December 2025 arXiv papers — page 112

Teri Rumble, Zbyněk Gazdík, Javad Zarrin, Jagdeep Ahluwalia

Recent advancements in Large Language Models (LLMs) have transformed Natural Language Processing (NLP), enabling complex information retrieval and generation tasks. Retrieval-Augmented Generation (RAG) has emerged as a key innovation, enhancing factual accuracy and contextual grounding by integrating external knowledge sources with generative models. Althoug

Mohammadmahdi Nouriborji, Morteza Rohanian, Omid Rohanian

Parameter sharing in recursive transformers reduces model size but collapses layer-wise expressivity. We propose Mixture of LoRAs (MoL), a lightweight conditional-computation mechanism that inserts Low-Rank Adaptation (LoRA) experts inside a shared feed-forward network (FFN). MoL enables token-conditional weight-space modulation of the shared FFN without unt

Romana Mikušincová, Alexandra Veledina, Fabio Muleri, Raul Ciancarella

We report on the average and orbital phase-resolved polarization of Cyg X-3 in the hard state during the 2023 Imaging X-ray Polarimetry Explorer (IXPE) observational campaign. We find the polarization degree of $ 21.2 \pm 0.4 \% $ and polarization angle of $ 92.2 \pm 0.5^\circ $, well compatible with the first hard-state IXPE observation in 2022. As the obse

Perry Abdulkadir

Large language models (LLMs) are increasingly deployed behind safety guardrails such as system prompts and content filters, especially in settings where product teams cannot modify model weights. In practice these guardrails are typically hand-tuned, brittle, and difficult to reproduce. This paper studies a simple but practical alternative: treat safety guar

Keaton Naff, Jonathan J. Zhu

This is the first of two articles in which we investigate the geometry of free boundary and capillary minimal surfaces in balls $B_R\subset\mathbb{S}^3$. In this article, we extend our previous half-space intersection properties to warped products, and extend (non-)umbilicity of discs and annuli to capillary minimal surfaces in high codimension. We establish

Ashim Mahara

Alpha-Root is a cybersecurity-focused dataset collected in a single shot from the Common Crawl web graph using community detection. Unlike iterative content-scoring approaches like DeepSeekMath, we mine quality domains directly from the web graph, starting from just 20 trusted seed domains.

Helmut Prodinger

We continue on a recent concept introduced by Kariuki and Okoth, about skew 2-Dyck paths, introducing an additional down-step $L$, together with the usual steps $U$ (up) and $D$ down. There is the syntactical condition that $UL$ and $LU$ can never occur. An automaton that checks these conditions is introduced, and the relevant generating functions are obtain

Weihan Xu, Kan Jen Cheng, Koichi Saito, Muhammad Jehanzeb Mirza

Joint editing of audio and visual content is crucial for precise and controllable content creation. This new task poses challenges due to the limitations of paired audio-visual data before and after targeted edits, and the heterogeneity across modalities. To address the data and modeling challenges in joint audio-visual editing, we introduce SAVEBench, a pai

Thomas R. Cameron, Kelvin Li

In 2018, the concept of a fort in graph theory was introduced as a non-empty subset of vertices satisfying the condition that no vertex outside the set has exactly one neighbor in the set. Since then, forts have played a significant role in characterizing zero forcing sets, modeling the zero forcing number as an integer program, and generating lower bounds f

Flávio P. Cruz, José A. Santos, Victor J. V. Otoya

In this article, we provide a new method for obtaining the propagator of two three-dimensional models of electrodynamics (Maxwell-Lee-Wick-Chern-Simons and Maxwell-Deser-Jackiw). This method introduce a new set of projection operators. Then we perform a causality and unitarity analysis.

Xiaojie Tao, Yaoyu Fan, Zhaoyi Ye, Rajit Gadh

The integration of heavy-duty electric vehicles (EVs) with Vehicle-to-Grid (V2G) capability offers a promising solution to enhance grid stability by providing primary frequency response in power systems. This paper investigates the potential of heavy-duty EVs to support the California power grid under different charging strategies: immediate, delayed, and co

Millend Roy, Agostino Capponi, Vladimir Pyltsov, Yinbo Hu

Electricity markets are under increasing pressure to maintain reliability amidst rising renewable penetration, demand variability, and occasional price shocks. Traditional capacity market designs often fall short in addressing this by relying on expected-value metrics of energy unserved, which overlook risk exposure in such systems. In this work, we present

Pouya Ahadi, Blair Winograd, Camille Zaug, Karunesh Arora

Active Learning (AL) has garnered significant interest across various application domains where labeling training data is costly. AL provides a framework that helps practitioners query informative samples for annotation by oracles (labelers). However, these labels often contain noise due to varying levels of labeler accuracy. Additionally, uncertain samples

Ashim Mahara

This project presents a unified detection framework that constructs a complete corpus of Microsoft Graph permissions, generates consistent LLM-based risk scores, and integrates them into a real-time detection engine to identify malicious OAuth consent activity.

Yongmin Yoo, Seungwoo Kim, Jingjiang Liu

Valuing intangible assets under uncertainty remains a critical challenge in the strategic management of technological innovation due to the information asymmetry inherent in high-dimensional technical specifications. Traditional bibliometric indicators, such as citation counts, fail to address this friction in a timely manner due to the systemic latency inhe

Furong Jia, Yuan Pu, Finn Guo, Monica Agrawal

Large language models (LLMs) excel on multiple-choice clinical diagnosis benchmarks, yet it is unclear how much of this performance reflects underlying probabilistic reasoning. We study this through questions from MedQA, where the task is to select the most likely diagnosis. We introduce the Frequency-Based Probabilistic Ranker (FBPR), a lightweight method t

Cashen Diniz, Mark D. Fuge

OptiWing3D is the first publicly available dataset of high-fidelity shape optimized 3D wing geometries. Existing aerodynamics datasets are either limited to 2D simulations, lack optimization, or derive diversity solely from perturbations to a single baseline design, constraining their application as benchmarks to inverse design approaches and in the study of

Oem Trivedi, Abraham Loeb

We develop a differential formulation of the Loeb Scale that extends the original static framework into a radially evolving, real time classification scheme for interstellar objects. By promoting each anomaly metric to a function of heliocentric distance and introducing a relaxation equation for the effective score, our method incorporates memory, hysteresis

Paulo Henrique da Costa, Alberto Ohashi, Francesco Russo

This paper provides the time-dependent $L^2$-martingale representation of the forward stochastic integral where the driving noise is the Riemann-Liouville fractional Brownian motion with parameter $\frac{1}{2} < H < 1$ and the integrand is a square-integrable adapted process. As a by-product, we obtain the exact $L^2$-isometry of the forward stochastic integ

Máximo Bañados, Daniela Bennett

The Gauss-Bonnet density `a la Palatini' is not a total derivative in four dimensions. We study spherically symmetric fields for the torsion-free theory. The resulting equations are highly complicated but we show the existence of unexpected hidden gauge symmetries, beyond diffeomorphisms and Weyl transformations.

John C. Boik, Kobus Esterhuysen, Jacqueline B. Hynes, Axel Constant

Advances in automated systems afford new opportunities for intelligent management of energy at household, local area, and utility scales. Home Energy Management Systems (HEMS) can play a role by optimizing the schedule and use of household energy devices and resources. One challenge is that the goals of a household can be complex and conflicting. For example

A. Della Corte, L. Guglielmi, M. Farotti

We study finite dimensional quantum systems with arbitrary collapse events, establishing, under no-information-erasure conditions, a structural no-go for operational irreversibility along single branches of the collapse dynamics. More precisely, we prove that, for every physically admissible selector of the collapse dynamics, there exists a topologically clo

Shyam Popat, Zhengyan Wu

In this paper, we establish the ergodicity of generalized Dean--Kawasaki equations with correlated noise and Dirichlet boundary conditions. In contrast to the ergodicity results of Fehrman, Gess, and Gvalani arXiv:2206.14789, our analysis accommodates irregular, square-root type noise coefficients. For such irregular coefficients, we prove that the law of th

Aritra Banik, Mano Prakash Parthasarathi, Venkatesh Raman, Diya Roy

The Minimum Consistent Subset (MCS) problem arises naturally in the context of supervised clustering and instance selection. In supervised clustering, one aims to infer a meaningful partitioning of data using a small labeled subset. However, the sheer volume of training data in modern applications poses a significant computational challenge. The MCS problem

Petr Koklev

Financial institutions face a trade-off between predictive accuracy and interpretability when deploying machine learning models for credit risk. Monotonicity constraints align model behavior with domain knowledge, but their performance cost - the price of monotonicity - is not well quantified. This paper benchmarks monotone-constrained versus unconstrained g

Marina Echeverria-Ferrero, Nicolas Moreno, Marco Ellero

Blood coagulation is governed by tightly regulated reaction networks that unfold within a flowing, heterogeneous microvascular environment. Reduced kinetic models of the intrinsic and extrinsic pathways have seen limited in vitro validation, and their behavior within spatially resolved flow fields remains largely unexplored. Here, we embed two established re

Sonal Prabhune, Balaji Padmanabhan, Kaushik Dutta

Large Language Models (LLMs) are increasingly deployed in business-critical domains such as finance, education, healthcare, and customer support, where users expect consistent and reliable recommendations. Yet LLMs often exhibit variability when prompts are phrased with minor differences, even when semantically equivalent. Such inconsistency undermines trust

Cristian Parra-Aldana, Juan Sosa

Bayesian hierarchical linear models provide a natural framework to analyze nested and clustered data. Classical estimation with Markov chain Monte Carlo produces well calibrated posterior distributions but becomes computationally expensive in high dimensional or large sample settings. Variational Inference and Stochastic Variational Inference offer faster op

Saad Alqithami

As generative agents become increasingly sophisticated and deployed in long-term interactive scenarios, their memory management capabilities emerge as a critical bottleneck for both performance and privacy. Current approaches either maintain unlimited memory stores, leading to computational intractability and privacy concerns, or employ simplistic forgetting

Patrick Kostelac, Xuerui Wang, Anahita Jamshidnejad

Modern engineering systems, such as autonomous vehicles, flexible robotics, and intelligent aerospace platforms, require controllers that are robust to uncertainties, adaptive to environmental changes, and safety-aware under real-time constraints. RL offers powerful data-driven adaptability for systems with nonlinear dynamics that interact with uncertain env

Enrique Otarola, Daniel Quero, Matias Sasso

We consider a bilinear optimal control problem with pointwise tracking for a semilinear elliptic PDE in two and three dimensions. The control variable enters the PDE as a (reaction) coefficient and the cost functional contains point evaluations of the state variable. These point evaluations lead to an adjoint problem with a linear combination of Dirac measur

Carlos Saji, Mario A. Castro, Vagson L. Carvalho-Santos, Eduardo Saavedra

Topological magnetic textures confined to two-dimensional (2D) non-orientable manifolds exhibit behaviors absent in planar systems. We investigate bimerons on M\"obius surfaces and show that the lack of global orientation alters conservation laws, yielding geometry-dependent topology and dynamics. Micromagnetic simulations reveal that the helical twist and n

Nir Lavee, Nati Linial

Consider the random process that starts with $n$ vertices and no edges, where the edges of $K_n$ are added one at a time in a uniformly chosen random order $e_1, e_2,\ldots, e_{\binom{n}{2}}$. Let $T$ be the earliest time at which $e_1$ belongs to a cycle in this evolving random graph. By solving the appropriate graph enumeration problem we show that $\mathb

Junyi Peng, Jin Li, Johan Rohdin, Lin Zhang

This paper presents the BUT submission to the WildSpoof Challenge, focusing on the Spoofing-robust Automatic Speaker Verification (SASV) track. We propose a SASV framework designed to bridge the gap between general audio understanding and specialized speech analysis. Our subsystem integrates diverse Self-Supervised Learning front-ends ranging from general au

Gangbin Yan, Pierfrancesco Ombrini, Zhichu Tang, Shakul Pathak

Ionic diffusion in solids is central to energy storage, electronics, and catalysis, yet its chemical origins are difficult to resolve because conventional diffusion models struggle with effects of confinement, crystallographic disorder, lattice distortions, and coupling to electronic or phononic carriers. These challenges are especially pronounced in battery

Jonathan Herbst, Michael Pellauer, Sherief Reda

We introduce a high-throughput neural network accelerator that embeds most network layers directly in hardware, minimizing data transfer and memory usage while preserving a degree of flexibility via a small neural processing unit for the final classification layer. By leveraging power-of-two (Po2) quantization for weights, we replace multiplications with sim

Yue Pan, Shany Danieli, Jenny E. Greene, Jiaxuan Li

We present a statistical census of bright, star-forming satellite galaxies around Milky Way (MW) analogs using the first data release of the Merian Survey. Our sample consists of 393 MW analogs with stellar masses $10^{10.5} < M_{\star, \rm host} < 10^{10.9} M_\odot$ at redshifts $0.07 < z < 0.09$, all central galaxies of their own dark matter halos. Using p

Davor Dragicevic

The robustness property of exponential dichotomies refers to the stability of this notion under small linear perturbations. In recent work~\cite{PPX}, the authors have identified a new class of perturbations under which the notion of a nonuniform exponential dichotomy persists. In the present paper, we show that it is possible to extend this class. Moreover,

Md. Najib Hasan, Imran Ahmad, Sourav Basak Shuvo, Md. Mahadi Hasan Ankon

Medical image classifiers detect gastrointestinal diseases well, but they do not explain their decisions. Large language models can generate clinical text, yet they struggle with visual reasoning and often produce unstable or incorrect explanations. This leaves a gap between what a model sees and the type of reasoning a clinician expects. We introduce a fram

Yunpeng Xu, Wenge Guo, Zhi Wei

Reliable uncertainty quantification is essential for deploying machine learning systems in high-stakes domains. Conformal prediction provides distribution-free coverage guarantees but often produces overly large prediction sets, limiting its practical utility. To address this issue, we propose \textit{Selective Conformal Risk Control} (SCRC), a unified frame

Aleksei Egorov, Lei Zhang, Erik van der Giessen, Francesco Maresca

Steel is a key structural material because of its considerable strength and ductility. However, when exposed to hydrogen, it is prone to embrittlement. Mechanistic understanding of the origin of hydrogen embrittlement is hampered by the lack of reliable interatomic potentials. Here, we perform large-scale molecular dynamics simulations of crack propagation a

Kuan Fang, Yuxin Chen, Xinghao Zhu, Farzad Niroui

We present SAGA, a versatile and adaptive framework for visuomotor control that can generalize across various environments, task objectives, and user specifications. To efficiently learn such capability, our key idea is to disentangle high-level semantic intent from low-level visuomotor control by explicitly grounding task objectives in the observed environm

Gregory Dresden, Xiaoya Gao

The equation commonly known as Sury's identity is a deceptively simple summation formula that connects the Lucas numbers, Fibonacci numbers, and powers of two. Many authors have given extensions and generalizations over the years; in this paper, we take a different approach that allows us to produce a good number of new summation formulas, all from elementar

Dingyi Yang, Qin Jin

In this work, we conduct systematic research in a challenging area: the automatic evaluation of book-length stories (>100K tokens). Our study focuses on two key questions: (1) understanding which evaluation aspects matter most to readers, and (2) exploring effective methods for evaluating lengthy stories. We introduce the first large-scale benchmark, LongSto

Tim Santens

We prove the analogue of Malle's conjecture for the global function field $\F_q(t)$ with $q$ sufficiently large, including a precise formula for the leading constant. The main ingredients are the recent breakthrough of Landesman--Levy on the stable homology of Hurwitz spaces, a novel interpretation of the Frobenius fixed components of Hurwitz spaces in terms

Sahibpreet Singh, Manjit Singh

AI-powered greenwashing has emerged as an insidious challenge within corporate sustainability governance, exacerbating the opacity of environmental disclosures and subverting regulatory oversight. This study conducts a comparative legal analysis of criminal liability for AI-mediated greenwashing across India, the US, and the EU, exposing doctrinal lacunae in

Yves Matanga, Chunling Du, Etienne van Wyk

Context: The city of Kinshasa faces severe traffic congestion, requiring strategic infrastructure capacity enhancements. Although a comprehensive master plan has been proposed, its implementation requires substantial financial investment, which remains constrained in the Democratic Republic of the Congo (DRC), an emerging economy. This research proposes a tr

Harri Hakula, Oona Rainio, Matti Vuorinen

We study the conformal capacity ${\rm cap}(\Omega,K)$ where $\Omega$ is a bounded domain of $\mathbb{R}^2$ and $K$ is a compact connected set in $\Omega$. Because the exact numerical value of the capacity is known only in a handful of special cases, it is important to find estimates for the capacity in terms of domain functionals, simpler than the capacity i

Eyal Kaplan

In his study of the Ultrapower Axiom (UA), Goldberg revealed a connection between UA and the determinacy of certain games that witness Lipschitz reducibility between ultrafilters. In particular, he analyzed the relationship between the Ketonen and Lipschitz orders - two natural extensions of the Mitchell order from normal measures to arbitrary $\sigma$-compl

Shangxuan Luo, Joshua Reiss

Procedural Music Generation (PMG) is an emerging field that algorithmically creates music content for video games. By leveraging techniques from simple rule-based approaches to advanced machine learning algorithms, PMG has the potential to significantly improve development efficiency, provide richer musical experiences, and enhance player immersion. However,

Nathaniel Smith, Yu Wang

The technological advancements facilitating the rapid development of cyber-physical systems (CPS) also render such systems vulnerable to cyber attacks with devastating effects. Supervisory control is a commonly used control method to neutralize attacks on CPS. The supervisor strives to confine the (symbolic) paths of the system to a desired language via sens

Kaustav Chatterjee, Joshua Li, Kundan Parajulee, Jared Schwennesen

Steep-profiled Highway Railway Grade Crossings (HRGCs) pose safety hazards to vehicles with low ground clearance, which may become stranded on the tracks, creating risks of train vehicle collisions. This research develops a framework for network level evaluation of hang-up susceptibility of HRGCs. Profile data from different crossings in Oklahoma were collec

Marcelo Bongarti, Michael Hintermüller

A generalized Nash equilibrium problem (GNEP) in Banach space consists of $N>1$ optimal control problems with couplings in both the objective functions and, most importantly, in the feasible sets. We address the existence of equilibria for convex GNEPs in Banach space. We show that the standard assumption of lower semicontinuity of the set-valued constraint

Elizabeth Denne

Given a thin strip of paper, tie a knot, connect the ends, and flatten into the plane. This is a physical model of a folded ribbon knot in the plane, first introduced by Louis Kauffman. We study the folded ribbonlength of these folded ribbon knots, which is defined as the knot's length-to-width ratio. The {\em ribbonlength problem} asks to find the infimal f

Jose Luis Castanon Remy, Shouhuai Xu

Space infrastructures represent an emerging domain that is critical to the global economy and society. However, this domain is vulnerable to attacks, including cyber attacks and other kinds of attacks. To enhance the resilience of this domain, we must understand these attacks that can be waged against it and the defenses that can be employed to mitigate thes

Ajeet Kumar, Uditanshu Sadual

Mutually Unbiased bases has various application in quantum information procession and coding theory. There can be maximum d + 1 MUBs in C^d and d/2 +1 MUBs in R^d. But , over R^d MUBs are known to be non existent when d is odd and for most of the other even d there are mostly 3 Real MUBs. In case of C^d the construction for complete set of MUBs are known for

Samruddhi Baviskar

We evaluate adversarial robustness in tabular machine learning models used in financial decision making. Using credit scoring and fraud detection data, we apply gradient based attacks and measure impacts on discrimination, calibration, and financial risk metrics. Results show notable performance degradation under small perturbations and partial recovery thro

Mohammad Mahdi Razmjoo, Mohammad Mahdi Sharifian, Saeed Bagheri Shouraki

Despite their remarkable performance, deep neural networks exhibit a critical vulnerability: small, often imperceptible, adversarial perturbations can lead to drastically altered model predictions. Given the stringent reliability demands of applications such as medical diagnosis and autonomous driving, robust detection of such adversarial attacks is paramoun

Eric A. Carlen, David A. Huse, Joel L. Lebowitz

We study composite open quantum systems with a finite-dimensional state space ${\mathcal H}_A\otimes {\mathcal H}_B$ governed by a Lindblad equation $\rho'(t) = {\mathcal L}_\gamma \rho(t)$ where ${\mathcal L}_\gamma\rho = -i[H,\rho] + \gamma {\mathcal D} \rho$, and ${\mathcal D}$ is a dissipator ${\mathcal D}_A\otimes I$ acting non-trivially only on part $A

N. K. B. M. P. K. B. Narasinghe, Uthayasanker Thayasivam

Large-scale multimodal foundation models, particularly Contrastive Captioners (CoCa), have achieved state-of-the-art results by unifying contrastive alignment with generative captioning. While zero-shot transfer capabilities are well-documented, the adaptation of these generative-contrastive hybrids to downstream tasks with extreme data scarcity (few-shot le

Andrew J. Andis, Songlin Lyu, Bingwei Long, Sebastian König

In this work, we study the longitudinal response function of the deuteron up to next-to-next-to-leading order in chiral effective field theory (Chiral EFT). We use an approach that maintains exact renormalization group (RG) invariance at each order of the EFT expansion by treating all subleading corrections in perturbation theory. To that end, we extent the

Yongyuan Liang, Xiyao Wang, Yuanchen Ju, Jianwei Yang

Scaling large multimodal models (LMMs) to 3D understanding poses unique challenges: point cloud data is sparse and irregular, existing models rely on fragmented architectures with modality-specific encoders, and training pipelines often suffer from instability and poor scalability. We introduce Lemon, a unified transformer architecture that addresses these c

Congzhou M Sha

Flow matching has emerged as a powerful framework for generative modeling through continuous normalizing flows. We investigate a potential topological constraint: when the prior distribution and target distribution have mismatched topology (e.g., unimodal to multimodal), the optimal velocity field under standard flow matching objectives may exhibit spatial d

Sourav Nayak, Dhriti Sundar Patra, Vladimir Rovenski

The interest of mathematicians in metric $f$-manifolds, in particular, almost contact metric manifolds, is motivated by the study of the geometry and dynamics of contact foliations, as well as their applications in physics. Weak metric $f$-manifolds, defined by V. Rovenski and R. Wolak (2022), open a new perspective on classical theory of $f$-manifolds and d

M. Serra-Ricart, J. Licandro, M. R. Alarcon

We present observations of the detection of a faint high-latitude jet in the inner coma of comet 3I/ATLAS that coincides with the broad plume detected in visible images along PA $280 \pm 10^{\circ}$. A detailed analysis shows that the jet was clearly detected on seven nights (2025, August~3, 5, 18, 19, 21, 24, and~29). The jet maintains an almost, though not

Chris Latimer, Nicoló Boschi, Andrew Neeser, Chris Bartholomew

Agent memory has been touted as a dimension of growth for LLM-based applications, enabling agents that can accumulate experience, adapt across sessions, and move beyond single-shot question answering. The current generation of agent memory systems treats memory as an external layer that extracts salient snippets from conversations, stores them in vector or g

Mengqian Wu, Jiayi Zhang, Raymond Z. Zhang

Debate has been widely adopted as a strategy to enhance critical thinking skills in English Language Arts (ELA). One important skill in debate is forming effective argumentation, which requires debaters to select supportive evidence from literature and construct compelling claims. However, the training of this skill largely depends on human coaching, which i

Yihan Hong, Hengxiang Feng, Yinghan Wang, Boxuan Li

The approval of the Bitcoin Spot ETF in January 2024 marked a transformative event in cryptocurrency markets, signaling increased institutional adoption and integration into traditional finance. This study examines Bitcoin's changing relationships with traditional assets, including equities, gold, and fiat currencies, following this milestone. Using rolling

Yalong Yu, Hengwei Zhang, Xiaoyan Liu, Tao Chu

The magnetoelectric coupling effect serves as a crucial bridge between electrical and magnetic order parameters in condensed matter physics, forming the physical basis for the development of next-generation low-power information storage and sensing technologies. However, material systems exhibiting this effect at room temperature are extremely rare, and the

Nadara Hudson, Ryan Endsley, John Chisholm

JWST continues to reveal an astonishing number of massive quiescent galaxies at $z>4$, with number densities $\gtrsim10\times$ higher than model predictions. NIRSpec spectra imply that many of these systems underwent intense starburst episodes (SFR$\,\gtrsim300M_\odot$/yr), though direct evidence of such starbursts in the Gyr largely comes from exceptionally

Hanyu Cai, Binqi Shen, Lier Jin, Lan Hu

Prompt engineering has emerged as a critical factor influencing large language model (LLM) performance, yet the impact of pragmatic elements such as linguistic tone and politeness remains underexplored, particularly across different model families. In this work, we propose a systematic evaluation framework to examine how interaction tone affects model accura

Saeed Abdallah, Mahmoud A. Albreem, Bassel Al Homssi, Mohamed Saad

Ambient backscatter communication (AmBC) has emerged as a highly attractive paradigm for energy-efficient communication. Full-duplex multi-tag AmBC systems provide the scalability and efficient spectrum utilization essential for next generation Internet-of-Things (IoT) networks. However, the presence of multiple tags, self-interference and hardware impairmen

Qingyuan Bai, Peter J. Haine

Given an open-closed decomposition of the stratifying poset, we construct a new semi-orthogonal decomposition of the $\infty$-category of constructible sheaves on a stratified space admitting an exit-path $\infty$-category. From this we obtain a direct sum decomposition of the localizing invariants of the $\infty$-category of constructible sheaves. Since the

Junbo Jacob Lian, Mingyang Yu, Kaichen Ouyang, Shengwei Fu

Black-box optimization often relies on evolutionary and swarm algorithms whose performance is highly problem dependent. We view an optimizer as a short program over a small vocabulary of search operators and learn this operator program separately for each problem instance. We instantiate this idea in Operator-Programmed Algorithms (OPAL), a landscape-aware f

Li Lin, Siyuan Xin, Yang Cao, Xiaochun Cao

Watermarking large language models (LLMs) is vital for preventing their misuse, including the fabrication of fake news, plagiarism, and spam. It is especially important to watermark LLM-generated code, as it often contains intellectual property.However, we found that existing methods for watermarking LLM-generated code fail to address comment removal attack.

Jyotsana Sharma, Shivani Bhardwaj, Sudhir K Pandey

In this work, we investigate the microscopic origin of magnetism in $\mathrm{SrCoO_3}$ by incorporating electronic correlations within the dynamical mean-field theory (DMFT) framework. We note a remarkable agreement of the calculated magnetic observables ( saturation magnetization $\sim$2.4 $\mu_B$; magnetic transition temperature, $T_c$$\sim$350 K) with the

Anil Khachi

Distance-dependent phase shifts, amplitude functions, and radial wave functions for neutron-alpha elastic scattering are studied using the Variable Phase Approach. The microscopic KKNN potential is employed to calculate scattering properties for the $S_{1/2}$, $P_{3/2}$, and $P_{1/2}$ partial waves over a range of laboratory energies. The variable phase equa

Boyang Yan

The transition of Large Language Models (LLMs) from passive code generators to autonomous agents introduces significant safety risks, specifically regarding destructive commands and inconsistent system states. Existing commercial solutions often prioritize interactive user safety, enforcing authentication barriers that break the headless loops required for t

Anastasiia Alokhina, Pan Li

Transformers exhibit a notable property of \emph{size generalization}, demonstrating an ability to extrapolate from smaller token sets to significantly longer ones. This behavior has been documented across diverse applications, including point clouds, graphs, and natural language. Despite its empirical success, this capability still lacks some rigorous theor

Sander Beckers

I develop a novel semantics for probabilities of counterfactuals that generalizes the standard Pearlian semantics: it applies to probabilistic causal models that cannot be extended into realistic structural causal models and are therefore beyond the scope of Pearl's semantics. This generalization is needed because, as I show, such probabilistic causal models

Dong Liu, Juan S. Giraldo, Peter Palensky, Pedro P. Vergara

Centralised reinforcement learning (RL) for voltage magnitude regulation in distribution networks typically involves numerous agent-environment interactions and power flow (PF) calculations, inducing computational overhead and privacy concerns over shared data. Thus, we propose a distributed RL algorithm to regulate voltage magnitude. First, a dynamic Theven

Erik Hoel

Scientific theories of consciousness should be falsifiable and non-trivial. Recent research has given us formal tools to analyze these requirements of falsifiability and non-triviality for theories of consciousness. Surprisingly, many contemporary theories of consciousness fail to pass this bar, including theories based on causal structure but also (as I dem

Anurag Dutt, Young Won Choi, Avirup Sil, Anshul Gandhi

With the widespread adoption of Large Language Models (LLMs), energy costs of running LLMs is quickly becoming a critical concern. However, precisely measuring the energy consumption of LLMs is often infeasible because hardware-based power monitors are not always accessible and software-based energy measurement tools are not accurate. While various predictio

Yunlong He, Gwilherm Lesné, Ziqian Liu, Michaël Soumm

Recent advancements in image synthesis have enabled high-quality image generation and manipulation. Most works focus on: 1) conditional manipulation, where an image is modified conditioned on a given attribute, or 2) disentangled representation learning, where each latent direction should represent a distinct semantic attribute. In this paper, we focus on a

Zhe Liu, Runhui Huang, Rui Yang, Siming Yan

Although multi-modal large language models (MLLMs) have shown strong capabilities across diverse domains, their application in generating fine-grained 3D perception and prediction outputs in autonomous driving remains underexplored. In this paper, we propose DrivePI, a novel spatial-aware 4D MLLM that serves as a unified Vision-Language-Action (VLA) framewor

Bhoomika Maheshwari, Paul Stevenson, P. Van Isacker

Quantum computing offers a scalable approach to solving the nuclear shell model, a highly complex and exponentially scaled many-body problem. This work presents a numerical simulation of the subspace search variational quantum eigensolver (SSVQE) combined with an adaptive derivative-assembles pseudo-trotter (ADAPT) ansatz to obtain the low-lying states of an

Mohammad H. Taufik, Tariq Alkhalifah

Bayesian full waveform inversion (FWI) offers uncertainty-aware subsurface models; however, posterior sampling directly on observed seismic shot records is rarely practical at the field scale because each sample requires numerous wave-equation solves. We aim to make such sampling feasible for large surveys while preserving calibration, that is, high uncertai

Maher Boudabra

This paper continues our earlier work \cite{becher2025skorokhod} on variational questions arising from the planar Skorokhod embedding problem (PSEP). Given a centered probability measure $μ$ on $\mathbb R$ with finite second moment, PSEP asks for a simply connected domain $U\subset\mathbb C$ containing $0$ such that planar Brownian motion $(Z_t)$ started at

Mengying Yan, Ziye Tian, Siqi Li, Nan Liu

Clinical decision support tools built on electronic health records often experience performance drift due to temporal population shifts, particularly when changes in the clinical environment initially affect only a subset of patients, resulting in a transition to mixed populations. Such case-mix changes commonly arise following system-level operational updat

Yichen Liu, Hongyu Wu, Bo Liu

Smart grids rely on high-dimensional numeric telemetry and explicit operating rules to maintain reliable and secure operation. Recent large language models (LLMs) are increasingly considered as candidate decision-support components for power system operations, yet most deployments focus on textual logs, alerts, or operator messages and do not directly addres

Mizuho Aoki, Kohei Honda, Yasuhiro Yoshimura, Takeshi Ishita

This paper presents Vision-Language Global Localization (VLG-Loc), a novel global localization method that uses human-readable labeled footprint maps containing only names and areas of distinctive visual landmarks in an environment. While humans naturally localize themselves using such maps, translating this capability to robotic systems remains highly chall

Shivansh Sahni, Wenzhi Zhang

The Liquid Reasoning Transformer (LRT) is a transformer architecture designed for inference with adaptive depths using iterative changes, discard-based correction, and a learned stopping mechanism. Instead of relying on a single feedforward pass, the model updates a recurrent reasoning token across multiple internal steps, allowing it to correct early errors

Sreemaee Akshathala, Bassam Adnan, Mahisha Ramesh, Karthik Vaidhyanathan

Recent advances in agentic AI have shifted the focus from standalone Large Language Models (LLMs) to integrated systems that combine LLMs with tools, memory, and other agents to perform complex tasks. These multi-agent architectures enable coordinated reasoning, planning, and execution across diverse domains, allowing agents to collaboratively automate compl

Tiange Zhang, Zhimeng Huang, Xiandong Meng, Kai Zhang

Neural Video Compression has emerged in recent years, with condition-based frameworks outperforming traditional codecs. However, most existing methods rely solely on the previous frame's features to predict temporal context, leading to two critical issues. First, the short reference window misses long-term dependencies and fine texture details. Second, propa

Md. Hasib Ur Rahman

As Large Language Models (LLMs) become ubiquitous, the challenge of securing them against adversarial "jailbreaking" attacks has intensified. Current defense strategies often rely on computationally expensive external classifiers or brittle lexical filters, overlooking the intrinsic dynamics of the model's reasoning process. In this work, the Laminar Flow Hy

Rustem N. Garifullin

This paper examines the classification of hyperbolic equations. We study a class of equations of the form $$\frac{\partial^2 u}{\partial x\partial y}=F\left(\frac{\partial u}{\partial x},\frac{\partial u}{\partial y},u\right),$$ where $u(x,y)$ is the unknown function and $x,y$ are independent variables. The classification is based on the requirement for the

Manuel Bentele, Andreas Podelski, Axel Sikora, Bernd Westphal

Temporal HAL-API Dependencies (THADs) can be useful to capture an interesting class of correctness properties in embedded software development. They demand a moderate effort for specification (which can be done via program annotations) and verification (which can be done automatically via software model checking). In this sense, they have the potential to fo

Mohammad Abu-Shaira, Weishi Shi

Despite extensive focus on techniques for evaluating the performance of two learning algorithms on a single dataset, the critical challenge of developing statistical tests to compare multiple algorithms across various datasets has been largely overlooked in most machine learning research. Additionally, in the realm of Online Learning, ensuring statistical si

Álvaro Fernández Corral, Yahya Saleh

We enhance the approximation capabilities of algebraic polynomials by composing them with homeomorphisms. This composition yields families of functions that remain dense in the space of continuous functions, while enabling more accurate approximations. For univariate continuous functions exhibiting a finite number of local extrema, we prove that there exist

Mohammad Abu Shaira, Yunhe Feng, Heng Fan, Weishi Shi

Real-world data sets often exhibit temporal dynamics characterized by evolving data distributions. Disregarding this phenomenon, commonly referred to as concept drift, can significantly diminish a model's predictive accuracy. Furthermore, the presence of hyperparameters in online models exacerbates this issue. These parameters are typically fixed and cannot