Andrew Yates

Recent work has shown that directly fine-tuning large language models (LLMs) for dense retrieval yields strong performance, but their substantial parameter counts make them computationally inefficient. While prior studies have revealed significant layer redundancy in LLMs for generative tasks, it remains unclear whether similar redundancy exists when these m

Michael Newns, Shirley Xu, Mingyang Liu, Zike Cheng

Through their ability to achieve cryogenic levels of noise performance while operating at room temperature, optically-pumped, solid-state (OPSS) masers show great promise as quantum sensors, oscillators, and amplifiers. We here demonstrate maser oscillation in a microwave cavity containing a crystal of pentacene-doped para-terphenyl (ptc:ptp) pumped by a sin

Daewoon Kim, Si Young Yie, Jae Sung Lee

This paper proposes FedPOD, which ranked first in the 2024 Federated Tumor Segmentation (FeTS) Challenge, for optimizing learning efficiency and communication cost in federated learning among multiple clients. Inspired by FedPIDAvg, we define a round-wise task for FedPOD to enhance training efficiency. FedPIDAvg achieved performance improvement by incorporat

Yedi Zhang, Andrew Saxe, Peter E. Latham

Neural networks trained with gradient descent often learn solutions of increasing complexity over time, a phenomenon known as simplicity bias. Despite being widely observed across architectures, existing theoretical treatments lack a unifying framework. We present a theoretical framework that explains a simplicity bias arising from saddle-to-saddle learning

Seijin Kobayashi, Yanick Schimpf, Maximilian Schlegel, Angelika Steger

Large-scale autoregressive models pretrained on next-token prediction and finetuned with reinforcement learning (RL) have achieved unprecedented success on many problem domains. During RL, these models explore by generating new outputs, one token at a time. However, sampling actions token-by-token can result in highly inefficient learning, particularly when